2007, volume 67, issue 1

DRUGS 2007, Vol 67, No. 1

Drugs2007, Volume 67, Issue 1

Page # Article/Title

Current Opinion

1

Inhibiting Costimulatory Activation of T Cells: A Viable Treatment Option For Rheumatoid Arthritis?Louise C Pollard

·

Leading Article

11

Squalene Synthase Inhibitors: Clinical Pharmacology and Cholesterol-Lowering Potential.Valentine Charlton-Menys; Paul N Durrington

·

Therapy In Practice

17

Practical Issues and Challenges in the Diagnosis and Treatment of Pulmonary Sarcoidosis.Surinder K Jindal

·

Review Article

27

Serotonergic Drugs: Effects on Appetite Expression and Use for the Treatment of Obesity.Jason C G Halford; Joanne A Harrold; Emma J Boyland; Clare L Lawton; John E Blundell

·

57 Therapeutic Applications of Sildenafil Citrate in the Management of Paediatric Pulmonary Hypertension.Leah Leibovitch; Ilan Matok; Gideon Paret

·

75 Ocular Adverse Effects Associated with Systemic Medications: Recognition and Management.Ricardo M Santaella; Frederick W Fraunfelder

·

Adis Drug Profile

95

Fixed-Dose Combination Lercanidipine/Enalapril.Philip I Hair; Lesley J Scott; Caroline M Perry

·

Guest Commentaries

107

Fixed-Dose Combination Lercanidipine/Enalapril: A Viewpoint by Nicolas R. Robles.Nicolás R Robles

·

107 Fixed-Dose Combination Lercanidipine/Enalapril: A Viewpoint by Roland Asmar.Roland Asmar

·

Adis Drug Profile

109 Limaprost.Tracy Swainston Harrison; Greg L Plosker

·

Guest Commentaries

119

Limaprost: A Viewpoint by Akira Dezawa.Akira Dezawa

·

119 Limaprost: A Viewpoint by Shin-Ichi Konno.Shin-ichi Konno

·

Adis Drug Evaluation

121

Fenofibrate: A Review of its Use in Primary Dyslipidaemia, the Metabolic Syndrome and Type 2 DiabetesMellitus.Gillian M Keating; Katherine F Croom

·

Drugs 2007; 67 (1): 1-9CURRENT OPINION 0012-6667/07/0001-0001/$49.95/0

2007 Adis Data Information BV. All rights reserved.

Inhibiting Costimulatory Activation ofT CellsA Viable Treatment Option For Rheumatoid Arthritis?

Louise C. Pollard

Academic Department of Rheumatology, Kings College London, London, UK

There is now good evidence that T cells play a central role in the inflammatoryAbstractpathway that leads to the persistent synovitis that causes joint damage in rheuma-toid arthritis (RA). T cells require two signals to become activated. The secondstep in the activation of T cells involves costimulatory pathways, the bestdescribed pathway being the binding of CD28 on T cells to CD80/86 onantigen-presenting cells. This observation has led to the development of a newcategory of biological response modifier. Abatacept is a fusion protein (cytotoxicT-lymphocyte-associated antigen-4 immunoglobulin [CTLA4Ig]); which blocksthe binding of CD28 by avidly binding CD80/86. Without this costimulatoryactivation, the T cell becomes anergic.

Abatacept has consistently been shown to improve the signs and symptoms ofRA in phase II and phase III trials in patients with an inadequate response tomethotrexate and anti-tumour necrosis factor (TNF) therapy. Onset of action israpid and efficacy is maintained during the period of treatment. Recent trials havealso provided evidence of improvement in quality-of-life measures and radio-graphic progression. The safety profile to date has also been favourable andsupports the theory that targeting naive T cells early in the inflammatory pathwaywill lead to immunomodulation rather than immunosuppression. The evidenceproduced so far suggests that abatacept will be a useful addition to the availabletherapies for patients with RA.

Rheumatoid arthritis (RA) is the most common matory pathways involved. Although these drugschronic polyarthritis, and continuing inflammation have undoubtedly made a significant difference toleads to an increase in morbidity and mortality. RA RA treatment, they have not been without problems.causes disability through persistent synovitis, joint Currently there are three anti-tumour necrosis factordamage and high levels of pain and fatigue. RA has (TNF) agents available; and as a class of drugs thetraditionally been treated with disease-modifying predominant adverse effects are the increased risk ofantirheumatic drugs (DMARDs); however, the treat- infection and, in particular, the reactivation of latentment of RA has been revolutionised over the last 2 tuberculosis (TB). Anti-TNF therapy can also wors-decades with the introduction of ‘biological thera- en heart failure and not all patients respond to anti-pies’. This stemmed from advances in the under- TNF therapy. The other currently available biologi-standing of the pathogenesis of RA and the inflam- cal therapy is anakinra (anti-interleukin [IL]-1-re-

2 Pollard

ceptor-antagonist). While anakinra has been shown combination of release of matrix metalloproteinasesto improve the signs and symptoms of RA, it has not and activation of osteoclasts.[4]

proven as efficacious as anti-TNF therapy. There is Rather than a simple one-ligand activation path-ongoing research investigating other potential tar- way, T cells require two distinct but simultaneousgets in the inflammatory pathway. Abatacept targets signals to become activated. The first signal is thethe costimulatory pathways which are needed to presentation of a major histocompatibility complex-activate T cells. antigen complex on an antigen-presenting cell

(APC), binding to an antigen-specific receptor onThis review describes the rationale and evidencethe T-cell surface.[5] The second signal is via afor the use of costimulation inhibitors of T cellscostimulatory pathway, without this signal the T cell(abatacept) in the treatment of RA. To identify arti-becomes anergic.[6] The costimulatory pathway con-cles for this review a MEDLINE search was under-sists of a ligand on the APC binding with the recip-taken for publications since 1990 using the keyrocal receptor on the T-cell surface.[7] Several differ-words ‘abatacept’, ‘CD28 costimulation’ andent costimulatory pathways have been identified,‘BMS-188667’. In addition, articles presented atsome lead to the initial activation of T cells andannual scientific meetings were included; although Isome attenuate the T-cell response. One of the bestacknowledge that these have not been the subject ofdescribed costimulatory pathways is the binding ofrigorous peer review.CD28 ligand on T cells with the CD80/CD86 ligandon the APC.[8] This pathway is also felt to be one of

1. The Role of T Cells in the Pathogenesis the more important pathways and acts by initialof Rheumatoid Arthritis activation of the T cell. It is this pathway that is the

target for the drug abatacept. Cytotoxic T-lympho-There is now a wealth of evidence showing that T cyte-associated antigen-4 (CTLA4) is a T cell sur-

cells play a central role in the pathogenesis of RA. face protein which has a similar structure to CD28The Sakaguchi (SKG) mouse model supports this on T cells, and CTLA4 is quickly upregulated once aconcept; a point mutation in the gene encoding T cell is activated.[9,10] CTLA4 binds to CD80/ZAP-70, a key signal transduction molecule in T CD86, the same binding sites as CD28, and has acells, leads to the development of a chronic inflam- greater affinity for CD80/CD86, thus competingmatory arthritis in mice that is clinically and histo- with CD28. The binding of CTLA4 with CD80/logically similar to RA.[1] The point mutation leads CD86 leads to down-regulation of the immune re-to a positive selection of autoreactive T cells in the sponse and can prevent CD28-dependent T-cell acti-thymus that would otherwise have been deleted. vation.[11] Polymorphisms of CTLA4 can cause lossOther supporting data include the ability to transfer of self-tolerance[12] and, therefore, targeting costim-RA into immunodeficient mice using T cells taken ulation using CTLA4 may help regulate the immunefrom the synovium of patients with RA.[2] Clinical response in RA.observations also add to the weight of evidence, asactivated T cells are seen in abundance in the 2. Abataceptsynovium of RA patients.[3] T cells are dependent onactivation; once activated they differentiate, prolif- Abatacept (CTLA4Ig) is a soluble fusion proteinerate and migrate into the affected tissues where that consists of the extracellular domain of thethey produce cytokines and regulate other immune CTLA4 linked to the Fc domain of human immu-responses. In RA, activated T cells produce a num- noglobulin G1 (IgG1).[13] Like the naturally occur-ber of proinflammatory cytokines such as TNFα and ring CTLA4, abatacept binds to CD80/86 on APCs.IL-1, as well as activating B cells to produce immu- However, it differs from naturally occurringnoglobulins, including rheumatoid factor. Activated CTLA4, as abatacept is soluble rather than mem-T cells lead to bone loss and joint damage through a brane bound and so does not produce a signal to the

2007 Adis Data Information BV. All rights reserved. Drugs 2007; 67 (1)

T-Cell Activation Inhibitors in Rheumatoid Arthritis 3

T cell but blocks the binding of CD28 to CD80/ groups, including the placebo group. Fifteen seriousCD86, thus preventing full activation of the T cell adverse events were reported during the treatmentby interfering with the costimulatory signal. In vitro period, this comprised 4% in the active treatmentstudies have shown a reduction in T-cell prolifera- groups compared with 13% in the placebo group. Intion and inhibition of cytokines TNFα, interferon-γ most cases, the event was worsening of RA requir-and IL-2.[14] ing hospitalisation. There was one episode of septic

arthritis in the abatacept group diagnosed 88 daysAbatacept is administered by a 30 minute intra-after the last infusion, although the patient had alsovenous infusion given at baseline then at weeks 2had an intra-articular injection of corticosteroids.and 4, and 4-weekly thereafter. Abatacept has beenThere were five withdrawals; two patients from thelicensed for use in the US as monotherapy or inabatacept 0.5 mg/kg group (one due to worsening ofcombination with other DMARDs (except anti-TNFRA and one patient had breast cancer diagnosed attherapy).day 57), two patients withdrew from the abatacept

2.1 Clinical Trial Data 2 mg/kg group (one due to worsening of RA and onedue to probable anxiety attack after the second infu-

2.1.1 Pilot Study sion) and one patient withdrew from the belataceptIn the first pilot study published in 2002,[15] the 10 mg/kg group (due to viral upper respiratory in-

primary goal was to assess the safety, tolerability fection).and preliminary efficacy of pharmacological block- With regard to efficacy with both agents, a doseade of CD80 and CD86 in patients with RA. Two response was noted for the primary outcome, Amer-drugs were studied: abatacept and belatacept ican College of Rheumatology 20% improvement(LEA29Y). Belatacept is a second generation mole- criteria (ACR)20.[16] Both active treatments, particu-cule based on abatacept, which has greater avidity larly the higher doses, were associated with numericfor CD86 via the mutation of the residues of two improvements in ACR relative to placebo. Theamino acids on the parent molecule. The study was study was not powered to demonstrate efficacy;multicentre, randomised, double-blind and placebo- however, the difference between the ACR20 re-controlled, with seven treatment groups: abatacept sponse of the belatacept 10 mg/kg group and theor belatacept at 0.5 mg/kg, 2 mg/kg or 10 mg/kg, or placebo group reached clinical significance as didplacebo, with approximately 30 patients in each the ACR70 response of the abatacept 2 mg/kg groupgroup. Study medication was given on days 1, 15, 29 compared with the placebo group.and 57, with follow-up continued to day 169. Pa-tients had active RA diagnosed within the last 7 2.1.2 Phase II Studiesyears and had not responded to at least one Two papers have been published with phase IIDMARD. data for abatacept in RA. The first was a 6-month,

One hundred and seventy-four patients (81%) randomised, double-blind, placebo-controlled studycompleted the treatment period and 160 (75%) com- comparing the safety, efficacy and immunogenicitypleted to study end at day 169. Of the 40 patients of abatacept 2 mg/kg or 10 mg/kg with placebo inwho discontinued the study within the treatment patients with active RA who had an inadequateperiod, 12 were in the placebo group, 20 in the response to methotrexate.[17] The second publicationabatacept group and 8 in the belatacept group. The reported the results from the 6-month extension ofmost common reason for discontinuation was wors- the previous study.[18] In the first study, patients hadening of disease activity. established active RA and had been receiving

The study medication seemed well tolerated and methotrexate for at least 6 months; no othergenerally safe. There were no notable renal, hepatic DMARD was allowed. Abatacept or placebo wasor haematological adverse events during the study. administered by infusion on days 1, 15 and 30 andThe incidence of adverse events was similar in all monthly thereafter. A total of 544 patients were


4 Pollard

recruited and 339 patients were randomised. Base- methotrexate on a broad range of HR-QOL do-mains. General health status was measured using theline characteristics were similar across all groupsSF-36 and a health utility index (SF-6D) was alsowith mean disease duration of 8.9–9.7 years. A totalderived from 11 items of the SF-36.[21] Changes inof 259 patients completed the first 6 months ofthe SF-36 scale and summary measure scores andtreatment. Significantly more patients discontinuedthe SF-6D from baseline to 12 months were evalu-treatment because of lack of efficacy in the placeboated and compared between placebo and treatmentgroup compared with both treatment arms.groups. Overall, the changes in SF-36 scales andThe results for clinical efficacy showed that at 6summary measure scores differed between themonths the percentage of patients who achieved anabatacept 10 mg/kg and placebo group (multivariateACR20 response was significantly higher in theanalysis of variance [MANOVA] F = 4.7,abatacept 10 mg/kg group than placebo, and this wasp < 0.0001). Statistically significant improvementsevident from month 2 through to month 6. Therein mean score were observed in all eight scales ofwas no significant difference between the groupthe SF-36 and both summary measures, with out-receiving 2 mg/kg and the placebo group. The ratescomes favouring the abatacept group. A statisticallyof ACR50 (from day 90) and ACR70 (from day 30)significant difference in mean score change on theresponses were significantly higher in bothSF-6D was also observed in the abatacept group.abatacept groups compared with placebo at 6

Throughout the 12-month period that the twomonths. At 12 months, there was still a significantstudies cover, abatacept was well tolerated. Nodifference in ACR20 response in the abataceptdeaths or opportunistic infections were reported in10 mg/kg group compared with placebo. Again, atany of the patients who received abatacept over 1212 months there was no significant difference inmonths. The number of malignancies was similar inACR20 responses in the abatacept 2mg/kg groupthe abatacept group compared with placebo. Thecompared with placebo, suggesting that the lowersafety profile for both dosages of abatacept wasdose is not sufficient. ACR50 and ACR70 responsessimilar to placebo over 12 months.continued to be significantly higher in the abatacept

Patients in the 10 mg/kg group who completed10 mg/kg group compared with placebo at 12the 12-month study were eligible for enrolment in anmonths. At 12 months, statistically significant dif-open-label extension study for a further 12ferences were noted in all the components of themonths.[22] Of the 115 patients eligible, 73% enteredACR core set (tender joint count, swollen jointthe extension study and 89% of these patients com-count, pain, acute phase reactants, patient’s andpleted 2 years of treatment. Nearly 50% of thephysician’s global assessments and health assess-abatacept group achieved a Disease Activity Scorement questionnaire) in the abatacept 10 mg/kg groupfor 28 joints (DAS28) remission at 12 months,compared with placebo.which was sustained at 2 years (figure 1).

In the 6-month analysis, patients were also as-sessed using the SF-36 (short form 36-item quality 2.1.3 Phase III Studiesof life questionnaire)[19] score and patients in the The results from the first phase III trial evaluatingabatacept 10 mg/kg group had clinically and statisti- the efficacy and safety of abatacept were publishedcally significant improvements in all eight sub- in September 2005.[23] In this randomised, double-scales of the SF-36 compared with placebo; there blind study, abatacept was given to patients withwere non-significant improvements in the 2 mg/kg active RA and an inadequate response to anti-TNFαgroup. therapy (ATTAIN trial [Abatacept Trial in Treat-

Health-related quality of life (HR-QOL) data ment of Anti-TNF Inadequate Responders]). Twofrom the 12-month study were published in April groups of patients were enroled: those receiving2006.[20] This study examined the effect of abatacept anti-TNF therapy at the time of screening (currenttherapy (10 mg/kg) in combination therapy with users) and those who had previously received anti-



period. More patients in the abatacept groupachieved remission (as defined by a DAS28 of <2.6)than in the placebo group (10% vs 0.8%, p < 0.001)and 17.1% of patients in the abatacept group hadlow levels of disease (DAS28 ≤3.2) compared with3.1% of patients in the placebo group (p < 0.001).Improvements in physical function (improvementfrom baseline in HAQ of 0.3) were seen in 47.3% ofpatients in the abatacept group compared with23.3% in the placebo group (p < 0.001) at 6 months.Improvements in HR-QOL measures (SF-36) werealso observed. Improvements in scores for all sub-scales were seen in the abatacept group. There weresignificant improvements in the physical-compo-nent and mental-component summary scores com-pared with the placebo group.

Rem

issi

on r

ates

(D

AS

28 <

2.6)

[%]

60

50

40

30

20

10

01 90 180 270 360 450 540 630 720

Visit (days)

Blinded phase LTE phase

Abatacept 10 mg/kg (n = 84)Placebo (n = 67)

Fig. 1. Remission rates (Disease Activity Score for 28 joints[DAS28] <2.6) for patients receiving abatacept at 12 months (blind-ed phase) and at 2 years (open-label, long-term extension [LTE])based on as-observed data (reproduced from Dougados et al.,[22]

with permission).

The rates of adverse events and serious adverseTNF therapy (former users). All users had to stop events were similar in both groups, but a highertherapy before commencing the trial. Patients were number of infections were seen in the abataceptrandomly assigned in a 2 : 1 ratio to receive group (37.6% vs 32.3%, p = 0.3), but no unusual orabatacept or placebo and stratified according to use opportunistic infections were observed. One patientof anti-TNF therapy at the time of enrolment. in the abatacept group died from a myocardial in-Abatacept was given at a dose of 10 mg/kg, as a 30 farction and congestive cardiac failure.minute infusion on days 1, 15 and 29 and every 28

Further HR-QOL data from the ATTAIN studydays thereafter for 6 months. The two primaryhas recently been published.[25] In this study, HR-endpoints were the proportion of patients with anQOL was measured using the SF-36 (using both theACR20 response at 6 months and also the propor-scales and composite measures), HAQ and a visualtion of patients with at least a 0.3 improvement fromanalogue scale. The SF-36 was measured at baselinebaseline in the Health Assessment Questionnaireand at 1, 3 and 6 months (end-point) and the HAQ

(HAQ)[24] score. Secondary objectives includedwas measured at multiple points throughout the

ACR50 and ACR70 response rates at 6 months,study. At the end-point (6 months), the abatacept

changes in DAS28 scores and changes from baselinegroup had statistically significant improvements

in the SF-36 score at 6 months. A total of 738from baseline in all SF-36 subscales and composite

patients were screened, 393 underwent randomisa-scores, the HAQ and the fatigue measure, compared

tion and a total of 322 patients completed the study:with the placebo group. QOL was also examined for

86.4% in the abatacept group and 74.4% in thepatients with different severity levels at baseline.

placebo group. Lack of efficacy was the predomi-Those receiving abatacept with a lower DAS28 level

nant reason for withdrawal from both groups.at baseline (<7.59) improved significantly more in

At 6 months, the rate of ACR20 (50.4% vs most QOL measures (excluding role emotional [RE]19.5%, p < 0.001), ACR50 (20.3% vs 3.8%, in the SF-36) compared with those receiving place-p < 0.001) and ACR70 (10.2% vs 1.5%, p = 0.003) bo. Among those patients with a higher DAS28 atresponses was significantly higher in the abatacept baseline, all of the mental SF-36 scales, HAQ andgroup compared with placebo. The improvement fatigue scores showed similar improvements in bothover placebo was evident from day 15 onward and groups. Lastly, the study looked at change in QOL

over time. The results showed that the abataceptprogressively increased during the 6-month study


6 Pollard

group had significantly larger rates of change for all higher in the abatacept group compared with place-QOL outcomes (except for RE) compared with the bo (figure 2a–2c). Between 6 and 12 months, all theplacebo group. Most of the results indicated a signif- ACR responses continued to improve in patientsicant difference between treatment and placebo oc- receiving abatacept. At 1 year, ACR20 responsescurred by the 12th week. had significantly increased to 73.1%, ACR50 re-

sponses were 48.3% and ACR70 responses wereThe results from the phase III AIM (Abatacept in28.8% (figure 2a–2c). At 1 year, physical functionInadequate responders to Methotrexate) study have(HAQ) had improved statistically significantly morealso recently been published.[26] This 1-year,in the abatacept group compared with placeborandomised, double-blind, placebo-controlled trial(63.7% vs 39.3%; p < 0.001) [figure 2d]. Radio-aimed to compare the efficacy and safety ofgraphic data were collected for 92% of randomisedabatacept versus placebo in combination withpatients. At 1 year, patients who received abataceptmethotrexate in patients with RA and an inadequatedemonstrated statistically significant slowing ofresponse to methotrexate. Patients were randomly

assigned in a 2 : 1 ratio to receive either a fixed dose structural progression compared with placebo, withof abatacept or placebo. The infusion regimen for an approximately 50% reduction in change fromabatacept was the same as in the ATTAIN study. All baseline in Genant-modified Sharp scores comparedpatients received methotrexate 15mg per week, al- with placebo. The median change from baseline inthough 10mg per week was allowed if there was a erosion, joint-space and total score are shown inhistory of toxicity. No adjustment of methotrexate figure 3. At 6 and 12 months, 30.1% and 42.5% ofwas allowed in the first 6 months but after that time the abatacept group achieved a DAS28 of ≤3.2,increases of methotrexate were permitted, as was an compared with 10% and 9.9% of the placebo group,addition of a further DMARD or an increase in respectively (p < 0.001). At 1 year, DAS28 remis-corticosteroid if deemed necessary by the blinded sion (DAS28 ≤2.6) was seen in 14.8% of abataceptinvestigator. The three primary endpoints were the recipients compared with 1.9% of patients receivingproportion of patients achieving an ACR20 at 6 placebo (p < 0.001). The HR-QOL measures alsomonths, the proportion of patients with a clinically showed improvement in the abatacept group. At 6significant improvement (≥0.3) in HAQ score at 1

months, significant improvements from baseline inyear, and the radiographic progression of joint ero-

both physical (p < 0.001) and mental componentssions (assessed by comparing changes from baseline

(p = 0.009) were seen compared with placebo, andin the Genant-modified Sharp score[27]) at 1 year.they remained significant at 1 year.Secondary objectives included assessing ACR50

and ACR70 responses at 6 months and all ACR The overall incidence of adverse events was sim-responses at 1 year. Changes in DAS28 were also ilar in both abatacept and placebo groups. The inci-assessed, as were changes in HR-QOL by use of the dence of serious adverse events increased withSF-36. abatacept treatment, but rates of discontinuation due

to serious adverse events were similar betweenA total of 1250 patients were enrolled and 652groups. The incidence of infection reported as aunderwent randomisation with 452 assignedserious adverse event was higher with abataceptabatacept and 219 assigned placebo. Lack of effica-than with placebo and an increase in pneumoniacy was the most common reason for discontinuationwith abatacept treatment was observed. One possi-in the placebo group (18% vs 3%). Adverse eventsble case of TB was observed in each group and thewere the most common reason for discontinuation inincidence of neoplasms (benign or malignant) wasthe abatacept group (4% vs 2%). At 6 months thesimilar in both groups. No major autoimmune disor-rate of ACR20 (67.9% vs 39.7%; p < 0.001),ders such as multiple sclerosis or lupus were report-ACR50 (39.9% vs 16.8%; p < 0.001) and ACR70

(19.8% vs 6.5%; p < 0.001) were significantly ed. More infusion reactions were observed in the


T-Cell Activation Inhibitors in Rheumatoid Arthritis 7A

CR

20 r

espo

nses

(%

)1

80

70

60

50

40

30

20

10

00 15 29 57 85 11

314

116

322

528

136

5

AC

R70

res

pons

es (

%)1

80

70

60

50

40

30

20

10

00 15 29 57 85 11

314

116

322

528

136

5

AC

R50

res

pons

es (

%)1

80

70

60

50

40

30

20

10

00 15 29 57 85 11

314

116

322

528

136

5

HA

Q–D

I res

pons

es (

%)1

80

70

60

50

40

30

20

10

00 15 29 57 85 11

314

116

322

528

136

5

Abatacept + MTX (n = 424)Placebo + MTX (n = 214)2

a b

c d

Visit day Visit day

Fig. 2. Improvement in American College of Rheumatology % improvement criteria (ACR)20 (a), ACR50 (b) and ACR70 (c) responses over1 year in all patients who received at least one dose of the study medication; and (d) the percentage of patients who achieved a HealthAssessment Questionnaire Disability Index (HAQ-DI) response (≥0.3-unit improvement from baseline in HAQ-DI) determined over 1 year(reproduced from Kremer et al.,[26] with permission). 1 = intention-to-treat population where all patients who withdrew from the study wereconsidered to be ACR nonresponders subsequent to their withdrawal; 2 = because of adherence issues identified during the study, patientsfrom one site were excluded from all efficacy analyses before unblinding but were included in the analysis of safety; MTX = methotrexate.

abatacept group; two patients had to withdraw be- safe and well tolerated. However, a slight increase incause of severe acute infusion reactions. infection rate was observed in the patients who

received abatacept. The incidence of death and seri-2.2 Safety ous adverse events was similar in both groups. The

most commonly reported adverse effects were head-With the advent of the new biological therapies

ache, upper respiratory tract infections, nausea andan important consideration is the safety of these newnasopharyngitis.drugs, especially with the observed increased risk of

The ASSURE (Abatacept Study of Safety in Useinfection, in particular reactivation of TB[28] andwith other Rheumatoid Arthritis Therapies) trial wasopportunistic infections. Two abstracts have beendesigned to evaluate the safety profile of abataceptpublished specifically looking at the safety profile ofin routine clinic practice and, therefore, looked atabatacept.[29,30] Moreland et al.[29] looked at the safe-the safety of abatacept as an add-on treatment in RAty of abatacept from five randomised, double-blind,patients who were already taking one or moreplacebo-controlled trials. In this large populationDMARD and/or biological therapy.[30] Patient-re-(1955 patients who received abatacept vs 989 pa-ported outcomes were also assessed (HAQ, patienttients given placebo), abatacept seems generally


8 Pollard

system means that it is less likely to be immunosup-pressive as it targets naive T cells and spares theinnate immune system responses.

Phase II and phase III trials have provided goodevidence of the efficacy of abatacept in the treat-ment of RA. Abatacept therapy is associated withrapid and sustained improvements in both clinicaland patient-reported outcomes, including QOL mea-sures. As abatacept has a different mechanism ofaction to anti-TNF therapy, it would be reasonableto assume that abatacept would be effective in pa-tients resistant to ant-TNF therapy and indeed theimprovements seen in clinical outcomes were alsoseen in non-TNF responders. Abatacept treatmentalso resulted in a reduction in radiographic progres-

2

4

6

8

10

12

0

–2Sco

re c

hang

e fr

om b

asel

ine

Erosion score JSN score Total score

Abatacept + MTX (n = 391)Placebo + MTX (n = 195)Median interquartile

Fig. 3. Slowing of radiographic structural damage progression at 1year. Interquartile range changes from baseline in Genant-modifiedSharp erosion, joint space narrowing (JSN) and total scores wereevaluated at 1 year or at early termination (if applicable). The dottedline indicates the 10th and 90th percentiles. Data shown are fromrandomly assigned and treated patients with baseline and follow-upradiography (reproduced from Kremer et al.,[26] with permission).MTX = methotrexate.

sion and in significant rates of remission.global assessment and pain). This 1-year study The combination of the favourable safety profilerandomised and treated 1441 patients (959 received so far documented and the clinical benefits suggestabatacept and 482 received placebo), most of whom that abatacept will certainly be an important additionwere receiving combination non-biological thera- to the available therapies for patients with RA.pies. Improvements were seen in all patient-reported Abatacept is the first costimulatory blocker to beoutcomes in those patients treated with abatacept, licensed in RA, but there is potential for other drugswith the greatest benefit seen in those patients on in this class since new forms of CD28-CD80/86background non-biological therapies. With regards blockade are currently being developed, as areto safety, no difference was seen in the incidence of agents that block other costimulatory pathways suchneoplasms and serious infections between the treat- as CD40 and CD154.ment groups in those patients receiving background

Acknowledgementsnon-biological DMARDs; however, in the smallergroup of patients on background biological thera- No sources of funding were used to assist in the prepara-pies there was an increase in serious adverse events tion of this review. The author has no conflicts of interest that

are directly relevant to the content of this review.and infections observed. This suggests that the safe-ty profile for patients receiving background

ReferencesDMARDs appears favourable over 1 year, but given1. Sakaguchi N, Takahashi T, Hata H, et al. Altered thymic T-cellthe less favourable profile in those patients on back-

selection due to a mutation of the ZAP-70 gene causes autoim-ground biological therapies, this would not support mune arthritis in mice. Nature 2003; 426 (6965): 454-60

2. Mima T, Saeki Y, Ohshima S, et al. Transfer of rheumatoidthe use of abatacept in combination with other bio-arthritis into severe combined immunodeficient mice: the path-logical therapies at present. ogenetic implications of T cell populations oligoclonally ex-panding in the rheumatoid joints. J Clin Invest 1995; 96 (4):1746-583. Conclusions

3. Steiner G, Tohidast-Akrad M, Witzmann G, et al. Cytokineproduction by synovial T cells in rheumatoid arthritis. Rheu-matology 1999; 38 (3): 202-13Abatacept is the first drug in a new category of

4. Choy EH, Panayi GS. Cytokine pathways and joint inflamma-biological response modifiers; the CTLA4Ig fusiontion in rheumatoid arthritis. N Engl J Med. 2001; 344 (12):

protein modulates the CD80/86-CD28 costimulato- 907-165. VanderBorght A, Geusens P, Raus J, et al. The autoimmunery pathway and provides further evidence that T

pathogenesis of rheumatoid arthritis: role of autoreactive Tcells play a central role in the pathogenesis of RA. cells and new immunotherapies. Semin Arthritis Rheum 2001;The way in which abatacept modulates the immune 31 (3): 160-75



6. Schwartz RH. T cell anergy. Annu Rev Immunol 2003; 21: 21. Brazier J, Roberts J, Deverill M. The estimation of a preference-305-34 based measure of health from the SF-36. J Health Econ 2002

Mar; 21 (2): 271-927. Janeway Jr CA, Bottomly K. Signals and signs for lymphocyteresponses. Cell 1994; 76 (2): 275-85 22. Dougados M, Westhovens R, St Clair E, et al. Sustained remis-

sion and major clinical response at 2 years shown with8. Lenschow DJ, Walunas TL, Bluestone JA. CD28/B7 system ofabatacept (CTLA4Ig) in combination with methotrexate inT cell costimulation. Annu Rev Immunol 1996; 14: 233-58rheumatoid arthritis patients with an inadequate response to

9. Walunas TL, Lenschow DJ, Bakker CY, et al. CTLA-4 canmethotrexate [abstract no. 359]. Arthritis Rheum 2004; 50:function as a negative regulator of T cell activation. ImmunityS1851994; 1 (5): 405-13

23. Genovese MC, Becker JC, Schiff M, et al. Abatacept for rheu-10. Wang XB, Zheng CY, Giscombe R, et al. Regulation of surfacematoid arthritis refractory to tumor necrosis factor alpha inhi-and intracellular expression of CTLA-4 on human peripheral Tbition. N Engl J Med 2005; 353 (11): 1114-23cells. Scand J Immunol 2001; 54 (5): 453-8

24. Fries JF, Spitz PW, Kraines RG, et al. Measurement of patient11. Walunas TL, Bakker CY, Bluestone JA. CTLA-4 ligationoutcome in arthritis. Arthritis Rheum 1980; 23: 137-45blocks CD28-dependent T cell activation. J Exp Med 1996;

183 (6): 2541-50 25. Westhovens R, Cole JC, Li T, et al. Improved health-relatedquality of life for rheumatoid arthritis patients treated with12. Simmonds MJ, Gough SC. Genetic insights into disease mecha-

nisms of autoimmunity. Br Med Bull 2005; 71: 93-113 abatacept who have inadequate response to anti-TNF therapyin a double-blind, placebo-controlled, multicentre randomized13. Linsley PS, Brady W, Urnes M, et al. CTLA-4 is a secondclinical trial. Rheumatology (Oxford) 2006 Oct; 45 (10):receptor for the B cell activation antigen B7. J Exp Med 1991;

174: 561-9 1238-46

26. Kremer JM, Genant HK, Moreland LW, et al. Effects of14. Nadler S, Townsend R, Mikesell G, et al. Abatacept (CTLA4Ig;BMS-188667) significantly inhibits T-cell proliferation in vi- abatacept in patients with methotrexate-resistant active rheu-tro at clinically relevant concentrations [abstract no. matoid arthritis: a randomized trial. Ann Intern Med. 2006 JunTHU0103]. Ann Rheum Dis 2004; 63 Suppl. 1: 142 20; 144 (12): 865-76

15. Moreland LW, Alten R, Van den Bosch F, et al. Costimulatory 27. Genant HK, Jiang Y, Peterfy C, et al. Assessment of rheumatoidblockade in patients with rheumatoid arthritis: a pilot, dose- arthritis using a modified scoring method on digitized andfinding, double-blind, placebo-controlled clinical trial evaluat- original radiographs. Arthritis Rheum 1998 Sep; 41 (9):ing CTLA-4Ig and LEA29Y eighty-five days after the first

1583-90infusion. Arthritis Rheum 2002; 46 (6): 1470-928. Keane J, Gershon S, Wise RP, et al. Tuberculosis associated

16. Felson DT, Anderson JJ, Boers M, et al. American College ofwith infliximab, a tumor necrosis factor α-neutralizing agent.Rheumatology preliminary definition of improvement in rheu-N Engl J Med 2001; 345: 1098-104matoid arthritis. Arthritis Rheum 1995; 38: 727-35

29. Moreland L, Kaine J, Espinoza L, et al. Safety of abatacept in17. Kremer JM, Westhovens R, Leon M, et al. Treatment of rheu-rheumatoid arthritis patients in five double-blind, placebo-matoid arthritis by selective inhibition of T-cell activation withcontrolled trials [abstract no. 886]. Arthritis Rheum 2005; 52:fusion protein CTLA4Ig. N Engl J Med 2003; 349 (20):

1907-15 S350

30. Combe B, Weinblatt M, Birbara C, et al. Safety and patient-18. Kremer JM, Dougados M, Emery P, et al. Treatment of rheuma-toid arthritis with the selective costimulation modulator reported outcomes associated with abatacept in the treatmentabatacept: twelve-month results of a phase IIb, double-blind, of rheumatoid arthritis patients receiving background diseaserandomized, placebo-controlled trial [published erratum ap- modifying anti-rheumatic drugs (DMARDs): the ASSUREpears in Arthritis Rheum 2005 Oct; 52 (10): 3321]. Arthritis trial [abstract no. 1918]. Arthritis Rheum 2005; 52: S709Rheum 2005; 52 (8): 2263-71

19. Ware JE. The SF-36 health survey. In: Spilker B, editor. Qualityof life and pharmacoeconomics in clinical trials. 2nd ed. Lip- Correspondence and offprints: Dr Louise C. Pollard, Aca-pincott-Raven: Philadelphia, 1996 demic Department of Rheumatology, Kings College

20. Emery P, Kosinski M, Li T, et al. Treatment of rheumatoid London, Weston Education Centre, 10 Cutcombe Road,arthritis patients with abatacept and methotrexate significantly

London, SE5 9RJ, UK.improved health-related quality of life. J Rheumatol 2006 Apr;E-mail: [email protected] (4): 681-9


Drugs 2007; 67 (1): 11-16LEADING ARTICLE 0012-6667/07/0001-0011/$49.95/0


Squalene Synthase InhibitorsClinical Pharmacology and Cholesterol-Lowering Potential

Valentine Charlton-Menys and Paul N. Durrington

Division of Cardiovascular and Endocrine Sciences, Cardiovascular Research Group,Core Technology Facility, University of Manchester, Manchester, UK

HMG-CoA reductase inhibitors (statins) reduce cardiovascular disease mor-Abstractbidity and mortality with a high level of safety. Nonetheless, there are substantialnumbers of people who either do not tolerate statins or whose low-densitylipoprotein (LDL) levels are not lowered adequately. For these reasons, there is aneed to develop other cholesterol-lowering drugs. A target for these new agents isprovided by the enzymes distal to HMG-CoA reductase in the cholesterol biosyn-thesis pathway. Two classes of drugs have been developed: (i) squalene synthaseinhibitors, which act at the first committed step in cholesterol biosynthesis, distalto the mevalonate-farnesyl diphosphate pathway; and (ii) oxidosqualene cyclaseinhibitors, which act distal to the squalene intermediate. Of these, squalenesynthase inhibitors have received more attention and are the subject of this review.Squalene synthase inhibitors decrease circulating LDL-cholesterol by the induc-tion of hepatic LDL receptors in a similar manner to statins. They have fewersecondary effects mediated by a decrease in non-cholesterol products ofmevalonate metabolism distal to HMG-CoA reductase, but have the potential toincrease intermediates proximal to squalene. Squalene synthase inhibitors are justnow entering clinical trials and data on how effectively they lower LDL-cholester-ol and how they compliment the actions of statins and other agents is awaited withconsiderable interest.

The serum cholesterol level of an individual is lowering its concentration. Despite the upregulationof HMG-CoA reductase that ensues,[4] the moreone of the most important factors in predicting[1,2]

potent statins can lower LDL by ≥50%.[5] Statinsand preventing[3] coronary heart disease (CHD). Sta-have been consistently shown to reduce both CHD-tin drugs are competitive inhibitors of HMG-CoAand stroke-related morbidity and mortality,[6] andreductase (figure 1), the primary rate-limiting en-with the exception of cerivastatin, do so with a widezyme in the hepatic biosynthesis of cholesterol. Themargin of safety.[7]

resulting statin-induced decrease in intrahepaticcholesterol concentrations leads to the upregulation

1. Statinsof hepatic receptors for low-density lipoprotein(LDL), an effect mediated by the transcription factor The main problems with the use of statins relateSREBP (sterol response element binding protein). to dose and response. There is considerable inter-The increase in LDL receptor expression enhances individual variation in the magnitude of the effect ofthe fractional catabolic rate of circulating LDL, thus statins on LDL-cholesterol. There is evidence that

12 Charlton-Menys & Durrington

Acetyl CoA + H2O + Acetoacetyl CoA

HMG-CoA reductase

Mevalonate

Isopentanyl PP Dimethylallyl PP

Geranyl PP Prenylation of proteins

Farnesyl PP Dolichols

Squalene synthaseProtein glycosylation

Geranylgeranyl PP SMC proliferation

Squalene Ubiquinone (CoQ) Tumour cell growth

Squalene monooxygenase Electron transport

2,3 oxidosqualeneLipid antioxidant

Oxidosqualene cyclase

Lanosterol synthase 2,3;22,23 Diepoxysqualene

Lanosterol

Cholesterol 24,25 Epoxylanosterol

Oxysterols HMG CoA reductase

3-HMG-CoA

Fig. 1. The sites of action of HMG-CoA reductase inhibitors, squalene synthase inhibitors and oxidosqualene cyclase inhibitors in thehepatic biosynthesis of cholesterol. The putative functions of intermediates and derivatives of intermediates up to the first branch point in thebiosynthetic pathway for cholesterol synthesis are shown by the dashed arrows, and the stages involving multiple reactions are shown bythe green lines. CoQ = coenzyme Q (ubidecarenone); PP = pyrophosphate; SMC = smooth muscle cell.

people with low rates of cholesterol synthesis,[8] dose response remains relatively flat,[10] with furtherdecreases of LDL-cholesterol of only approximatelyperhaps due to increased absorption of dietary cho-6% resulting from doubling doses.[5,11]

lesterol,[9] may be most resistant. Accordingly, anindividual whose LDL-cholesterol is insufficiently

1.1 Safety of Statinslowered with any one statin at a given dose may notbenefit from an increased dose of that statin or from Although the margin of safety for most statinstreatment with a different statin. Even in people approved for clinical use is very wide,[7,12-14] eleva-whose initial response to statin therapy is good, the tion of liver aminotransferases (ALT and AST) and


Squalene Synthase Inhibitors 13

myositis or myopathy can occur. Their likelihood, 2.1 Safety and Efficacy of SqualeneSynthase Inhibitors Compared with Statinsalthough low, relates to the statin dose rather than to

the degree of reduction in LDL-cholesterol.[14] Stud-Preliminary studies with the squalene synthase

ies on rat and human myotube cultures showed that inhibitors BMS-187745 and BMS-188494 showedstatin-induced myopathy is probably due to de- that, at concentrations that markedly decreasedcreased geranylgeranylation of proteins secondary cholesterol synthesis, no myotoxicity was found into decreased mevalonate synthesis.[15,16]

vitro.[22] Thus, inhibition of squalene synthase didnot result in myotoxicity in vitro because in-Potentially, statins could have other effectstermediates formed before squalene and responsibledownstream of HMG-CoA reductase (figure 1).for prenylation of proteins were not depleted.[16,22] ItTheir consequences might prove favourable, en-was later shown that the squalene synthase inhibi-hancing the protection of statins against cardiovas-tors ER-27856[23] and TAK-475[24] and EP2302[25]

cular disease by mechanisms other than LDL-decreased circulating LDL levels, as with statins, bycholesterol lowering (pleiotropic effect), or un-inducing LDL receptors, assessed using HepG2favourable, leading to adverse events as in the casecells in culture. It was also shown in rhesus monkeys

of myositis. An example of a potentially favourablethat ER-27856 lacked the hepatotoxicity found with

statin effect, as yet unexplained, is the decrease in C-atorvastatin,[23] and the same was shown for TAK-

reactive protein associated with their use. [17] Impor- 475, in cynomolgus monkeys.[24] More recently, twotant effects of statins downstream of HMG-CoA new potent squalene synthase inhibitors (EP2306reductase include prenylation (post-translational and EP2302) have been described.[25] EP2302 inhib-modification of proteins by farnesyl pyrophosphate ited cholesterol synthesis dose-dependently with aor geranylgeranyl pyrophosphate), which regulates similar potency to that of simvastatin. In tests so far,the subcellular location of G-proteins influencing the degree of inhibition observed in vitro has notmany signalling cascades within the cell.[18] Oxys- been seen in vivo, probably because of upregulationterols and farnesyl pyrophosphate derived from the of HMG-CoA reductase, in the case of statins, or

through a compensatory increase in intestinalcholesterol biosynthetic pathway after mevalonate,cholesterol uptake with both statins and squalenebut before squalene synthase, also affect the activitysynthase inhibitors. Tavidrou et al.[25] also showedof nuclear orphan receptors such as liver X receptorthat oleate-induced apolipoprotein B secretion by(LXR) and farnesoid X-activated receptor (FXR),HepG2 cells was more markedly inhibited by simva-which are important in biliary cholesterol metabo-statin than by EP2306 or EP2302.lism, lipoprotein metabolism and excretion, and in

macrophage foam cell formation.[19]

2.2 Other Potential Toxicity of SqualeneSynthase Inhibitors

2. Squalene Synthase InhibitorsHMG-CoA reductase is the site of physiological

regulation of cholesterol biosynthesis, making it un-Squalene synthase is another enzyme in the cho- likely that accumulation of metabolites earlier in the

lesterol biosynthetic pathway (figure 1). Important pathway (figure 1) would be toxic, but this is notdifferences between cholesterol-independent effects necessarily true of squalene synthase inhibition.of squalene synthase inhibitors and those of HMG- Triparanol, another inhibitor of cholesterol biosyn-CoA reductase might be expected because squalene thesis downstream of mevalonate, was found tosynthase, which acts downstream of mevalonate, is cause cataract formation.[26] More recently, an asso-the first committed step of hepatic cholesterol bio- ciation between lanosterol synthase (figure 1) muta-synthesis at the final branch point of the cholesterol tions and cholesterol deficiency resulting in cataractbiosynthetic pathway (figure 1).[20,21] formation in a rat model was described.[27] However,



in the rats with cataracts, the cholesterol deficiency 4. Clinical Trials with Squalenewas confined largely to the lens. In contrast, hepatic Synthase Inhibitorsand serum cholesterol levels were not decreased.

Squalene synthase inhibitors are now enteringSince vertebrate eye lenses are not vascularised, declinical trials and, thus, it will be fascinating to seenovo cholesterol synthesis in the lens is necessaryhow effectively they lower LDL-cholesterol levelsfor normal proliferation of epithelial cells in theand the extent to which their role might complementlens. This suggests that the gene or isoform of thethose of statins and other lipid-lowering drugs.gene regulating lens cholesterol synthesis differsPharmacokinetic, and more especially pharmacody-from the gene regulating hepatic cholesterol synthe-namic and toxicological studies, will be required insis. There is evidence that statins, in much higherhumans to determine whether squalene synthasedoses than used in the clinic, can cause cataractinhibitors do in reality offer advantages over statins.formation in rats;[28] this would appear to be due to aThus far, one pharmacokinetic and pharmacody-failure of upregulation of lens HMG-CoA reductase,namic study with the squalene synthase inhibitorand decreased ubiquinone (ubidecarenone) levelsBMS-188494 has been performed,[32] in whichwith statin treatment may also be a factor.[29]

squalene synthase inhibition was quantified indi-rectly by assaying dicarboxylic acids in urine. This

3. Oxidosqualene Cyclase Inhibitors elegant approach to quantifying the pharmacody-namic response revealed encouragingly that farnesylAnother enzyme target in cholesterol biosynthe-pyrophosphate, which might be expected to accu-sis is 2,3-oxidosqualene cyclase (figure 1). Selectivemulate when squalene synthase was inhibited, wasinhibitors of oxidosqualene cyclase have been re-in fact metabolised through a series of oxidativeported to decrease cholesterol biosynthesis[21,30]

steps to dicarboxylic acids, readily excreted in urine.without influencing LDL catabolism,[31] unlike sta-tins, which principally decrease circulating LDL- 5. Future Directionscholesterol by upregulating hepatic LDL uptake.Inhibition of oxidosqualene cyclase with U18666A On a final note, another source of cholesterol isalso, unlike HMG-CoA reductase inhibition, con- from intestinal absorption. Ezetimibe is a selectivesistently induces cataract formation in rats. Howev- inhibitor of intestinal cholesterol absorption and iser, this may not be wholly due to inhibition of increasingly finding use in the clinic as an LDL-cholesterol synthesis because U18666A has been cholesterol-lowering agent, although randomised,shown to have a direct toxic effect of lens epithelial clinical events trials are yet to be reported. Eze-cells.[31] Inhibition of oxidosqualene cyclase results timibe has a limited LDL-cholesterol-lowering ef-in redirection of 2,3-oxidosqualene, which in turn fect of around 20% either alone or in the presence ofresults in increased formation of oxysterols, and a statin.[33] It acts by decreasing the intestinal choles-downregulates HMG-CoA reductase;[21] this would terol supply to the liver, lowering hepatic cholester-not occur with inhibition of squalene synthase (fig- ol levels and thus inducing LDL-receptor expres-ure 1). These findings suggest that squalene sion. It is effective because it not only decreases thesynthase inhibitors will prove to have less toxicity absorption of dietary cholesterol but also interruptsthan either oxidosqualene cyclase inhibitors[21,30,31] the enterohepatic circulation of cholesterol enteringor statins.[28,29] This is because inhibition of squalene the intestine in bile. Its limitation is the upregulationsynthase would not result in increased oxysterol of hepatic cholesterol biosynthesis. Nonetheless, theformation or decreased ubiquinone levels. However, 20% additional decrease over and above thatit remains for more detailed clinical assessment to achieved with a statin is clinically worthwhile nowbe undertaken to determine whether the newer that therapeutic targets for LDL-cholesterol havesqualene synthase inhibitors, such as TAK-475 and been lowered.[34,35] The LDL-cholesterol-loweringEP2302 cause lens opacities or other toxicity. goal can also be achieved with lower doses of the


Squalene Synthase Inhibitors 15

8. Naoumova RP, Marais AD, Mountney J, et al. Plasma mevalon-statin, potentially avoiding the adverse effects asso-ic acid, an index of cholesterol synthesis in vivo, and respon-

ciated with high statin doses. Squalene synthase siveness to HMG-CoA reductase inhibitors in familialhypercholesterolaemia. Atherosclerosis 1996; 119: 203-13development will thus occur against this back-

9. Miettinen TA, Gylling H. Ineffective decrease of serum choles-ground. terol by simvastatin in a subgroup of hypercholesterolemiccoronary patients. Atherosclerosis 2002; 164: 147-52

10. Illingworth DR. Management of hypercholesterolemia. Med6. ConclusionsClin North America 2000; 84: 23-42

11. Roberts WC. The rule of 5 and the rule of 7 in lipid-lowering bySqualene synthase inhibitors represent an inter-statin drugs. Am J Cardiol 1997; 80: 106-7

esting group of drugs for lowering LDL-cholesterol. 12. Colhoun HM, Betteridge DJ, Durrington PN, et al. Primaryprevention of cardiovascular disease with atorvastatin in theThe potential will depend on their LDL-loweringCollaborative Atorvastatin Diabetes Study (CARDS); mul-efficacy compared with and in combination with ticentre randomised placebo-controlled trial. Lancet 2004, 96

existing treatment, particularly the statin group. The 13. LaRosa JC, Grundy SM, Waters DD, et al. Intensive lipidlowering with atorvastatin in patients with stable coronaryinhibition of squalene synthase is likely to havedisease. N Engl J Med 2005; 352: 1425-35

fewer downstream effects on other pathways than 14. McKenney JM, Davidson MH, Jacobson TA, et al. Final conclu-sions and recommendations of the national lipid associationseen with statins. This might mean that some of thestatin safety assessment task force. Am J Cardiol 2006; 97adverse effects associated with statin therapy can beSuppl.: 89-94C

avoided. However, they might also lack some of the 15. Flint OP, Masters BA, Gregg RE, et al. HMG CoA reductaseinhibitor-induced myotoxicity: pravastatin and lovastatin in-favourable pleiotropism of statins and could havehibit the geranylgeranylation of low-molecular molecularother adverse effects related to accumulation of me- weight proteins in neonatal rat muscle cell culture. Toxicol

tabolites upstream of squalene synthase. The out- Appl Pharmacol 1997; 145: 99-11016. Johnson TE, Zhang X, Bleicher KB, et al. Statins inducecome of clinical trials is eagerly awaited.

apoptosis in rat and human myotube cultures by inhibitingprotein geranylgeranylation but not ubiquinone. Toxicol ApplPharmacol 2004; 200: 237-50Acknowledgements

17. Halcox JPJ, Deanfield JE. Beyond the laboratory: clinical impli-cations for statin pleiotropy. Circulation 2004; 109 Suppl. II:No sources of funding were used to assist in the prepara-II42-8tion of this review. The authors have no conflicts of interest

18. Edwards PA, Ericsson J. Sterols and isoprenoids: signallingthat are directly relevant to the content of this review.molecules derived from the cholesterol biosynthetic pathway.Annu Rev Biochem 1999; 68: 157-85

19. Edwards PA, Kast HR, Anisfeld AM. BAREing it all: theReferencesadoption of LXR and FXR and their roles in lipid homeostasis.1. Simons LA. Interrelations of lipids and lipoproteins with coro-J Lipid Res 2002; 43: 2-12nary artery disease mortality in 19 countries. Am J Cardiol

20. Gibbons GF, Mitropoulos KA, Myant NB. Biochemistry of1986; 57: 5-10cholesterol. Amsterdam: Elsevier Biomedical Press, 1982:2. Law MR, Thompson SG, Wald NJ. Assessing possible hazards131-88of reducing serum cholesterol. BMJ 1994; 308: 373-9

21. Mark M, Muller P, Maier R, et al. Effects of a novel 2,3-oxidos-3. Grundy SM. Cholesterol and heart disease: a new era. JAMAqualene cyclase inhibitor on the regulation of cholesterol bio-1986; 256: 2849-58synthesis in HepG2 cells. J Lipid Res 1996; 37: 148-584. Roitelman J, Masson D, Avner R, et al. Apomine, a novel

22. Flint OP, Masters BA, Gregg RE, et al. Inhibition of cholesterolhypocholesterolemic agent, accelerates degradation ofsynthesis by squalene synthase inhibitors does not induce3-hydroxy-3-methylglutaryl-coenzyme A reductase and stimu-myotoxicity in vitro. Toxicol Appl Pharmacol 1997; 145: 91-8lates low density lipoprotein receptor activity. J Biol Chem

23. Hiyoshi H, Yangimachi M, Ito M, et al. Effect of ER-27856, a2004; 279: 6465-73novel squalene synthase inhibitor, on plasma cholesterol in5. Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statinsrhesus monkeys: comparison with 3-hydroxy-3-methyl-on low density lipoprotein cholesterol, ischaemic heart dis-glutaryl-CoA reductase inhibitors. J Lipid Res 2000; 41:ease, and stoke: systematic review and meta-analysis. BMJ1136-442003; 326: 1423-7

24. Nishimoto T, Amano Y, Tozawa R, et al. Lipid-lowering6. Baigent C, Keech A, Kearney PM, et al. Cholesterol Treatmentproperties of TAK-475, a squalene synthase inhibitor: in vivoTrialists’ (CTC) Collaborators. Efficacy and safety of choles-and in vitro. Br J Pharmacol 2003; 139: 911-8terol-lowering treatment: prospective meta-analysis of data

25. Tavridou A, Kaklamanis L, Megaritis G, et al. Pharmacologicalfrom 90,056 participants in 14 randomised trials of statins.characterization in vitro of EP2306 and EP2302, potent inhibi-Lancet 2005; 366: 1267-78tors of squalene synthase and lipid biosynthesis. Eur J7. Gaw A, Packard CJ. Comparative chemistry, pharmacology andPharmacol 2006; 535: 34-42mechanism of action of the statins. In: Gaw A, Packard CJ,

Shepherd J, editors. Statins: the HMG CoA reductase inhibi- 26. Laughlin RC, Carey TF. Cataracts in patients treated withtors in perspective. London: Martin Dunitz, 2000: 47-61 triparanol. J Amer Med Assoc 1962; 181: 339-40



27. Mori M, Li G, Abe I, et al. Lanosterol synthase mutations cause inhibitor, BMS-188494, in healthy volunteers. J Clincholesterol deficiency-associated cataracts in the Shumiya cat- Pharmacol 1998; 38: 1116-21aract rat. J Clin Invest 2006; 116: 395-404 33. Gupta EK, Ito MK. Ezetimibe: the first in a novel class of

28. Cenedella RJ, Kuszak JR, Al-Ghoul KJ, et al. Discordant ex- selective cholesterol absorption inhibitors. Heart Diseasepression of the sterol pathway in lens underlies simvastatin- 2002; 4: 399-409induced cataracts in Chbb:Thom rats. J Lipid Res 2003; 44: 34. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent198-211 clinical trials for the National Education Program Adult Treat-

29. Cenedella RJ, Neely AR, Sexton P. Concentration and distribu- ment Panel III guidelines. Circulation 2004; 110: 227-39tion of ubiquinone (coenzyme Q), the endogenous lipid antiox- 35. JBS2: Joint British Societies’ guidelines on prevention of car-idant, in the rat lens: effect of treatment with simvastatin. Mol diovascular disease in clinical practice. Heart 2005; 91: 5Vis 2005; 11: 594-602 (Suppl.): v1-52

30. Eisele B, Budzinski R, Muller P, et al. Effects of a novel2,3-oxidosqualene cyclase inhibitor on cholesterol biosynthe-sis and lipid metabolism in vivo. J Lipid Res 1997; 38: 564-75 Correspondence and offprints: Dr Valentine Charlton-

31. Cenedella RJ, Jacob R, Borchman D, et al. Direct perturbation Menys, Division of Cardiovascular and Endocrine Sciences,of the lens membrane structure may contribute to cataracts Cardiovascular Research Group, University of Manchester,caused by U18666A, an oxidosqualene cyclase inhibitor. J

Core Technology Facility (3rd Floor), 46 Grafton Street,Lipid Res 2004; 45: 1232-41Manchester, M13 9NT, UK.32. Sharma A, Slugg PH, Hammett JL, et al. Clinical pharmacoki-

netics and pharmacodynamics of a new squalene synthase E-mail: [email protected]


Drugs 2007; 67 (1): 17-26THERAPY IN PRACTICE 0012-6667/07/0001-0017/$49.95/0


Practical Issues and Challenges in theDiagnosis and Treatment ofPulmonary SarcoidosisSurinder K. Jindal

Department of Pulmonary Medicine, Postgraduate Institute of Medical Education & Research,Chandigarh, India

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171. Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.1 Suspecting the Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181.2 Exclusion of ‘Clinical Mimics’: Differential Diagnosis from Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . 19

1.2.1 Tuberculin Anergy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191.2.2 Radiographic Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1.3 Establishing the Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201.3.1 Syndromic Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201.3.2 Pulmonary Function Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201.3.3 Biochemical Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211.3.4 Haematological Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211.3.5 Histocytological Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211.3.6 Bronchoalveolar Lavage and Induced Sputum Examination . . . . . . . . . . . . . . . . . . . . . . . . 21

1.4 Disease Extent and Severity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211.5 Disease Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2. Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222.1 Initial Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222.2 Treatment of Asymptomatic Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.3 Standard Treatment for Patients with Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.3.1 Corticosteroids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.4 Treatment of Refractory and Recurrent Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.4.1 Methotrexate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.4.2 Azathioprine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.4.3 Other Cytotoxic Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.4.4 Toxicity and Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.4.5 Other Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Sarcoidosis is a granulomatous disease with multisystem involvement. Diag-Abstractnosis is generally easy to establish from the characteristic clinical and radiograph-ic features. In India and other developing countries, tuberculosis is the closestclinical mimic and needs to be excluded before therapy for sarcoidosis is institut-ed. Tuberculin anergy and histopathological demonstration of characteristic com-pact granulomas help in the diagnosis of sarcoidosis. Corticosteroids constitute

18 Jindal

the mainstay of therapy for symptomatic pulmonary and most other forms ofextrapulmonary sarcoidosis. Asymptomatic disease does not require any treat-ment, but milder forms may be treated with topical corticosteroids and sympto-matic therapy. Alternative drugs such as cytotoxic agents, hydroxychloroquineand other agents are used either alone or in combination for the treatment ofrelapses and recurrences and refractoriness or in the presence of complications ofcorticosteroids. Treatment is usually continued for about a year, but it may need tobe prolonged in patients with disease that persists and the response to therapy isdelayed.

Sarcoidosis is a systemic disease with primary other immunosuppressant agents in patients withinvolvement of the thoracic contents, especially the healed or latent TB. Therefore, it is all the morelungs, and the hilar and mediastinal lymph nodes. important to satisfactorily exclude TB before start-The skin and eyes are the other common sites of ing therapy for sarcoidosis. These issues are of greatinvolvement. Organs of almost all other systems, significance not only in countries with a highersuch as the cardiovascular, central nervous, gastro- prevalence of TB but also in other populationsintestinal and rarely genitourinary systems, may also predisposed to TB, including patients with HIVbe affected. infection and healthcare workers.

Sarcoidosis, histologically characterised by the1. Diagnosispresence of noncaseating granulomas, has been con-

fused with tuberculosis (TB) almost since its discov-There are five important steps in the diagnosis of

ery. The possibility of an aetiological relationshipsarcoidosis: (i) suspecting the diagnosis; (ii) exclud-

between the two diseases has also been proposeding the mimics, i.e. other common causes of similar

from time to time. Both the epidemiological and theclinical features (e.g. TB); (iii) establishing the diag-

immunological evidence clearly show sarcoidosis asnosis; (iv) determination of disease extent and sever-

a distinct disease with a distinct aetiopathogenesis,ity; and (v) assessment of activity and treatment

type of treatment and natural history.[1] A recentresponsiveness.

analysis of surveillance data from Japan on 460 000employees between 1941 and 1996 and a nationwide 1.1 Suspecting the Diagnosissurvey of the general population (1959–91) has re-vealed strong epidemiological dissimilarities be- The algorithm for the diagnosis of sarcoidosistween sarcoidosis and TB, which do not support a essentially starts from the symptoms and clinicalTB aetiology of sarcoidosis.[2] Despite these facts, features with which a patient presents. A patientdifficulties prevail for practicing physicians in dis- may report with either general constitutional or re-tinguishing between sarcoidosis and TB in several spiratory symptoms to a general physician, pulmo-clinical situations, especially in countries with a nologist or other medical practitioner. Often, pa-high prevalence of TB.[3,4] This is generally attribu- tients seen in dermatology, ophthalmology, cardiol-table to similarities in several clinical features and ogy, neurology or other specialties are suspected ofthe histopathological appearance of TB and sar- having sarcoidosis on the basis of abnormalitiescoidosis. TB is the most important pathological dif- detected on a chest radiograph. Sometimes, a patientferential diagnosis of sarcoidosis in biopsy and sur- may be completely asymptomatic and present withgical pathology of the lungs, lymph nodes, skin, abnormalities on chest radiograph such as hilar andliver, bone marrow and other biopsy sites.[5] mediastinal adenopathy, or with pulmonary infil-

Added to the difficulty of diagnosis is the fear of trates detected on a routine examination for medicalactivating old lesions with corticosteroids and/or fitness or insurance purposes.


Management of Pulmonary Sarcoidosis 19

Isolated pulmonary sarcoidosis is rarely diag- ic (CT) scans. Peripheral supraclavicular, axillary,epitrochlear and inguinal lymph nodes are palpablenosed from history or physical examination. Mostin about one-third of patients.patients with pulmonary sarcoidosis will present

with general symptoms of cough; breathlessness on Almost similar groups of lymph nodes are alsoexertion; heaviness, tightness or pain in the chest involved in TB. Enlargement in both conditions isand/or general constitutional symptoms of malaise, painless, but tubercular lymph nodes are generally

matted on palpation. On the other hand, sarcoidfatigue, joint pains, aches and fever. Other clinicalnodes are discrete, movable and firm in consistency.manifestations are non-specific for several aetio-Differentiation between TB and sarcoid lymph nodelogical diagnoses.enlargement is not always possible on the basis ofPhysical examination of the chest is mostly nor-clinical examination. Fluctuations, ulceration andmal, since lung involvement generally occurs in theformation of sinuses is rare and almost rules out theform of isolated hilar and/or mediastinal lymph nodediagnosis of sarcoidosis.involvement. Parenchymal lung involvement, which

is characteristically present as diffuse inflammatory1.2 Exclusion of ‘Clinical Mimics’: Differential

infiltration, is generally silent. Very rarely, there isDiagnosis from Tuberculosis

the presence of fine crackles, which may suggest thepresence of pulmonary fibrosis. Occasionally, the Of the many clinical conditions that may simu-airways are involved, manifesting as airway ob- late a similar overall clinical and/or radiologicalstruction and wheezing on auscultation. Lung cavi- picture, TB tops the list in developing countriesties, cysts or bullae are rarely seen. Pleural effusion where the general prevalence of TB is high. Butand thickening may also occur. Effusion, if any, is lymphomas, fungal infections, vasculitides and oth-generally serous and exudative. Other rare pleural er granulomatous disorders are also important. Sev-manifestations include pneumothorax, chylothorax eral clinical features may help in differentiatingand haemothorax.[6] diagnosis from TB (table I). It is in only the 5–10%

of patients who present with overlapping featuresThere are several published reports on the fre-that the diagnosis may not be easy to make. Demon-quency of involvement of organs in sarcoidosis.[3-8]

stration of tuberculin anergy is one of the mostThe lungs are affected in >90% of patients. Involve-important tests to exclude TB. Radiological andment of extrapulmonary organs is common, alongother laboratory investigations are generally essen-with that of thoracic structures. Some of the moretial to establish a more definitive diagnosis and tocharacteristic features from which the diagnosis ofexclude other causes of similar clinical features.sarcoidosis is suspected are the dermatological man-

ifestations, such as erythema nodosum, lupus pernio 1.2.1 Tuberculin Anergyand other skin lesions; uveitis, keratoconjunctivitis Sarcoidosis patients show less reactivity tosicca, dacryocystitis and retinal vasculitis; and bilat- agents that cause delayed skin reactions.[16] Tuber-eral lacrimal, parotid and salivary gland enlarge- culin hypersensitivity is also depressed in aboutment.[9-11] Several other non-pulmonary organs that two-thirds of patients with sarcoidosis.[17] Anergy tomay also be involved either in isolation or along tuberculin in sarcoidosis is best demonstrated withwith the pulmonary disease include the liver, spleen, an intradermal skin test. The Mantoux test is one ofgastrointestinal tract, heart, peripheral nerves and the most important features to distinguish sar-kidneys.[12-15]

coidosis from TB. A positive tuberculin test as aThoracic lymph node enlargement, especially bi- marker of tubercular infection has a specificity of

lateral hilar lymphadenopathy, is the most common >85%. A positive tuberculin test in a patient withfeature diagnosed on chest radiograph. Enlarged hilar lymphadenopathy strongly suggests the aetio-retroperitoneal lymph nodes, rarely palpable, are logical diagnosis of TB. Even in countries with adetected on ultrasound or computerised tomograph- high prevalence of TB, a positive Mantoux test


20 Jindal

Table I. Summary of the differences in clinical features and other parameters between tuberculosis and sarcoidosis

Clinical feature Tuberculosis Sarcoidosis

Constitutional symptoms Fever, chills, malaise, weight loss Fatigue, myalgias, mild fever

Respiratory symptoms Cough, sputum, haemoptysis Dry cough, breathlessness, chest tightness/heaviness

Thoracic involvement

Mediastinal and hilar nodes Less common Most common

Lung parenchyma Fibrocavitary, miliary Nil or interstitial, miliary

Pleural Common Rare

Heart and pericardium Common Less common

Skin Lupus vulgaris, sinuses Erythema nodosum, lupus pernio, plaques

Lymphadenopathy

Most common Cervical Thoracic

Others Thoracic Peripheral

Eyes Choroiditis Uveitis, chorioretinitis

Other glands Rare Parotid, lacrimal, salivary

(irrespective of the size of reaction) in sarcoidosis is groups, e.g. centrilobular, perilymphatic or randomvery rare, and the incidence of Mantoux negativity in relation to the secondary lobule.[20] The distribu-in sarcoidosis is not influenced by the high rates of tion in sarcoidosis is typically perivascular. A ‘gal-Mantoux positivity in the general population.[18] A axy sign’ indicating large pulmonary nodules com-negative Mantoux test almost excludes the diagnosis posed of coalescent small nodules seen in sar-of TB except in seriously ill patients (e.g. those with coidosis has been described in TB as well.disseminated miliary disease or TB meningitis) or in Radiological findings are also useful in diagnos-heavily immunosuppressed patients such as those ing the involvement of extrapulmonary organs suchwith HIV infection. as the liver, spleen, kidneys, heart and CNS.[21]

1.2.2 Radiographic Investigations 1.3 Establishing the DiagnosisRadiological findings on a plain chest radiograph

may also be characteristic of sarcoidosis more than 1.3.1 Syndromic Presentationany other diagnosis. The most typical findings in The clinico-radiological picture is fairly charac-sarcoidosis include bilateral hilar lymphadenopathy teristic of sarcoidosis in patients with syndromicwith or without lung interstitial involvement. Over presentations such as Lofgren’s syndrome (fever,95% of 100 consecutive asymptomatic patients with bilateral hilar adenopathy, erythema nodosum andbilateral hilar lymphadenopathy were diagnosed arthralgia) and Heerfordt syndrome of chronicwith sarcoidosis.[19] The hilar lymph nodes are char- uveoparotid fever (chronic febrile illness, bilateralacteristically large, symmetrical and sharply out- parotidomegaly and uveitis). Biopsy diagnosis, al-lined and are known as ‘potato nodes’. Difficulty though useful, may not be considered essential inarises in patients with atypical features such as uni- such patients.lateral or asymmetrical massive lymph nodes or

1.3.2 Pulmonary Function Testsassociated large paratracheal and mediastinal lymphnode enlargement, or associated with pulmonary Assessment of pulmonary function tests such ascavitation, pleural effusion or miliary distribution. spirometry, carbon monoxide diffusion capacity,Miliary lung involvement is also commonly seen in cardiopulmonary exercise test and blood gases, al-TB, pneumoconioses, hypersensitivity pneumonias though useful in the overall assessment of respirato-and metastases. High-resolution CT (HRCT) images ry involvement and disability, is rarely helpful in thehelp in characterising the distribution in different differential diagnosis. The abnormalities depend up-



on the extent of involvement and whether the air- Granulomas can also be demonstrated on biop-ways are also affected. The spirometric pattern is sies of lymph nodes, skin plaques or nodules, en-generally variable from restrictive to obstructive or larged lacrimal or salivary glands, liver, musculo-mixed functional defect and from mild to severe in skeletal tissues or other clinically involved organs.degree. Biopsy of the accessory salivary gland was found to

be quite useful in elderly patients.[26] Although skin1.3.3 Biochemical Investigations is the most easily available site, it is futile to do aHypercalcaemia and hypercalciuria are the only biopsy of the erythema nodosum lesions that occur

biochemical changes that are fairly characteristic of as a result of a hypersensitivity reaction and do notsarcoidosis. Elevated liver enzymes such as AST, show any granulomas.ALT and serum alkaline phosphatase may occur in On histological examination, the sarcoid granulo-hepatic sarcoidosis. Hyperuricaemia may occur in mas are typically compact, discrete and do not showabout 40% of patients. Elevation of blood urea and any caseation. On the other hand, granulomas seencreatinine, which indicates renal dysfunction, is in TB or other infections are loose, ill formed andmore often a result of renal calculi as a result of often show caseation. Special staining for acid-fasthypercalcaemia and/or hyperuricaemia than renal bacilli and fungi may be required for granulomassarcoidosis. with atypical features.[5]

Serum ACE is frequently elevated in over half1.3.6 Bronchoalveolar Lavage and Inducedthe patients. Serum ACE estimation has been advo-Sputum Examinationcated as a measure of granuloma load and, therefore,Total and differential cell count in thea marker of disease activity.

bronchoalveolar lavage (BAL) fluid and occasional-1.3.4 Haematological Examination ly the induced sputum has been used as an index ofA number of haematological alterations are de- disease activity in sarcoidosis, but is rarely helpful

scribed on peripheral blood examination, but none in the differential diagnosis. There is a lymphocyticof these is characteristic of sarcoidosis. The com- predominance in BAL fluid in sarcoidosis. A logmonly reported abnormalities include anaemia in up transformation of the ratio of lymphocytes and poly-to 20% of patients and leukopenia in up to 40% of morph neutrophils is rarely found to be useful inpatients. Eosinophilia and thrombocytopenia are rel- differentiating sarcoidosis from idiopathic pulmona-atively rare. Most of the haematological abnormali- ry fibrosis. The pattern of differential gene expres-ties are mild and rarely severe. Bone marrow exami- sion on alveolar macrophages obtained from BALnation is rarely required in patients in whom fluid has been examined to distinguish the origin ofhaematological alterations are severe or pose diffi- granulomas, but a large number of granuloma-asso-culties in attributing a cause. ciated features were found to be common.[27]

1.3.5 Histocytological Diagnosis 1.4 Disease Extent and SeverityMost patients with extrapulmonary sarcoidosis,

most with stage II–IV pulmonary sarcoidosis and a The extent of the pulmonary disease is decidedfew with stage I pulmonary sarcoidosis need histo- based on symptomatology and chest radiographlogical confirmation of diagnosis. The lung is the findings. On the basis of a plain posteroanteriormost fruitful site for tissue biopsy. Transbronchial chest radiograph, pulmonary sarcoidosis is classi-lung biopsy provides a diagnostic yield from 40% to fied into five stages: 0, no visible intrathoracic le->90% when four or five specimens are ob- sions; I, bilateral hilar lymphadenopathy; II, bilater-tained.[22,23] Endobronchial biopsies were also posi- al hilar lymphadenopathy along with pulmonarytive for granulomas in about 35% of patients.[24] infiltration; III, pulmonary infiltrations without bi-Combining endobronchial and transbronchial lung lateral hilar lymphadenopathy; and IV, pulmonarybiopsies is shown to improve the yield even more.[25] fibrosis with honeycombing cysts and bullae forma-


22 Jindal

tion.[5] Stage IV disease is generally the most severe rectly correlate with corticosteroid therapy. Analysisand irreversible form of pulmonary sarcoidosis, but of BAL fluid as well as serological markers havenot all patients are always symptomatic. Some of the been employed for this assessment.[29] There is lym-patients with radiological stage IV disease may have phocytic predominance in BAL fluid but there is nomild or even asymptomatic presentation. direct correlation with disease activity. Radioiso-

topic scanning with Gallium-67 localises the site ofThe disease extent and severity are also judgedactive inflammation, but does not distinguish be-from the number of involved organs and the densitytween activity due to sarcoidosis, TB, malignanciesof granuloma within the organs. These are assessedor other causes.from symptomatology, number and type of organ

Scanning procedures such as HRCT have beeninvolvement, and radiographic investigations.used to find soft infiltrates suggesting activity in theSymptomatic involvement of vital organs such aslungs. Contrast enhancement with gadolinium im-the eyes (especially the posterior uveal tract), lungproves this sensitivity. Positron emission tomogra-parenchyma, liver, kidneys, and cardiovascular orphy scanning with various markers may also proveneurological systems is considered to be serious andto be useful in detecting sarcoidosis activity.[30]requires urgent systemic therapy. Lupus pernio, cys-

tic bone lesions, involvement of the upper respirato- Pulmonary function tests have sometimes beenry tract, e.g. nasal mucosa and larynx, hypercal- used to decide on the disease progression and tocaemia and calcinosis, are also forms indicating an monitor the course and response to therapy.[31] Al-adverse prognosis. Isolated skin and lymph node though a rough correlation was reported betweeninvolvement or anterior uveitis alone are considered forced vital capacity and overall histological score,to be milder forms. Investigations for individual there was no direct relationship with radiographicorgan dysfunction are helpful in assessing their in- features.[32]

volvement, but the levels of ACE or other non-2. Managementspecific markers are rarely useful.

Fortunately, the therapeutic response to several1.5 Disease Activity anti-inflammatory drugs is good in sarcoidosis. The

recommended treatments vary from none to mul-Activity implies an active ongoing disease ortidrug therapy with a combination of cytotoxicinflammation. It should not be confused with thedrugs.[5] Corticosteroids, the use of which in sar-disease extent, severity or unfavourable progno-coidosis was first reported in 1951, remain the cor-sis.[28] Unfortunately, there is no single test to assessnerstone of therapy.[33,34] However, more importantdisease activity. The presence of systemic constitu-are the issues about whom to treat and when to starttional symptoms (e.g. fever, aches and pains, weight(and stop) treatment. In general, there is a broadloss, nausea, vomiting and headaches), active uvei-consensus on most of these issues.tis, hypercalcaemia and organ inflammation such as

serositis, myocarditis, hepatitis and neurosarcoido-2.1 Initial Evaluation

sis, indicate the presence of active disease. StageI–III pulmonary disease is generally considered to As discussed in section 1, before deciding onbe active, but stage IV disease on radiography is treatment, each patient requires evaluation for theusually seen as end-stage fibrosis from progressive indication as well as the form of therapy. A detailedpulmonary disease. history and clinical examination, including eyes, a

Laboratory assessments, such as a raised erythro- posteroanterior chest radiograph, electrocardiogramcyte sedimentation rate, leukocytosis, thrombo- and tuberculin skin test, are important. Routinecytopenia or raised serum ACE levels indicate ongo- haematological and biochemical investigations areing inflammation. A raised serum ACE level is a also carried out for their baseline values. Lung CTgood indicator of granuloma load but does not di- scans are required in patients with atypical findings,



normal chest radiograph but strong clinical suspi- the presence of adverse effects, contraindications orcion, or in the presence of complications. non-responsiveness to corticosteroids. However,

each drug is likely to have its own advantages anddisadvantages.2.2 Treatment of Asymptomatic Patients

2.3.1 CorticosteroidsMost investigators agree that asymptomatic pul-monary disease rarely requires any therapy, but pa- Efficacy of treatment with corticosteroids wastients with persistent pulmonary infiltrates, even if recently evaluated in a meta-analysis of 12 ran-asymptomatic, may require treatment.[5,33-36] Pa- domised, controlled trials that included 1051 pa-tients with milder forms of disease, such as skin tients.[38] The treatment was shown to improve thedisease (erythema nodosum) alone or anterior uvei- chest radiographic findings and a global score oftis, may need only topical treatments, whereas pa- chest radiograph, symptoms and spirometry, buttients with stage I pulmonary sarcoidosis may be little improvement in lung function. Long-term ben-observed without any treatment. Patients manifest- efits beyond 2 years were not clearly demonstrable.ing with only cough as a symptom may be treated

Although the earlier studies reported the use ofwith inhaled corticosteroid therapy. But all newcorticotropin (adrenocorticotrophic hormone;patients with stage IV radiographic changes, even ifACTH) in the initial treatment, prednisolone orasymptomatic, should be given a treatment trial withprednisone remain the most commonly used drug.corticosteroids for 8–12 weeks before being labelledThe initial dosage of 20–40 mg/day is continued foras having end-stage fibrosis.1–3 months depending upon symptomatic response.Treatment is generally started on the basis ofHigher dosages are required for cardiac- andsymptoms and disease severity. Patients with vitalneurosarcoidosis. Re-evaluation of the condition isorgan involvement (such as the heart, liver or CNS),required if no response is seen after 3 months ofsymptomatic pulmonary disease or hypercalcaemiatherapy. In patients who show an adequate response,definitely require oral treatment. Although severalthe dosage is tapered to 5–10 mg/day and continuedlaboratory parameters, as discussed in section 1.5,for at least a year as maintenance therapy. Alternatehave been used to assess the stage of disease, noneday therapy in equivalent dosages may also be giv-has been found to be as useful as clinical featuresen. Some patients show recurrent relapses and needand end organ involvement for making a treatmentlong-term, low-dose treatment. Inhaled corticoste-decision.roids have been used in some studies with a limitedThere are no clear guidelines for the treatment ofbenefit.[39]patients who are strongly suspected of having sar-

coidosis according to various criteria but who exhib- Adverse effects of corticosteroids are common.it a positive tuberculin test. We generally practice Some of the effects, such as fluid retention, mildadditional anti-tubercular chemoprophylaxis with elevations in blood pressure and glucose intoler-rifampicin and isoniazid in such patients during ance, usually disappear when treatment is taperedtreatment for sarcoidosis. and discontinued, but the problems may persist,

especially with prolonged corticosteroid use and in2.3 Standard Treatment for Patients patients with pre-existing diseases such as hyperten-with Symptoms sion and diabetes mellitus. Several other important

complications include weight gain, cataracts, osteo-porosis and bone fractures, which require additionalThere are several options of treatment available.treatment measures. Careful monitoring and controlCorticosteroids, usually oral, constitute the firstof corticosteroid-induced complications is thereforechoice of treatment for stage II and III disease.[33-37]

important in all patients.Alternative or additional drugs are required only in


24 Jindal

2.4 Treatment of Refractory and trointestinal symptoms, mucosal ulcers and leuko-Recurrent Disease penia are common adverse effects. Folic acid sup-

plementation at a dosage of 1mg every day is helpfulCytotoxic agents are required in patients with a to lessen these adverse effects of cyclophosphamide.

recurrence of disease, which is seen in almost one- Hepatic toxicity may occur with a cumulative dos-third of patients within 2 years of the discontinua- age of about 1g or more; however, pulmonary toxic-tion of corticosteroid therapy. Cytotoxic agents are ity is rare. Cyclophosphamide is commonly associ-also used in the presence of non-responsiveness to ated with haemorrhagic cystitis.corticosteroids or with the development of compli- Regular monitoring for hepatic, haematologicalcations with corticosteroids. and renal toxicity is required in patients receiving

cytotoxic agents. Complete blood counts and2.4.1 Methotrexateurinalyses are advised every 4–6 weeks in patientsMethotrexate, the most frequently employedreceiving azathioprine. Patients receiving cy-agent, is effective in about two-thirds of patientsclophosphamide or methotrexate need similar moni-with cutaneous, pulmonary, joint, ocular or neuro-toring with urinalysis, complete blood counts andlogical disease.[36,40] Methotrexate is generally usedliver function tests every 3–4 weeks.in a 7.5–15mg dose every week for a period of about

Gonadal damage and teratogenicity is an impor-6–12 months. Response is usually slow and may nottant toxicity observed with the use of cytotoxicappear up to 6 months after the start of therapy.drugs. Cyclophosphamide can cause early meno-

2.4.2 Azathioprine pause and aspermia. Recovery from suppression ofExperience with azathioprine in sarcoidosis is spermatogenesis as well as from ovarian failure is

limited. It is generally used in patients with chronic unpredictable. Azathioprine is relatively less terato-disease at an oral dose of 2–3 mg/kg. Its efficacy is genic.[5] Use of cytotoxic drugs is therefore bestgenerally similar to that of methotrexate.[41]

avoided for women of childbearing age, but if theindication for the use of cytotoxic drugs is impor-2.4.3 Other Cytotoxic Agentstant, pregnancy is better postponed until at least 6Chlorambucil was used in the past but is nomonths after the last dose of any of these agents.longer favoured.[5] Cyclophosphamide is useful forThere is an increased opportunity now to preserverefractory disease, especially neurosarcoidosis. It issperm as well as ova before administration of chem-used either orally in a dosage of 50–150mg daily orotherapy with cytotoxic drugs, so as to enable preg-500–2000mg intravenously every 2 weeks. In viewnancy at a later stage. This has been made possibleof an increased risk of malignancy when used on awith the advancements in cryopreservation and as-daily basis for >1 year, it is recommended to use thissisted reproduction techniques such as in vitro fer-agent on alternate days.[42] Increased daily fluidtilisation with intracytoplasmic sperm injections.intake is recommended with its use to minimise the

Readers interested in cryopreservation in patientsoccurrence of haemorrhagic cystitis. Leflunomide,with non-malignant systemic diseases are referred toan analogue of methotrexate, is less toxic but moreHallak et al.,[44] Ranganathan et al.,[45] Mattle etcostly.[43]

al.[46] and Sonmezer et al.[47]

2.4.4 Toxicity and Monitoring2.4.5 Other AgentsMost cytotoxic drugs can cause nausea and

vomiting, which are generally dose dependent. With Both chloroquine and hydroxychloroquine havemethotrexate, other commonly affected organs are been used with some success in patients with sar-the bone marrow, liver, kidney and the gonads. coidosis, especially for cutaneous lesions, as well asAzathioprine may cause leukopenia and gastrointes- for hypercalcaemia and neurosarcoidosis. Oculartinal effects; however, its hepatic toxicity is less than toxicity, frequently seen with chloroquine, is notthat of methotrexate. With cyclophosphamide, gas- seen with hydroxychlorquine, even with prolonged



6. Soskel NT, Sharma OP. Pleural involvement in sarcoidosis.use.[48] Hydroxychloroquine can also be combinedCurr Opin Pulm Med 2000; 6: 455-68

with corticosteroids. 7. Loddenkemper R, Kloppenborg A, Schoenfeld N, et al. Clinicalfindings in 715 patients with newly detected pulmonary sar-Several other agents that have been used forcoidosis: results of a cooperative study in former West Germa-patients with refractory sarcoidosis include ci- ny and Switzerland. WATL Study Group. Sarcoidosis VascDiffuse Lung Dis 1998; 15: 178-82closporin, radiation therapy and immunomodulators

8. James DG, Neville E, Siltzbach LE, et al. A worldwide reviewsuch as thalidomide, pentoxifylline and infliximab.of sarcoidosis. Ann N Y Acad Sci 1976; 278: 321-34

Experience with most of these agents is limited and 9. Hutchinson J. Case of livid papillary psoriasis. In: Illustrationsof clinical surgery. Vol. 1. London: J&A Churchill, 1877: 42-3anecdotal.[36,42,49] It is important to screen patients

10. Sharma OP. Cutaneous sarcoidosis: clinical features and man-for occult TB and lymphoproliferative disordersagement. Chest 1972; 61: 320-5

before starting therapy with any of these agents. 11. Usui Y, Kaiser ED, See RF, et al. Update of ocular manifesta-tions in sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis 2002;19: 167-753. Conclusion

12. Sharma O, Kadakia J, Sharma O. Gastrointestinal sarcoidosis.Semin Respir Med 1992; 13: 442-9

The management of sarcoidosis over the last few 13. White ES, Lynch III JP. Sarcoidosis involving multiple systems:diagnostic and therapeutic challenges. Chest 2001; 119:decades has seen many new developments. One1593-7

important aspect that has emerged recently is to 14. Stern BJ. Neurological complications of sarcoidosis. Curr OpinNeurol 2004; 17: 311-6avoid treatment for several forms of mild or asymp-

15. Gullapalli D, Phillips II LH. Neurologic manifestations of sar-tomatic disease. On the other hand, valuable strate-coidosis. Neurol Clin 2002; 20: 59-83

gies, alternative drugs and treatment guidelines have 16. Scadding JG. Skin reaction of delayed type. Acta Med Scand1965; 425: 191-4now become available for routine decision making

17. James DG, Nevile E, Walker A. Immunology of sarcoidosis.in the presence of recurrence, complications or asso-Am J Med 1975; 59: 388-94

ciated medical conditions and co-morbid ill- 18. Gupta D, Chetty M, Kumar N, et al. Anergy to tuberculin insarcoidosis is not influenced by high prevalence of tuberculinness.[5,34-36,50,51]

sensitivity in the population. Sarcoidosis Vasc Diffuse LungDis 2003; 20: 40-5

Acknowledgements 19. Winterbauer RH, Belic N, Moores KD. A clinical interpretationof bilateral hilar adenopathy. Ann Intern Med 1973; 79: 65-71

20. Andreu J, Mauleon S, Pallisa E, et al. Miliary lung diseaseThe author acknowledges the help of Drs Ritesh Agarwalrevisited. Curr Probl Diagn Radiol 2002; 31: 189-97and Dheeraj Gupta in providing some of the material for the

21. Koyama T, Ueda H, Togashi K, et al. Radiologic manifestationspaper, and Ms Manju Aggarwal for preparing and typing theof sarcoidosis in various organs. Radiographics 2004; 24:

manuscript. The author has not received any funding or other 87-104sponsorship from any source for writing this paper. The 22. Gilman MJ, Wang KP. Transbronchial lung biopsy in sar-author has no conflicts of interest that are directly relevant to coidosis: an approach to determine the optimal number of

biopsies. Am Rev Respir Dis 1980; 122: 721-4the content of this review.23. Jindal SK, Gupta D, Aggarwal AN. Sarcoidosis in developing

countries. Curr Opin Pulm Med 2000; 6: 448-5424. Gupta D, Mahendran C, Aggarwal AN, et al. Endobronchial visReferences

a vis transbronchial involvement on fiberoptic bronchoscopy1. James DG. Sarcoidosis 2001. Postgrad Med J 2001; 77: 177-80in sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis 2001; 18:2. Hosoda Y, Sasagawa S, Yamaguchi T. Sarcoidosis and tubercu-91-2losis: epidemiological similarities and dissimilarities: a review

25. Torrington KG, Shorr AF, Parker JW. Endobronchial diseaseof a series of studies in a Japanese work populationand racial differences in pulmonary sarcoidosis. Chest 1997;(1941-1996) and the general population (1959-1984). Sar-111: 619-22coidosis Vasc Diffuse Lung Dis 2004; 21: 85-93

26. Chevalet P, Clement R, Rodat O, et al. Sarcoidosis diagnosed in3. Jindal SK, Gupta D, Aggarwal AN. Sarcoidosis in India: practi-elderly subjects: retrospective study of 30 cases. Chest 2004;cal issues and difficulties in diagnosis and management. Sar-126: 1423-30coidosis Vasc Diffuse Lung Dis 2002; 19: 176-84

27. Gaede KI, Mamat U, Muller-Quernheim J. Differential gene4. Sharma SK, Mohan A. Sarcoidosis: global scenario & Indianexpression pattern in alveolar macrophages of patients withperspective. Indian J Med Res 2002; 116: 221-47sarcoidosis and tuberculosis. J Mol Med 2004; 82: 206-105. Statement on sarcoidosis. Joint statement of the American Tho-

28. Costabel U, Guzman J, Drent M. Diagnostic approach to sar-racic Society (ATS), the European Respiratory Society (ERS)coidosis. Eur Res Mon 2005; 32: 259-64and the World Association of Sarcoidosis and other Granulo-

matous Disorders (WASOG) adopted by the ATS Board of 29. Ziegenhagen MW, Rothe ME, Schlaak M, et al. Bronchoalveo-Directors and the ERS Executive Committee, February 1999. lar and serological parameters reflecting the severity of sar-Am J Respir Crit Care Med 1999; 160: 736-55 coidosis. Eur Respir J 2003; 21: 407-13


26 Jindal

30. Lebtahi R, Crestani B, Belmatoug N, et al. Somatostatin recep- 43. Baughman RP, Lower EE. Leflunomide for chronic sarcoidosis.tor scintigraphy and gallium scintigraphy in patients with Sacoidosis Vasc Diffuse Lung Dis 2004; 21: 43-8sarcoidosis. J Nucl Med 2001; 42: 21-6

44. Hallak J, Hendin BN, Thomas AJ, et al. Investigation of fertiliz-31. Bradivik I, Wollmer P, Blom-Bulow B, et al. Lung mechanics ing capacity of cryopreserved spermatozoa from patients with

and gas exchange in steroid treated pulmonary sarcoidosis: acancer. J Urol 1998; 159: 1217-9seven year follow up. Sarcoidosis 1991; 8: 105-14

45. Ranganathan P, Mahran AM, Hallak J, et al. Sperm cry-32. Gupta D, Jorapur V, Bambery P, et al. Pulmonary sarcoidosis:opreservation for men with nonmalignant systemic diseases: aspirometric correlation with transbronchial biopsy. Sarcoidosis

Vasc Diffuse Lung Dis 1997; 14: 77-80 descriptive study. J Andrology 2002; 23: 71-533. Siltzbach LE, Posner A, Medine MM. Cortisone therapy in 46. Mattle V, Behringer K, Engert A, et al. Female fertility after

sarcoidosis. JAMA 1951; 147: 927-9 cytotoxic therapy -protection of ovarian function during chem-34. Gibson GJ. Sarcoidosis: old and new treatments (editorial). otherapy of malignant and nonmalignant diseases. Eur J

Thorax 2001; 56: 336-9 Hematol Suppl 2005; 66: 77-8235. Fazzi P. Pharmacotherapeutic management of pulmonary sar-

47. Sonmezer M, Shamonki MI, Oktay K. Ovarian tissue cry-coidosis. Am J Respir Med 2003; 2: 311-20

opreservation: benefits and risks. Cell Tissue Res 2005; 322:36. Baughman RP, Lower EE. Therapy for sarcoidosis. Eur Respir

125-32Mon 2005; 32: 301-1548. Jones SK. Ocular toxicity and hydroxychloroquine: guidelines37. Gibson GJ, Prescott RJ, Muers MF, et al. British Thoracic

for screening. Br J Dermatol 1999; 140: 3-7Society sarcoidosis study: effects of long term corticosteroidtreatment. Thorax 1996; 51: 238-47 49. Doty JD, Mazur JE, Judson MA. Treatment of sarcoidosis with

38. Paramothayan S, Jones PW. Corticosteroids for pulmonary sar- infliximab. Chest 2005; 127: 1064-71coidosis. Cochrane Database Syst Rev 2005; (2): CD001114

50. Reynolds HY. Sarcoidosis: impact of other illnesses on the39. Pietinalho A, Tukiainen P, Haahtela T, et al. Oral prednisolone presentation and management of multi-organ disease. Lung

followed by inhaled budesonide in newly diagnosed pulmona-2002; 180: 281-99ry sarcoidosis: a double-blind, placebo-controlled multicenter

51. Sharma OP. Pulmonary sarcoidosis: management. J Postgradstudy. Finnish Pulmonary Sarcoidosis Study Group. Chest1999; 116: 424-31 Med 2002; 48: 135-41

40. Lower EE, Baughman RP. Prolonged use of methotrexate forsarcoidosis. Arch Intern Med 1995; 155: 846-51

Correspondence and offprints: Dr Surinder K. Jindal, De-41. Muller-Quernheim J, Kienast K, Held M, et al. Treatment ofchronic sarcoidosis with an azathiprine/prednisolone regimen. partment of Pulmonary Medicine, Postgraduate Institute ofEur Respir J 1999; 14: 1000-1

Medical, Education & Research, Sector 12, Chandigarh,42. Doty JD, Mazur JE, Judson MA. Treatment of corticosteroid-

160012, India.resistant neurosarcoidosis with a short-course cyclophos-phamide regimen. Chest 2003; 124: 2023-6 E-mail: [email protected]


Drugs 2007; 67 (1): 27-55REVIEW ARTICLE 0012-6667/07/0001-0027/$49.95/0


Serotonergic DrugsEffects on Appetite Expression and Use for theTreatment of Obesity

Jason C.G. Halford,1,2 Joanne A. Harrold,1,3 Emma J. Boyland,1 Clare L. Lawton2 andJohn E. Blundell2

1 Kissileff Laboratory for the Study of Human Ingestive Behaviour, School of Psychology,University of Liverpool, Liverpool, UK

2 Institute of Psychology, University of Leeds, Leeds, UK3 Department of Medicine, Diabetes and Endocrinology Research Group, University of

Liverpool, Liverpool, UK

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

1. Endogenous Serotonin and Serotonin Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292. The Serotonin Satiety System: Interaction with Other Neuropeptide Signalling Pathways . . . . . . . . . 30

2.1 Serotonin and Neuropeptide Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312.2 Serotonin and the Melanocortin System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312.3 Serotonin and Orexin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3. Endogenous Serotonin and Altered Nutritional Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.1 Reduced Energy Intake or Bodyweight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.2 Cachexic and Endogenous Serotonin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343.3 Increased Energy Intake or Bodyweight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4. Effects of Serotonergic Drugs on Food Intake in Rodents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.1 Serotonin Receptors and Food Intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.2 Serotonin and Feeding Behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364.3 Serotonin Receptor Knockouts, Food Intake and Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

5. Serotonergic Drugs, Food Intake and Feeding Behaviour in Humans . . . . . . . . . . . . . . . . . . . . . . . . . . 385.1 5-Hydroxytryptophan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385.2 Fenfluramine and Dexfenfluramine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385.3 Fluoxetine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405.4 Sibutramine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415.5 Preferential and Selective Serotonin Receptor Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

6. Serotonergic Drugs and Obesity: Rodent Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446.1 Dexfenfluramine and Selective Serotonin Reuptake Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446.2 Serotonin Receptor Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

7. Serotonergic Drugs and Weight Loss: Clinical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457.1 Fenfluramine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457.2 Dexfenfluramine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467.3 Sibutramine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467.4 Selective Serotonin Reuptake Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477.5 Serotonin Precursors and Receptor Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

28 Halford et al.

8. Current and Future Serotonergic Anti-Obesity Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489. Novel Serotonergic Targets for Weight Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

10. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Over 35 years of research suggests that endogenous hypothalamic serotoninAbstract(5-hydroxytryptamine) plays an important part in within-meal satiation andpost-meal satiety processes. Thus, the serotonin system has provided a viabletarget for weight control, critical to the action of at least two effective anti-obesitytreatments, both producing clinically significant weight loss over a year or more.Numerous serotonin receptor subtypes have been identified; of these, serotonin5-HT1B and 5-HT2C receptors have been specifically recognised as mediators ofserotonin-induced satiety.

A number of serotonergic drugs, including selective serotonin reuptake inhibi-tors (SSRIs), dexfenfluramine and 5-HT2C receptor agonists, have been shown tosignificantly attenuate rodent bodyweight gain. This effect is strongly associatedwith marked hypophagia and is probably mediated by the hypothalamic mela-nocortin system. Additionally, sibutramine, dexfenfluramine, fluoxetine and the5-HT2C receptor agonist chlorophenylpiperazine (mCPP) have all been shown tomodify appetite in both lean and obese humans, resulting in reduced caloricintake. Clinical studies demonstrate serotonergic drugs specifically reduce appe-tite prior to and following the consumption of fixed caloric loads, and cause areduction in pre-meal appetite and caloric intake at ad libitum meals. Weight lossin the obese has also been produced by treatment with both the serotonin precursor5-hydroxytryptophan and the preferential 5-HT2C receptor agonist mCPP.

A new generation of 5-HT2C receptor selective agonists have been developedand at least one, lorcaserin (APD356), is currently undergoing clinical trials. Inaddition, 5-HT6 receptor antagonists such as PRX-07034 and BVT74316 havebeen shown to potently reduce food intake and bodyweight gain in rodent modelsand have recently entered clinical trials. However, the role of the 5-HT6 receptorin the expression of appetite remains to be determined. The hope is that thesedrugs will not only be free of their predecessors’ adverse effect profiles, but willalso be equally or more effective at regulating appetite and controllingbodyweight.

To understand the potential for serotonin (5- length and frequency, and govern the duration ofhydroxytryptamine)-targeted treatment for obesity it between-meal intervals. Signals are generated fromis necessary to consider the nature of appetite regu- the very commencement of consumption, the short-lation and the role of serotonin within it. Appetite is term consequences of which serve to terminate eat-an expression of numerous regulatory processes that ing behaviour and act as powerful inhibitors ofdetermine the initiation and termination of meals, further intake. This process signals to the brain anthe amount and types of foods consumed, and meal estimation of a meal as opposed to an accurate


Serotonergic Drugs in Obesity 29

analysis of content. There is an important distinction other recent reviews and are best discussed by thosemore qualified.[2,3] This review instead emphasisesbetween the short-term satiety signals produced bythe research identifying both the role of serotoninthe physiological consequences of meal intake (epi-and the functioning of the endogenous serotoninsodic) and the longer-term signals created by thesatiety system as a whole. The review then focusesbody’s constant metabolic need for energy (tonic).on the effects of serotonergic drugs in human feed-Episodic signals are a crucial factor in the meal-by-ing behaviour studies as well as clinical weight lossmeal regulation of energy intake, and are critical totrials.both the appetite fluctuations and patterns of eating

The monoamine neurotransmitter serotonin wasbehaviour we undertake throughout the day. Thelinked to the control of food intake and of feedingmonoamine neurotransmitter serotonin influencesbehaviour for the first time 30 years ago. Earlythis episodic, meal-by-meal, regulation of food in-studies on this subject used serotonin precursorstake through its role in satiety.[1] Tonic inhibitory(such as tryptophan and 5-hydroxytryptophan [5-signals, by contrast, are not a result of this flux ofHTP]) to increase CNS serotonin levels and demon-sensory, cognitive, pre-absorption and post-absorp-strated a significant reduction in the food intake oftive factors that are crucial to the satiety develop-laboratory animals in response to this manipulation.ment process. Tonic factors are generated instead byThis hypophagic response could also be generatedthe storage and metabolism of energy. The adiposevia other mechanisms to increase serotonin activity,tissue hormone leptin is a key example of this.such as the direct administration of serotonin intoLeptin is secreted in response to excess fat deposi-the CNS, directly stimulating (agonising) serotonintion and also acts as a potent inhibitor of food intake.receptors or blocking synaptic serotonin break-We have argued previously that serotonin and leptindown.[4-8] Additionally, it was shown that preventingconstitute separate aspects of appetite regulation andserotonin synthesis (using parachlorophenlyanineare both generated by markedly distinct processes,[pCPA]), which depletes neuronal serotonin, or thebut both ultimately act to inhibit food intake via theneurotoxic lesioning of serotonin neurons (with 5,7-same hypothalamic circuitry.[1]

dihyroxytryptamine [5,7-DHT]), resulted in not on-Ultimately, for any anti-obesity drug that acts on ly a prevention of serotonin-induced hypophagia but

reducing food intake to be successful it must influ- an increase in food intake.[9] Relative to now, theence subjective feelings of appetite experiences pharmacological tools available to these early re-before, during and after food intake, enabling the searchers were nonspecific; nevertheless, it becameindividual to consume less food. With regard to evident that CNS serotonin was implicated in theserotonin, there is a greater wealth of data on the control of food intake. Blundell[10] proposed in 1977effects on human appetite and rodent eating beha- that the serotonin system had not only an inhibitoryviour than for any other peripheral or central target. role in feeding but was also a key satiety factor (i.e.Therefore, the purpose of this review is to explore a factor in the natural energy intake control mecha-the efficacy of serotonergic drugs at inhibiting appe- nism).tite and consequently food intake. These are theeffects which should ultimately induce the clinically 1. Endogenous Serotonin andsignificant weight loss required to treat obesity. This Serotonin Receptorsis not an in-depth review of the serotonin receptorsubtype pharmacology or their role in feeding. Neuronal serotonin is synthesised from the essen-These topics have been covered comprehensively in tial amino acid tryptophan. In the cell body cyto-


30 Halford et al.

plasm, the enzyme tryptophan hydroxylase hydrox- the serotonin 5-HT1A, 5-HT1B and 5-HT2C recep-tors. The postsynaptic 5-HT1B and 5-HT2C recep-ylates dietary I-tryptophan to 5-HTP. 5-HTP is thentors are generally thought to be involved in therapidly decarboxylated at the terminal, by the en-serotonin satiety system.[12] However, recent atten-zyme I-amino acid decarboxylase, to produce sero-tion has focused on 5-HT6 receptors with data fromtonin. The majority of serotonin produced is takena number of sources demonstrating 5-HT6 receptorup via a vesicle membrane transport mechanism andantagonism reduces food intake and bodyweightstored in presynaptic vesicles. After release, synap-gain (see section 9). Whether these effects are con-tic serotonin continues to stimulate pre- and post-sistent with satiety remains to be determined.synaptic receptors until it is either converted to 5-

hydroxyindole acetic acid (5-HIAA) by monoamine2. The Serotonin Satiety System:oxidase (MAO) or reabsorbed into the presynapticInteraction with Other Neuropeptideneuron for reuse (a sodium-dependent process) [fig-Signalling Pathwaysure 1]. The late 1980s and early 1990s were a time

for significant advances in the discovery and identi-Energy balance is regulated by a number of pe-fication of novel serotonin receptors.[11] The seroto-

ripherally generated signals. Many of these signalsnin receptors currently thought to be most directlyconverge on the hypothalamic nuclei known to beimplicated in the feeding control mechanisms arecritical to the control of feeding behaviour expres-sion. These nuclei contain a diverse selection ofneuropeptide systems which either inhibit (anorexi-genic) or stimulate (orexigenic) food intake. Theextent to which serotonin-induced hypophagia in-volves other neuropeptide signalling systems linkedto the hypothalamic regulation of appetite is not yetfully understood and has been explored by a numberof studies. Of particular interest is the antagonisticrelationship between serotonin and the food intakestimulating neuropeptide neuropeptide Y (NPY).Similarly, links between the serotonergic neuronsand another orexigenic system, orexin, may also becritical to the expression of appetite. It seems thatthe interaction between these orexigenic neuropep-tides and serotonin may be critical in determiningthe expression of appetite, i.e. whether to eat or not.However, perhaps the more critical neuropeptide isthe anorexigenic melanocortin system. It seems thata functioning hypothalamic melanocortin system isrequired for serotonergic drugs to alter feeding be-haviour, and presumably for feeding-inducedchanges in endogenous serotonin to influence appe-tite. Interestingly, these systems are critical in medi-ating the effects of tonic peripheral signals such as

N

CH 2 C OH

OHO

N

CH 2 CH 2 NH 2HO

N

CH 2 CH NH 2

COOHHO

N

CH2 CH NH 2

COOH

5-Hydroxyindoleacetic acid (5-HIAA)

Monoamineoxidase

Serotonin (5-HT)

Amino acid decarboxylase

5-Hydroxytryptophan (5-HTP)

Tryptophan hydroxylase

Tryptophan

Fig. 1. The synthesis and metabolism of serotonin (reproducedfrom Blundell and Halford,[12] with permission).



leptin and insulin. The serotonin system itself ap- and bodyweight. It appears to be a critical site for thepears to be more sensitive to peripheral episodic integration of afferent episodic and tonic signals ofsignals such as those produced by a number of gut intake and energy status. Neuroanatomical, molecu-peptides released in response to the ingestion of lar and electrophysiological methods have beenfood.[12] Thus, it has been proposed that the seroto- used in combination to show that serotonergic drugsnin system is a key link in the integration of episodic are dependent on a functioning melanocortin systemsignals into the hypothalamic energy regulatory sys- to exert their effects on feeding. It has been showntem.[1] that anorectic serotonin drugs activate pro-opiome-

lanocortin (POMC) neurons in the ARC. This effect2.1 Serotonin and Neuropeptide Y was demonstrated by specifically examining FOS-

like immunoreactivity in response to dexfen-NPY plays a significant role in the control offluramine administration.[17] Confirmation thatfood intake and energy balance. An interaction be-dexfenfluramine directly activates these neuronstween NPY and serotonin was suggested by earlycomes from electrophysiological studies that inves-studies in which NPY-induced hyperphagia wastigated the expression of green fluorescent proteinshown to be blocked by the serotonin receptor ago-under control of the POMC promoter.[17] POMC isnist fenfluramine. In addition to this, hypothalamicthe peptide precursor of a variety of molecules,NPY levels have been reported to decrease afterincluding adrenocorticotrophic hormone (ACTH),treatment with serotonin receptor agonists, and toendorphins and enkephalins, all of which play im-increase after administration of serotonin receptorportant roles in metabolic processes. Many of theseantagonists.[13,14] Reduced serotonin availability hasproducts can generate pronounced effects on feed-also been found to decrease the density of NPYing behaviour. However, α-melanocyte-stimulatingneurons.[15] NPY is synthesised in the arcuate nucle-hormone (α-MSH) appears to be the major POMCus (ARC) and released in a number of hypothalamicproduct in terms of an interaction with serotonin.areas, predominantly the paraventricular nucleusMoreover, α-MSH is the main breakdown product(PVN) and ventromedial hypothalamic nucleusof POMC. Up to 80% of α-MSH-containing neu-(VMH). Subsequent studies have narrowed theirrons express mRNA for the 5-HT2C receptor[17] andfocus to specifically examine the potential interac-it is probable that these receptors play a part in thetive relationship of PVN NPY and serotonin, and the

manner in which this interaction exerts influence serotonin-mediated activation of POMC neurons.over the control of feeding. It is worth mentioning Furthermore, previous reports indicate that seroton-that only administration of 5-HT2A receptor agonists ergic compounds cause the release of α-MSH fromand antagonists into the PVN has been shown to superfused hypothalamic slices. As a final point, it ismodulate NPY-induced hyperphagia.[16] Despite a worth noting that a blockade of melanocortin 3 and 4lack of confirmation over the nature of the serotonin receptors, either by pharmacological or geneticreceptor expressed by PVN NPY neurons, the find- means, is sufficient to attenuate the anorectic effectsings highlighted do provide continued support for of serotonergic drugs.[18] These findings indicatethe interaction of NPY and serotonin mechanisms. that serotonin targets downstream melanocortin

pathways and acts via the 5-HT2C receptor to reduce2.2 Serotonin and the Melanocortin System food intake and bodyweight.

In a recent paper, Heisler et al.[19] have furtherThe melanocortin system is well known to be akey system involved in the regulation of food intake demonstrated the role of downstream melanocortin


32 Halford et al.

MC4R

α-MSHAgRP

POMC

5-HT2CR

GABA

5-HT5-HT

5-HT1BR

NPY/AgRP

Fig. 2. Schematic diagram of proposed serotonin sites of action on melanocortin pathways. This shows the inhibition of agouti-relatedpeptide (AgRP) neuron and the inhibition of the AgRP axonal projection by serotonin through serotonin 5-HT1B receptor (5-HT1BR) action,and the activation of the pro-opiomelanocortin (POMC) neuron through the 5-HT2C receptor (5-HT2CR) [reproduced from Heisler et al.,[19]

with permission]. MC4R = melanocortin 4 receptor; MSH = melanocyte-stimulating hormone; NPY = neuropeptide Y

4 receptors (MC4R) in mediating serotonin-induced lease of α-MSH. It would be interesting to deter-mine whether blockade of MC4R reverses the effecthypophagia. It seems that activation of both 5-HT2C

of 5-HT2C and 5-HT1B receptor agonism on beha-and 5-HT1B receptors produce hypophagia by pro-viour indicators of satiety such as the behaviouralmoting the release of the endogenous agonist andsatiety sequence. Certainly, the effects of dexfen-inhibiting the release of the endogenous antagonistfluramine, a drug that reduces food intake throughof the MC4R. This is achieved through three mecha-the activation of 5-HT2C and 5-HT1B receptors, isnisms (see figure 2). Firstly, Heisler et al., postulateineffective at reducing food intake at normal anorec-that serotonin inhibits orexigenic agouti-related pep-tic doses in mice with ectopic expression of endoge-tide (AgRP) neurons in the ARC via 5-HT1B recep-nous AgRP.[19] Reductions in food intake are onlytor activation. This inhibits the releases of theobserved at doses previously shown to produce se-

MC4R antagonist AgRP. Secondly, activation ofdation.

axonal 5HT1B receptors on AgRP neuronal projec-tions also decreases their inhibitory effect on adja-

2.3 Serotonin and Orexincent anorexigneic POMC neurons. This promotesthe release of the MC4R agonist α-MSH. Finally, as Serotonin and orexin are both involved in feedingprevious studies have shown, serotonin also acti- behaviour regulation, in addition to their roles in thevates these anorexignic POMC neurons via activa- sleep-wake cycle. Orexin neurons project to a vasttion of 5-HT2C receptors, again promoting the re- majority of areas in the brain from their specific



location in the lateral hypothalamus. Notably, in to cause perturbations in nutritional status. Thismonoaminergic nuclei (such as the noradrenergic allows an insight into the regulation (both nutritionallocus ceruleus, the serotonergic dorsal raphe nucleus and physiological) of neurotransmitter systems thatand the dopaminergic ventral tegmental area) there function to control feeding and energy balance.are dense projections of orexin neurons. Of these, Therefore, if endogenous serotonin is a key appetitethe serotonergic dorsal raphe nucleus is one of the regulation factor, its levels and functioning shoulddensest projection sites.[20] In addition to this, within reflect the organism’s nutritional status, fed orthe dorsal raphe nucleus, serotonergic neurons ex- fasted.press both orexin OX1R and OX2R receptors[21,22]

and are excited by orexin-A neuropeptide.[23,24] 3.1 Reduced Energy Intake or BodyweightMoreover, orexin has also been shown to alter sero-

It would be logical to assume that a system thattonin release in the hypothalamus.[25] An involve-inhibits food intake would be inactivated in situa-ment of a serotonergic pathway in orexin-inducedtions of energy need, allowing the organism to ad-behaviour is indicated by these anatomical projec-dress this with feeding behaviour. It has been showntions and interactions.[26] However, there is to date, athat malnourished animals demonstrate a significantlack of published studies focusing on this interactionreduction in CNS serotonin levels in several areas ofbetween orexin and serotonin that influences feed-the brain.[32] Furthermore, intense food-seeking be-ing behaviour. As it is known that serotonin inhibitshaviour and increased food intake (when food isfood intake, it is probable that orexin projection tomade available) has been noted in fasting animals.the dorsal raphe nucleus establishes part of an inhib-Food deprivation is thought to increase turnover ofitory feedback loop which ultimately connects to theserotonin within the hypothalamus.[33] Additionally,hypothalamus, damping down hypophagic signal-it is linked to a significant reduction in the ratio ofling.[27] This is likely to occur via the 5-HT1A andneurons in the VMH that respond to serotonin,[34] an5-HT2C receptor mechanisms.[28,29] Support for thiseffect that may be due to alterations in the expres-argument is provided by 5-HT1A receptor immu-sion of serotonin receptor subtypes. This reductionnoreactivity results showing that orexin neurons inin responsiveness to serotonin seen in these rodentthe lateral hypothalamus contain 5-HT1A recep-models may provide a partial explanation for thetors.[30] Additionally, serotonin has been shown toincreased motivation for feeding after a period ofhyperpolarise orexin neurons via the 5-HT1A recep-fasting. For instance, fasting-induced reductions intor.[31] It is worth noting that agonism of theserotonin would allow increased NPY activity and apresynaptic 5-HT1A autoreceptor could potentiallypotent feeding response.stimulate rather than inhibit food intake. It is likely

With regard to human studies, when female vol-that this inhibitory serotonergic contribution is im-unteers were placed on a 4-week hypocaloric diet,portant for the physiological regulation of the orexinplasma levels of tryptophan, the serotonin precursor,system.were reduced.[35] These reductions in tryptophancorrelated with the degree of weight loss observed in3. Endogenous Serotonin and Alteredthe women over the 28-day study period. SeveralNutritional Statusstudies have noted the effect of dieting on trypto-

If a system is critical to the control of energy phan, which suggests that dieting in humans maybalance, it should fluctuate in response to nutritional result in central serotonin function being dimin-status. There are a variety of models that can be used ished. In one study, prolactin response to the 5-


34 Halford et al.

HT2C receptor preferential agonist 3.3 Increased Energy Intake or Bodyweight

chlorophenylpiperazine (mCPP) was increased fol-It is also logical to assume that a system thatlowing dieting.[36] The data provided by these stud-

inhibits food intake would be activated in situationsies suggest that the sensitivity of the 5-HT2C recep-

of energy excess, allowing the organism to addresstor is increased in response to chronic restrictions of

other needs by inhibiting feeding behaviour. Thus,food intake, an effect that results from the decline in

intake should promote increases in serotonin whichthe actual transmitter. In summary, it has been normally prevent weight gain. Numerous studiesshown that inadequate nutrition or hypocaloric diets have demonstrated the effects of dietary carbohy-seem to both reduce endogenous serotonin levels drate on serotonin precursor entry into the CNS.and increase serotonin receptor sensitivity. Moreover, central serotonergic systems appear to

mediate the effects of episodic satiety signallinghormones such as cholecystokinin (CCK).[1] Fur-3.2 Cachexic and Endogenous Serotoninthermore, it is possible to hypothesise that animalsunable to control their bodyweight show deficien-Not only do enforced reductions in food availa-cies in their endogenous serotonin system. Studies

bility cause changes in bodyweight; weight loss alsousing obese Zucker rats have been instrumental in

results from malaise-induced reductions of food in-providing insights into the effects of obesity and

take. An obvious question is whether endogenous chronic hypophagia on the regulation of serotonin.serotonin activation contributes to the suppression In this leptin-resistant obesity model, numerousof appetite produced by illness. Rat cancer models studies have shown that abnormal hypothalamic se-are useful for providing chronic anorexic conditions. rotonin activity contributes to both hyperphagia andParticularly in the VMH, the hypothalamic seroto- weight gain. These animals demonstrate an unal-nin system plays a crucial role in the anorexia.[37] tered pattern of serotonin release associated withAfter injection of interleukin (IL)-1α into the VMH, food deprivation and re-feeding. Moreover, baseline

levels of serotonin, specifically, are significantlyserotonin levels within the VMH are increased inlower in obese than lean animals.[41] It has beenboth normal and tumour-bearing rats.[38] Corre-shown that such rats have lower hypothalamic levelsspondingly, elevations of tryptophan (the serotoninof 5-HIAA, the serotonin metabolite.[42] These dataprecursor) in tumour-bearing rats positively corre-indicate that obese Zucker rats have reduced rates oflate with the degree of anorexia.[39] In humans thereendogenous serotonin turnover. Furthermore, thisis evidence that plasma tryptophan levels normalisedefect in the endogenous serotonin system’s func-

following tumour resection in patients, resulting intioning appears to permit weight gain.

improved food intake, which strengthens the rela-Dietary-induced obesity following exposure to

tionship between serotonin, tumour and anorexia.[40]highly palatable food is the rodent model that resem-

These studies again strongly support the notion that bles human obesity the most closely. The severity ofserotonin is a key anorectic agent. Interestingly, obesity varies between individual animals in thisanother system heavily implicated in cachexia is the model; approximately 50% of a population developmelanocortins. Therefore, it is possible to hypothe- obesity when provided with a palatable diet.[43] Suchsise that the effects of serotonin on food intake in rats, demonstrably obesity-prone, exhibit abnormal-these models are mediated by the hypothalamic me- ities of diurnal and fasting-induced alterations inlanocortin system. brain serotonin turnover. Specifically, the obesity-



prone rats, in contrast to obesity-resistant animals, lease of serotonin and/or preventing its reuptake).fail to show diurnal variation in serotonin turnover The resulting suppression of food intake was thenin the PVN and ARC. Also, these rats demonstrate challenged pharmacologically, using selective sero-significant reductions in VMH nucleus serotonin tonin receptor antagonists to identify the specificturnover in comparison to their obesity-resistant serotonin receptor subtypes fundamental to drug-equivalents.[44] It is possible that these abnormalities induced hypophagia. Fenfluramine or dexfen-constitute a component of a genetic programming, fluramine-induced food suppression models werepredisposing prone rats to become obese when ex- used in a majority of the studies to have employedposed to palatable food. Moreover, these abnormali- antagonists to identify these receptors. This is likelyties have been shown to normalise after obesity has to be at least partially as a result of the robustness ofdeveloped. This could account for the persistence of the models of fenfluramine or dexfenfluramine-in-the obesity in these rats, whereby the animals avidly duced reduction in food intake and bodyweight, butdefend their increased bodyweight against the food also partly a reflection of the known effectiveness, atrestriction. Therefore, in rodents, reduced serotonin that time, of their treatment for human obesity. Theturnover or an unresponsive endogenous serotonin use of selective serotonin receptor antagonists ofsystem is associated with proneness to obesity. various 5-HT1 and 5-HT2 receptor subtypes had

Given the rodent data, it seems logical to ask if an indicated that dexfenfluramine-induced hypophagiaimpaired serotonin system contributes to human is mediated by 5-HT1B receptors.[46] The foundationobesity. In obese humans, plasma tryptophan levels for this was evidence gathered from numerous stud-are also reported to be low. This could contribute to ies showing that the actions of dexfenfluraminea reduction in satiety response to food and conse- were blocked with 5-HT1A/1B receptor antagonists,quent overconsumption. Moreover, low tryptophan but that the actions of dexfenfluramine were notlevels are not normalised by considerable weight blocked by 5-HT2A/2C receptor antagonists.[47-49] Inreductions.[45] Therefore, in humans, a reduction in 2002, Simansky and Nicklous[50] infused the highlyCNS serotonin levels in the obese could contribute selective 5-HT1B receptor antagonist directly intoto an inability to exert sufficient control over their the parabrachial nucleus of the rat and blockedown daily caloric intake. Although it is difficult to dexfenfluramine-induced reduction in food intake.prove any role for serotonin deficiency in the devel- Nevertheless, the central role of 5-HT1B receptors isopment of obesity, it seems appropriate to assume it not confirmed by all experimental evidence. Vickersmay be critical in preventing the obese individual et al.[51] showed that dexfenfluramine-induced hy-successfully losing weight. pophagia could be blocked by pretreatment with the

highly selective 5-HT2C receptor antagonist SB-4. Effects of Serotonergic Drugs on Food242084, whereas this effect was not seen followingIntake in Rodentspretreatment with the highly selective 5-HT1B re-ceptor antagonists GR-127935 and SB-224289.

4.1 Serotonin Receptors and Food IntakeFluoxetine, a selective serotonin reuptake inhibi-

tor (SSRI) produces a consistent food intake reduc-Following the initial identification of serotonin astion that cannot be blocked easily by serotonin an-a key inhibitor system, the next step was to revealtagonists.[52-54] Nevertheless, the non-selective 5-the receptors responsible for mediating the serotoninHT1/2 receptor antagonist metergoline has been suc-satiety response. In the beginning, studies used sero-

tonin manipulations (i.e. drugs promoting the re- cessful in both partially and fully blocking fluoxe-


36 Halford et al.

tine-induced hypophagia.[55,56] These data suggest intake but distinct effects on feeding behaviour.[67]

that at the doses used, fluoxetine-induced hypopha- More specifically, amfetamine seemed to fragmentgia may be mediated by the 5-HT1B and 5-HT2C normal feeding behaviour (the duration of feeding isreceptor subtypes that are critical to the mediation of decreased but its frequency is increased and thethe effects of dexfenfluramine on food intake. The decline in eating behaviour indicative of satiety dis-hypophagia induced by other SSRIs such as ser- appears), whereas fenfluramine appeared to enhancetraline,[57] in contrast to fluoxetine, seems to be satiety by reducing meal size. Subsequent studiesblocked more reliably by antagonists of the 5-HT1B have shown drugs that either stimulate serotoninand 5-HT2C receptors.[58] release or inhibit serotonin reuptake produce

Alongside the development of more selective changes in feeding behaviour, as measured by theserotonin receptor antagonists, further serotonin re- behavioural satiety sequence (BSS) or otherceptor agonists have also become available. More behavioural assays, which are consistent with therecently, scientists have been able to work with operation of satiety. Specifically, numerous seroton-selective serotonin receptor agonists that have been ergic drugs have been shown to enhance the BSS.[68]

developed specifically as novel obesity treatments. The BSS is a stochastic progression of behaviourDirect agonism of serotonin receptors in rodents whereby as satiety develops, the initial feeding be-produces consistent reductions in food intake. Fur- haviour is replaced with activity, followed bythermore, agonists of the 5-HT1B and 5-HT2C recep- grooming, and then terminated with a prolongedtor subtypes generate changes in feeding behaviour period of resting/inactivity. If the effects of a drugthat are consistent with the operation of satiety (see cause a reduction in food intake but also cause asection 4.2). These agonist agents include mCPP disruption or delay in this temporal behavioural(preferential 5-HT1B/2C receptor agonist), TFMPP pattern, it is probable that drug-induced hypophagia(N-[3-trifloromethyl)phenyl]piperazine) [preferen- is at least partially due to disturbance of this normaltial 5-HT1B/2C receptor agonist], CP-93129 (selec- feeding behaviour by mechanisms other than satiety,tive 5-HT1B receptor agonist), CP-94 253 (selective for example, hyperactivity, malaise, sedation or nau-5-HT1B receptor agonist), and Ro 60-0175 (selec- sea. However, if a drug is shown to induce hypopha-tive 5-HT2C receptor agonist).[29,59-66] While these gia but does not disrupt this sequence of behaviour,food intake studies seem to have confirmed the role

it is reasonable to assume that this drug does notof both receptor subtypes in mediating the serotonin

interfere with the natural development of satiety.hypophagic response, the question remains; was

Alternatively, if a drug causes a reduction in fooddrug-induced hypophagia due to enhanced satiety?

intake and appears to enhance the BSS (i.e. feedingbehaviour is terminated early and there is an earlier

4.2 Serotonin and Feeding Behaviouronset of each of the other phases of the sequence), itcan be concluded that satiety is the primary mecha-It is important to clarify a key distinction betweennism of action.[68]

drugs that reduce food intake by exerting an effectPrevious studies have shown that both fen-on natural satiety mechanisms and those that cause a

fluramine and dexfenfluramine adjust feeding beha-reduction in food intake by inducing nausea, seda-viour in a manner consistent with the operation oftion, hyperactivity or malaise. For example, whensatiety.[69-71] Sertraline, fluoxetine and other SSRIsfenfluramine and amfetamine (amphetamine) werehave produced similar effects to dexfenfluramine oncompared in early clinical studies it became clear

that the two drugs produced similar effects on food a number of behavioural indices of appetite in ro-



Table I. Effect of various serotonergic drugs on rodent energy intake and expression of the behavioural satiety sequence (BSS)

Drug Mechanism (and receptors mediating effect if known) Effect

on intake on BSS

Dexfenfluramine Serotonin releaser and reuptake inhibitor (5-HT1B ↓ ↑and 5-HT2C)

Fluoxetine Serotonin reuptake inhibitor (5-HT1 and 5-HT2) ↓ ↑Sertraline Serotonin reuptake inhibitor (5-HT1B and 5-HT2C) ↓ ↑Paroxetine Serotonin reuptake inhibitor (receptor involvement unknown) ↓ ↑Femoxitine Serotonin reuptake inhibitor (receptor involvement unknown) ↓ ↑DOI 5-HT2 agonist ↓ Disrupts (hyperactivity)

MK-212 5-HT2 agonist ↓ Disrupts (sedation)

mCPP 5-HT1B/2C agonist ↓ ↑TFMPP 5-HT1B/2C agonist ↓ ↑RU-24969 5-HT1A/1B agonist ↓ Disrupts (hyperactivity)

CP-93,129 Selective 5-HT1B agonist ↓ ↑CP-94,253 Selective 5-HT1B agonist ↓ ↑Ro 60-0175 Selective 5-HT2C agonist ↓ ↑Org 12962 Selective 5-HT2C agonist Unknown Unknown

VER-3323 Selective 5-HT2C agonist Unknown Unknown

BTV-933 Selective 5-HT2C agonist Unknown Unknown

YM348 Selective 5-HT2C agonist Unknown Unknown

Lorcaserin (APD356) Selective 5-HT2C agonist Unknown Unknown

Sibutramine Serotonin and noradrenaline reuptake inhibitor ↓ ↑(NAα1/β2 and 5-HT2A/2C)

DOI = 2,5-dimethoxy-4-iodoamphetamine; mCPP = m-chlorophenylpiperazine; TFMPP = N-[3-trifloromethylphenyl]piperazine; ↑ indicatesincreases; ↓ indicated decreases.

dents, i.e. inducing the BSS.[56,68,72-75] Additionally, 4.3 Serotonin Receptor Knockouts, FoodIntake and Obesitythe current globally licensed anti-obesity treatment

sibutramine (serotonin and noradrenaline reuptake‘Knockout’ mice provide the final and possibly

inhibitor [SNRI]) has also been shown to enhancestrongest argument for the vital role of 5-HT2C

the BSS.[68] Several selective agonists of serotonin receptors in mediating the hypophagic effects ofreceptors produce changes in feeding behaviour endogenous serotonin. A breed of mice was success-consistent with the operation of satiety, notably the fully produced by Tecott et al.[76] that did not pos-5-HT1B receptor agonist CP-94 253, the preferential sess a single functional 5-HT2C receptor. In these 5-5-HT1B/2C receptor agonists mCPP and TFMPP, HT2C knockout mice, marked hyperphagia wasand the selective 5-HT2C receptor agonists.[62,66,68,73]

demonstrated from 5 weeks after birth, an effectImportantly, drugs that have also been shown to accompanied by marked hyperactivity. Hyperactivi-agonise other serotonin receptor subtypes, for exam- ty declined later in the life of these mice, whereasple DOI (2,5-dimethoxy-4-iodoamphetamine) [5- hyperphagia persisted, leading to the developmentHT2A/2C] or RU-24 969 (5-HT1A/1B), are unsuitable of obesity.[77] Interestingly, these mice were alsofor further development because they disrupt the partly resistant to dexfenfluramine-inducedBSS by inducing hyperactivity, a critical side effect. hypophagia.[78] However, reductions in food intakeFor a summary of the effects of serotonergic drugs could still be produced in the mutant mice by the useon rodent food intake and the BSS see table I. of dexfenfluramine. There was also some evidence


38 Halford et al.

of enhanced satiety in the BSS paradigm in response to follow for the first 6 weeks of the study, butto this drug. These responses were distinctly lesser during the second 6-week period the participantsin magnitude in the knockout mice than in the con- were given a diet regimen. Each participant record-trol wild-type mice. Therefore, it seems that there is ed their food intake at regular intervals in a diary,a deficiency in the endogenous serotonin satiety both before treatment commenced and during bothsystem of obese 5-HT2C receptor knockout mice. the initial (non-diet) and latter (diet) phases. FromAdditionally, 5-HT1B receptor knockout mice are baseline, those receiving 5-HTP lost 5.0kg insignificantly heavier than wild types,[79] an effect bodyweight (1.7kg in the non-diet phase and 3.3kgassociated with significantly greater food consump- in the diet phase) in comparison to 1.2kg, a non-tion. However, these animals merely appear to be significant reduction in bodyweight, seen in thoselarger rather than fully obese.[2] who had received placebo for 12 weeks. Those

receiving the drug reported significant decreases in5. Serotonergic Drugs, Food Intake and daily energy intake of 41% (5636kJ) in the non-dietFeeding Behaviour in Humans phase and 60% (8202kJ) in the diet phase from

baseline. Conversely, reductions in self-reportedStudies first employing racemic fenfluraminefood intake in the placebo group were 14% andand then dexfenfluramine provided a majority of the24%, respectively. The same research group thenoriginal data on the effects of pharmacological ma-replicated their original findings in a shorternipulation of serotonin on food intake in humans.study.[96] In this study, 25 overweight individualsSubsequently, studies employing the SSRI fluoxe-with type 2 diabetes mellitus were randomly allocat-tine and then sibutramine were also published. Overed to receive either placebo or 5-HTP (750 mg/day)the past 13 years, numerous other studies have beenfor 2 weeks. Diet diaries were also utilised in thispublished, using the serotonin precursor 5-HTP andstudy. In the group treated with 5-HTP, weight losspreferential serotonin receptor subtype agonists.from baseline over the 2-week study period wasThese studies have shown that by increasing synap-2.1kg. Significant decreases in self-reported energytic or neuronal levels of serotonin or directly agonis-expenditure accompanied this weight loss effect,ing specific serotonin receptor subtypes, robustwith a 21% (1700kJ) reduction reported on day 7hypophagia can be produced. Furthermore, is it clearand 22% (1760kJ) on day 14. It is likely, from anfrom the meticulous description of the effects of theexamination of these data, that participants in thismanipulation on behaviour and subjective ratings oftrial underreported energy intake on a number ofhunger and satiety that these pharmacological inter-occasions. Nevertheless, it does appear thatventions have produced hypophagia by modifyinghypophagia is associated with weight loss.human appetite expression. The effects of ser-

otonergic drugs on appetite are shown in table II.5.2 Fenfluramine and Dexfenfluramine

5.1 5-HydroxytryptophanIn the late 1970s, the effects of fenfluramine on

The serotonin precursor 5-HTP has been shown human food intake were demonstrated. For exam-to produce potent effects on self-reported food in- ple, in 1979 Rogers and Blundell[82] published a keytake in obese individuals. In one study, 20 obese study showing that a single dose of fenfluramineindividuals (defined as ‘hyperphagics’) were treated (60mg), when given to lean healthy males, couldwith either 5-HTP (900 mg/day) or placebo over 12 reduce food intake at a lunchtime meal by 789kJweeks.[81] The participants were not prescribed a diet (26%). The effect of reduced caloric intake was also



Table II. Summary of studies that have investigated the effects of serotonergic drugs on food intake in humans

Drug Study design No. of study Duration of Hunger and food Caloric intake Referenceparticipants treatment preference(status)

Tryptophan Laboratory study 16 (healthy) 1d NR ↓Intake 80

5-HTP Laboratory study 20 (obese) 6wk ↓Carbohydrate intake, ↓Intake and bodyweight 81early post-meal satiety of 6wk period

Fenfluramine Laboratory study 12 (normal 1d NR ↓Test meal intake and 82bodyweight) eating rate

Dexfenfluramine Laboratory study 8 (obese) 3d ↓Hunger, ↑fullness after ↓Intake after second 83first meal meal (11–19%)

Dexfenfluramine Laboratory study 24 (obese) 2wk ↓Carbohydrate snack NR 84intake

Dexfenfluramine Laboratory study 20 (obese) 8d ↓Carbohydrate intake at ↓Intake 85meal, ↓snack intake

Dexfenfluramine Laboratory study 13 (healthy men) 1d ↓Hunger ↓Intake 86

Dexfenfluramine Laboratory study 10 (lean women), 1d ↓Hunger ↓Intake 8711 obese women

Dexfenfluramine Laboratory study 12 (healthy men) 1d ↓Fat intake No overall effect 88

Fluoxetine Laboratory study 11 (healthy men) 2wk ↓Hunger on days 8 and ↓Intake on days 1 8915, but not on day 1 and 8, but not on day 15

Fluoxetine Laboratory study 12 (obese women) 14d No macronutrient ↓Intake (22.4%) 90preference, ↓hunger

Sertraline Relapse 53 (obese women) 6wk ↓Hunger, ↓food NR 91prevention trial preoccupation

mCPP Laboratory study 12 (healthy 1d NR ↓Test meal intake 92women)

mCPP Laboratory study 18 (obese) 2wk ↓Hunger, ↓bodyweight NR 93

Sumatriptan Laboratory study 15 (healthy 1d No change in hunger, ↓Intake 94women) ↓fat intake

Sibutramine Laboratory study 12 (obese women) 14d ↓Hunger ↓Daily intake 955-HTP = 5-hydroxytryptophan; mCPP = m-chlorophenylpiperazine; NR = not reported; ↑ indicates increase; ↓ indicates decrease.

accompanied by significant decreases in eating rate compensatory weakening of satiety states betweenmeals.and desire to eat, both indicative of a drug-induced

modulation of normal appetite. The desire to eat A number of differing laboratory-based feedingbeing significantly lower prior to the meal and the paradigms have successfully shown dexfen-

fluramine-induced hypophagia in humans. Not alllow rate of consumption at the start of the mealcan be detailed here but it is useful to consider somesignify that fenfluramine had retarded normal hun-of these studies, which have indicated the efficacy ofger by enhancing the pre-meal satiety state. Similar-both short- and long-term dexfenfluramine on foodly, Foltin et al.[97] found that both male and female,intake, in both lean and obese individuals. Acute

healthy, normal weight individuals reduced theirdoses of dexfenfluramine were administered to lean

total daily caloric intake when given fenfluraminehealthy males in a study by Goodall and Silver-

40 mg/day. Specifically, the participants reduced stone.[86] The results of this study showed that atheir meal size rather than the number of meals, single 30mg dose of dexfenfluramine produced aindicating that fenfluramine had enhanced within- significant reduction in cumulative intake and eatingmeal satiation processes but had not produced any rate of approximately 23% over a 2-hour period.


40 Halford et al.

Dexfenfluramine, as with fenfluramine, in this ad were employed in a double-blind, placebo-con-libitum feeding paradigm produced a significant trolled, crossover design. Food intake and eatingdecrease in pre-meal hunger ratings. In another behaviour were assessed on three distinct probe daysstudy, Blundell and Hill[83] also noted that dexfen- (days 1, 8 and 15). Over the 2-week treatment peri-fluramine produced significant reductions in hunger od, participants in the drug treatment group lostprior to a meal in both the lean and the obese. significantly more weight than those in the controlAdditionally, the effect on hunger was greater in group (1.07 vs 0.15kg). Fluoxetine treatment re-magnitude in the obese than the lean. In this study, duced cumulative intake at the 2-hour buffet mealthe drug also reduced prospective consumption in on day 1 (by 15.7%) and day 8 (by 12.6%) but notthe obese and the lean, and solely in the lean en- day 15. Again in the fluoxetine group, a significanthanced feelings of fullness were reported. reduction in hunger was reported on day 8 only.

Subsequently, Drent et al.[98] administered From this study, it is not possible to ascertain wheth-dexfenfluramine (30 mg/day) to overweight and er day 15 was the last day of fluoxetine treatment orobese individuals for 9 weeks, in a randomised, the first day of the washout period.placebo-controlled, double-blind trial. Throughout

Another study examined the effects of 14 days’the trial, dexfenfluramine produced a marked reduc-

administration of fluoxetine (60 mg/day) on mealtion in self-reported food intake. The placebo group

patterns.[99] The group treated with fluoxetine lostactually gained 0.2kg over the study period despite

3.6kg in comparison to a weight gain of 0.3kg seenself-reporting a 15% reduction in daily food intake

in the placebo control group. When compared withfrom baseline. Conversely, the group treated with

the placebo control, fluoxetine treatment also signif-dexfenfluramine lost 3.1kg and reduced self-report-

icantly reduced reported meal and snack intake. Aed daily food intake by a significant 30% from

more detailed study of the effects of fluoxetine (60baseline. It is possible that, as the placebo group

mg/day) on the eating behaviour and food intake offailed to lose weight, there may have been an under-

obese females was carried out by Lawton et al.[90]reporting of energy consumed by participants in this

Volunteers in this study, as with the McGuirk andtrial. Despite this, the effects shown on bodyweight

Silverstone study,[89] were treated for 14 days withand food intake are still impressive. Analysis of theboth drug and placebo in a randomised, double-food diaries indicated that dexfenfluramine treat-blind, crossover design. Patients returned to the lab-ment reduced energy intake by reducing meal andoratory on days 7 and 14 to receive fixed test meals.snack size but not number. This study also demon-This was designed to investigate whether fluoxe-strated that the dexfenfluramine reductions in foodtine-induced hypophagia was intensified or weak-intake were associated with the effect of the drug onened by equi-caloric test meals that differed in mac-bodyweight. Moreover, even when the drug is givenronutrient composition. The results of this studylong term to obese individuals, dexfenfluramine-were similar to those of previous studies, as over theinduced reductions in food intake appear to be a2-week treatment period significantly more weightrobust phenomenon.was lost in the drug group than the controls (1.97 vs0.04kg). There was also a significant associated5.3 Fluoxetinereduction in post-test meal hunger in the drug group.Irrespective of their macronutrient composition, itA study by McGuirk and Silverstone[89] investi-was shown that fluoxetine increased the satiety im-gated the effect of 2 weeks’ fluoxetine (60 mg/day)pact of the fixed equi-caloric test meals. Directtreatment on food intake. Healthy male participants



measurement of subsequent food intake at the ad used to investigate the effect of a single 15mg doselibitum evening meal that followed the test meal on of sibutramine on total daily energy consumption.days 7 and 14 showed that fluoxetine produced a The drug was taken prior to a fixed load breakfast. It27% (198kcal) reduction in energy consumption. was shown to reduce total caloric intake on the studyDuring this study participants were also asked to by 1304 kcal, approximately 12%. This figure iscomplete daily food diaries. Analysis of these dia- achieved through significant reductions in caloricries showed that during the 14 days of treatment, intake at both lunch and dinner (637 and 393kJ,fluoxetine produced a self-reported reduction in dai- respectively). The number of food items eaten perly energy intake of 22.4% (421 kcal/day). day was reduced by an average of 1.6 items (approx-

It appears that a tolerance to fluoxetine-induced imately 10%). This effect occurred mainly at lunch.hypophagia develops after 2 weeks of administra- In addition to the pronounced hypophagia andtion, as suggested by the data from McGuirk and changes in feeding behaviour seen at lunch, theSilverstone.[89] However, Ward et al.,[100] carried out 15mg dose of sibutramine significantly reduceda 16-week outpatient study in the obese, in which hunger 4 hours after administration, an effect thatparticipants attended the laboratory at weeks 7 and coincided with the start of the lunch.16 of treatment. On these days, long-term fluoxetine

Rolls et al.[95] examined the effects of sibu-administration was shown to reduce participants’

tramine on food intake and appetite in obese individ-total food intake. Specifically, the mean number of

uals. In this study, the double-blind, placebo-con-total eating occasions within a study day was re-

trolled, crossover design used was similar to that ofduced by fluoxetine. Unlike drug-induced hypopha-

previous fluoxetine studies.[89,90] It was conductedgia, tolerance did appear to develop to drug-induced

over 14 days, with participants invited into the labo-weight loss in this study. At the week 7 timepoint,

ratory to have their eating behaviour assessed direct-fluoxetine-induced weight loss was significantly

ly on days 7 and 14. Two daily dosages of the druggreater than with placebo, but this was not the case

were used, 10mg as is normally prescribed, and aat the week 16 study endpoint.

higher 30mg dose. Participants were instructed totake the drug before breakfast. On study days, par-

5.4 Sibutramineticipants attended the laboratory for breakfast, lunchand dinner, all of which were ad libitum meals. TheIn addition to fluoxetine and dexfenfluramine,30mg dose of sibutramine showed effects on foodacute doses of sibutramine, an SNRI, have beenintake, subjective measures of appetite andshown to reduce food intake[101] and appetite in leanbodyweight that were observed earlier in the studymale volunteers.[101,102] Hansen et al.,[102] initiallyand were greater in magnitude than the effects of thestudied the effect of sibutramine (30 mg/day) onlower dose. Specifically, the 30mg dose reducedenergy expenditure but also noted that the drugcaloric intake by 1763kJ (23%) by day 7 and byproduced an enhancement of the inhibition of appe-2079kJ (26%) on day 14 (vs placebo). The 10mgtite resulting from the set breakfast (2.1kJ) given todose also significantly reduced total caloric intakeall participants. Although this study did not measure

the effects on subsequent ad libitum food intake, by 1290kJ (19%), but only on day 14. The signifi-sibutramine treatment was shown to increase the cant reductions shown in total caloric intake (onsatiety impact of a fixed load of food. days 7 and 14 for the 30mg dose, on day 14 only for

10mg) resulted from significant reductions in ener-In a subsequent study by Chapelot et al.,[101] agy intake at both lunch and dinner, but not at break-placebo-controlled counterbalanced design was


42 Halford et al.

fast. It is not clear whether sibutramine-induced However, it was noted that both daily hunger ratingshypophagia was greater at lunch or at dinner, but the and prospective food consumption were significant-effect does appear to have been accompanied by ly decreased on the first (day 1) and the last (day 56)reductions in pre-meal hunger and prospective con- days of sibutramine treatment. The results of thissumption ratings at the 30mg dose. Interestingly, study also showed that sibutramine produced a sig-both the Chapelot and Rolls studies[2,9,101] demon- nificant decrease in bodyweight of 2.4kg comparedstrate that sibutramine-induced significant effects on with a slight rise of 0.3kg seen in the placebo groupappetite were observed before the ad libitum meal, over the course of the study. Sibutramine was shownbut not after. Following the ad libitum meal, drug- to have little effect on energy expenditure; therefore,induced reductions appear to ‘normalise’ the subse- the weight loss observed is likely to have resultedquent post-meal appetite ratings, whereas after a from the effect the drug had on food intake.fixed load these ratings are suppressed. Importantly,

Barkeling et al.[104] recently provided perhaps thein both studies the use of sibutramine treatment in

most persuasive argument to demonstrate the linkthe lean and the obese resulted in the same post-meal

between sibutramine-induced hypophagia and sibu-satiety being attained by significantly less food con-

tramine-induced weight loss. This multiphase studysumption.

was devised to examine the effects of sibutramineTolerance to the hypophagic effects of sibu- on appetite in the obese, particularly to see whether

tramine did not appear in the 14 days’ treatment of sibutramine still reduced food intake after 10the Rolls et al.[95] study. Moreover, it was observed months of treatment and how predicted weight lossthat changes in appetite and food intake were ac- progressed on a long-term trial. Obese volunteerscompanied by significant drug-induced weight loss were recruited to a 14-day fully randomised, place-in this study. The 10mg dose of sibutramine had bo-controlled, crossover study. Following 14 dayscaused a reduction in bodyweight by 0.7kg at day 7, of treatment with sibutramine 15 mg/day or placebo,and by day 14 weight was reduced by 0.8kg com- the participants were invited into the laboratory topared with placebo. The higher 30mg dose of sibu- consume an ad libitum lunch. During the initialtramine had reduced bodyweight by 0.6kg on day 7 double-blind study, a 16% kcal reduction in energyand by 1.2kg on day 14. The results of this study intake at the test lunch was observed. Subsequently,appear to suggest that in the obese, sibutramine the participants were placed on a 10-month open-treatment produces both hypophagia and associated label treatment with sibutramine. At the end of thisweight loss. period, the participants returned to the laboratory for

a final visit at which they were provided with theHansen et al.[103] conducted an 8-week randomis-same ad libitum meal they had received on previoused, double-blind, placebo-controlled study in theoccasions. Intake at this lunch was reduced by 27%obese to examine the effects of sibutramine 15 mg/when compared with their pre-weight loss trial pla-day on energy expenditure. The study design in-cebo intake. Sibutramine also significantly in-volved two laboratory visits, one at the start ofcreased ratings of fullness and decreased prospec-treatment and one at the end, in which the partici-tive consumption after the fixed breakfast but notpants were required to live in a respiration chamberafter the ad libitum lunch. Interestingly, this studyfor 32 hours. Participants ate freely at set mealsdemonstrated that the appetite response to sibu-during this time and their appetite was assessed.tramine was undiminished after 10 months of treat-Disappointingly, the authors did not report the ef-ment. It is particularly important to note that thefects of sibutramine on food intake at these visits.



initial effect of sibutramine on appetite in the 14-day pared with placebo (0.04kg). Participants were in-trial was predictive of the effect of sibutramine on vited to the laboratory on the penultimate day ofbodyweight during the subsequent 10-month open- treatment and blood samples were taken to assesslabel weight loss trial. mCPP concentrations and prolactin response. Dur-

ing this procedure, hunger rating scales were also5.5 Preferential and Selective Serotonin completed by the participants. Analysis of the scalesReceptor Agonists subsequently showed that drug treatment produced a

significant decrease in hunger ratings.Direct agonism of serotonin receptors also

Collectively, the three studies discussed make itpotently reduces food intake. In the rat, mCPP (a

evident that mCPP can effectively reduce food in-5-HT1B/2C receptor preferential agonist) has been

take and appetite in lean individuals, and body-shown to reduce food intake via activation of

weight and appetite in obese individuals. However,5-HT2C receptors. A number of studies have also

it is important to note that in lean participants,shown that mCPP reliably reduces food intake in

mCPP also produced transient but significant in-humans. The effect of an acute dose of mCPP (0.4

creases in the self-reported subjective ratings ofmg/kg) was initially investigated in a double-blind,light-headedness, anxiety and nausea.[92,105] It hasplacebo-controlled, crossover design study, con-also been observed that transient increases in bloodducted with lean, healthy female volunteers.[92] Eachpressure and heart rate can occur in response toparticipant was given either mCPP or placebo orallyacute doses of mCPP.[106] If more specific 5-HT2C150 minutes prior to the presentation of a buffetreceptor agonists could be developed (particularly iflunch. At this ad libitum meal, food intake in thosethey avoid some of the aforementioned transientreceiving mCPP was reduced by 30% (approximate-adverse effects produced by the less specificly 1000kJ). This study also showed that the effect ofmCPP), this could be an important advancement inmCPP on food intake was significantly associatedobesity treatment. However, it still remains to bewith pre-meal hunger ratings being reduced. Subse-demonstrated whether the hypophagic effects ofquently, this study was replicated in a larger groupmCPP and more selective 5-HT2C receptor agonistsof both male and female lean, healthy volun-occur in obese individuals.teers.[105] The drug was again effective at reducing

The 5-HT2C receptor is, of course, not the onlyfood intake in women (28%, 1205kJ) and also inreceptor that has been implicated in mediating themen (20% reduction, 1219kJ). Again, the drug pro-effects of the endogenous serotonin satiety system.duced significantly reduced hunger ratings prior toSumatriptan, a novel 5-HT1B/1D receptor agonist,the meal in both men and women (150 minutes afterhas also been found to produce a significant reduc-administration). This effect occurred marginally af-tion in food intake in healthy women.[94] This studyter peak plasma mCPP concentrations (120 minutesused a double-blind, placebo-controlled, crossoverafter administration) and just prior to the lunch.design to examine the effects of an acute 6mg doseA study in obese individuals examined the effectsof sumatriptan on food intake at a buffet style lunch.of mCPP on appetite and bodyweight but not foodA 23% reduction in food intake (approximatelyintake.[93] In this double-blind, placebo-controlled,850kJ) was produced by the sumatriptan injectioncrossover trial, participants were treated for 14 dayscompared with placebo. Importantly, a 34% de-with mCPP (20mg twice daily for women, 25mgcrease in fat intake was also observed at the lunchtwice daily for men). Significantly more weight

(0.8kg) was lost from baseline with mCPP com- following drug treatment. In this study, no signifi-


44 Halford et al.

cant effects on ratings of nausea and light headed- gic drugs has also produced these effects onness were observed. However, as there were no bodyweight gain and food intake. For example, SS-significant effects on appetite reported, it is probable RIs such as fluoxetine,[110,111] sertraline,[112] fluvox-that this study was statistically underpowered. amine[113] and paroxetine[114] have also all been

shown to attenuate bodyweight gain in various ro-6. Serotonergic Drugs and Obesity: dent models, an effect normally associated withRodent Studies significant hypophagia.

6.2 Serotonin Receptor Agonists6.1 Dexfenfluramine and Selective SerotoninReuptake Inhibitors

In rodent models it also appears that direct agon-It has been proven in a number of rodent models ism of 5-HT2C receptors affects weight gain. In their

that drugs which promote serotonin release or inhib- first study, for example, Vickers et al.[108,109] haveit reuptake do inhibit bodyweight gain. In one study, shown the inhibitory effects of the preferentialdaily injections of dexfenfluramine (10 mg/kg) over 5-HT2C receptor agonist mCPP on rodenta 12-day period were shown to be particularly effec- bodyweight gain. mCPP 12 mg/kg/day was deliv-tive at abolishing weight gain associated with expo- ered by implanted mini-pumps during the initialsure to high-fat diets in Osborne-Mendel rats (a study, the results of which demonstrated that thestrain particularly susceptible to dietary-induced effects of mCPP on bodyweight were associated, atobesity).[107] This reduction in bodyweight gain was least partly, with hypophagia. Furthermore, toler-associated with a marked reduction in caloric intake. ance to the effects of mCPP on rodent bodyweight

It has also been noted in a separate study that no gain did not develop during this study and animalstolerance developed to the bodyweight gain attenu- treated with mCPP weighed 8% less than controlsating effects of dexfenfluramine (6 mg/kg/day inby the end of the study.fused peripherally via a mini-pump) during a 14-day It is known that mCPP also agonises several otherstudy in Lister hooded rats.[108] At the end of the serotonin receptors (e.g. 5-HT2A and 5-HT2B recep-study, the animals treated with dexfenfluramine tors); therefore, it is theoretically possible that theweighed 5% less than controls. As with the previ- drug-induced attenuation of bodyweight gain couldously mentioned study,[107] these effects on be due to activation of any of these receptors. How-bodyweight were associated with drug-induced re- ever, using selective antagonists, it has been demon-ductions in food intake. However, it was observed strated that mCPP-induced hypophagia is a result ofthat in the second week of this study dexfen- activation of the 5-HT2C receptor specifically.[115]

fluramine-induced hypophagia was less pro- Consequently, it is probable that the hypophagicnounced. component of mCPP-induced weight loss, at least, is

In a longer 28-day study, dexfenfluramine was due to activation of the 5-HT2C receptor. In a secondadministered twice daily (2.5 mg/kg).[109] Again, the study,[109] mCPP (10 mg/kg/day) was given orallydrug reduced food intake and attenuated bodyweight for 28 days. Again, mCPP induced significant atten-gain. Animals treated with dexfenfluramine for 28 uation of bodyweight gain and reductions in dailydays weighed approximately 50g (12%) less than food intake. Animals treated with mCPP over thecontrols, and no tolerance developed to the 28-day period subsequently weighed approximatelyhypophagic effects of dexfenfluramine in this study. 50g (12%) less than controls. No tolerance appearsThe administration of a number of other serotoner- to develop to the hypophagic effects of the 5-HT2C



receptor agonist. Pair-feeding was carried out in mechanism by which serotonin induces weight loss.order to match the food intake of mCPP-treated Correspondingly, serotonergic drugs must restrainanimals, which produced the same degree of weight the motivation to eat and maintain lower levels ofgain, suggesting that the drug’s hypophagic effects food consumption for substantial periods of time. Aare responsible for the attenuation of bodyweight majority of the early clinical data on weight lossgain. originates from studies that employed fenfluramine

and dexfenfluramine, drugs that have since beenThe effects of several more selective 5-HT2C

withdrawn from the market. Clinical data on thereceptor agonists on rodent bodyweight have recent-effects of fenfluramine on bodyweight have beenly been studied. In the study by Vickers et al.,[108]

collected and published since the late 1960s, butRo 60-0175 (26 mg/kg/day) was infused into thegiven that this drug is now withdrawn, these studiesanimals via implanted mini-pumps for 14 days. Theneed not be detailed. However, Haddock et al.[118]results demonstrate that Ro 60-0175 produced a sig-have produced an extremely useful meta-analysis ofnificant reduction in bodyweight gain; the animalsboth early and more recent drug trials.treated with Ro 60-0175 weighed 10% less than

controls at the end of the study. As with mCPP,7.1 FenfluramineRo 60-0175 reduced food intake over the treatment

period; however, by day 11 tolerance to theThere are numerous early therapeutic trials ofhypophagic effect was evident. Similarly, YM348,

fenfluramine, and for those interested in this area,another potent and highly selective 5-HT2C receptorwe recommend the extensive review published inagonist, also produced an attenuation of bodyweight1975 by Pinder and colleagues.[119] Closer examina-gain over a 2-week treatment period (at dosages of 3tion of the review demonstrates that the majority ofand 20 mg/kg/day).[116] Animals treated with thethe early fenfluramine studies were a comparison ofhigher dose of YM348 weighed 21.5% less thanthe effects of fenfluramine with either placebo orcontrols at the end of the study. Tolerance to drug-other anorectic drugs available at the time (for ex-induced hypophagia appeared in the second week ofample, phentermine, dexamfetamine, mazindol andthis treatment also.amfepramone [diethylpropion]). Moreover, theMultiple doses of the novel selective 5-HT2Cstudies involved treatment over periods of 12 weeksreceptor agonist lorcaserin (APD356 [4.5, 9, 18 andor less and were small scale (participant numbers in36 mg/kg]) have recently been shown to inhibit theeach condition were frequently <30). The level ofdevelopment of dietary-induced obesity.[117] Lor-fenfluramine-induced weight loss reported in thesecaserin significantly reduced bodyweight gain instudies depended on the dose of fenfluramine given,both male and female rats. This effect was associat-the duration of the trial, the additional dietary ad-ed with an initial episode of marked hypophagia.vice/regimen given to the participants and the differ-Tolerance appeared to develop to the hypophagicences between the target population included in eacheffects of all dosages of lorcarserin during thisof these trials, and inevitably varied between 1.2 andstudy.11.9kg. Despite this variation, fenfluramine-inducedweight loss was shown to be a robust clinical effect.7. Serotonergic Drugs and Weight Loss:

Of the 14 trials of fenfluramine to meet inclusionClinical Datacriteria from Haddock and colleagues’[118] meta-

Examination of the clinical data available ap- analysis, an average of 5.06kg weight loss (placebopears to suggest that hypophagia is the primary subtracted = 2.41kg) was produced by the drug. The


46 Halford et al.

data were obtained from studies which varied in take reduction of 6–10% over the year-long studydosage (from 39 to 120 mg/day), duration (the long- period.[121]

est trial was just 18 weeks, average trial length was Dexfenfluramine has also been shown to inhibitonly 9.7 weeks), dietary advice, participant numbers human food intake and weight regain after treatment(the maximum number of participants in the fen- with a very low calorie diet (VLCD). A VLCD wasfluramine groups of the included trials was 58, with used in a study by Finer et al.[122] to reduce weight inan average of just 20) and other characteristics. the obese by 14kg in 8 weeks. After adherence to a

diet and such rapid weight loss, there is often a7.2 Dexfenfluramine strong disposition to resume overeating and regain

the weight lost. In this study, following terminationThe European INDEX (INternational DEXfen- of the 8-week VLCD, patients given dexfen-

fluramine) trial was a key study of the clinical fluramine 15mg twice daily for 26 weeks continuedefficacy of dexfenfluramine.[120] INDEX was a mul- to lose weight. Compared with the placebo group,ticentre, randomised, double-blind, placebo-con- who regained an average of 2.9kg, the dexfen-trolled trial carried out with 822 obese volunteers. fluramine group lost an additional 5.8kg, bringingOf these, 404 individuals received dexfenfluramine total weight loss to 21.3kg over the entire 34 weeks15mg twice daily and the remaining participants of the study. Therefore, in this study,[122] dexfen-received placebo. After 12 months, of the partici- fluramine overcame the physiological and psycho-pants to complete the study, 52% of those receiving logical drive to eat following rapid, substantialdexfenfluramine had lost 10% or more of their ini- weight loss.tial bodyweight compared with only 30% of those Returning to Haddock and colleagues’[118] meta-receiving placebo. Average weight loss from base- analysis, of the 14 trials to meet inclusion criteria,line with dexfenfluramine during this trial was average dexfenfluramine-induced weight loss from9.82kg (10.26%), a figure significantly greater than baseline was 8.9kg. For any anti-obesity drug in-that achieved with placebo (7.15kg; 7.18%). cluded within their analysis, this is the largest aver-

Interestingly, withdrawal from dexfenfluramine age effect observed. In part, this could be attributa-following completion of the 12-month INDEX trial ble to the longer trial length (an average of 33led to an immediate rise in daily energy consump- weeks) in a majority of the dexfenfluramine studiestion which was accompanied by rapid weight gain analysed and also the successful lifestyle interven-over a 2-month period.[121] This appears to suggest tions. Despite these factors, the average placebo-that despite weight loss stopping after 6 months, subtracted weight loss produced by dexfenfluraminedexfenfluramine had maintained a strong influence was equal to or greater than that with any other drug,over food intake for the entire 12-month period. The including current obesity treatments such as orlistatdrug had reduced both self-reported hunger and and sibutramine. These meta-analyses did not in-bodyweight to a point of physiological resistance at clude dexfenfluramine, but Haddock et al.[118] didwhich equilibrium had been reached between hun- find exfenfluramine-induced weight loss to be inger urges and drug anorectic activity. After the drug line with that from sibutramine and orlistat.withdrawal at 12 months, the rebound in hungerdemonstrated a lack of tolerance to the hypophagic 7.3 Sibutramineeffects of dexfenfluramine. Despite the reduction infood intake being predominantly in the first 6 The STORM (Sibutramine Trial of Obesity Re-months, dexfenfluramine maintained an energy induction and Maintenance) trial demonstrated the



clinical efficacy of the SNRI sibutramine. Obese analysis of 11 fluoxetine trials demonstrates that anpatients were prescribed sibutramine 10 mg/day in average bodyweight reduction from baseline ofaddition to a low calorie diet over 6 months and lost 4.1kg (3.3kg placebo subtracted) was achieved by11.3kg in this randomised, double-blind trial.[123] this drug.Following this open-label run-in period, the diet However, following the trajectory of fluoxetine-phase ended and patients were entered into an

induced weight loss after a 6-month period shows18-month randomised, double-blind, placebo-con-

that the drug’s effects are not sustained.[131] Early introlled study. Within this study, participants were

the trial, participants treated with fluoxetine dis-randomly allocated to either placebo or sibutramine

played significant weight loss (at week 24, maxi-(10 mg/day). Of the group to be maintained on sibu-mum weight loss of approximately 4.9kg) and eventramine for 18 months, there was little weight regainby the end of the 60-week trial weight loss was stillobserved (bodyweight gain at study end was 9.3kgsignificantly greater in the fluoxetine group than inlower than at pre-run-in baseline 24 months previ-the placebo group. However, at week 60 those re-ously). In contrast, those in the placebo group ap-ceiving placebo had lost 1.5kg and those receivingpeared to begin regaining weight within 2 months offluoxetine just 2.2kg. As had been noted in previous,entering the trial phase. Although weight loss wassmaller-scale, 1-year trials and in analysis of subsetsnot observed after the first 6 months, there was littleof these data, both groups were seen to begin regain-indication of any tolerance developing to sibu-ing weight halfway through the study.[132,133] Theretramine over the 24 months of treatment. A number

of other 1- and 2-year studies have demonstrated the is little evidence to date that any other SSRI wouldweight loss-inducing efficacy of sibutramine.[124-127] be any more efficacious. For example, sertraline, inIrrespective of variation in specific protocol or pa- contrast to dexfenfluramine, appears to be ineffec-tient populations (for example, diabetic or non-dia- tive at preventing weight regain after a brief periodbetic, hypertensive, etc.), when examining data from of VLCD.[91] However, early reports on the effectsthese studies conducted over a year or more, it is of another serotonin reuptake inhibitor, zimeldineagain evident that the dynamic phase of sibu- have been interesting. One study has demonstratedtramine-induced weight loss occurs within the first 6 that 8 weeks’ treatment with zimelidine (100mgmonths of treatment. Following this period, sibu- twice daily) can produce significant placebo-sub-tramine stabilises bodyweight at a level significantly tracted weight loss, an effect associated with a sig-lower than at the pretreatment point (i.e. baseline). nificant reduction in appetite ratings.[134] It should beSibutramine has produced a placebo-subtracted

noted that while most psychiatric medications pro-weight loss of 4.45 or 4.3kg, as shown by two meta-

duce pronounced weight gain (including lithium,analyses of clinical data.[128,129]

tricyclic antidepressants and antipsychotic medica-tion) and, moreover, that many of the untreated

7.4 Selective Serotonin Reuptake Inhibitorspsychiatric population are also liable to gain weight,the evidence for SSRI-induced weight gain in theIn terms of anti-obesity properties, the most com-SSRI-treated psychiatric population is inconsistent.prehensively studied SSRI is fluoxetine. EarlyThis appears consistent with the reported weightclinical studies, of 6–8 weeks in duration, suggestedloss-inducing effects of SSRIs in non-psychiatricthat treatment with fluoxetine was able to producepopulations. It is necessary to fully re-examine theweight loss of 0.5kg (on average) per week.[130]

weight loss-inducing effects of currently known SS-Accordingly, Haddock and colleagues’[118] meta-


48 Halford et al.

RIs and their active metabolites, as this may produce 8. Current and Future Serotonergicsuitable candidates for a new anti-obesity drug. Anti-Obesity Drugs

Dexfenfluramine was voluntarily withdrawn in7.5 Serotonin Precursors and1997 because of the risk of primary pulmonaryReceptor Agonistshypertension, a move which dealt a blow to thedevelopment of serotonergic anti-obesity com-Currently, no large-scale clinical trial data existpounds.[137] The heart valve abnormalities discov-on the effects of serotonin precursors or receptorered in a few patients in the postmarketing periodagonists on weight loss in the obese. The previouslyled to the very rapid withdrawal of this drug.[138,139]mentioned studies by Cangiano et al.[81,96] haveSubsequently, the SNRI sibutramine was approvedshown that 5-HTP can induce weight loss for up tofor the treatment of obesity. It has been suggested12 weeks in the obese (6% reduction in initial bodythat sibutramine reduces weight by inducing bothmass). In addition, the preferential 5-HT2C receptorsatiety and thermogenesis. It has been difficult toagonist mCPP has been shown to induce weight lossdemonstrate the latter effect in humans. Studiesover a 2-week period in the obese.[93] Several selec-quoted in this review suggest that the primary actiontive 5-HT2C receptor agonists have reached or areof sibutramine is on satiety in humans; therefore, itabout to reach phase II clinical trials, the data fromis likely that despite what the limited preclinical datawhich have not been made widely available, al-suggest, this effect is mediated by serotonin activa-though some have been presented. For example,tion. Without doubt, the effects of sibutramine ondetails have been provided on the efficacy ofrodent feeding behaviour and human appetite areOrg 12962, a 5-HT2C receptor agonist. The effectsindistinguishable from those of dexfenfluramine,of Org 12962 (10mg twice daily) were studied in 40fluoxetine and the preferential and selective 5-HT2Cobese participants over a 12-week study.[135] Partici-agonists. There have been adverse effect issues as-pants treated with Org 12962 lost 13.7kg (14%) ofsociated with sibutramine itself, which are currentlytheir initial body mass. However, this studyunder investigation.presented a strikingly large placebo effect, as those

in the placebo group lost the same proportion of As a consequence, the focus from pharmaceuticalweight during the treatment period. There was no companies has been on developing 5-HT2C receptordetailing of the effects of drug treatment on appetite. agonists. This is partially because of the increasingInterestingly, compliance to Org 12962 appeared to evidence that this receptor subtype was critical toincrease towards the end of the treatment phase the dexfenfluramine mechanism of action but alsocompared with placebo. Org 12962 treatment because these receptors are not thought to be widelyseemed to help the participants adhere to the rigor- distributed outside the CNS so any issue of primaryous but effective weight loss measures that had been pulmonary hypertension is avoided.[2] Numerous se-prescribed to all participants in this study, which lective 5-HT2C receptor agonists have been devel-could be attributable to the agonist’s efficacy at oped including Org 12962 from Organon, Ro 60-suppressing the hunger increases caused by dieting. 0175 from Roche and Vernalis, VER-3323 fromAs a consequence, it would have been interesting to Vernalis, BVT-933 from Biovitrum and Glax-examine if extending the trial would have shown the oSmithKline and YM348 from Yamanouchicontinued appetite modulation of Org 12962 to Pharmaceuticals.[2] Some of the compounds havetranslate into a significantly greater weight loss than passed into phase I and (in the case of BVT-933)that seen in the placebo group.[136] phase II trials; it is regrettable that their effects on



human appetite, food intake and bodyweight remain • reduce meal size and the number of between-largely unknown. One issue may have been drug meal eating episodes while the patient exper-affinity to serotonin receptors other than 5-HT2C, iences greater and longer-lasting satisfaction forcausing adverse effects during the clinical trial stud- those still remainingies. • selectively reduce the intake of energy-dense

high-fat foods most associated with obesity andHowever, lorcaserin (Arena Pharmaceuticals)ill health (those generally being snack and so-is known to be currently undergoing clinical tri-called convenience foods).als[140-142] and, according to the company, in a phaseAll future serotonergic drugs should be measuredIa study the drug significantly reduced meal size. A

by these criteria, above and beyond those for anti-single 10mg dose produced a statistically significantobesity drugs in general. Furthermore, in order to10.7% (122.5 kcal) mean reduction in meal sizemake a sizeable impact on the market, any drug willrelative to placebo. The drug also completed a phaseneed to produce placebo-subtracted weight loss overIb safety dose-escalation study, and no effect on1 year that is greater than that currently produced byheart valves or pulmonary artery pressure was ob-anti-obesity drugs orlistat (less than 4kg) and sibu-served, so phase II trials were carried out in 2005.tramine (over 4kg), and that reported to be producedThe results of these, again according to the compa-by novel anti-obesity agent rimonabant (approxi-ny, are promising.[140] In phase IIa trials, a 15mgmately 5–6kg).[144]daily dose produced a statistically significant mean

weight loss of 1.3kg (compared with 0.4kg in the9. Novel Serotonergic Targets forplacebo group) over a 28-day treatment period.[142]

Weight ControlRecently, results of the phase IIb trials have alsobeen published, showing that treatment with lor-

While most of the focus on serotonin and weightcaserin was associated with a highly significant av-control has been on drugs agonising the 5-HT2Cerage weight loss of 1.8, 2.6 and 3.6kg at daily dosesreceptor, other targets in the serotonin system exist.of 10, 15 and 20mg, respectively, over the 12-weekThe 5-HT6 receptor is one of the most recent addi-treatment period. In comparison, those in the place-tions to the serotonin receptor family. It is almostbo group lost just 0.3kg in that time. This drug isexclusively expressed within the CNS with highexpected, funding permitted, to enter phase IIIlevels in cortical and limbic regions.[145] The recentclinical trials in the second half of 2006. The struc-development of specific 5-HT6 receptor ligands hasture of lorcaserin is undisclosed, but it is likely toindicated potential roles for this receptor in a num-have come from a series of novel 3-benzazepineber of physiological processes, including feed-derivatives.[141]

ing.[146] Selective 5-HT6 receptor antagonists haveHalford et al.[143] described the ideal attributes of been reported to produce significant reductions in

any appetite suppressant anti-obesity drug in 2003. food intake when administered to ob/ob mice, withAn ultimate appetite suppressant anti-obesity drug these hypophagic effects being accompanied by sig-should ideally: nificant reductions in bodyweight and insulin• reliably alter feeding behaviour and food choice levels.[147] Such results identify 5-HT6 receptor an-

to produce a reduction in caloric intake sus- tagonists as potential anti-obesity agents. Notably ataining the period of weight loss number of companies have recently indicated they

• enable the establishment of healthier eating pat- have preclinical 5-HT6 receptor-based anti-obesityterns drug development programmes and presumably a


50 Halford et al.

number of suitable 5-HT6 receptor antagonists now and block weight gain in rodents. Whether drug-exist. 5-HT6 antagonists such as PRX-08034 (Epix induced hypophagia and associated effects onPharmaceuticals) and BVT 74316 (Biovitrum AB) weight are due to selective action on appetite ishave recently entered clinical trials.[148,149] unclear.

A new generation of selective 5-HT2C receptor10. Summary agonists have been developed and some have passed

into clinical testing. The selectivity of these com-Anti-obesity treatments have targeted and will pounds should ensure that they avoid the adverse

continue to target the endogenous serotonin satiety effects associated with their predecessors. However,system. As levels of endogenous serotonin respond it is essential that these new drugs produce markedto both deprivation and energy excess, and a reduced effects on appetite and feeding behaviour, and arecaloric intake lowers CNS serotonin levels and turn- able to induce substantial and sustained hypophagiaover, serotonin appears to play a key role in appetite. that is sufficient to produce clinically significantMoreover, susceptibility to weight gain in both ro- weight loss. When assessing the effects of thesedents and humans may be attributable to low levels drugs, it is important to examine not just kilocalorieof endogenous serotonin and serotonin dysfunction. or gram reduction in intake, but also observedThe hypothalamic serotonin satiety system is known changes in appetite (such as hunger, prospectiveto interact with orexigenic systems such as orexin consumption, fullness, etc.) and feeding behaviourand NPY. Serotonin may inhibit feeding behaviour (eating rate, meal size, daily meal and snack num-by, among other means, blocking these hunger sig- ber). As obesity is known to be linked to the con-nals. In addition, the hypophagic effects of ser- sumption of highly palatable, energy-dense foodsotonergic drugs appear to be mediated by the an- (high in fat and/or sugar) it is necessary to establishorexigenic melanocortin system. Therefore, as an how a drug modifies the type of food chosen inepisodic satiety transmitter, serotonin (like the tonic terms of energy density, macronutrient compositionadiposity signal leptin) influences feeding behaviour and palatability. Any drug able to reduce the likingvia both stimulatory and inhibitory effects on sever- for or the wanting of highly palatable foods that areal regulatory neuropeptide systems in the hypothala- known to promote overconsumption and weightmus. gain is likely to be of significant therapeutic value.

Drugs that either directly or indirectly stimulate There are also a number of non-serotonergic treat-hypothalamic 5-HT2C receptors in rodents produce ments undergoing clinical trials. Many of these areboth changes in the structure of feeding behaviour also entering phase II studies or preparing for phaseand reductions in food intake that are consistent with III trials.[150] Ultimately, it is against these drugs, asthe satiety process. These drugs cause an enhance- well as existing treatments, that the efficacy of anyment of the post-meal satiety potency of fixed calor- serotonin-targeted anti-obesity treatment will beic loads and reduce pre-meal appetite and food in- judged.take at ad libitum meals in both lean and obesehumans. Reductions in bodyweight gain and de- Acknowledgementscreases in bodyweight from baseline are strongly

The authors would like to thank Miss Lisa D.M. Richardsassociated with the hypophagic action of these drugsfor her help in preparing this manuscript and Dr Steve Vick-

in both rodent dietary-induced obesity models and ers (RenaSci Consultancy Ltd) for providing much usefulhuman clinical trials. Moreover, antagonism of material for inclusion within this review. The authors would5-HT6 receptors also appears to reduce food intake like in particular to thank Lora Heisler (University of Cam-



complete serotonergic depletion in the rat. Synapse 1996; 24:bridge), Joel Elmquist (Harvard Medical School) and87-96Michael Cowley (Oregon Health Sciences University) for

16. Currie PJ. Integration of hypothalamic feeding and metabolicproviding updates on their most recent research and allowingsignals: focus on neuropeptide Y. Appetite 2003; 41: 335-7

us to reproduce their figure. The website for the Kissileff 17. Heisler LK, Cowley MA, Tecott LH, et al. Activation of centralLaboratory is http://www.liv.ac.uk/Psychology/kissilefflab/ melanocortin pathways by fenfluramine. Science 2002; 297:Home.html. 609-11

18. Heisler LK, Cowley MA, Kishi T, et al. Central serotonin andThe laboratory would like to thank corporate donors Glax-melanocortin pathways regulating energy homeostasis. N YoSmithKline, NJ, USA, for providing funds for postgraduateAcad Sci 2003; 994: 169-74training. Drs Harrold and Halford have received research

19. Heisler LK, Jobst EE, Sutton GM, et al. Serotonin reciprocallyfunding from Predix Pharmaceuticals. Professor Blundell and regulated melanocortin neurons to modulate food intake. Neu-Dr Halford have received research funding from Sanofi- ron 2006; 51: 239-49Aventis. 20. Nambu T, Sakurai T, Mizukami K, et al. Distribution of orexin

neurons in the adult rat brain. Brain Res 1999; 827: 243-6021. Hervieu GJ, Cluderay JE, Harrison DC, et al. Gene expression

and protein distribution of the orexin-1 receptor in the rat brainReferencesand spinal cord. Neuroscience 2001; 103: 777-971. Halford JCG, Blundell JE. Separate systems for serotonin and

22. Marcus JN, Aschkenasim CJ, Lee CE, et al. Differential expres-leptin in appetite control. Ann Med 2000; 32: 222-32sion of orexin receptors 1 and 2 in the rat brain. J Comp Neurol2. Vickers SP, Dourish CT. Serotonin receptor ligands and the2001; 435: 6-25treatment of obesity. Curr Opin Invest Drugs 2004; 5: 377-88

23. Brown RE, Sergeeva OA, Eriksson KS, et al. Convergent exci-3. Bentley JM, Adams DR, Bebbington D. Indoline derivatives astation of dorsal raphe serotonin neurons by multiple arousal5-HT2C receptor agonists. Bioorg Med Chem Let 2004; 14:systems (orexin/hypocretin, histamine and noradrenaline). J2367-70Neurosci 2002; 22: 8850-94. Bray GA, York DA. Studies on food intake of genetically obese

24. Cai XJ, Liu XH, Evans M, et al. Orexins and feeding: specialrats. Am J Physiol 1972; 233: 176-9occasions or everyday occurrence? Reg Peptides 2002; 104: 1-5. Jesperson S, Scheel-Kruger J. Evidence for a difference in9mechanism of action between fenfluramine- and ampheta-

25. Orlando G, Brunetti Ll, Di Nisio C, et al. Effects of cocaine- andmine-induced anorexia. J Pharm Pharmacol 1973; 22: 637-8amphetamine-regulated transcript peptide, leptin and orexins6. Barrett AM, McSherry L. Inhibition of drug-induced anorexia inon hypothalamic serotonin release. Eur J Pharmacol 2001;rats by methysergide. J Pharm Pharmacol 1975; 27: 889-95430: 269-727. Pinder BM, Brogden RN, Sawyer PR, et al. Fenfluramine: a

26. Matsuzaki I, Sakurai T, Kunii K, et al. Involvement of thereview of its pharmacological properties and therapeutic effi-serotonergic system in orexin-induced behavioral alterations incacy in obesity. Drugs 1975; 10: 241-323rats. Reg Peptides 2002; 104: 119-238. Garattini S, Samanin R. Anorectic drugs and brain neurotrans-

27. Brown RE, Sergeeva O, Eriksson KS, et al. Orexin A excitesmitters. In: Silverstone T, editor. Food intake and appetite.serotonergic neurons in the dorsal raphe nucleus of the rat.Berlin: Dahlem Konferenzen, 1976: 82-208Neuropharmacology 2001; 40: 457-99. MacKenzie RG, Hoebel BG, Ducret RP, et al. Hyperphagia

28. Simansky KJ. Serotoninergic control of the organisation offollowing intraventricular p-chlorophenylalanine-, leucine- orfeeding and satiety. Behav Brain Res 1996; 73: 37-42tryptophan-methyl esters: lack of correlation with whole brain

serotonin levels. Pharmacol Biochem Behav 1979; 10: 951-5 29. Hewitt KN, Lee MD, Dourish CT, et al. Serotonin 2C receptoragonists and the behavioural satiety sequence in mice.10. Blundell JE. Is there a role for serotonin (5-hydroxytryptamine)Pharmacol Biochem Behav 2002; 71: 691-700in feeding? Int J Obesity 1977; 1: 15-42

30. Collin M, Backberg M, Onnestam K, et al. 5-HT1A receptor11. Hoyer D, Martin G. 5-HT receptor classification and nomencla-immunoreactivity in hypothalamic neurons involved in bodyture: towards a harmonization with the human genome.weight control. Neuroreport 2002; 13: 945-51Neuropharmacology 1997; 36: 419-28

31. Muraki Y, Yamanaka A, Tsujino N, et al. Serotonergic regula-12. Blundell JE, Halford JCG. Serotonin and appetite regulation:tion of the orexin/hypocretin neurons through the 5-HT1Aimplications for the treatment of obesity. CNS Drugs 1998; 9:receptor. J Neurosci 2004; 24: 7159-66473-95

32. Barrafan-Majia MG, Castilla-Serna L, Calderon-Guzman D, et13. Rogers P, McKibbin PE, Williams G. Acute fenfluramine ad-al. Effect of nutritional status and ozone exposure on rat brainministration reduces neuropeptide Y concentrations in specificserotonin. Arch Med Res 2002; 33: 15-9hypothalamic regions of the rat: possible implications for the

anorectic effect of fenfluramine. Peptides 1991; 12: 251-5 33. Xie QW. Experimental studies on changes of neuroendocrine14. Brown CM, Coscina DV. Ineffectiveness of hypothalamic sero- functions during starvation and refeeding. Neuroendocrinolo-

tonin to block neuropeptide Y-induced feeding. Pharmacol gy 1991; 53: 52-9Biochem Behav 1995; 51: 641-6 34. Nishimura F, Nishihara M, Torii K, et al. Changes in respon-

15. Compan V, Dusticier N, Nieoullon A, et al. Opposite changes in siveness to serotonin on rat ventromedial hypothalamic neu-striatal neuropeptide Y immunoreactivity after partial and rons after food deprivation. Physiol Behav 1996; 60: 7-12


52 Halford et al.

35. Wolfe BW, Metzger ED, Stollar C. The effects of dieting on serotonin antagonists. Pharmacol Biochem Behav 1988; 31:plasma tryptophan concentration and food intake in healthy 475-9women. Physiol Behav 1997; 61: 537-41 53. Grignaschi G, Samanin R. Role of serotonin and catecholamines

36. Cowen PJ, Clifford EM, Walsh AES, et al. Moderate dieting in brain in feeding suppressant effects of fluoxetine.causes 5-HT2C supersensitization. Psychol Med 1996; 26: Neuropharmacology 1992; 31: 445-91156-9 54. Lightowler S, Wood M, Brown T, et al. An investigation of the

37. Varma M, Torelli GF, Meguid MM, et al. Potential strategies for mechanism responsible for fluoxetine-induced hypophagia inameliorating early cancer anorexia. J Surg Res 1998; 81: 69-76 rats. Eur J Pharmacol 1996; 296: 137-43

38. Yang ZJ, Blaha V, Meguid MM, et al. Interleukin-1α injection 55. Lee MD, Clifton PG. Partial reversal of fluoxetine anorexia byinto ventromedial hypothalamic nucleus of normal rats de- the 5-HT antagonist metergoline. Psychopharmacology 1992;presses food intake and increases release of dopamine and 107: 359-64serotonin. Pharmacol Biochem Behav 1999; 62: 61-5 56. Halford JCG, Blundell JE. Metergoline antagonizes fluoxetine

39. Cangiano C, Laviano A, Muscaritoli M, et al. Cancer anorexia: induced suppression of food intake but not changes in thenew pathogenic and therapeutic insights. Nutrition 1996; 12: behavioural satiety sequence. Pharmacol Biochem BehavS48-51 1996; 54: 745-51

40. Cangiano C, Testa U, Muscaritoli M, et al. Cytokines, trypto- 57. Koe BK, Weissman A, Welch WM, et al. Sertraline, 1S,4S-N-phan and anorexia in cancer patients before and after surgical methyl-4(3,4-dichlorophenyl)-1,2,3,4-tetrahydro-1-naphathy-tumor ablation. Anticancer Res 1994; 14: 1451-5 lamine, a new uptake inhibitor with selectivity for serotonin. J

Pharmacol Exp Ther 1983; 266: 686-70041. Meguid MM, Fetissov SO, Blaha V, et al. Dopamine andserotonin VMN release is related to feeding status in obese and 58. Lucki I, Kreider MS, Simansky KJ. Reduction of feeding beha-lean Zucker rats. Neuroreport 2000; 11: 2069-72 viour by the serotonin uptake inhibitor sertraline. Psychophar-

macology 1988; 96: 289-9542. Svec F, Thompson H, Porter J. Levels of hypothalamic neuro-transmitters in lean and obese Zucker rats. Nutr Neurosci 2002; 59. Kennett GA, Curzon G. Evidence that the hypophagia induced5: 321-6 by mCPP and TFMPP requires 5-HT1C and 5-HT1B recep-

tors; hypophagia induced by RU-24969 only requires 5-HT1B43. Harrold JA, Widdowson PS, Clapham JC, et al. Individualreceptors. Psychopharmacology 1988; 96: 93-100severity of dietary obesity in unselected Wistar rats: relation-

ship with hyperphagia. Am J Physiol 2000; 279: E340-7 60. Kennett GA, Curzon G. Evidence that mCPP may havebehavioural effects mediated by central 5-HT1C receptors. Br44. Hassanain M, Levin BE. Dysregulation of hypothalamic seroto-J Pharmacol 1988; 94: 137-47nin turnover in diet-induced obese rats. Brain Res 2002; 929:

175-80 61. Kennett GA, Curzon G. Potencies of antagonists indicate that 5-HT1C receptors mediate 1-3(chlorophenyl)piperazine-induced45. Breum L, Rasmussen MH, Hilsted J. Twenty-four–hour plasmahypophagia. Br J Pharmacol 1991; 10: 2016-20tryptophan concentrations and ratios are below normal in

obese subjects and are not normalized by substantial weight 62. Halford JCG, Blundell JE. The 5-HT1B receptor agonist CP-reduction. Am J Clin Nutr 2003; 77: 1112-8 94,253 reduces food intake and preserves the behavioural

satiety sequence. Physiol Behav 1996; 60: 933-946. Clifton PG. The neuropharmacology of meal patterning. In:Cooper SJ, editor. Ethology and psychopharmacology. Chich- 63. Lee MD, Simansky KJ. CP-94,253: a selective serotonin1B (5-ester: Wiley, 1994: 313-28 HT1B) agonist that promotes satiety. Psychopharmacology

1997; 131: 264-7047. Neill JC, Cooper SJ. Evidence that d-fenfluramine anorexia ismediated by 5-HT1 receptors. Psychopharmacology 1989; 97: 64. Schreiber R, Selbach K, Asmussen M, et al. Effects of seroto-213-8 nin1/2 receptor agonists on dark-phase food and water intake

in rats. Pharmacol Biochem Behav 2000; 67: 291-30548. Samanin R, Mennini T, Bendotti C, et al. Evidence that central5-HT2C receptors do not play an important role in anorectic 65. Clifton PG, Lee MD, Dourish CT. Similarities in the action ofactivity of d-fenfluramine in the rat. Neuropharmacology Ro 60-0175, a 5-HT2C receptor agonist, and d-fenfluramine1989; 28: 465-9 on feeding patterns in the rat. Psychopharmacology 2000; 152:

256-6749. Neill JC, Bendotti C, Samanin R. Studies on the role of 5-HTreceptors in satiation and the effect of d-fenfluramine in the 66. Lee MD, Kennett GA, Dourish CT, et al. 5-HT1B receptorsrunway test. Eur J Pharmacol 1990; 190: 105-12 modulate components of satiety in the rat: behavioural and

pharmacological analyses of the selective serotonin1B agonist50. Simansky JJ, Nicklous DM. Parabrachial infusion of -fen-CP-94,253. Psychopharmacology 2002; 164: 49-60fluramine reduces food intake: blockade by the 5-HT1B antag-

onist SB-216641. Pharmacol Biochem Behav 2002; 71: 681- 67. Blundell JE, Latham CJ. Pharmacological manipulation of feed-90 ing behaviour: possible influences of serotonin and dopamine

on food intake. In: Garattini S, Samanin R, editors. Central51. Vickers SP, Dourish CT, Kennett GA. Evidence that hy-mechanisms of anorectic drugs. New York (NY): Raven Press,pophagia induced by d-fenfluramine and d-norfenfluramine in1978: 83-109the rat is mediated by 5-HT2C receptors. Neuropharmacology

2001; 41: 200-9 68. Halford JCG, Wanninayake SCD, Blundell JE. Behavioural52. Wong DT, Reid LR, Threlkeld PG. Suppression of food intake satiety sequence (BSS) for the diagnosis of drug action on food

in rats by fluoxetine: comparison of enantiomers and effects of intake. Pharmacol Biochem Behav 1998; 61: 159-68



69. Blundell JE, Latham CJ. Characteristic adjustments to the struc- 86. Goodall E, Silverstone T. Differential effect of d-fenfluramineture of feeding behaviour following pharmacological treat- and metergoline on food intake in human subjects. Appetitements: effects of amphetamine and fenfluramine and the antag- 1988; 11: 215-88onism by pimozide and metergoline. Pharmacol Biochem 87. Blundell JE, Hill AJ. On the mechanism of action of dexfen-Behav 1980; 12: 717-22 fluramine: effect on alliesthesia and appetite motivation in lean

70. Blundell JE, McArthur RA. Behavioural flux and feeding: con- and obese subjects. Clin Neuropharmacol 1988; 11 Suppl. 1:tinuous monitoring of food intake and food selection, and the 121-34Svideo-recording of appetitive and satiety sequences for the 88. Goodall EM, Cowen PJ, Franklin M, et al. Ritanserin attenuatesanalysis of drug action. In: Samanin R, Garattini S, editors. anorectic endocrine and thermic responses to d-fenfluramineAnorectic agents: mechanisms of action and tolerance. New in human volunteers. Psychopharmacology 1993; 112: 461-6York (NY): Raven Press, 1981: 19-43 89. McGuirk J, Silverstone T. The effect of 5-HT re-uptake inhibi-

71. Halford JCG, Blundell JE. 5-Hydroxytryptaminergic drugs tor fluoxetine on food intake and body weight in healthy malecompared on the behavioural sequence associated with satiety. subjects. Int J Obesity 1990; 14: 361-72Br J Pharmacol 1993; 100: 95 90. Lawton CL, Wales JK, Hill AJ, et al. Serotoninergic manipula-

72. Clifton PG, Barnfield AMC, Philcox L. A behavioural profile of tion, meal-induced satiety and eating patterns: effects of fluox-fluoxetine induced anorexia. Psychopharmacology 1989; 97: etine in obese female subjects. Obes Res 1995; 3: 345-5689-95 91. Wadden TA, Bartlet SJ, Foster GD, et al. Sertraline and relapse

73. Simansky KJ, Viadya AH. Behavioural mechanisms for the prevention following treatment by a very low calorie diet: aanorectic actions of the serotonin (5-HT) uptake inhibitor controlled clinical trial. Obes Res 1995; 3: 549-57sertraline in rats: comparison with directly acting agonists. 92. Walsh AE, Smith KA, Oldman AD. m-ChlorophenylpiperazineBrain Res Bull 1990; 25: 953-60 decreases food intake in a test meal. Psychopharmacology

74. McGuirk J, Muscat R, Willner P. Effects of the 5-HT uptake 1994; 116: 120-2inhibitors femoxetine and parpexetine, and the 5-HT1A ago- 93. Sargent PA, Sharpley AL, Williams C, et al. 5-HT2C receptornist cltoprazine, on the behavioural satiety sequence. activation decreases appetite and body weight in obese sub-Pharmacol Biochem Behav 1992; 41: 801-5 jects. Psychopharmacology 1997; 133: 309-12

75. Kitchener SJ, Dourish CT. An examination of the behavioural 94. Boeles S, Williams C, Campling GM, et al. Sumatriptan de-specificity of hypophagia induced by 5-HT1B, 5-HT1C and 5- creases food intake and increases plasma growth hormone inHT2 receptor agonists using the post-prandial sequence in rats. healthy women. Psychopharmacology 1997; 129 (Pt 2): 179-Psychopharmacology 1994; 113: 368-77 82

76. Tecott LH, Sun LM, Akanna SF, et al. Eating disorder and 95. Rolls BJ, Shide DJ, Thorward ML, et al. Sibutramine reducesepilepsy in mice lacking 5-HT2C serotonin receptors. Nature food intake in non-dieting women with obesity. Obes Res1995; 374: 542-6 1998; 6: 1-11

77. Nonogaki K, Abdullah L, Goulding EH, et al. Hyperactivity and 96. Cangiano C, Laviano A, Del Ben M, et al. Effects of oral 5-reduced energy cost of physical activity in serotonin 5-HT2C hydroxy-tryptophan on energy intake and macronutrient selec-receptor mutant mice. Diabetes 2003; 52: 315-20 tion in non-insulin dependent diabetic patients. Int J Obesity

78. Vickers SP, Clifton PG, Dourish CT, et al. Reduced satiating 1988; 22: 648-54effect of d-fenfluramine in serotonin 5-HT2C receptor mutant 97. Foltin RW, Haney M, Comer S, et al. Effect of fenfluramine onmice. Psychopharmacology 1999; 143: 309-14 food intake, mood, and performance of humans living in a

79. Bouwknecht JA, van der Guten J, Hijsenm TH, et al. Male and residential laboratory. Physiol Behav 1996a; 59: 295-305female 5-HT1B receptor knockout mice have higher body 98. Drent ML, Zelissen PMJ, Kopperchaar HPF, et al. The effect ofweights than wildtypes. Physiol Behav 2001; 74: 507-16 dexfenfluramine on eating habits in a Dutch ambulatory an-

80. Silverstone T, Goodall E. The clinical pharmacology of appetite droid overweight population with an overconsumption ofsuppressant drugs. Int J Obes 1984; 8 (1): 23-33 snacks. Int J Obesity 1995; 19: 299-304

81. Cangiano C, Ceci F, Casinco A, et al. Eating behaviour and 99. Pijl H, Koppeschaar HPF, Willekens FLA, et al. Effect ofadherence to dietary prescription in obese adult subjects treat- serotonin re-uptake inhibition by fluoxetine on body weighted with 5-hydroxytryptophan. Am J Clin Nutr 1992; 56: 863-7 and spontaneous food choice in obesity. Int J Obesity 1991; 15:

237-4282. Rogers PJ, Blundell JE. Effect of anorexic drugs on food intakeand the micro-structure of eating in human subjects. 100. Ward AS, Comer SD, Haney M, et al. Fluoxetine-maintainedPsychopharmacology 1979; 66: 159-65 obese humans: effect on food intake and body weight. Physiol

Behav 1999; 66: 815-2183. Hill AJ, Blundell JE. Sensitivity of the appetite control systemin obese subjects to nutritional and serotoninergic challenges. 101. Chapelot D, Mamonier C, Thomas F, et al. Modalities of theInt J Obesity 1990; 14: 219-33 food intake-reducing effect of sibutramine in humans. Physiol

Behav 2000; 68: 299-30884. Wurtman JJ, Wurtman RJ, Growdon JH, et al. Carbohydratecraving in obese people: suppression by treatments affecting 102. Hansen DL, Toubro S, Stock MJ, et al. Thermogenic effects ofserotonergic transmission. Int J Eat Disord 1982; 1: 2-15 sibutramine in humans. Am J Clin Nutr 1998; 68: 1180-6

85. Wurtman JJ, Wurtman RJ, Mark S, et al. d-Fenfluramine selec- 103. Hansen DL, Toubro S, Stock MJ, et al. The effect of sibutraminetively suppresses carbohydrate snacking in obese subjects. Int on energy expenditure and appetite during chronic treatmentJ Eat Disord 1985; 4: 89-99 without dietary restriction. Int J Obesity 1999; 23: 1016-24


54 Halford et al.

104. Barkeling B, Elfhag K, Rooth P, et al. Short-term effects of 120. Guy-Grand B, Apfelbaum M, Creoaldi C, et al. Internationalsibutramine (Reductil) on appetite and eating behaviour and trial of long-term dexfenfluramine in obesity. Lancet 1989;the long-term therapeutic outcome. Int J Obesity 2003; 27: Nov 11: 1142-5693-700 121. Guy-Grand B. Clinical studies with d-fenfluramine. Am J Clin

Nutr 1992; 55: 173-6s105. Cowen PJ, Sargent PA, Williams C, et al. Hypophagic, endo-crine and subjective responses to m-chlorophenylpiperazine in 122. Finer N, Finer S, Naoumova RP. Drug therapy after very-low-healthy men and women. Hum Psychopharmacol 1995; 10: calorie-diets. Am J Clin Nutr 1992; 56: 195-8s385-91 123. James WPT, Astrup A, Finer N, et al. Effect of sibutramine on

weight maintenance after weight loss: a randomised trial.106. Ghaziuddin N, Welch K, Greden J. Central serotonergic effectsLancet 2000; 356: 2119-25of m-chlorophenylpiperazine (mCPP) among normal control

adolescents. Neuropsychopharmacology 2003; 28: 133-9 124. McMahon FG, Fujioka K, Singh BN, et al. Efficacy and safetyof sibutramine in obese white and African American patients107. Fisler JS, Undernerger SJ, York DA, et al. d-Fenfluramine in awith hypertension: a 1-year, double-blind, placebo-controlled,rat model of dietary fat-induced obesity. Pharmacol Biochemmulticenter trial. Arch Intern Med 2000; 160: 2185-91Behav 1993; 45: 487-93

125. Smith IG, Goulder MA. Randomized placebo-controlled trial of108. Vickers SP, Benwell KR, Porter RH, et al. Comparative effectslong-term treatment with sibutramine in mild to moderateof continuous infusion of mCPP, Ro 60-0175 and d-fen-obesity. J Fam Pract 2001; 50: 505-12fluramine on food intake, water intake, body weight and loco-

126. McNulty SJ, Ur E, Williams G. A randomized trial of sibu-motor activity in rats. Br J Pharmacol 2000; 130: 1305-14tramine in the management of obese type 2 diabetic patients109. Vickers SP, Easton N, Webster LJ, et al. Oral administration oftreated with metformin. Diabetes Care 2003; 26: 125-33the 5-HT2C receptor agonist, mCPP, reduces body weight gain

127. Poston WSC, Reeves RS, Haddock CK, et al. Weight loss inin rats over 28 days as a result of maintained hypophagia.obese Mexican Americans treated for 1-year with orlistat andPsychopharmacology 2003; 167: 274-80lifestyle modification. Int J Obesity 2003; 27: 1486-93

110. Yen TT, Wong DT, Bemis KG. Reduction of food consumption128. Arterburn DE, Crane PK, Veenstra DL. The efficacy and safetyand body weight of normal and obese mice by chronic treat-

of sibutramine for weight loss: a systematic review. Archment with fluoxetine: a serotonin reuptake inhibitor. Drug DevIntern Med 2004; 164: 994-1003Res 1987; 10: 37-45

129. Padwal R, Li SK, Lau DCW. Long-term pharmacotherapy for111. Fuller RW, Wong DT. Fluoxetine: a serotonergic appetite sup-overweight and obesity: a systematic review and meta-analysispressant drug. Drug Dev Res 1989; 17: 1-15of randomized controlled trials. Int J Obesity 2003; 27: 1437-

112. Nielsen JA, Chapin DS, Johson JL, et al. Sertraline, a serotonin- 46uptake inhibitor, reduces food intake and body weight in lean

130. Wise SD. Clinical studies with fluoxetine in obesity. Am J Clinrats and genetically obese mice. Am J Clin Nutr 1992; 55: 185-

Nutr 1992; 55: 181-4s8s

131. Goldstein DJ, Rampey AH, Roback PJ, et al. Efficacy and safety113. Wieczorek I, Schulz C, Jarry H, et al. The effects of the selective of long-term fluoxetine treatment of obesity -maximising suc-

serotonin reuptake-inhibitor fluvoxamine on body weight in cess. Obes Res 1995; 3 Suppl. 4: 481-90sZucker rats are mediated by corticotropin-releasing hormone.

132. Darga LL, Carroll-Michals C, Botsford SJ, et al. Fluoxetine’sInt J Obesity 2001; 25: 1566-9

effect on weight loss in obese subjects. Am J Clin Nutr 1991;114. Konkle ATM, Sreter KB, Bajer SL, et al. Chronic paroxetine 54: 315-21

infusion influences macronutrient selection in male 133. Goldstein DJ, Rampey AH, Enas GG, et al. Fluoxetine: aSprague–Dawley rats. Pharmacol Biochem Behav 2003; 74: randomized clinical trial in the treatment of obesity. Int J883-90 Obesity 1994; 18: 129-35

115. Kennett GA, Wood MD, Bright F, et al. SB 242084, a selective 134. Simpson RJ, Lawton DJ, Watt MH, et al. Effect of zimelidine, aand brain potent 5-HT2C receptor. Neuropharmacology 1997; new antidepressant, on appetite and body weight. Br J Clin36: 609-20 Pharmacol 1981; 11: 96-8

116. Hayashi A, Sonoda R, Kimura Y, et al. Antiobesity effect of 135. Van Baak M, Lentjes M, Mujakovic S, et al. Behavior modifica-YM348, a novel 5-HT2C receptor agonist, in Zucker rats. tion and societal change in the prevention of obesity. Obes ResBrain Res 2004; 1011: 221-7 2003; 11 (1 Suppl.): A111

117. Bjenning C, Williams J, Whelan K, et al. Chronic oral adminis- 136. Halford JCG. Serotonin (5-HT) drugs: effects on appetite ex-tration of APD356 significantly reduces body weight and fat pression and use for the treatment of obesity. Curr Drugmass in obesity-prone (DIO) male and female rats. Int J Targets 2004; 5: 637-46Obesity 2004; 28 (1 Suppl.): 214s 137. Abeniam L, Moride Y, Brenot F, et al. Appetite suppressant

drugs and the risk of primary pulmonary hypertension. N Engl118. Haddock CK, Poston WSC, Dill PL, et al. Pharmacotherapy forJ Med 1996; 335: 609-16obesity: a quantitative analysis of four decades of published

randomized clinical trials. Int J Obesity 2002; 26: 262-73 138. Greenway FL, Caruso MK. Safety of obesity drugs. Expert OpinDrug Saf. In press119. Pinder RM, Brogden RN, Sawyer PR, et al. Fenfluramine: a

review of the pharmacological properties and therapeutic effi- 139. Bays HE, Dujovene CA. Anti-obesity drug development. Expertcacy in obesity. Drugs 1975; 10: 241-323 Opin Invest Drugs 2002; 11: 1189-204



140. Bays HE. Current and investigational antiobesity agents and 147. Shacham S, Marantz Y, Senderowitz H, et al. Novel 5-HT6obesity therapeutic treatment targets. Obes Res 2004; 12: receptor antagonists for the treatment of obesity. Obes Res1197-211

2005; 13: A192141. Smith BM, Smith JM, Tsai JH, et al. Discovery and SAR of new

148. EPIX Pharmaceuticals. EPIX Pharmaceuticals announces find-benzazapines and potent and selective 5-HT2C receptor ago-

ings from obesity and cognitive impairment studies at Societynist for the treatment of obesity. Bioorgan Med Chem Lett2005; 12: 1467-70 for Neuroscience Meeting [online]. Available from URL:

http://investor.epixpharma.com/phoenix.zhtml?c=91717&p=142. Smith S, Anderson J, Frank A, et al. The effects of APD356, aselective 5-HT2C agonist, on weight loss in a 4 week study in irol-newsArticle&ID=917339&highlight [Accessed 2006 Novhealthy obese patients [Abstract]. Obes Res 2005; 13 Suppl.: 13]101-R

149. Biovitrum’s project portfolio within obesity advances [online].143. Halford JCG, Cooper GD, Dovey TM, et al. Pharmacological

Available from URL: http://www.stockholmbioregion.com/approaches to obesity treatment; current medical chemistry.templates/page____517.aspx [Accessed 2006 Nov 13]CNS Agents 2003; 3: 283-310

150. Halford JCG. Obesity drugs in clinical development. Curr Opin144. Van Gaal LF, Rissanen AM, Scheen AJ, et al. Effects of thecannabinoid-1 receptor blocker rimonabant on weight reduc- Invest Drugs 2006; 7: 312-8tion and cardiovascular risk factors in overweight patients: 1-year experience from the RIO-Europe study. Lancet 2005;365: 1389-97

Correspondence and offprints: Dr Jason C.G. Halford, School145. Woolley ML, Marsden CA, Fone KC. 5-HT6 receptors. Current

of Psychology, University of Liverpool, Liverpool, L69drug targets: CNS Neurol Disord 2004; 3: 59-79

7ZA, UK.146. Vickers SP, Dourish CT. Serotonin receptor ligands and the

treatment of obesity. Curr Opin Invest Drugs 2004; 5: 377-88 E-mail: [email protected]




Therapeutic Applications of SildenafilCitrate in the Management ofPaediatric Pulmonary HypertensionLeah Leibovitch,1,2 Ilan Matok1 and Gideon Paret1

1 Department of Pediatric Critical Care, Safra Children’s Hospital, The Chaim Sheba MedicalCenter, Tel-Hashomer and the Sackler Faculty of Medicine, Tel Aviv University,Tel Aviv, Israel

2 Division of Neonatology and Developmental Biology, Department of Pediatrics, David GeffenSchool of Medicine, University of California at Los Angeles, Los Angeles, California, USA

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571. Definition of Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592. Subcategories of Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

2.1 Idiopathic and Familial Pulmonary Arterial Hypertension (PAH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602.2 Pulmonary Hypertension Secondary to Underlying Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

3. Diagnosis and Assessment of Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604. Treatment Modalities for Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4.1 Conventional Supportive Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.2 Specific Pharmacological Therapies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4.2.1 Calcium Channel Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.2.2 Nitric Oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.2.3 Prostaglandins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.2.4 Endothelin Receptor Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634.2.5 Phosphodiesterase Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5. Sildenafil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636. Clinical Applications of Sildenafil in Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

6.1 Idiopathic and Familial PAH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646.2 PAH Associated with Congenital Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676.3 Pulmonary Hypertension Associated with Lung Diseases and/or Hypoxaemia . . . . . . . . . . . . . . . 686.4 Sildenafil and Inhaled Nitric Oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696.5 Persistent Pulmonary Hypertension of the Newborn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

7. Adverse Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Pulmonary hypertension is characterised by a progressive increase in pulmona-Abstractry vascular resistance and a poor prognosis. The exact underlying mechanisms arestill poorly understood; however, it is hypothesised that pulmonary medial hyper-trophy and endothelial dysfunction lead to impaired production of vasodilatorssuch as nitric oxide (NO) and prostacyclin, and increased expression of vasocon-strictors such as endothelin-1. The current treatment modalities for pulmonaryhypertension include conventional supportive therapies and more specific phar-

58 Leibovitch et al.

macological therapies that are targeted at abnormalities of endothelial function.NO and phosphodiesterase type 5 (PDE5) inhibitors induce pulmonary vasodila-tion by increasing intracellular cyclic guanosine monophosphate (cGMP) concen-trations.

Sildenafil citrate is a highly selective inhibitor of PDE5. Investigations inanimal models and recent clinical case reports with some studies in the paediatricpopulation suggest that sildenafil may be a promising agent in treating pulmonaryhypertension. The effect of sildenafil on pulmonary vasculature appears to beindependent of the underlying cause, thereby providing a role in idiopathicpulmonary arterial hypertension (PAH), PAH associated with congenital heartdisease, pulmonary hypertension secondary to lung disease or persistent pulmona-ry hypertension of the newborn. It may also be beneficial in postoperativepulmonary hypertension and in neonates who are difficult to wean from inhaledNO. It is easily administered and effective, and has minimal systemic adverseeffects. Although the reported results in children with pulmonary hypertension arepromising, it is an experimental drug and large-scale randomised controlledstudies are required to validate the safety, efficacy and dosage in the paediatricpopulation.

Pulmonary hypertension is an important clinical sion was poor.[2] Fortunately, significant advances inpharmacological treatments have improved qualitypresentation of various paediatric and neonatal dis-of life and patient survival.[3]eases, including congenital heart disease, lung dis-

ease and liver disease, all of which cause high The most recent classification of pulmonary hy-morbidity and mortality. The disease is character- pertension was established during the 2003 Worldised by a progressive increase in pulmonary vascular Symposium (table II) and applies to both adults andresistance, leading to right ventricular failure. Pul- children.[4] This classification separates the causesmonary hypertension has similar characteristics in of pulmonary hypertension into five main catego-adults and children; however, some differences ex- ries: pulmonary arterial hypertension (PAH), pul-ist[1] and they are summarised in table I. Previously, monary venous hypertension associated with left-the prognosis of patients with pulmonary hyperten- sided heart disease, pulmonary hypertension associ-

Table I. Main differences between adults and children with pulmonary hypertension

Characteristic Adults Children

Main aetiologies Idiopathic and familial PAH, thromboembolic Most common aetiologies outside immediateand autoimmune diseases neonatal period are chronic lung disease and

congenital heart disease

Histopathology Advanced pulmonary vascular obstructive Severe pulmonary arterial medial hypertrophydisease with intimal fibrosis and plexiform with marked intimal proliferationlesions

Presenting symptoms Exertional dyspnoea, chest pain Failure to thrive, poor appetite, cyanoticevents, syncope, epilepsy-like attacks

The positive response rate to acute About 12% About 40%vasodilator testing with orally administeredvasodilator drugs

Gene mutation (BMPR2) Up to 50% in familial PAH have mutation in About 8% of the children with familial historythe gene have mutation in the gene

PAH = pulmonary arterial hypertension.


Sildenafil Citrate in Paediatric Pulmonary Hypertension 59

ated with lung disease and/or hypoxaemia, pulmo-nary hypertension due to chronic thrombotic and/orembolic disease, and miscellaneous.

In the paediatric population, PAH and pulmonaryhypertension associated with lung disease and hyp-oxia are the two most common categories.

PAH can be either idiopathic, familial or associ-ated with other underlying causes, such as connec-tive tissue disease, congenital heart disease, portalhypertension, drugs and toxins, and persistent pul-monary hypertension of the newborn.[3] Lung dis-eases associated with pulmonary hypertension in-clude respiratory distress syndrome, meconium as-piration syndrome, bronchopulmonary dysplasia,congenital diaphragmatic hernia, pulmonary hypo-plasia, interstitial lung disease and others.[3]

In the last three decades, there has been signifi-cant progress in understanding the pathogenicmechanisms involved in the development of PAH.As a result of this extensive research, new andeffective drugs, such as prostanoids, endothelin re-ceptor inhibitors and phosphodiesterase inhibitors,were developed. These drugs show improvement inexercise capacity, quality of life and survival ofthese patients.[5,6]

Sildenafil is a highly selective inhibitor ofphosphodiesterase (PDE)-5 which was recently dis-covered as a convenient and effective oral treatmentfor patients with pulmonary hypertension, mainly

Table II. Revised clinical classification of pulmonary hypertensiona

Classification

Pulmonary arterial hypertension

Idiopathic

Familial

Associated with collagen vascular disease, congenital systemic-to-pulmonary shunts, portal hypertension, HIV infection, drugsand toxins, thyroid disorders, glycogen storage disease,Gaucher’s disease, hereditary haemorrhagic telangiectasia,haemoglobinopathies, myeloproliferative disorders

Associated with significant venous or capillary involvement:pulmonary veno-occlusive disease, pulmonary capillaryhaemangiomatosis

Persistent pulmonary hypertension of the newborn

Pulmonary hypertension with left heart disease

Left-sided atrial or ventricular heart disease

Left-sided valvular heart disease

Pulmonary hypertension associated with lung diseases and/or hypoxaemia

Chronic obstructive pulmonary disease

Interstitial lung disease

Sleep-disordered breathing

Alveolar hypoventilation disorders

Chronic exposure to high altitude

Developmental abnormalities

Pulmonary hypertension due to chronic thrombotic and/orembolic disease

Thromboembolic obstruction of proximal pulmonary arteries

Thromboembolic obstruction of distal pulmonary arteries

Non-thrombotic pulmonary embolism (tumour, parasites, foreignmaterial)

Miscellaneous

Sarcoidosis, histiocytosis X, lymphangiomatosis, compression ofpulmonary vessels (adenopathy, tumour, fibrosing mediastinitis)

a This classification was adapted from the 2003 WorldSymposium on Pulmonary Hypertension (Venice 2003).[4]

those with idiopathic PAH or PAH secondary to anunderlying disease. This review highlights the re-

1. Definition of Pulmonary Hypertensioncent therapeutic applications of sildenafil in pulmo-nary hypertension of children and neonates. The definition of pulmonary hypertension in chil-

We reviewed the relevant medical literature by dren is the same as for adult patients. It is defined assearching MEDLINE (1995 to April 2006), using a mean pulmonary artery pressure ≥25mm Hg at restthe keyword ‘sildenafil’ and the combinations with or ≥30mm Hg during exercise, with normal pulmo-the following keywords: ‘pulmonary hypertension nary artery wedge pressure (≤15mm Hg) and anin children’, ‘pulmonary hypertension in neonates’, increased pulmonary vascular resistance index (≥3‘pulmonary arterial hypertension in children’, ‘pul- Wood units/m2).[2] The inclusion of exercise-in-monary arterial hypertension in neonates’ and ‘pul- duced haemodynamic abnormalities in the defini-monary hypertension of the newborn’. Our search tion of pulmonary hypertension is important, sincewas mainly on human studies, as well as experimen- the vasoactive response to triggers such as exercisetal models in animals, and preference was given to and hypoventilation is greater in children than inmanuscripts published in English. adults.[3,7] This difference probably stems from the



characteristic increased vascular medial hypertro- have abnormalities in endothelial function whichphy that is more pronounced in young children and results in impaired release of vasodilators, such asresults in dynamic vasoactivity. The increase in nitric oxide (NO) and prostacyclin, or an enhancedintimal fibrosis and plexiform lesions reduces this production of vasoconstrictors, such as endothelindynamic component in adults.[7] and thromboxane.[9]

2.2 Pulmonary Hypertension Secondary to2. Subcategories ofUnderlying DiseasesPulmonary Hypertension

Pulmonary hypertension in children and neonates2.1 Idiopathic and Familial Pulmonary is usually secondary to an underlying disease suchArterial Hypertension (PAH) as congenital heart disease, respiratory disease, per-

sistent pulmonary hypertension of the newborn, orIdiopathic PAH is a rare disease that is diagnosedother less common causes.after exclusion of all known causes.[5] It is a new

There is a wide spectrum of congenital heartterminology for the subcategory that previously wasdiseases that result in PAH[9] (table III), and the ageknown as primary pulmonary hypertension, and wasat which these pathologies cause irreversible pulmo-officially changed in the 2003 World Pulmonarynary vascular disease varies significantly. It seemsHypertension Symposium.[4,5]

that a combination of high pressure and high flowThe estimated incidence of idiopathic PAHcauses more rapid development of severe vascularranges from 1 to 2 cases per 1 million people in theremodelling.[9]

general population.[8,9] Idiopathic PAH occurs mostHypoxia plays an important role in the develop-frequently in young adult women[8,10] and is

ment of pulmonary hypertension. Respiratorycharacterised by progressive, sustained elevation inpathologies that result in hypoxia can be eitherpulmonary arterial pressure without a defined cause.bronchial (obstructive or restrictive) or parenchymalLike other pulmonary hypertensive disorders, thediseases (table III).average time to diagnosis is 1–2 years. The major

Persistent pulmonary hypertension of the new-difficulties in establishing the diagnosis of the idio-born is a syndrome characterised by increased pul-pathic type include the subtle presentation andmonary vascular resistance, extrapulmonary right-crossover of symptoms with other diseases. Symp-to-left shunting and severe hypoxaemia. Symptomstoms include dyspnoea, syncope, fatigability andusually present within 12 hours of birth.[3] It ischest pain. In infants, children and young adults,frequently associated with meconium aspiration,moderate pulmonary hypertension can be toleratedpneumonia, respiratory distress syndrome, sepsis,well without manifesting these symptoms.congenital diaphragmatic hernia, or pulmonary hy-The familial form of what was previously calledpoplasia.[15]

‘primary pulmonary hypertension’, and which ac-counts for only a small percentage of cases, is now 3. Diagnosis and Assessment ofcategorised as familial PAH. This form is inherited Pulmonary Hypertensionas an autosomal dominant trait, with a pattern ofgenetic anticipation. It is found in about 6% of The presenting signs and symptoms of patientspatients with primary pulmonary hypertension.[8,11] with pulmonary hypertension are age-specific andRecently, the gene for the familial form was mapped consist of nonspecific presentations.[9] Newbornsto chromosome 2q33 and named BMPR2 (bone with persistent pulmonary hypertension present withmorphogenic protein receptor-2).[12,13] cyanosis and respiratory distress.[15,16] Infants with

Although the cause of idiopathic PAH is un- pulmonary hypertension often present with signs ofknown, several physiological and pathological fea- low cardiac output, such as poor appetite, failure totures are commonly observed. Nearly all patients thrive, lethargy, diaphoresis, tachypnoea, tachycar-



Table III. Causes of pulmonary hypertension in the paediatric population (adapted from Tulloh,[14] with permission)

Cause

Neonates

persistent pulmonary hypertension of the newborn (idiopathic)

respiratory distress syndrome and bronchopulmonary dysplasia

structural disease: congenital diaphragmatic hernia, pulmonary hypoplasia, alveolar capillary dysplasia

interstitial disease: meconium aspiration syndrome, infection

Cardiac

congenital heart disease: left-to-right shunt (ASD, VSD, AVSD, PDA)

transposition of the great arteries

obstructive anomalies (TAPVC, MS, HLHS, HOCM, DCM)

Acquired

interstitial lung disease

cystic fibrosis

scoliosis

neuromuscular disease

chronic obstructive pulmonary disease

vasculitis

hypercoagulability states (protein C and protein S deficiency, factor V Leiden)

sleep-disordered breathing

Idiopathic (sporadic, familial)ASD = atrial septal defect; AVSD = atrioventricular septal defect; DCM = dilated cardiomyopathy; HLHS = hypoplastic left heart syndrome;HOCM = hypertrophic obstructive cardiomyopathy; MS = mitral stenosis; PDA = persistent ductus arteriosus; TAPVC = total anomalouspulmonary venous connection; VSD = ventricular septal defect.

dia and irritability. Toddlers and children may have tenol, inhaled NO, intravenous adenosine or inhaledcyanotic events with exertion, which are caused by iloprost) that is performed during the catheterisa-right-to-left shunting through a patent foramen tion.[5] Patients who are responsive to acute vasodi-ovale. Without adequate shunting, the patients may lator testing (reduction of 20% in the mean pulmo-present with syncope. After early childhood, chil- nary artery pressure with no change or increase indren present with similar symptoms to those in cardiac output) are likely to have a favourable re-adults, mainly exertional dyspnoea and chest pain.[7] sponse to calcium channel antagonist therapy.[7]

Patient history, physical examination and routine The modified New York Heart Associationtests such as electrocardiography and chest radiog- (NYHA) functional classification or the WHO clas-raphy provide important clues to the diagnosis. sification is used to clinically assess the severity ofHowever, the final diagnosis is established by right pulmonary hypertension (table IV). These classifi-(and left) heart catheterisation, including haemody- cations determine the degree of a patient’s dailynamic testing. Echocardiography confirms the pres- function and are helpful in guiding treatment.[17]

ence of pulmonary hypertension by demonstrating Another clinical test is the 6-minute walk test (6-increased right ventricular pressure and rules out or MWT), which is an independent predictor of deathconfirms the presence of congenital heart disease, in adult patients with idiopathic PAH and has beenpulmonary venous disease or left ventricular dys- used as the primary endpoint in many clinical trials.function.[9] Cardiac catheterisation remains the diag- It measures the distance that the patient can walk innostic gold standard, as it allows accurate measure- 6 minutes, and it is used to compare the exercisement of pulmonary artery pressure and pulmonary capacity during the course of the disease and thevascular resistance.[5,14] The degree of the pulmona- response to treatment. The WHO classification andry hypertension reversibility is determined by a the 6-MWT were adapted from adults and are usedshort-acting vasodilator test (intravenous epopros- as clinical comparison tools.



4. Treatment Modalities for ity of life and survival in patients with pulmonaryPulmonary Hypertension hypertension, mainly in patients with PAH.[6]

4.2.1 Calcium Channel Antagonists4.1 Conventional Supportive Therapy Calcium channel antagonists are used as the ini-

tial treatment option in children with pulmonaryThe conventional treatments for pulmonary hy- hypertension who are acute responders to the short-

pertension include diuretics (to control symptoms of acting vasodilator test (about 40% of the childrenvolume overload from right ventricular failure), sup- and the percentage of responders falls with time).[3]

plemental oxygen and digoxin (which has a benefi- The frequent occurrence of significant adverse ef-cial effect in some patients).[18] Anticoagulant ther- fects prevents the use of these drugs empirically orapy in children is based on studies in adults. Clinical in patients who are not responders.[7,19]

data on their long-term use in children is limited, butsupportive.[3]

4.2.2 Nitric OxideSyncope and intractable right-sided heart failure NO is an inhaled agent which acts as a selective

are indications for atrial septostomy in patients who and potent vasodilator of the pulmonary vascula-are treated with vasodilators, but whose pulmonary ture.[20] Inhaled NO is additional to or replacementhypertension remains refractory to treatment. In pa- of the endogenous NO. It diffuses to the pulmonarytients who do not respond to prolonged vasodilator arterial smooth muscle cells and activates solubletreatment, and in those with certain types of lesions, guanylate cyclase, resulting in an increase in cyclicsuch as pulmonary vein stenosis, lung transplanta- guanosine monophosphate (cGMP), which activatestion has been offered as the last treatment option.[7]

a cascade of events leading to smooth muscle relax-ation[3,9,20] (figure 1). Studies show that inhaled NOis safely used and effective in treating different4.2 Specific Pharmacological Therapiesforms of pulmonary hypertension, including exacer-bations of idiopathic PAH, persistent pulmonaryIn the last three decades, new and effective drugshypertension of the newborn, and pulmonary hyper-for pulmonary hypertension have emerged as a re-tension following cardiac surgery.[21,22]sult of improved understanding of the pathogenesis

of the disease. The main categories of pharmacolog-4.2.3 Prostaglandinsical agents include calcium channel antagonists, in-

haled NO, prostaglandins, endothelin receptor in- Epoprostenol (prostacyclin) has been used forhibitors and phosphodiesterase inhibitors. These two decades for the treatment of patients with PAHdrugs show improvement in exercise capacity, qual- and has shown improvement in haemodynamics,

Table IV. World Health Organization classification of pulmonary hypertension (adapted from Hoeper,[6] with permission)

Classification

Class I Patients with pulmonary hypertension but with no limitations of usual physical activityOrdinary physical activity does not cause increased dyspnoea or fatigue, chest pain or near syncope

Class II Patients with pulmonary hypertension resulting in mild limitation of physical activityThey are comfortable at restOrdinary physical activity causes increased dyspnoea or fatigue, chest pain or near syncope

Class III Patients with pulmonary hypertension resulting in marked limitation of physical activityThey are comfortable at restLess than ordinary activity causes increased dyspnoea or fatigue, chest pain or near syncope

Class IV Patients with pulmonary hypertension with inability to perform any physical activity without symptomsThese patients manifest signs of right heart failureDyspnoea and/or fatigue may even be present at restDiscomfort is increased by any physical activity



Bosentan (approved in the US and Europe) is anoral dual endothelin receptor antagonist that wasshown to improve exercise capacity, quality of lifeand cardiopulmonary haemodynamics in adult pa-tients with PAH.[24] There are recent publications onbosentan in children with PAH, suggesting that it issafely used and efficacious, and resulting inhaemodynamic improvement.[23,25]

Sitaxsentan sodium is an experimental drug thatacts as a highly selective antagonist of the A recep-tor, thereby blocking the vasoconstricting effects ofthe A receptor, while maintaining vasodilation

ProstacyclinInhaled NO

L-citrulline

L-arginineGuanylylcyclase

Sildenafil GMP

cGMP Protein kinase G

Vasodilatation

K+ channelsCa2+ channels

eNO

++

+

−

+

+

+P

DE

5

Fig. 1. Schematic illustration of the pathways involving endogenousnitric oxide (eNO) and cyclic guanosine monophosphate (cGMP)and the interaction with the pharmacological agents. GMP = gua-nosine monophosphate; NO = nitric oxide; PDE5 = phosphodiester-ase type 5.

through the B receptor.[2] The experience with thismedication in the paediatric population is very limit-quality of life and exercise capacity in these pa-ed.tients.[3,5,9]

The chemical instability of this medication at 4.2.5 Phosphodiesterase Inhibitorsneutral pH and room temperature (half-life of 1–2 PDEs are a large family of enzymes that catalyseminutes) requires a continuous intravenous delivery hydrolytic cleavage of the 3′-phosphodiester bondsystem with cold packs to maintain stability. These of the cyclic nucleotides (e.g. cyclic adenosinerequirements make the drug delivery cumbersome, monophosphate [cAMP], cGMP), controlling theirinconvenient and associated with complications. intracellular levels.[26] PDE inhibitors block this hy-Therefore, a search for alternative routes of drug drolytic cleavage, causing an accumulation ofdelivery led to the clinical investigation of oral cAMP or cGMP, which finally leads to vasodilation.(beraprost, not approved in the US or Europe), in- Nonspecific PDE inhibitors include caffeine, the-haled (iloprost, approved in the US and UK) and ophylline, dipyridamole and pentoxifylline. Milri-subcutaneous (treprostinil, not registered in Europe, none is a specific inhibitor of PDE3, whereasbut approved recently in the US for long-term ad- zaprinast and sildenafil are specific inhibitors ofministration) epoprostenol analogues. Thus far, PDE5.[15] There is a high concentration of PDE5 innone has been proven to be as efficacious as intrave- smooth muscle cells of the pulmonary vascula-nous epoprostenol.[3]

ture.[27] Therefore, specific inhibition of this enzymeenhances the accumulation of cGMP and results in4.2.4 Endothelin Receptor Antagonistspulmonary vasodilation.Endothelin-1 is a very potent vasoconstrictor and

a promoter of cell proliferation. Plasma endothelin-1 5. Sildenafillevels are increased in patients with idiopathic PAH,

Sildenafil is a potent and highly selective inhibi-both in adults and paediatric patients,[23] and aretor of PDE5.[28] Since pulmonary arterial medialcorrelated inversely with prognosis.[3] Thus, en-hypertrophy and vasoconstriction is the main mech-dothelin receptor antagonists are promising drugsanism of PAH in the paediatric population, inhib-for the treatment of PAH. To date, there are at leastiting PDE5 (thereby increasing the concentration oftwo different known receptor subtypes; A and B.[3]

secondary messenger cGMP and enhancing NO-The A receptors are localised on smooth musclemediated vasodilation) is a promising mode ofcells and mediate vasoconstriction and prolifera-targeting the medial muscular hypertrophy and in-tion,[3] while B receptors are found predominantlyducing vasodilation.[29]on endothelial cells and are associated with vasodi-

lation through the release of vasodilators and clear- Although sildenafil can be used intravenously, itance of endothelin-1.[3] is unique for its oral preparation.[28] Oral sildenafil,



20mg three times daily, was recently approved by 6. Clinical Applications of Sildenafil inPulmonary Hypertensionthe US FDA for the treatment of adults with PAH

without functional class restriction.[18] In children, it With the increased understanding of the mecha-is used as rescue therapy or in experimental proto- nisms involved in the pathogenesis of pulmonarycols. There is no recommended dosage for children hypertension, and particularly in PAH, novel treat-with PAH and the current dosages used are extrapo- ment options become available. Sildenafil repre-lated from the adult dose range. Nevertheless, dos- sents one of the most recently recognised drugs withages of 0.5–2.0 mg/kg/dose might be considered as a significant therapeutic potential in pulmonary hy-therapeutic in children, as the plasma concentrations pertension.[33] This drug has been used in various

pathologies associated with PAH in children, andmeasured 1 hour after oral ingestion were similar tothe outcome reported in a number of comparativethe peak plasma concentrations achieved in adultsstudies, case series and case reports.after a single dose of 25–100mg, and the patients

demonstrated clinical improvement.[30,31]

6.1 Idiopathic and Familial PAHSildenafil is rapidly absorbed after oral adminis-

Fewer than 50% of patients with idiopathic PAHtration, with absolute bioavailability of aboutare found to be responsive to the vasodilator test and40%.[32] In healthy adults, maximum observed plas-these patients will benefit from calcium channelma concentrations are reached within 30–120 min-antagonist therapy. In all other patients, intravenousutes (median 60 minutes) of oral administration ininfusion of epoprostenol is the mainstay of ther-

the fasted state.[6,32] Karatza and colleagues showedapy.[3] However, continuous intravenous infusion is

that the maximum serum concentrations of the drugnot the optimal modality in the paediatric population

in children are reached an hour after administration and is associated with potential complications suchand are dose-dependent.[31] Sildenafil is eliminated as line infections, catheter dislodgement and pumppredominantly by hepatic microsomal isoenzymes malfunction.[3] For these reasons, oral sildenafil of-of cytochrome P450 (CYP) [3A4 (major route) and fers an attractive and effective alternative.[28,34]

CYP2C9 (minor route)]. It is converted to N- Several studies and case reports were publisheddesmethyl sildenafil, which is an active metabolite recently, investigating the effect of oral sildenafil inwith properties similar to sildenafil. Plasma concen- patients with PAH[10,35-39] (table V). Kothari and

Duggal[35] conducted a prospective, uncontrolledtrations of this metabolite account for about 20% ofstudy in nine children aged 5–18 years; six hadthe pharmacological effects of sildenafil. Bothprimary pulmonary hypertension, and three had sec-sildenafil and its major circulating N-desmethyl me-ondary causes. In all six children with primary pul-tabolite are almost completely bound to plasma pro-monary hypertension, who were graded as NYHAteins (96%). Protein binding is independent of totalfunctional class 3 or 4, sildenafil was added todrug concentrations. Both sildenafil and its metabo-conventional therapy. The study showed a remarka-lite have terminal half-lives of about 4 hours.[6] Theble and sustained benefit of sildenafil in these pa-

concomitant use of potent CYP3A4 inhibitors (e.g.tients. The improvement was subjective as well as

erythromycin, ketoconazole, itraconazole), as well objective, as reflected in the NYHA functional classas the nonspecific inhibitor of CYP, cimetidine, is and 6-MWT. The authors found that the mean pul-associated with increased plasma sildenafil concen- monary artery pressure decreased in three of thetrations. After either oral or intravenous administra- patients who underwent re-catheterisation after ation, sildenafil is excreted as metabolites predomi- mean period treatment with sildenafil of 7 months.nantly in the faeces (approximately 80% of adminis- In all patients, sildenafil was well tolerated and wastered oral dose) and to a lesser extent in the urine associated with only minor adverse effects. Sastry et(approximately 13% of the administered oral dose). al.[36] evaluated the efficacy of sildenafil in six chil-


Sildenafil C

itrate in Paediatric Pulm

onary Hypertension

65

2007 A

dis D

ata

Info

rma

tion

BV. A

ll righ

ts rese

rved

.D

rug

s 2007; 67 (1)

Table V. Summary of studies on sildenafil citrate in paediatric patients with pulmonary arterial hypertension (PAH)

Study Study type No. of Age (y) Route of Main indications Treatment Outcome Follow-up Adverse reactions(year) pts administration duration (mo) period (mo)

and dosage

Kothari & Prospective, 6 5–18 Oral, 5.5 mg/ Primary 3.5–8 Improved NYHA functional 3.5–8 Dizziness, flushing and

Duggal[35] uncontrolled kg/day– 150 pulmonary class headache (when dose

(2002) mg/day hypertension Improved 6-MWT >125mg)

(divided in 3 Improved haemodynamics No visual symptoms

doses) No inconvenience due to

erectile effect

Sastry et Prospective, 6 4–16 Oral, 25–100 Primary 5–20 Improved NYHA functional 5–20 Minor headache, flushing,

al.[36] uncontrolled mg/dose pulmonary class abdominal discomfort

(2002) (8-hourly) hypertension Improved 6-MWT

Improved haemodynamics

Oliveira & Prospective, 6 3–19 Oral, 2–8 mg/ Idiopathic PAH 4–36 Improved NYHA functional 4–36 No adverse effects

Amaral[37] uncontrolled kg/day or class One death after abrupt and

(2005) 100–500 mg/ Improved systemic unplanned withdrawal of

day (4–6 saturation medication

doses/day)

Humpl et Prospective, 4 5–18 Oral, 0.25–1 Idiopathic PAH 12 Improved 6-MWT 12 Self-limited nosebleeds,

al.[38] uncontrolled mg/kg/dose Improved haemodynamics menstrual losses with the

(2005) (6-hourly) menarche, facial flushing,

headache and dizziness

Abrams et Case report 1 4 Oral, 2 mg/kg Primary 3 Improved exercise 3 None

al.[10] (4-hourly) pulmonary capacity

(2000) hypertension

Karatza et Case report 1 14 Oral, 0.5 mg/ Primary 6 Increased exercise 6 None

al.[39] kg, increased pulmonary capacity

(2004) to 2 mg/kg hypertension Increased oxygen

(4-hourly) saturation

NYHA = New York Heart Association; pts = patients; 6-MWT = 6-minute walk test.

66Leibovitch et al.

2007 A

dis D

ata

Info

rma

tion

BV. A

ll righ

ts rese

rved

.D

rug

s 2007; 67 (1)

Table VI. Summary of studies on sildenafil citrate in paediatric patients with congenital heart disease and pulmonary arterial hypertension (PAH)

Study (year) Study type No. of Age (y) Route of Main indications Treatment Outcome Follow-up Adverse reactionspts administration and duration period

dosage

Schulze-Neick Prospective, 12 0.2–15.7 IV, 1 mg/kg divided in Preoperative Stat doses Decreased pulmonary No follow-up No significant adverseet al.[40] uncontrolled 2 unequal doses, catheterisation resistance reactions(2003) each over 10 min

12 0.11–0.65 IV, 0.025, 0.1 & 0.25 Postoperative PAH Stat doses Decreased pulmonary No follow-up No significant adversemg/kg (in 10–15 min resistance reactionssteps) Increased

intrapulmonaryshunting

Stocker et al.[41] Prospective, 15 0.11–0.7 IV, 0.35 mg/kg over Stable infants at risk 20 min Sildenafil augmented No follow-up Sildenafil produced(2003) randomised 20 min + iNO of PAH, early after single dose pulmonary vasodilator systemic hypotension

cardiac surgery for effect of iNO and impairedVSD or ASD oxygenation

Humpl et al.[38] Prospective, 10 6–18 Oral, 0.25–1 mg/kg/ PAH persisted after 12mo Improved 6-MWT 12mo Self-limited(2005) uncontrolled dose (6-hourly) correction of Improved nosebleeds,

congenital heart haemodynamics menstrual losses withdisease, or if the menarche, facialconsidered inoperable flushing, headache

and dizziness

Kothari & Prospective, 3 5–18 Oral, 5.5 mg/kg/day– After cardiac surgery 3.5–8mo Improved NYHA 3.5–8mo Dizziness, flushingDuggal[35] uncontrolled 150 mg/day (divided functional class and headache (when(2002) in 3 doses) Improved 6-MWT dose >125mg)

Improved No visual symptomshaemodynamics No inconvenience due

to erectile effect

Karatza et al.[31] Case 2 6.5, 10.5 Oral, 0.5–2 mg/kg/ Pulmonary 6mo Increased exercise 6mo No adverse effects(2005) reports dose (4-hourly) hypertension capacity and noted

associated with increased saturationcongenital heartdisease

Carroll & Case report 1 7 Oral, 0.5–2 mg/kg/ Pulmonary Unknown Significant Unknown NoneDhillon[28] dose (6-hourly) hypertension improvement of(2003) associated with exercise test

congenital heartdisease

Continued next page


dren (aged 4–16 years) with primary pulmonaryhypertension. They found an overall improvementin the echocardiographic and haemodynamic mea-surements in all patients, as well as improvements inthe NYHA functional class and the 6-MWT. Theirstudy also suggested a possible survival advantage,but this could not be confirmed based on their studydesign and small group of patients.

In a more recent prospective case series, Oliveiraand Amaral[37] reported six patients aged 3–19 yearswith idiopathic PAH who had no response to previ-ous conventional treatment. They were treated withoral sildenafil (2–8 mg/kg/day) during a period of4–36 months. All patients achieved a good therapeu-tic response, with improvement by at least one func-tional class, and an increase in systemic arterialoxygen saturation. No major adverse effects wereobserved, but one patient died suddenly after un-planned withdrawal of sildenafil. Humpl et al.[38]

conducted a 12-month single-drug, open-labelclinical trial using oral sildenafil in 14 children; 4 ofthem with idiopathic PAH. Similar to the previousstudies, this study also found an improvement inboth exercise tolerance and pulmonary vascularhaemodynamics. No adverse effects were reported.Although their study was not controlled, the au-thors’ feeling was that these patients, treated withsildenafil, had a better survival curve than patientswho were treated in their institution in the past.

Although all these studies are relatively smallscale and uncontrolled, they suggest that sildenafilmay be useful in the management of idiopathic PAHin the paediatric population. Large randomised andcontrolled studies will enable definite conclusionsand optimal dosage recommendations.

6.2 PAH Associated with CongenitalHeart Disease

There are several possible applications forsildenafil in children with congenital heart disease.In the immediate postoperative phase, sildenafil canbe used to decrease the pulmonary arterial pressure.Another application is in patients who develop PAHassociated with congenital heart disease, later in thecourse of the disease[28,31,35,38,40-44] (table VI).


Tab

le V

I. C

ontd

Stu

dy (

year

)S

tudy

typ

eN

o. o

fA

ge (

y)R

oute

of

Mai

n in

dica

tions

Tre

atm

ent

Out

com

eF

ollo

w-u

pA

dver

se r

eact

ions

pts

adm

inis

trat

ion

and

dura

tion

perio

ddo

sage

Say

gili

et a

l.[42]

Cas

e re

port

19

Ora

l, 0.

75 m

g/kg

/dos

eP

AH

ass

ocia

ted

with

3wk

Ena

bled

to

wea

n iN

O12

mo

Non

e(2

004)

(6-h

ourly

)V

SD

cor

rect

ion

and

pulm

onar

y ar

tery

pres

sure

dec

reas

ed

Kul

karn

i et

Cas

e re

port

14

Ora

l, 0.

25–

0.5

mg/

kg/

Afte

r he

art

8wk

Dec

reas

ed p

ulm

onar

y8w

kN

one

al.[4

3]do

se (

4-ho

urly

)tr

ansp

lant

atio

n fo

rpr

essu

res

(200

4)di

late

dca

rdio

myo

path

y

Kno

dere

r et

Cas

e re

port

11

Ora

l, 0.

25–0

.5 m

g/kg

Pul

mon

ary

11m

oN

orm

al p

ulm

onar

y3y

Non

eal

.[44]

(6-h

ourly

)hy

pert

ensi

onpr

essu

res

and

gain

ing

(200

5)as

soci

ated

with

wei

ght

cong

enita

l hea

rtdi

seas

e

AS

D =

atr

ial s

epta

l def

ect;

iNO

= in

hale

d ni

tric

oxi

de;

IV =

intr

aven

ous;

NY

HA

= N

ew Y

ork

Hea

rt A

ssoc

iatio

n; p

ts =

pat

ient

s; V

SD

= v

entr

icul

ar s

epta

l def

ect;

6-M

WT

= 6

-min

ute

wal

k te

st.


Intravenous sildenafil has been used in pre-cardi- ic vasodilation. The investigators terminated thestudy early, because they noticed a consistent deteri-ac surgery catheterisation to evaluate the reversibili-oration in oxygenation and increased alveolar-arteri-ty of PAH. Schulze-Neick et al.[40] compared theal gradient, which indicated an increase in intrapul-effects of intravenous sildenafil with inhaled NO inmonary shunting in the post-bypass lung.preoperative and postoperative patients with con-

Humpl et al.[38] performed a 12-month, open-genital heart disease and elevated pulmonary vascu-label clinical trial of oral sildenafil in 10 childrenlar resistance (table VI). They showed that intrave-with PAH secondary to congenital heart diseasenous sildenafil may be equal or superior to a stan-(table VI). They found that sildenafil was well toler-dard dose of inhaled NO in reducing the elevatedated and had no haematological, renal or hepaticpulmonary vascular resistance in these patients, bothfunctional impairments, as well as no vision-relatedduring routine cardiac catheterisation and afterabnormalities. Sildenafil was found to improve bothopen-heart surgery. They found no significant effectexercise tolerance and pulmonary vascular haemo-of sildenafil on intrapulmonary shunting in the pre-dynamics. Kothari and Duggal[35] reported three pa-operative group who underwent cardiac catheterisa-tients who underwent surgery for congenital hearttion. In the postoperative mechanically ventilateddisease (ventricular septal defect or patent ductuspatients, the measured intrapulmonary shunting wasarteriosus); all had severe pulmonary hypertensionstatistically higher in the sildenafil-treated patients;and received oral sildenafil in addition to conven-however, this was not clinically significant. This istional therapy. The authors concluded that the studypresumably related to vasodilation of pulmonaryshowed a remarkable and sustained benefit ofarterioles supplying non-ventilated areas of lung, asildenafil, and it was well tolerated.well described complication of intravenous, non-

These studies show that the use of an efficaciousselective pulmonary vasodilation. These findingsoral pulmonary vasodilator, such as sildenafil, mayare in contrast to those with inhaled NO, where itshave a role in the clinical setting of PAH associatedeffect is localised only to adequately ventilated areaswith congenital heart disease. It may allow weaningof lung.[40]

of intravenous pulmonary vasodilator support andPulmonary endothelial dysfunction is failure of can be administered conveniently for several days,

the endothelium to produce adequate amounts of weeks or months, and result in improvement in bothendogenous NO.[45] Open heart surgery with car- exercise tolerance and pulmonary vascular haemo-diopulmonary bypass is known to amplify this fail- dynamics.ure,[45] and inhaled NO is an important therapeuticmodality in these patients when they return from 6.3 Pulmonary Hypertension Associated withsurgery. Stocker et al.[41] conducted a prospective, Lung Diseases and/or Hypoxaemiarandomised trial with intravenous sildenafil and in-haled NO in infants after cardiac surgery. They Chronic lung disease in children can be compli-investigated the acute effects of intravenous silde- cated by severe chronic pulmonary hypertension.[28]

nafil on haemodynamics and oxygenation, and its Sildenafil can offer a possible therapeutic option ininteraction with inhaled NO in stable infants at risk these patients because of its effect on the pulmonaryof pulmonary hypertension, early after cardiac sur- vasculature, which is independent of the underlyinggery with cardiopulmonary bypass. Their study re- cause.[28] There are several single case reports in thevealed very similar conclusions to Schulze-Neick et literature that describe the use of sildenafil in vari-al.,[40] showing that intravenous sildenafil reduced ous lung pathologies such as interstitial pneumoni-

tis, pertussis, bronchopulmonary dysplasia and dia-pulmonary vascular resistance and enhanced thephragmatic hernia.pulmonary vasodilator effects of NO. They also

demonstrated that intravenous sildenafil significant- Carroll and Dhillon published their experiencely reduced the systemic blood pressure with system- with oral sildenafil in two infants with chronic pul-



monary hypertension secondary to pneumonitis.[28] 6.4 Sildenafil and Inhaled Nitric OxideOne infant was treated with sildenafil as palliative

Severe rebound hypoxaemia and pulmonary hy-treatment to allow him some time at home. In the pertension are significant problems when weaningsecond patient, sildenafil was commenced because inhaled NO in some patients, mainly after prolongedintravenous prostacyclin was impractical for home inhaled NO therapy and abrupt cessation of treat-use. He showed improvement with no significant ment.[15] This clinical issue is frequently seen in theadverse effects. However, both patients died from paediatric population, and is also well defined inprogression of their underlying pulmonary disease. newborns with persistent pulmonary hypertension.McEniery et al.[46] described an infant with pertussis Atz and Wessel[50] hypothesised that sildenafilwho developed pulmonary hypertension. Bordetella may add to the pulmonary vasodilation effect of

inhaled NO or blunt the rebound effect of abruptpertussis produces a number of active toxins thatdiscontinuation of inhaled NO. They reported theirimpair cardiovascular function and inhibit endothe-experience in three infants (aged 1 day, 6 weeks andlial NO production; hence, pertussis is associated4 months) with pulmonary hypertension followingwith pulmonary hypertension. This patient wassurgery for congenital heart disease. Sildenafil wastreated with a combination of sildenafil and inhaledused after failed attempts to discontinue inhaled NO,NO, and his haemodynamic condition wasand resulted in a rise in cGMP levels within 90

stabilised. The authors postulated that this drugminutes of administration, with an associated drop

combination may offer potential treatment in pa-in pulmonary arterial pressure, allowing successful

tients with pertussis toxaemia. Hon et al.[47] reported weaning from NO. They found no adverse systemicthe use of oral sildenafil in a 5-month-old preterm effects. In another recent study,[15] oral sildenafilinfant with severe bronchopulmonary dysplasia and was used in five neonates and younger infants (agePAH refractory to inhaled NO. The infant was treat- range 2–28 weeks) with refractory suprasystemiced with oral sildenafil for 6 months until complete pulmonary hypertension after gradual withdrawal ofresolution of PAH and oxygen supplement was inhaled NO, despite alkalinisation and inotropicweaned off. There were no adverse effects during support. Sildenafil permitted discontinuation of in-

haled NO within 4–6 hours, without haemodynamicthe treatment period. Chaudhari et al.[48] reported oninstability, allowing extubation in four of them with-a neonate with severe pulmonary hypertension asso-in 48 hours. Mychaskiw et al.[51] reported a 17-year-ciated with impaired alveolarisation and plexiformold patient who had replacement of a biventricularpulmonary arteriopathy. The neonate was treatedassist device to support heart failure secondary to awith oral sildenafil and inhaled NO, and this resultedviral cardiomyopathy. In this case, sildenafil ame-in significant clinical improvement with recoveryliorated the rebound pulmonary hypertension that

from the pulmonary hypertensive crisis.developed following withdrawal of inhaled NO.

Animal models may suggest another role of The fact that sildenafil seems to be as effective assildenafil in chronic lung disease. Ladha et al.[49]

inhaled NO in improving pulmonary vasodilation,demonstrated that, in addition to its known vasodi- and may even have an additive effect,[52-54] makes itlatory effect, sildenafil also preserved alveolar a useful tool in weaning patients after prolongedgrowth and lung angiogenesis in rats. These findings therapy with inhaled NO.suggest a role for the NO or cGMP pathway during

6.5 Persistent Pulmonary Hypertension ofalveolar development. If this hypothesis is validat-the Newborned, it might explain the beneficial effect of long-

term treatment with sildenafil in patients with Inhaled NO is currently regarded as the goldbronchopulmonary dysplasia, impaired alveolar standard therapy for pulmonary hypertension of thestructures and other parenchymal lung diseases. newborn, whereas the extracorporal membrane oxy-



genation (ECMO) is considered as a rescue ther- drome, intravenously administered sildenafil com-pletely reversed the increased pulmonary vascularapy.[15] However, there are some difficulties whichresistance without affecting systemic haemodynam-limit the use of NO: about 30% of the patients do notics.[55] These experimental models also showed thatrespond to this drug,[15] there is a risk of developingwhen sildenafil was added to inhaled NO it pro-rebound pulmonary hypertension upon its with-duced unacceptable deterioration in oxygenation, asdrawal, even after relatively short periods of use,a result of an increase in intrapulmonary shunt-and its mode of delivery usually requires endotra-ing.[60] These models point to a possibly higher riskcheal intubation.[15] Other important considerationsinvolved in using sildenafil in parenchymal lungare the development of methaemoglobinaemia anddisease, such as meconium aspiration syndrome. Incell membrane damage from peroxynitrites.[15] Thesuch situations, sildenafil might induce or worsenhigh cost involved and the lack of infrastructure inventilation perfusion mismatch, resulting in lowersome parts of the world makes NO and ECMO evenarterial oxygenation. Alternative administration inless available. These limitations led to further stud-an aerosolised form may lower the risk of ventila-ies on the possible role of sildenafil, which mighttion perfusion mismatch and the systemic adverseovercome some of the difficulties mentioned.effects.[59] It is also important to remember that

Juliana and Abbad[55] published a successful caseinfants are at risk for sepsis and septic shock and

of sildenafil treatment for persistent pulmonary hy-developing associated pulmonary hypertension. In-

pertension in an infant, where inhaled NO or ECMOflammatory mediators that are produced during the

were not available. Just recently, a pilot randomised,septic state increase the concentrations of endoge-

blinded study on oral sildenafil in infants with per-nous NO, which cause changes in systemic vasomo-

sistent pulmonary hypertension of the newborn wastor tone.[61] In these circumstances, sildenafil may

published.[56] This study was performed in a neona-theoretically worsen the circulatory status.

tal intensive care unit in Colombia. It was wellequipped but there was no inhaled NO, high fre-

7. Adverse Eventsquency oscillating ventilators or ECMO available.Seven infants were treated with oral sildenafil 1 mg/ Most of the studies and the reports on sildenafilkg every 6 hours, whereas six infants received place- in children and infants with pulmonary hypertensionbo. The main outcome variable was the effect of oral suggest that there is a very low incidence of adversesildenafil on oxygenation. Oxygenation index im- effects, and these are usually minor. Adverse effectsproved in all infants within 6–30 hours. Sildenafil that were reported in adults treated for pulmonarywas well tolerated. In the treatment group no notice- hypertension include headache, flushing, dizziness,able effect on blood pressure was observed. Six postural hypotension, dyspepsia, constipation, back-infants survived in the treatment group, whereas ache, blurred vision, giddiness and short-lived erec-only one infant survived in the placebo group. The tions.[34] In most of the paediatric patients reviewedstudy showed that sildenafil may be effective in the here, no major adverse effects were noted. However,treatment of persistent pulmonary hypertension of a few major concerns should be considered beforethe newborn. treating patients with pulmonary hypertension with

Sildenafil has been studied in animal models of sildenafil. Intravenous sildenafil was found to sig-neonatal pulmonary hypertension. It was shown to nificantly reduce the systemic blood pres-be a selective pulmonary vasodilator with no effect sure.[40,41,60] In some cases, the drop in systemicon the systemic arterial pressure, potentiating the blood pressure was corrected with volume expand-effects of inhaled NO when given orally,[57] as an ers,[55] whereas in others it was not clinically signifi-intravenous infusion[58] or in an aerosolised form.[59] cant and did not require any specific treatment.[38]

In an experimental model of pulmonary hyperten- Another important consideration is the deteriorationsion associated with meconium aspiration syn- in arterial oxygenation and the increase in alveolar-



arterial gradient, which is probably due to an inefficacy, safety and optimal dosage of sildenafil ascrease in intrapulmonary shunting.[40,41,58] This can treatment of pulmonary hypertension in the paedia-be unacceptable in infants with parenchymal lung tric population.disease and low reserves.

AcknowledgementsVisual and ocular abnormalities were also report-ed with the use of sildenafil in pulmonary hyperten-

We thank Professor Sherin Devaskar for reviewing oursion. Blurred vision, increased perception of lightmanuscript and for her insightful comments. The authors

and blue-green colour-tinged vision are described as have no conflicts of interest that are directly relevant to thepossible adverse effects.[45] In premature babies, re- contents of this review and received no funding for its prepa-

ration.tinopathy of prematurity (ROP) is a concern, butonly one case report suggested that the use ofsildenafil induced severe ROP in a preterm baby Referenceswho had severe pulmonary hypertension.[62] Howev- 1. Haworth SG. Primary pulmonary hypertension in childhood.

Arch Dis Child 1998; 79: 452-5er, that baby had many other risk factors for the2. D’Alonzo GE, Barst RJ, Ayres SM, et al. Survival in patients

development of ROP, as was discussed by Pierce et with primary pulmonary hypertension. Results from a nationalprospective registry. Ann Intern Med 1991; 115: 343-9al.[63]

3. Rosenzweig EB, Widlitz AC, Barst RJ. Pulmonary arterialhypertension in children. Pediatr Pulmonol 2004; 38: 2-22

8. Conclusions 4. Simonneau G, Galie N, Rubin LJ, et al. Clinical classification ofpulmonary hypertension. J Am Coll Cardiol 2004; 43: 5-12S

5. Rosenzweig EB, Barst RJ. Idiopathic pulmonary arterial hyper-Pulmonary hypertension in the paediatric popula-tension in children. Curr Opin Pediatr 2005; 17: 372-80

tion remains an incurable disease, but the recent 6. Hoeper MM. Drug treatment of pulmonary arterial hyperten-advances in understanding the mechanisms involved sion. Drugs 2005; 65: 1337-54

7. Widlitz A, Barst RJ. Pulmonary arterial hypertension in chil-in the pathogenesis of the disease are an importantdren. Eur Respir J 2003; 21: 155-76

step in developing new and effective therapies.[6]8. Rubin LJ. Primary pulmonary hypertension. N Engl J Med

1997; 336: 111-7These newly available medications improve exer-9. Ivy D. Diagnosis and treatment of severe pediatric pulmonarycise capacity, quality of life and survival.[6]

hypertension. Cardiol Rev 2001; 9: 227-37Sildenafil, which is considered an experimental 10. Abrams D, Schulze-Neick I, Magee AG. Sildenafil as a selective

pulmonary vasodilator in childhood primary pulmonary hyper-drug for children with pulmonary hypertension, ap-tension. Heart 2000; 84: E4pears as one of the most promising pulmonary vaso- 11. Loyd JE, Butler MG, Foroud TM, et al. Genetic anticipation and

dilator agents; it is relatively safe and simple to abnormal gender ratio at birth in familial primary pulmonaryhypertension. Am J Respir Crit Care Med 1995; 152: 93-7use,[52] and the effect it has on the pulmonary vascu-

12. Lane KB, Machado RD, Pauciulo MW, et al. Heterozygouslature is independent of the underlying cause. There germline mutations in BMPR2, encoding a TGF-beta receptor,

cause familial primary pulmonary hypertension. The Interna-are reports of successful use of sildenafil in childrentional PPH Consortium. Nat Genet 2000; 26: 81-4with idiopathic PAH, PAH that results from congen-

13. Deng Z, Morse JH, Slager SL, et al. Familial primary pulmonaryital heart disease, or pulmonary hypertension secon- hypertension (gene PPH1) is caused by mutations in the bone

morphogenetic protein receptor-II gene. Am J Hum Genetdary to lung disease; it may also be beneficial in2000; 67: 737-44

postoperative pulmonary hypertension and in pa- 14. Tulloh RM. Congenital heart disease in relation to pulmonarytients who are difficult to wean from inhaled hypertension in paediatric practice. Paediatr Respir Rev 2005;

6: 174-80NO.[45,50,51]15. Travadi JN, Patole SK. Phosphodiesterase inhibitors for persis-

Although the data from the literature are promis- tent pulmonary hypertension of the newborn: a review. PediatrPulmonol 2003; 36: 529-35ing, they are mainly from small case series and

16. Greenough A, Khetriwal B. Pulmonary hypertension in thesingle case reports. Treatment regimens were not newborn. Paediatr Respir Rev 2005; 6: 111-6uniform, the dosage of sildenafil varied between 17. Rich S, editor. Primary pulmonary hypertension: executive

summary from the World Symposium–Primary Pulmonarystudies, and the follow-up period was not alwaysHypertension 1998. Geneva: World Health Organization, 1998

sufficient. Therefore, large-scale randomised, con- 18. Lee SH, Rubin LJ. Current treatment strategies for pulmonaryarterial hypertension. J Intern Med 2005; 258: 199-215trolled clinical trials are required to confirm the



19. Packer M, Medina N, Yushak M, et al. Detrimental effects of 39. Karatza AA, Narang I, Rosenthal M, et al. Treatment of primaryverapamil in patients with primary pulmonary hypertension. pulmonary hypertension with oral sildenafil. Respiration 2004;Br Heart J 1984; 52: 106-11 71: 192-4

20. Frostell C, Fratacci MD, Wain JC, et al. Inhaled nitric oxide. A 40. Schulze-Neick I, Hartenstein P, Li J, et al. Intravenous sildenafilselective pulmonary vasodilator reversing hypoxic pulmonary is a potent pulmonary vasodilator in children with congenitalvasoconstriction. Circulation 1991; 83: 2038-47 heart disease. Circulation 2003; 108 Suppl. 1: II167-73

21. The Neonatal Inhaled Nitric Oxide Study Group. Inhaled nitric 41. Stocker C, Penny DJ, Brizard CP, et al. Intravenous sildenafiloxide in full-term and nearly full-term infants with hypoxic and inhaled nitric oxide: a randomised trial in infants afterrespiratory failure. N Engl J Med 1997; 336: 597-604 cardiac surgery. Intensive Care Med 2003; 29: 1996-2003

22. Clark RH, Kueser TJ, Walker MW, et al. Low-dose nitric oxide 42. Saygili A, Canter B, Iriz E, et al. Use of sildenafil with inhaledtherapy for persistent pulmonary hypertension of the newborn. nitric oxide in the management of severe pulmonary hyperten-Clinical Inhaled Nitric Oxide Research Group. N Engl J Med sion. J Cardiothorac Vasc Anesth 2004; 18: 775-62000; 342: 469-74 43. Kulkarni A, Singh TP, Sarnaik A, et al. Sildenafil for pulmonary

23. Barst RJ, Ivy D, Dingemanse J, et al. Pharmacokinetics, safety, hypertension after heart transplantation. J Heart Lung Trans-and efficacy of bosentan in pediatric patients with pulmonary plant 2004; 23: 1441-4arterial hypertension. Clin Pharmacol Ther 2003; 73: 372-82 44. Knoderer CA, Ebenroth ES, Brown JW. Chronic outpatient

24. Rubin LJ, Badesch DB, Barst RJ, et al. Bosentan therapy for sildenafil therapy for pulmonary hypertension in a child afterpulmonary arterial hypertension. N Engl J Med 2002; 346: cardiac surgery. Pediatr Cardiol 2005; 26: 859-61896-903 45. Raja SG, Nayak SH. Sildenafil: emerging cardiovascular indica-

25. Rosenzweig EB, Ivy DD, Widlitz A, et al. Effects of long-term tions. Ann Thorac Surg 2004; 78: 1496-506bosentan in children with pulmonary arterial hypertension. J 46. McEniery JA, Delbridge RG, Reith DM. Infant pertussis deathsAm Coll Cardiol 2005; 46: 697-706 and the management of cardiovascular compromise. J Paediatr

26. Essayan DM. Cyclic nucleotide phosphodiesterases. J Allergy Child Health 2004; 40: 230-2Clin Immunol 2001; 108: 671-80

47. Hon KL, Cheung KL, Siu KL, et al. Oral sildenafil for treatment27. Rabe KF, Tenor H, Dent G, et al. Identification of PDE of severe pulmonary hypertension in an infant. Biol Neonate

isozymes in human pulmonary artery and effect of selective 2005; 88: 109-12PDE inhibitors. Am J Physiol 1994; 266: 536-43

48. Chaudhari M, Vogel M, Wright C, et al. Sildenafil in neonatal28. Carroll WD, Dhillon R. Sildenafil as a treatment for pulmonary pulmonary hypertension due to impaired alveolarisation and

hypertension. Arch Dis Child 2003; 88: 827-8 plexiform pulmonary arteriopathy. Arch Dis Child Fetal Neo-29. Galie N, Ghofrani HA, Torbicki A, et al., for the Sildenafil Use natal Ed 2005; 90: 527-8

in Pulmonary Arterial Hypertension (SUPER) Study Group.49. Ladha F, Bonnet S, Eaton F, et al. Sildenafil improves alveolar

Sildenafil citrate therapy for pulmonary arterial hypertension.growth and pulmonary hypertension in hyperoxia-induced

N Engl J Med 2005; 353: 2148-57lung injury. Am J Respir Crit Care Med 2005; 172: 750-6

30. Boolell M, Allen MJ, Ballard SA, et al. Sildenafil: an orally50. Atz AM, Wessel DL. Sildenafil ameliorates effects of inhaledactive type 5 cyclic GMP-specific phosphodiesterase inhibitor

nitric oxide withdrawal. Anesthesiology 1999; 91: 307-10for the treatment of penile erectile dysfunction. Int J Impot Res51. Mychaskiw G, Sachdev V, Heath BJ. Sildenafil (Viagra) facili-1996; 8: 47-52

tates weaning of inhaled nitric oxide following placement of a31. Karatza AA, Bush A, Magee AG. Safety and efficacy ofbiventricular-assist device. J Clin Anesth 2001; 13: 218-20sildenafil therapy in children with pulmonary hypertension. Int

52. Michelakis E, Tymchak W, Lien D, et al. Oral sildenafil is anJ Cardiol 2005; 100: 267-73effective and specific pulmonary vasodilator in patients with32. Nichols DJ, Muirhead GJ, Harness JA. Pharmacokinetics ofpulmonary arterial hypertension: comparison with inhaled ni-sildenafil after single oral doses in healthy male subjects:tric oxide. Circulation 2002; 105: 2398-403absolute bioavailability, food effects and dose proportionality.

53. Watanabe H, Ohashi K, Takeuchi K, et al. Sildenafil for primaryBr J Clin Pharmacol 2002; 53 Suppl. 1: 5-12Sand secondary pulmonary hypertension. Clin Pharmacol Ther33. Zaiman A, Fijalkowska I, Hassoun PM, et al. One hundred years2002; 1: 398-402of research in the pathogenesis of pulmonary hypertension.

54. Ghofrani HA, Wiedemann R, Rose F, et al. Combination ther-Am J Respir Cell Mol Biol 2005; 33: 425-31apy with oral sildenafil and inhaled iloprost for severe pulmo-34. Sastry BK, Narasimhan C, Reddy NK, et al. Clinical efficacy ofnary hypertension. Ann Intern Med 2002; 136: 515-22sildenafil in primary pulmonary hypertension: a randomized,

55. Juliana AE, Abbad FC. Severe persistent pulmonary hyperten-placebo-controlled, double-blind, crossover study. J Am Collsion of the newborn in a setting where limited resourcesCardiol 2004; 43: 1149-53exclude the use of inhaled nitric oxide: successful treatment35. Kothari SS, Duggal B. Chronic oral sildenafil therapy in severewith sildenafil. Eur J Pediatr 2005; 164: 626-9pulmonary artery hypertension. Indian Heart J 2002; 54: 404-9

56. Baquero H, Soliz A. Neira F, et al. Oral sildenafil in infants with36. Sastry BK, Narasimhan C, Reddy NK, et al. A study of clinicalpersistent pulmonary hypertension of the newborn: a pilotefficacy of sildenafil in patients with primary pulmonary hy-randomized blinded study. Pediatrics 2006; 117: 1077-83pertension. Indian Heart J 2002; 54: 410-4

57. Weimann J, Ullrich R, Hromi J, et al. Sildenafil is a pulmonary37. Oliveira EC, Amaral CF. Sildenafil in the management ofvasodilator in awake lambs with acute pulmonary hyperten-idiopathic pulmonary arterial hypertension in children andsion. Anesthesiology 2000; 92: 1702-12adolescents. J Pediatr (Rio J) 2005; 81: 390-4

38. Humpl T. Reyes JT, Holtby H, et al. Beneficial effect of oral 58. Shekerdemian LS, Ravn HB, Penny DJ. Intravenous sildenafilsildenafil therapy on childhood pulmonary arterial hyperten- lowers pulmonary vascular resistance in a model of neonatalsion: twelve-month clinical trial of a single-drug, open-label, pulmonary hypertension. Am J Respir Crit Care Med 2002;pilot study. Circulation 2005; 111: 3274-80 165: 1098-102



59. Ichinose F, Erana-Garcia J, Hromi J, et al. Nebulized sildenafil 63. Pierce CM, Petros AJ, Fielder AR. No evidence for severeis a selective pulmonary vasodilator in lambs with acute pul- retinopathy of prematurity following sildenafil. Br Jmonary hypertension. Crit Care Med 2001; 29: 1000-5 Ophthalmol 2005; 89: 250

60. Shekerdemian LS, Ravn HB, Penny DJ. Interaction betweeninhaled nitric oxide and intravenous sildenafil in a porcine

Correspondence and offprints: Dr Leah Leibovitch, Divisionmodel of meconium aspiration syndrome. Pediatr Res 2004;of Neonatology, Safra Children’s Hospital, The Chaim She-55: 413-8

61. Symeonides S, Balk RA. Nitric oxide in the pathogenesis of ba Medical Center, Tel Hashomer, 52621, Israel.sepsis. Infect Dis Clin North Am 1999; 13: 449-63 E-mail: [email protected]

62. Marsh CS, Marden B, Newsom R. Severe retinopathy of prema-turity (ROP) in a premature baby treated with sildenafil acetate(Viagra) for pulmonary hypertension. Br J Ophthalmol 2004;88: 306-7




Ocular Adverse Effects Associatedwith Systemic MedicationsRecognition and Management

Ricardo M. Santaella and Frederick W. Fraunfelder

Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, USA

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751. Categorising Adverse Drug-Related Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772. Medications and Adverse Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

2.1 Bisphosphonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792.2 Antiepileptic Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

2.2.1 Topiramate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812.2.2 Vigabatrin and Tiagabine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

2.3 Isotretinoin and Other Retinoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832.4 Ethambutol and Isoniazid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832.5 Amiodarone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842.6 Hydroxychloroquine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852.7 Erectile Dysfunction Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872.8 Tamoxifen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882.9 Cyclo-Oxygenase-2 Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882.10 Nicotinic Acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892.11 Herbal Medications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

2.11.1 Canthaxanthine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.2 Chamomile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.3 Datura . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.4 Echinacea purpurea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.5 Ginkgo biloba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.6 Liquorice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.7 Vitamin A (Retinol) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902.11.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

This article reviews several retrospective case series and reported adverseAbstractevents regarding common ocular adverse effects related to systemic therapy. It isnot intended as a comprehensive summary of these well described adverse drugreactions, nor is it intended to cover the complete spectrum of all ocular adverseeffects of systemic therapy. Many systemic drugs may produce ocular toxicity,including bisphosphonates, topiramate, vigabatrin, isotretinoin and other reti-noids, amiodarone, ethambutol, chloroquine and hydroxychloroquine, tamoxifen,quetiapine, cyclo-oxygenase (COX)-2 inhibitors, erectile dysfunction agents andsome herbal medications. For this review, the certainty of the adverse effect

76 Santaella & Fraunfelder

profile of each medication was evaluated according to the WHO CausalityAssessment Guide.

A certain relationship has been established for pamidronate and alendronate ascauses of scleritis, uveitis, conjunctivitis and blurred vision. Topiramate has beenestablished as adversely causing symptoms consistent with acute angle-closureglaucoma, typically bilateral. Vigabatrin has been shown to cause bilateral irre-versible visual field defects attributed to underlying medication-induced retinalpathology. Isotretinoin should be considered in the differential diagnosis of anypatient with pseudotumour cerebri. Patients taking amiodarone and hydroxy-chloroquine should be monitored and screened regularly for development of opticneuropathy and maculopathy, respectively. Sildenafil has been reported to causeseveral changes in visual perception and is a possible, not yet certain, cause ofanterior ischaemic optic neuropathy. Patients taking tamoxifen should also bemonitored for development of dose-dependent maculopathy and decreased colourvision. COX-2 inhibitors should be included in the differential diagnosis ofreversible conjunctivitis. Several herbal medications including canthaxanthine,chamomile, datura, Echinacea purpurea, Ginkgo biloba and liquorice have alsobeen associated with several ocular adverse effects.

It is the role of all healthcare professionals to detect, treat and educate thepublic about adverse reactions to medications as they are an important healthproblem.

The term ‘side effect’ usually refers to an unde- er, if undetected, toxic effects may progress andsired or negative effect of medication that is extrane- cause irreversible ocular damage often with an asso-ous to the intended therapy. When the effect is ciated reduction in visual function.[3,4]

negative, the term ‘adverse effect’ is used. Drug- For ophthalmic drugs to be effective, they mustinduced ocular adverse effects are the second most reach ocular tissue in relatively high concentrations.frequent reason for claims against ophthalmolo- There are several different administration routes forgists.[1,2] This may not be surprising given that pre- ophthalmic drugs, including the topical, oral, paren-scribing medications is the most common therapeu- teral, periocular, intracameral (intraocular adminis-tic service provided by physicians. According to the tration into the anterior segment) and intravitrealNational Center for Health Statistics, new or contin- routes. Topical application is the most commonued medications are ordered or provided at 41% of route of administration because it is simple, lessvisits to an ophthalmologist’s office. Because seri- invasive and does not involve the passage of drugsous injury can occur, drug-related adverse effects through the blood-aqueous barrier. However, somecan be costly to defend, indemnify or settle.[2] disorders require systemic drug administration to

achieve adequate therapeutic levels of the drug inThe rich blood supply and relatively small massand around the ocular tissues.of the eye make it particularly susceptible to drug-

induced adverse reactions. Adverse ocular reactions Certain factors increase the probability of anto drugs are diverse. Drug molecules present in the adverse ocular reaction. One such factor is use of asystem may become selectively deposited in specif- medication over long periods of time, for example,ic ocular tissues such as the cornea, lens and retina, in cases of arthritic and cardiovascular diseases. Incausing varied symptoms of drug toxicity. Fortu- some patients, it may be difficult to establish wheth-nately, most adverse reactions induced by systemic er ocular pathology is caused by the condition beingmedications are reversible if detected early. Howev- treated or by a drug used to treat the condition.


Ocular Adverse Effects Associated with Systemic Medications 77

Patient age is also a significant factor in the preva- We performed a MEDLINE literature search us-ing the following keywords: ‘ocular’, ‘visual’, ‘eye’,lence of ocular drug reactions. Older patients are‘side effects’, ‘adverse effects’, ‘medication’ andmore likely to have used medications for protracted‘treatment’. Some of the medication adverse effectsperiods. Also, the metabolism and excretion of aobtained through this search are summarised below;drug can be affected by decreased efficiency of thethese were selected at the authors’ discretion, takingkidney and liver secondary to the patient’s age, or byinto account some of the more recent publishedconditions adversely affecting these organs.[3,4]

medication adverse effects, and are not presented inSome drug responses cannot be predicted from any particular order. This brief review is not intend-

the drug’s pharmacological mode of action. A ge- ed as a comprehensive summary of these well de-netic basis may underlie many of these unpredict- scribed adverse drug reactions, nor is it intended toable responses, as observed in the rapid rise in cover the complete spectrum of all ocular adverseintraocular pressure reported with topical corticoste- effects of systemic therapy. Interested readers are

encouraged to refer to textbooks cited within theroids.[5]

references.The prevalence of adverse reactions is closelyassociated with drug dosage. Most reported ocular

1. Categorising Adversereactions occur when the dose is beyond the thera-Drug-Related Eventspeutic range. It is essential that clinicians try to

establish whether the ocular problem coincided with The WHO Causality Assessment Guide of Sus-the start of drug therapy or with a change in drug pected Adverse Reactions was used to classify thedosage. A useful marker is seen when the onset of reported adverse drug-related events into the follow-the reaction coincides with commencement of the ing categories: certain, probable/likely, possible, un-medication, but reactions can occur at any time likely, conditional/unclassified and unassessable/during or after a course of medication, and can unclassifiable.[6] The ‘certain’ category includescontinue for years after cessation. plausible time relationship to drug administration

Table I. WHO definitions: causality assessment of suspected adverse reactions (reproduced from Brick,[2] with permission)

Certain

A clinical event, including a laboratory test abnormality, occurring in a plausible time relationship to drug administration, and whichcannot be explained by concurrent disease or other drugs or chemicals. The response to withdrawal of the drug (dechallenge) shouldbe clinically plausible. The event must be definitive pharmacologically or phenomenologically, using a satisfactory rechallenge procedureif necessary

Probable/Likely

A clinical event, including laboratory test abnormality, with a reasonable time sequence to administration of the drug, unlikely to beattributed to concurrent disease or other drugs or chemicals, and which follows a clinically reasonable response on withdrawal(dechallenge). Rechallenge information is not required to fulfill this definition

Possible

A clinical event, including laboratory test abnormality, with a reasonable time sequence to administration of the drug, but which couldalso be explained by concurrent disease or other drugs or chemicals. Information on drug withdrawal may be lacking or unclear

Unlikely

A clinical event, including laboratory test abnormality, with a temporal relationship to drug administration which makes a causalrelationship improbable, and in which other drugs, chemicals or underlying disease provide plausible explanations

Conditional/Unclassified

A clinical event, including laboratory test abnormality, reported as an adverse reaction, about which more data is essential for a properassessment or the additional data are under examination

Unassessable/Unclassifiable

A report suggesting an adverse reaction which cannot be judged because information is insufficient or contradictory, and which cannotbe supplemented or verified



Table II. Classification of adverse ocular effects associated with medication usage

Medication Certain Probable Possible

Pamidronate Blurred vision Periocular, lid, and/or orbital DiplopiaPain oedema Visual hallucinationsPhotophobia Yellow visionOcular irritation Retrobulbar neuritisNonspecific conjunctivitis Cranial nerve palsyAnterior uveitis (rare posterior)Anterior scleritis (rare posterior)Episcleritis

Alendronate Blurred vision Diplopia GlaucomaPainConjunctivitisUveitisScleritis

Risedronate No effects Conjunctivitis DiplopiaPain PapilloedemaScleritis EpiscleritisUveitisBlurred vision

Etidronate Blurred vision Conjunctivitis Diplopia

Topiramate Acute glaucoma (mainly bilateral) Blepharospasm ScleritisAnterior chamber shallowing Oculogyric crisis Teratogenic effects, includingIncreased ocular pressure Retinal bleeds ocular malformationsMydriasis UveitisSuprachoroidal effusions

Isotretinoin Abnormal meibomian gland secretion Decreased colour vision Corneal ulcersBlepharoconjunctivitis Permanent loss of dark DiplopiaCorneal opacities adaptation Eyelid oedemaDecreased dark adaptation Optic neuritisDecreased tolerance for contact lens wear Idiopathic intracranialDecreased vision hypertension with optic discIncreased tear osmolarity oedemaKeratitis Permanent sicca-like syndromeMeibomian gland atrophy Subconjunctival haemorrhageMyopiaOcular siccaOcular discomfortPhotophobiaPseudotumour cerebriTeratogenic ocular abnormalities

Amiodarone Aggravated sicca (drug in tears) Anterior subcapsular lens Autoimmune reaction (dry mouth,Blepharoconjunctivitis opacities dry eyes, peripheral neuropathyBright lights Corneal ulceration and pneumonitis)Coloured haloes around lights Loss of eyelashes orCorneal microdeposits eyebrowsGlare Non-arteritic ischaemic opticHazy vision neuropathyPhotosensitivity Pseudotumour cerebriPeriocular skin pigmentationThyroid eye diseaseVisual sensations

Continued next page



Table II. Contd

Medication Certain Probable Possible

Sildenafil Changes in colour perception No effects Mydriasis (emotional effect?)coloured tinge Retinal vascular accidentsdecreased colour vision (secondary to exertion?)dark colours appear darker Subconjunctival haemorrhage

Blurred vision Anterior ischaemic opticcentral haze neuropathytransitory decreased vision

Changes in light perceptionincreased perception of brightnessflashing lights, especially when blinking

ERG changesConjunctival hyperaemiaOcular painPhotophobia

Tamoxifen Corneal opacities No effects No effectsRetinal opacities, degeneration, pigmentarychanges, haemorrhageLoss of visual acuity

COX-2 inhibitors Conjunctivitis No effects No effectsBlurred vision

Nicotinic acid No effects Cystoid macular oedema Decreased vision, dry eyes,discoloration of the eyelids,eyelid oedema,ProptosisLoss of eyebrows and eyelashes,and superficial punctate keratitis

Canthaxanthine Crystalline retinopathy No effects No effects

Chamomile Allergic conjunctivitis No effects No effects

Datura Mydriasis No effects No effects

Echinacea purpurea No effects Conjunctivitis No effects

Ginkgo biloba No effects No effects Spontaneous hyphemaRetinal haemorrhage

Liquorice No effects No effects Vasospasm, visual lossassociated with migraine-likesymptoms

Vitamin A Intracranial hypertension (when taken in No effects No effectslarge doses)

COX = cyclo-oxygenase; ERG = electroretinogram.

and inability to explain the adverse effect by concur- 2. Medications and Adverse Effectsrent disease or other drugs or chemicals. Dechal-

lenge data are necessary and rechallenge should be 2.1 Bisphosphonates

positive. ‘Probable’ is the same as ‘certain’ withoutBisphosphonates inhibit bone resorption by bind-

positive rechallenge data. ‘Possible’ is an adverse ing to hydroxyapatite crystals and inhibiting theirevent in a reasonable time sequence to administra- dissolution.[7] Different bisphosphonates vary great-

ly in their efficacy and their adverse-effect profilestion of the drug, but could also be explained bydepending on the structure of the individual drug.

concurrent disease or other drugs or chemicals. Pos-These medications are associated with ocular ad-

itive dechallenge data are lacking or unclear in this verse effects that are mainly inflammatory, i.e. con-category (table I). junctivitis, uveitis and episcleritis.[8,9] Recent studies



have proved that pamidronate can cause scleri- clodronate.[17] Reports of clodronate-induced uveitistis.[10,11] have a ‘probable’ association and blurred vision is

classified as having a ‘possible’ association[18] (tablePamidronate disodium, an intravenous bisphos-II).phate, has been reported to cause anterior uveitis and

nonspecific conjunctivitis.[9-12] Its most striking as- Pamidronate is the first medication ever reportedsociation is that of being the first medication report- to cause scleritis. Bisphosphonates are high molecu-ed to cause scleritis.[10] Case reports have also asso- lar weight drugs that have been described as poten-ciated pamidronate with episcleritis,[13] nerve pal- tially causing immune complex formation and pos-sy,[14] ptosis[14] and retrobulbar neuritis[15] (table II). sibly being secreted by the lacrimal gland, thus

Pamidronate bears a ‘certain’ relationship to uve- causing transient irritation to mucus membranes.[3]

itis, conjunctivitis, episcleritis and scleritis when Pamidronate stimulates the production of a distincttaken at standard doses between 30 and 90mg intra- subgroup of T cells to inhibit bone resorption. Asvenously. The onset of ocular signs and symptoms analogues of pyrophosphate, they can activate re-was usually noted within 48 hours. ceptors in T cells leading to cytokine release.[18,19]

This may contribute to an immunological reaction inAlendronate is an oral bisphosphonate widelypatients who develop uveitis and/or scleritis. Someprescribed for the treatment and prevention of osteo-authors have contended that the nitrogen-containingporosis, in particular postmenopausal osteoporosis,bisphosphonates (alendronate, pamidronate, rise-and for the treatment of Paget’s disease of thedronate) are more likely to cause uveitis.[20-23] Nitro-bone.[16] This bisphosphonate has been associatedgen-containing bisphosphonates are known to causewith a ‘certain’ relationship to blurred vision, oculartransient pyrexia, a flu-like syndrome and serologi-pain, conjunctivitis, uveitis and scleritis when takencal changes resembling a typical acute phase re-in dosages ranging from 5 to 40mg daily (table II).sponse.[3,18,19] The cytokines released with this acuteOnset of ocular signs and symptoms was noted anphase response could act as adjuvants in an immuneaverage of 2 days to 2 weeks (range 1 day to 1 year)reaction with the uvea of the eye as the target organ.after starting therapy[11] (table II).

Risedronate is taken orally and is indicated in the However, a non-nitrogen-containing bisphospho-treatment of Paget’s disease of the bone.[16] The data nate, clodronate, has also been reported as inducingare not sufficiently complete to classify any ocular uveitis.[18] Still, there are many more reports ofadverse effect as ‘certain’. However, there are posi- uveitis associated with nitrogen-containing bisphos-tive rechallenge data on a single report of scleritis phonates than there are for the bisphosphonates thatassociated with this medication, which indicates do not contain nitrogen (etidronate, clodronate,there could be a cause and effect relationship.[10] tiludronate).

Etidronate is an oral medication indicated in the In summary, bisphosphonates are associated withtreatment of symptomatic Paget’s disease, or for ocular inflammation. Ocular pain, photophobia,heterotopic ossification in hip replacement and spi- blurred vision, periorbital changes and glaucomanal cord injury patients. This medication has been can be due to uveitis, scleritis and other types ofassociated with conjunctivitis and blurred vision[11] ocular inflammation. Ocular inflammation, espe-(table II). cially scleritis and uveitis, is of the greatest concern

in the eye care of patients taking bisphosphonates.Clodronate is used in Europe and Canada forScleritis causes a severe deep eye pain, and can leadtumour-induced bone disease.[7] It has not been ap-to structural damage of the globe and loss of vision.proved by the US FDA. This medication is also usedPatients typically present with a piercing ocular painby some to prolong survival in breast cancer pa-that is worse at night and awakens them from sleep.tients. Although some studies indicate there are few-The sclera assumes a violaceous hue in natural sun-er metastases with clodronate therapy, there is nolight and scleral vessels become inflamed and canevidence of prolonged survival in patients taking



have a crisscross pattern. Scleral oedema develops the drug may need to be discontinued for uveitis toresolve. Episcleritis may require topical ocular med-acutely and scleral thinning can occur, along withication; however, the bisphosphonate may be con-vision loss, if necrotising scleritis persists or inflam-tinued. In all patients studied by Fraunfelder etmation is left unchecked. Ocular complications fromal.,[11] the bisphosphonate had to be discontinued forscleritis can include keratitis, cataract, uveitis andthe scleritis to resolve, even on full medical therapy.glaucoma.[12,24]

In uveitis associated with bisphosphonate ther-2.2 Antiepileptic Drugs

apy, the early symptoms may be mild or severe,depending on which part of the uvea is affected and

2.2.1 Topiramateon the amount of inflammation. Anterior uveitis

Topiramate, a sulfamate-substituted monosac-usually has the most dramatic symptoms, typically

charide, is structurally unrelated to any otherpresenting with severe pain in the eye, redness of the

antiepileptic drug (AED). This AED is also used inconjunctiva, sensitivity to bright light and a decrease

the management of migraine, depression and neuro-in vision. The examiner may see prominent blood

pathic pain. Off label, it has gained popularity as avessels on the conjunctiva near the edge of the iris,

weight reduction agent, to treat migraine headacheswhite blood cells floating in the aqueous humour and to treat bipolar disorder.and deposits of white blood cells on the inside

Topiramate has been associated with acute angle-surface of the cornea. Intermediate uveitis is typical-

closure glaucoma. Findings associated with topira-ly painless. Vision may be decreased and the patient

mate-associated acute, bilateral, secondary angle-may see floaters (irregular floating black spots).

closure glaucoma syndrome include blurred vision,Posterior uveitis typically produces decreased vision conjunctival hyperaemia, corneal oedema, shallowwith or without floaters. There may also be retinal anterior chamber, cataracts, pupil changes, elevateddetachment and inflammation of the optic nerve intraocular pressure, visual field defects and blind-(symptoms include loss of vision, which may vary ness. Topiramate has also been associated withfrom a small blind spot to total blindness). Diffuse causing bilateral myopia, bilateral suprachoroidaluveitis may produce any or all of these symptoms. effusions, blepharospasm, myokymia, nystagmusUveitis can rapidly damage the eye and can produce and diplopia. Scleritis has also been reported.[25]

long-term, vision-threatening complications, such as Topiramate-associated visual adverse effects andswelling of the macula, glaucoma and cataracts. their WHO classifications are listed in table II. TheThose affected may have only one episode or peri- entity described as topiramate-associated acute, bi-odic recurrences over months to years. lateral, secondary angle-closure glaucoma[26,27] can

Suggestions for treatment of bisphosphonate-in- present in the same manner as an acute angle closureduced ocular adverse effects are as follows. If there glaucoma attack. All the findings of acute glaucoma,is a persistent decrease in vision or if ocular pain such as ocular pain, headache, nausea and vomiting,occurs, examination by an ophthalmologist is neces- pupillary changes, hyperaemia, corneal oedema,sary. Nonspecific conjunctivitis seldom requires cataracts, retinal and vascular accidents, visual fieldtreatment and usually decreases in intensity or may defects and blindness have been reported.[25] In mostbe absent on subsequent treatments. In rare inpatients, this is a bilateral process. If not recognisedstances, a nonsteroidal anti-inflammatory eye drop as a drug-related event, this condition can easily bemay be needed. More than one ocular adverse effect confused with acute, pupillary block, narrow anglecan occur at the same time, i.e. episcleritis with glaucoma for which a peripheral iridectomy is indi-uveitis. Bilateral anterior uveitis or, rarely, posterior cated. However, if the drug is stopped and medicalor bilateral uveitis may occur and can vary markedly management instituted, pressure may return to nor-in severity. Many patients require intensive topical mal in hours to days without the need for an iridecto-ocular or systemic medication. In some instances, my. The acute pressure elevation usually occurs



within the first 2 weeks after starting topiramate the last decade, the marked efficacy of this medica-therapy, but it has been reported within hours after tion and its low toxic effects prompted widespreaddoubling the dose. If pressure elevation goes un- use in Europe. The manufacturers of vigabatrintreated, serious outcomes are possible, including (Hoechst Marion Roussel) had received 28 reportsblindness. of visual field abnormalities worldwide by January

1997 in an estimated 140 000 patients treated.[38]Acute myopia up to 8.75 diopters may occur in amatter of hours after starting topiramate use, but This information appears to have not been releasedmay take weeks to fully resolve on or off medica- during that time. Since 1997, numerous reports havetion. Sulfa-containing medications, such as topira- appeared of visual field abnormalities in adults andmate, are known to cause transient myopia. Lenticu- children treated with this AED. In most documentedlar swelling, forward rotation of the lens-iris dia- cases, the visual field defect seems to be a specific,phragm, ciliary body swelling causing increased bilateral, symmetrical and irreversible peripheralcurvature of the lens surfaces, and spasm of accom- constriction.[38-40] The fact that most patients aremodation have all been proposed as the mechanism asymptomatic with normal visual acuity may havebehind this occurrence.[3,25,28-31] contributed to the late recognition of these visual

field defects that apparently occur in more than 30%The management of topiramate-related acutepressure elevation requires stopping the drug in con- of patients but were initially estimated to affectsultation with the prescribing physician, since de- fewer than 0.1%.[41-44]

creasing the dosage as little as 50mg may exacerbate The site of toxicity is proposed to be the retina,pre-existing systemic conditions. Instituting maxi- where GABA is an important modulatory neuro-mal medical therapy, including oral and topical transmitter. Vigabatrin increases GABA levels byaqueous suppressants, is indicated. Laser iridotomy inhibiting the GABA transaminase enzyme. GABAor peripheral iridectomy may not be beneficial if the is an inhibitory neurotransmitter in bipolar cells andglaucoma is only associated with topiramate ther- some amacrine cells and may have a role in theapy, but would be considered adequate in the setting modulation of phototransduction from the retinalof angle closure if one is not certain that it is photoreceptor cells to the ganglion cells.[45] System-completely medication-induced. Topical miotics are ic vigabatrin has been shown to cross the blood-probably contraindicated, as these could precipitate retinal barrier and can be detected immuno-a relative pupillary block that would exacerbate the cytochemically in the retina.[46] Vigabatrin causescondition. white matter microvacuolation and intramyelinic

oedema in the brains of rodents and dogs but not in2.2.2 Vigabatrin and Tiagabinemonkeys and humans.[47] It also has been shown toVigabatrin is a selective, enzyme-activated, irre-cause accumulation of GABA in the retinal Mullerversible GABA aminotransferase inhibitor.[32] It is acells.[48] How an increase in retinal GABA levelscustom-made AED that is particularly useful in themay produce visual field constriction is not clear,management of drug-resistant partial seizures andbut the lower density of ganglion cells in the periph-infantile spasms, especially those secondary to tu-eral retina or a toxic effect on the retinal Muller cellsberous sclerosis.[33-36] The antiepileptic effect is pre-has been suggested as being possibly related.[49]

sumably mediated by elevation of GABA levels ofThus far, the understanding of the role played bythe brain caused by inhibition of GABA metabo-GABA in retinal transmission is not sufficient tolism.[36]

allow for a mechanistic explanation of the adverseInitially, only relatively minor adverse effectseffects of vigabatrin.were attributed to vigabatrin use.[37] It was first

There is no consensus regarding screening visualintroduced into clinical practice in the UK on a trialfield examinations for patients taking vigabatrin.[49]basis in the mid 1980s and granted licence in 1989.

The FDA has not approved its use in the US. Over The incidence, higher than previously estimated, of



this complication, particularly in asymptomatic pa- acitretin and etretinate have been reported as havingtients, suggests that screening visual field examina- a ‘probable’ causal relationship to IH.[59]

tions may be necessary. Daneshvar et al.[49] have The mechanism of how retinoids cause IH hasrecommended visual field examinations covering been postulated as being through a common path-the peripheral 60° of the visual field for all patients way.[61] It is possible that high doses of this class oftaking vigabatrin, before or soon after starting treat- drug induce a secretion of cerebrospinal fluid andment and at regular intervals thereafter. Any repro- alter the lipid constituents of the arachnoid villi.ducible visual field constriction not present at the This may then disrupt the normal transport systemsbaseline examination may be associated with and impede the absorption of cerebrospinal fluid atvigabatrin. The decisions on continuing treatment in the arachnoid villi.[64]

these patients must be individualised based on a Isotretinoin has been shown to be secreted inbenefit-risk assessment. tears by the lacrimal gland and has been associated

Tiagabine has a mechanism of action that resem- with causing meibomian gland dysfunction and thusbles that of vigabatrin. Thus, suspicions have been causing eye and contact lens discomfort, dry eyepresented that tiagabine might produce similar complaints, blepharoconjunctivitis, transient blur-changes in visual function to those associated with ring of vision and acute transient refractivevigabatrin.[50] However, findings have not been rep- changes.[3]

licated and several studies have found normal visual For patients receiving retinoids, we suggest thefunction in patients treated with tiagabine. Although following management guidelines. In patients onno significant association with visual field changes retinoid therapies who develop otherwise unex-or acuities has yet been found with tiagabine, cases plained headaches or blurred vision, a prompt oph-of deteriorating colour vision have been report- thalmology consultation to rule out papilloedemaed.[51-53] should be performed. Periodic ophthalmology ex-

aminations should also be completed to rule out2.3 Isotretinoin and Other Retinoids papilloedema even in asymptomatic patients on reti-

noid therapy for 6 months or more because IH canRetinoids are used to treat severe recalcitrant develop without symptoms. Concomitant use of tet-

nodular acne, acne vulgaris and severe recalcitrant racyclines or vitamin A should be avoided, as theypsoriasis, and to induce remission of leukaemia. may potentiate the development of IH.[65,66]

Ocular adverse effects are dose related and probablythe most frequent adverse reactions of these drugs. 2.4 Ethambutol and IsoniazidIsotretinoin-related ocular adverse effects are listedwith their WHO classifications in table II. Ethambutol is still considered a primary therapy

The retinoid family includes vitamin A and syn- against Mycobacterium tuberculosis and has syner-thetic derivatives such as isotretinoin, etretinate and gistic actions when combined with other agents.[67]

retinoin. Isotretinoin, the most widely prescribed It can cause a multitude of dose- and time-dependentretinoid, has been associated with a ‘certain’ WHO ocular adverse effects including colour visionclassification for inducing intracranial hypertension changes, visual field defects and, of most impor-(IH) due to positive rechallenge data, a characteris- tance, unilateral or bilateral optic neuritis which cantic pattern of rapid IH onset after exposure, and the continue to progress for 1–2 months after the drug isfact that this agent is a degradative product of all- discontinued.[68] The Physicians’ Desk Referencetrans retinoic acid (tretinoin), a known contributor (PDR) discusses the possibility of optic neuropathyto IH.[54-60] Vitamin A has been well documented as and attempts to provide guidelines for clinicians toa cause of IH, and other retinoids such as tretinoin screen for optic nerve toxicity.[16] However, it re-can cause IH.[61-63] The incidence of IH in patients mains unclear what tests should be performed andtaking tretinoin has been quoted as 9%.[64] Tretinoin, how often. The literature describes a dose-related



incidence of ocular adverse effects with develop- axonal transport in the optic nerve and lead to opticneuropathy.ment of optic neuropathy in 50% of patients at a

dose of 60–100 mg/kg/day, 5–6% at 25 mg/kg/day The US PDR recommends monthly ophthalmicexaminations for patients taking doses ofand 1% with dosages ≤15 mg/kg/day.[3]

ethambutol >15 mg/kg/day.[16] No official standardEthambutol optic neuropathy is usually bilateralof care exists regarding how often to see patients andand can be asymmetric. Ethambutol toxicity maywhich tests to perform for those taking >15 mg/kg/affect only the small calibre papillomacular bundleday. Outside the US, especially in some developingaxons, and optic atrophy will not develop untilcountries, monthly exams are not practical andmonths after the fibres are lost. This means objectivewould create a large burden for eye care providers.

findings on the fundus exam are frequently absent.Because of medicolegal concerns, ophthalmologists

The literature suggests that optic neuropathy mayin the US should use the PDR as a guideline, as the

occur, on average, 2–8 months after starting therapy. package insert published in the PDR was agreedThe earliest ophthalmological findings in toxic optic upon by the FDA and industry alike.neuropathy from ethambutol may be loss of visual

The authors of this article recommend obtainingacuity, colour vision loss or central scotomas. informed consent prior to assuming care for patientsEthambutol also has an affinity for the optic chiasm taking ethambutol, explaining that optic neuropathywith bitemporal visual field defects manifesting can occur at any dose despite regular ophthalmicwith toxicity.[3,69]

exams and that vision loss can be severe and irre-Isoniazid is frequently prescribed concurrently versible. A baseline examination should be per-

with ethambutol for tuberculosis owing to multiple formed including visual acuity, visual field and col-cases of drug resistance to single-agent therapy. our vision testing, and a dilated fundus and opticIsoniazid also has been associated with optic neu- nerve exam. If any visual symptoms occur, the med-

ication should be discontinued and the patient seenropathy, and differentiating toxicity due to ethambu-by an ophthalmologist. Frequency of examinationstol versus isoniazid can be challenging.[70,71] In gen-should be monthly for dosages >15 mg/kg/dayeral, the toxicity from isoniazid is less frequent, less(PDR); however, monthly exams at lower dosessevere and is usually reversible.[3] When in doubt, itmay be necessary for patients at increased risk formay be necessary to undertake dechallenge withtoxicity, such as those with diabetes mellitus, chron-isoniazid and/or ethambutol after consultation withic renal failure or alcoholism, and for the elderly,the primary care physician.children and those with other ocular pathologies or

The pathophysiology underlying ethambutol op-ethambutol-induced peripheral neuropathy. Discon-

tic neuropathy remains unclear. However, one pos-tinuation of ethambutol should be considered after

sible explanation is that ethambutol chelates copperany signs of loss of visual acuity or colour vision or

in the retinal ganglion cells and their fibres in the for a visual field defect. Optical coherence tomogra-optic nerve. Ethylenediiminodibutyric acid, a me- phy or contrast sensitivity testing should also betabolite of ethambutol, is a strong chelator of cop- considered in these patients, as these tests could pickper. Copper is required as a cofactor for cytochrome up early ethambutol toxicity not detected with thec oxidase, a key enzyme in the electron transport baseline exam.chain and cellular oxidative metabolism within mi-tochondria. It is possible that ethambutol decreases 2.5 Amiodaronethe levels of copper available for cytochrome coxidase, and therefore also decreases the required Amiodarone is a well known benzofuran deriva-energy for axonal transport around the optic tive that is used in the treatment of various cardiacnerve.[72,73] Thus, mitochondrial insufficiency in the arrhythmias. Its half-life ranges from 26 to 160optic nerve fibres may underlie the impairment of days.[3] Ocular adverse effects produced by this drug



are common and are time- and dose-dependent. Be- drug-induced lipid storage disease (primary lipido-sis) in optic nerve axons.[76] There is a selectivecause of this, ophthalmic examinations should beaccumulation of intracytoplasmic lamellar inclu-conducted for a baseline and every 6–12 months, orsions in large optic nerve axons, and this may de-more frequently depending on findings. The corneacrease axoplasmic flow biochemically or mechani-is the most commonly affected ocular structure, withcally. Many of these patients may already have asuperficial punctuate opacities occurring incompromised optic nerve secondary to vascular dis-69–100% of patients receiving amiodarone. There isease, and the amiodarone deposition in the axonsoften a loss of eyelashes or eyebrows along withfurther impedes neural function, causing vision loss.eyelid and periocular tissue photosensitivity. Fur-The result is optic nerve head oedema, which canthermore, the conjunctiva may develop yellow-persist as long as transport is compromised. Resul-brown deposits. Lenticular changes can include an-tant optic nerve head oedema may persist as long asterior, subcapsular, small, yellow-white punctuatetransport is inhibited, i.e. as long as several monthsopacities and cortical changes.[3] One of the mostfollowing discontinuation of amiodarone, which, assevere adverse effects of amiodarone use is opticnoted earlier in this section, has a half-life of up toneuropathy.160 days.Amiodarone-associated optic neuropathy has

Since it may be impossible to distinguish NAIONbeen reported to have an incidence of 1.79%.[74] Thefrom amiodarone-induced optic neuropathy in manyexpected incidence of nonarteritic ischaemic opticpatients, those who experience any visual distur-neuropathy in the general population aged ≥50 yearsbance should see an ophthalmologist promptly. Pa-is 0.3%.[74]

tients should have a baseline ophthalmic exam andMacaluso et al.[75] characterised amiodarone-as- may be seen every 6 months for monitoring, al-

sociated optic neuropathy as having an insidious though this is controversial. If optic neuropathy isonset, slow progression, bilateral visual loss and suspected, discontinuation of the drug in consulta-protracted disc swelling that stabilised within tion with the patient’s cardiologist at the first signsmonths of discontinuing the medication. Because of optic nerve involvement must be considered un-patients taking amiodarone often have severe vascu- less the ophthalmologist is very confident of thelar disease, the incidence of nonarteritic anterior diagnosis of NAION.ischaemic optic neuropathy (NAION) in these pa-tients is probably higher than that of the age- 2.6 Hydroxychloroquinematched population. At present there are no reportedcases of amiodarone neuropathy causing vision of This aminoquinoline is used in the treatment ofno light perception. Macaluso and colleagues pro- malaria and extraintestinal amoebiasis. Hydroxy-posed that the insidious onset, bilateral pathology, chloroquine is also used for the treatment of rheu-less severe visual loss (20/20 to 20/200 in ami- matoid arthritis and lupus erythematosus, dermato-odarone-induced versus 20/20 to no light perception logical conditions and various inflammatory disor-in NAION) and slower improvement of optic disc ders.oedema of amiodarone-induced optic neuropathy Ocular adverse effects can be numerous, includ-compared with that of acute NAION helps the oph- ing an enhanced Hudson-Stahli line, whorl-like cor-thalmologist in differentiating these two causes of neal depositions (cornea verticillata), transient oede-decreased vision associated with optic disc oedema. ma, decreased sensitivity, retina parafoveal granu-The WHO classification of other established visual larity of the retinal pigment epithelium (early in theadverse effects related to this medication are listed disease), bull’s-eye appearance of the macula (latein table II. in the disease), attenuation of the vascular tree,

The cause of amiodarone-induced optic neuropa- peripheral fine granular pigmentary changes, promi-thy is unknown, but this condition may be due to a nent choroidal pattern filling defects in late phase of



fluorescein angiography, and other angiography factors such as being aged >60 years and having achanges.[3] high body fat level, concomitant kidney or liver

disease, and/or concomitant retinal disease also jus-Though the side effects of hydroxychloroquinetify more frequent examinations.[19] A recent prelim-are numerous, the most adverse are mainly dose-inary paper by Shroyer et al.[78] suggests that indi-related, seen today usually in overdose situations,viduals with an ABCR mutation (Stargadt’s disease)obese patients or extremely thin individuals. Themay also be predisposed to develop retinal toxicitymost serious of these effects is a form of retinalwhen exposed to chloroquine/hydroxychloroquine.toxicity known as hydroxychloroquine bull’s-eye

maculopathy. This name stems from the characteris- The goal of these recommendations is to findtic ring-like or bull’s-eye pattern of retinal pigment early changes, i.e. relative scotomas, and is aimed atepithelium atrophy and hyperfluorescence on fluo- detection and not prevention, as discontinuation ofrescein angiography. The mechanism of this toxicity the drug is the only way to prevent possible adverseis not well understood but it seems to be related to effects and this is not an option for some patients.the affinity for this medication for melanin within Early paracentral relative scotomas seldom advancethe retinal pigment epithelium. Patients may have when the drug is discontinued. Later findings in-abnormal sensory testing responses and distorted clude retinal changes, colour vision loss, absolutecolour vision (late phase of disorder) with a yellow, scotoma or decreased vision. Even if the drug isgreen or blue tinge to objects and coloured haloes stopped, once these occur, changes are irreversible,around lights. This maculopathy is bilateral and and many patients may continue to lose some visionreproducible by Amsler grid and visual field testing. and/or peripheral fields.

Corneal deposits are also a characteristic finding How to best follow patients on hydroxychloro-in patients taking this medication. Vortex keratopa- quine was summarised in an article by Marmor etthy or cornea verticillata is not specific to hydrox- al.[77] These recommendations, although not univer-ychloroquine but a finding of several ‘amphophilic’ sally adopted, follow the overall guidelines of thedrugs that can form complexes with cellular American Academy of Ophthalmology for examin-phospholipids, which cannot be metabolised by ing patients. Patients aged 20–29 years should havelysosomal phospholipases and, thus, are visible clin- one examination; 30–39, two examinations; 40–64,ically as deposits in the superficial cornea.[3] This an examination every 2–4 years; and those ≥65 anmedication also has been found to be excreted in the examination every 1–2 years.tear film and can aggravate dry eye along with To summarise these recommendations, the base-possibly decreasing contact lens tolerance. line examination should be performed within the

For patients taking antimalarial drugs, the Ameri- first year after starting this drug. Patients shouldcan Academy of Ophthalmology (2002) recom- have a complete, dilated ophthalmic examination,mends baseline dilated fundus exams and testing the including the informed consent mentioned earlier incentral visual field by Amsler grid or a central 10- this section, warning of possible permanent visualdegree radius visual field using automated field test- problems in rare instances. This baseline examina-ing (Humphrey 10-2 or similar).[77] Ideally, these tion should include visual acuity testing, Amslertests should be done before starting the antimalarial grid testing (with instructions for monthly homemediations (or at the latest within 1 year after start- use), and optional colour vision testing (preferablying the medications). Low-risk patients are consid- including the blue-yellow axis, such as the pseudo-ered to be those taking <6.5 mg/kg/day for less than isochromatic plates for colour [American Optical5 years. These patients should have a follow-up Corporation]). If macular abnormalities are evident,examination every 5 years. Higher-risk patients, it would be ideal to obtain fundus photographs. Ifsuch as those taking >6.5 mg/kg/day for more than 5 any progressive ocular abnormality is suspected, ayears, should be examined more frequently. Risk baseline Humphrey 10-2 visual field or other auto-



mated perimetry concentrated on the central visual leads to increasing levels of cGMP, which causesfield should be considered. Multifocal electroreti- smooth muscles in the corpus cavernosum to relax,nogram (ERG) is optional. allowing blood inflow. This medication is known to

cause changes in colour perception as well asAdditional complete follow-up examinationsblurred vision. These agents may cause changes inmay be done at 2- to 4-year intervals, in patientslight perception, increased perception of brightnessyounger than 40 years who are not deemed high risk.

High-risk patients are those who are obese, extreme- and/or a sensation of seeing flashing lights, especial-ly thin, frail, elderly or have significant renal or ly when blinking.[79] Sildenafil has also been shownhepatic disease or macular disease of any type. to cause ERG changes. Ocular adverse effects arePatients should be seen sooner if they experience uncommon, dosage dependent and thus far have allany persistent visual symptoms, if they do not meet been fully reversible. The WHO classification forthe above criteria, or if their dosage exceeds 6.5 mg/ these medications is listed in table II.kg. These ocular adverse effects are dose dependent

For patients aged 40–64 years, follow-up every with all three drugs. The incidence with sildenafil2–4 years is adequate as long as the above high-risk use has been quoted to be 3% in those taking 50mg,criteria are not met. Patients aged ≥64 years should 10% with 100mg and 40–50% in those takingbe followed more closely, with follow-up every 1–2 200mg.[80,81] The incidence is the same for all ages,years. Annual eye examinations should be consid- and the incidence and severity of reported ocularered if patients have been on hydroxychloroquine adverse effects is directly proportional to blood drugtherapy for longer than 5 years. Annual examina- concentrations. The adverse effects based on dosagetions are also appropriate for patients considered start 15–30 minutes after ingestion of the drug andhigh risk or whose dosage exceeds 6.5 mg/kg. Fol- usually peak 60 minutes after ingestion.[81]

low-up exams should include repeat baseline exami-Sildenafil can be used safely in patients withnation and fundus photography if any macular ab-

glaucoma and macular degeneration. Postmarketingnormalities are noted. Fluorescein angiographysurveillance of sildenafil, vardenafil and tadalafilshould be considered in the presence of suspicioushas produced several reports associating these medi-pigmentary changes. Automated central visualcations as possible culprits for anterior ischaemicfields are optional. Multifocal ERG is helpful inoptic neuropathy in several patients. So far there areselected patients.no data which confirm a ‘certain’ relationship be-Patients taking chloroquine should undergo thetween ischaemic optic neuropathy and erectile dys-same tests mentioned above for baseline and follow-function medications. Present evidence from post-up exams, as described above. Patients should bemarketing surveillance suggests visual adverse ef-seen at least annually if dosage is <3.0 mg/kg offects, due to this class of medication, are benign andideal bodyweight or every 6 months if dosage istransitory. From these poorly documented data, the>3.0 mg/kg bodyweight. Patients who are short,association between sildenafil and NAION is ‘possi-obese or have renal and/or liver impairment shouldble’ according to WHO criteria requiring that aalso be seen every 6 months.clinical event occur within a reasonable time fromdrug administration. There is no well researched

2.7 Erectile Dysfunction Agentsexplanation as to how sildenafil therapy could causeNAION. Over 27 million men have used sildenafil.Sildenafil, vardenafil and longer-acting tadalafilMost are vasculopathic and fall into an age group inare selective inhibitors of cyclic guanosinewhich they are already at risk for NAION.[79]

monophosphate (cGMP) specific phosphodiesteraseWe feel that the only patients who should nottype 5 (PDE5). In the treatment of erectile dysfunc-

take PDE5 inhibitors are those who have previouslytion, inhibition of PDE5, which is responsible forexperienced NAION in one eye. These patients maythe degradation of cGMP in the corpus cavernosum,



be more prone to developing NAION in the same or segments in combination with a handheld indirect orfellow eye if sildenafil or other medicines in this contact lens should be performed. Baseline colourclass are ingested. vision testing is important. In keeping with the

American Academy of Ophthalmology’s currentrecommendations, in healthy adults, a complete eye2.8 Tamoxifenexamination at least every 2 years should be com-

Tamoxifen is an antiestrogen agent primarily pleted. More frequent examinations should be car-used in the treatment of estrogen-receptor positive ried out if ocular symptoms occur. The discovery ofbreast cancer, ovarian cancer, pancreatic cancer and a limited number of intraretinal crystals in the ab-malignant melanoma. The standard dosage is ≤20 sence of macular oedema or visual impairment doesmg/day and the ocular adverse event incidence not seem to warrant discontinuation of the drug.levels are about 1–2%. Dose-dependent ocular ad- Consultation with the oncologist is essential if sig-verse effects include corneal opacities (whorl-like nificant ocular findings occur. Presence of age-relat-subepithelial calcium map-dot changes), and retinal ed maculopathy is not a contraindication to the useor macular yellow or white refractile opacities, oe- of tamoxifen. However, informed consent may bedema, degeneration, pigmentary changes and advisable in our litigious society. Presence of poste-haemorrhages.[3] rior subcapsular cataracts is not an indication to stop

Tamoxifen retinopathy most commonly occurs the drug, since the condition usually progresses evenafter more than 1 year of therapy when a total of if the drug is discontinued. Significant colour loss>100g of the drug has been taken. The retinopathy may be a valid reason to consider discontinuing thecan include cystoid macular oedema, punctuate drug. Gorin et al. recommend considering discontin-macular retinal pigment epithelial changes, uing the drug for 3 months (in patients on prophylac-parafoveal haemorrhages and peripheral retinal pig- tic therapy) and retesting.[82,83] If colour vision re-ment epithelium changes. Refractile bodies located turns to normal, then restart the drug and retest in 3in the inner retina histologically may be the products months. If, at any time, there is no rebound fromof axonal degeneration. Visual acuity loss and the stopping the drug, or continued progression, then itretinal lesions do not appear to regress in chronic may be necessary to consult an oncologist and re-form of tamoxifen toxicity if the drug is discontin- evaluate the benefit-risk.ued unless cystoid macular oedema or haemor-rhages are the cause of vision loss. 2.9 Cyclo-Oxygenase-2 Inhibitors

An acute form of toxicity has been described butis not well defined and usually occurs within a few Cyclo-oxygenase (COX)-2 inhibitors are indicat-weeks after therapy is instituted. Symptoms include ed in the treatment of osteoarthritis, rheumatoidvision loss, retinal oedema, retinal haemorrhage and arthritis, acute pain and dysmenorrhoea through se-optic disc swelling. This was thought to result from lective blockade of COS-2 receptors. The anti-in-tamoxifen estrogenic activity, which may cause ve- flammatory activity appears to be effective withnous thromboembolism. However, Gorin et al.[82,83] fewer gastrointestinal adverse effects than olderconfirmed the risk of tamoxifen in the development NSAIDs.of posterior subcapsular cataracts, but found mini- Selective COX-2 inhibitors include rofecoxib,mal effects of this drug on retinal small vessel celecoxib, valdecoxib, lumiracoxib, nimesulide andocclusive disease, with no more a risk factor than etodolac. With these medications, blurred vision andhypertension or glaucoma. conjunctivitis appear to be identified most frequent-

We offer the following recommended guidelines. ly as adverse effects associated with COX-2 inhibi-Baseline ophthalmic examination within the first tors, especially for rofecoxib and celecoxib, the twoyear of starting tamoxifen, which should include slit most widely prescribed medications in this class[84]

lamp biomicroscopy of the anterior and posterior (table II).



Coulter et al.[85] propose a possible mechanism 2.10 Nicotinic Acidbehind the blurred vision some patients experience

Nicotinic acid (niacin) taken orally has beenwhen receiving COX-2 inhibitors. Inhibition of syn-proven in the treatment of cardiovascular and cer-thesis of prostaglandins and other related com-ebrovascular disease because of its cholesterol- andpounds that control retinal blood flow could lead totriglyceride-lowering effects.[91-93] A comprehensivevisual changes. The vascular endothelium of retinalreview of ocular adverse effects from nicotinic acidblood vessels produces compounds such as prosta-indicates a possible association with decreased vi-cyclin, thromboxane A2 and prostaglandin H2. Bothsion, cystoid macular oedema, dry eyes, discolora-COX-1 and COX-2 mediate synthesis of prosta-tion of the eyelids, eyelid oedema, proptosis, loss ofcyclin. Inhibition of either COX-1 or COX-2 mayeyebrows and eyelashes, and superficial punctatealter the COX pathway and in turn alter regulation ofkeratitis.[91] The most serious of these is cystoidretinal blood flow with potential changes in vi-macular oedema (CME). CME occurs primarily insion.[86] Other NSAIDs have been reported to causepatients receiving 3 g/day, although it has been seenblurred vision consistent with this theory.[87,88]

in patients taking as little as 1.5 g/day. A characteris-In the case of selective COX-2 inhibitors, anothertic, but not pathognomonic, petalloid hyperfluores-mechanism may contribute to the blurred visioncence without leakage is revealed by fluoresceinsome patients experience. Four of the six COX-2angiography. The mechanism behind CME is un-inhibitors are sulfonamides: celecoxib, rofecoxib,known but the condition usually resolves within 2valdecoxib and nimesulide. Sulfonamide medica-weeks after stopping treatment with nicotinic acid.tions are known to cause blurred vision throughA possible aetiology could be a prostaglandin-in-what is thought to be transient myopia. The mecha-duced toxic response on the Muller cells or intracel-nism is not fully understood but could include len-lular accumulation of fluid secondary to some as yetticular swelling, forward rotation of the lens-irisundocumented derangement of intracellular metabo-diaphragm, ciliary body swelling causing increasedlism.[93,94]

curvature of the lens surfaces, and spasm of accom-modation.[25,87-89] The sulfonamide property of these

2.11 Herbal MedicationsCOX-2 inhibitors could contribute to some cases ofblurred vision. Curiously, the two non-sulfur-con-

Herbal medicines and nutritional supplementstaining selective COX-2 inhibitors do not have theare not marketed to treat specific diseases, are ex-bulk of data to support a ‘certain’ association withempt from the interstate commerce law and fallblurred vision including a paucity of positive rechal-under the purview of the Dietary Supplement andlenge reports.Health Education Act of 1994 in the US. No efficacy

Conjunctivitis also has a ‘certain’ association or safety has to be proven to sell these agents, ofwith some COX-2 inhibitors. This may not be unu- which there are 700 botanicals and 1000 nutritionalsual, as many medications are secreted in the tears. It products.[16,95]

is possible these medications are secreted in the These products are of clinical importance to oph-tears as well, leading to transient inflammation of thalmologists, since many of the therapies can causethe conjunctiva which resolves upon discontinuation adverse effects and some interfere with prescribedof the drug. Many examples of this exist, such as medications. Approximately 40% of patients whoirritative conjunctivitis from oral diazepam or oral use alternative therapies do not disclose this infor-isotretinoin.[3,55,90]

mation to their doctor.[96]

Discontinuation of therapy leads to resolution Canthaxanthine, chamomile, datura, Echinaceawithout long-term sequelae. In every instance stud- purpurea, Ginkgo biloba, liquorice (licorice) andied, the ocular adverse effect resolved within 72 vitamin A are all associated with clinically signifi-hours of discontinuation of COX-2 inhibitor. cant ocular adverse effects[94,97] (table II).



2.11.1 Canthaxanthine reaction.[101] There is evidence that echinacea mayThis agent is used in cosmetics, as a food colour- activate the body’s autoimmune response, and the

ing and to produce an artificial suntan when admin- PDR suggests that its use be avoided in patients withistered orally. It is naturally occurring and is found autoimmune diseases.[16]

in crustaceans and chanterelle mushrooms. When2.11.5 Ginkgo bilobataken orally, canthaxanthine can deposit in all layersGinkgo biloba has been shown to act as a bloodof the retina, especially the superficial layers of the

thinner and increase bleeding times.[101] Spontane-macula. Clinically, the retinal findings are slowlyous hyphaema and retinal haemorrhages have oc-reversible and patients are usually asymptomatic,curred in patients taking this agent. Thus, ginkgobut visual changes may be seen in static thresholdshould be used with caution in patients who are alsoperimetry, electroretinography and changes in darktaking warfarin or aspirin, as the effects are additive.adaptation.[16,98]

2.11.6 Liquorice2.11.2 ChamomileLiquorice (Glycyrrhiza glabra) root derivesChamomile (Matricaria chamomilla) is used

some of its medicinal properties from the isoflavo-usually in the form of tea to treat eye disease as wellnoid glabridin. Glabridin inhibits COX activity, andas insomnia, indigestion, migraine headaches, bron-has both anti-inflammatory and antiplatelet effects.chitis, fevers, colds, inflammation and burns.[95,99]

It may help patients with gastric ulcers and pepticThere is strong evidence that this tea, when appliedulcers. Ocular migraine-like visual symptoms havetopically in or around the eye, can cause severebeen reported that occur without a headache.[94] Theconjunctivitis. A possible mechanism for these pa-authors of this article postulate that vasospasm oftients’ conjunctivitis could be sensitivity to the aller-the brain, retinal and/or optic nerve blood vesselsgens present in M. chamomilla pollen. Because pa-plays a role in these visual symptoms, as there istients are using chamomile to treat their eyes, clini-strong evidence that liquorice, through its glucocor-cians should recognise the possibility of M.ticoid and noradrenaline effects, can cause vaso-chamomilla sensitivity in patients with what appearsspasm throughout the body.[102] Clinically, it appearsto be allergic conjunctivitis, especially in patientsthat individuals need to consume large amounts ofwho already have an atopic history.[94]

liquorice for ocular adverse effects to occur. How-2.11.3 Datura ever, caution is advised if a patient has a history ofThe dried leaves of datura are used to treat eye migraine headaches, as the effects could be additive.

inflammation as well as asthma, bronchitis, influen-2.11.7 Vitamin A (Retinol)za and coughs. The leaves contain alkaloids, inVitamin A is used primarily as an oral dietaryextremely varying concentrations, which are anti-

supplement and in the treatment of vitamin A defi-cholinergic and parasympatholytic. Jimson weedciency and acne. Vitamin A plays a vital role in(Datura stramonium) is the main member of thisvision and is an essential fat-soluble vitamin forgenus utilised for its potential therapeutic value.many biological activities. Vitamin A deficiencyThese botanicals have been shown to contain sco-can result in night blindness, xerosis of the conjunc-polamine and have been associated with pupillarytiva and cornea, and alopecia of the eyelashes.[94]

dilation.[94]

There is strong evidence that retinoids, particularly2.11.4 Echinacea purpurea if taken in excess, can cause IH.[54,55] PhysiciansEchinacea is a herbal product used to treat the should be aware that large doses of vitamin A can

common cold, cough, fevers, urinary tract infec- cause IH and special caution should be exercisedtions, burns and influenza.[94,100] An allergic con- with patients who are taking vitamin A in addition tojunctivitis has been associated with this product. another type of retinoid, such as isotretinoin, as theThe mechanism may be due to an anaphylactic effects could be additive in the causation of IH. This



authors have no conflicts of interest that are directly relevantcondition resolves in the majority of patients whento the content of this review.vitamin A is discontinued.

2.11.8 SummaryReferencesHerbal medicines and nutritional supplements

1. Bettman JW. Seven hundred medicolegal cases in ophthalmolo-are being used by a large segment of the population, gy. Ophthalmology 1990; 97: 1379-84

2. Brick DC. Medication errors result in costly claims for ophthal-often without strong evidence on efficacy or safety.mologists. Surv Ophthalmol 1995; 40: 232-6Fortunately, if the clinician recognises an ocular or 3. Fraunfelder FT, Fraunfelder FW. Drug-induced ocular side ef-

systemic adverse effect from one of these agents, the fects. 5th rev. ed. Boston (MA): Butterworth-Heinemann,2001symptoms are usually reversible. Clinicians should

4. Chiou GCY. Ocular toxicology. 2nd ed. Philadelphia (PA):remain vigilant in recognising adverse ocular reac- Taylor & Francis, 1999: 43-86, 225-354tions as well as enquiring whether these alternative 5. Kersey JP, Broadway DC. Corticosteroid-induced glaucoma: a

review of hte literature. Eye 2006 Apr; 20 (4): 407-16treatments are being used, as patients frequently do6. Edwards R, Biriell C. Harmonisation in pharmacovigilance.

not disclose this information to their physicians. Drug Saf 1994; 10: 93-1027. Fleish H. Bisphosphonates. Pharmacology and use in the treat-

ment of tumour-induced hypercalcaemic and metastatic bone3. Conclusiondisease. Drugs 1991; 42: 919-44

8. Fraunfelder FW, Rosenbaum JT. Drug-induced uveitis. Inci-Adverse reactions to medications are an impor-dence, prevention and treatment. Drug Saf 1997; 17: 197-207tant health problem and it is the role of healthcare 9. Macarol V, Fraunfelder FT. Pamidronate disodium and possible

professionals to detect, treat and educate the public ocular adverse drug reactions. Am J Ophthalmol 1994; 118:220-4about them. Many systemic drugs may produce ocu-

10. Fraunfelder FW, Fraunfelder FT. Scleritis and other ocular sidelar toxicity. effects associated with pamidronate disodium. Am J

Ophthalmol 2003; 135: 219-22Any drug in any form may cause an adverse11. Fraunfelder FW, Fraunfelder FT. Bisphosphonates and ocularocular reaction if it reaches the eye. Fortunately,

side effects. N Engl J Med 2003; 348: 1187-8most ocular changes induced by drugs are reversible 12. Foster CS, Forstot SL, Wilson LA. Mortality rate and rheuma-

toid arthritis patients developing necrotizing scleritis or pe-if detected in the early stages of toxicity, although ifripheral ulcerative keratitis: effects of systemic immunosup-they are not detected early, some adverse reactions pression. Ophthalmology 1984; 91: 1253-63

may progress to cause serious and irreversible ocu- 13. Stewart GO, Stuckey BG, Ward LC. Iritis following intravenouspamidronate. Aust N Z J Med 1996; 26: 414-5lar damage associated with significant loss of visual

14. Ghose K, Waterworth R, Trolove P. Uveitis associated withperformance. pamidronate. Aust N Z J Ophthalmol 1994; 24: 320

15. Des Grottes JM, Schrooyen M, Dumon JC. Retrobulbar opticDiagnosing drug reactions requires a high indexneuritis after pamidronate administration in a patient with aof suspicion and careful review of the patient’shistory of cutaneous porphyria. Clin Rheumatol 1997; 16: 93-5

history, including recent use of a drug. Withdrawal 16. Medical Economics Company. Physicians’ Desk Reference.56th ed. Montvale (NJ): Medical Economics Company, 2002of the suspected drug should result in improvement

17. Diel IJ, Solomayer EF, Costa SD, et al. Reduction in new(dechallenge), and reinstitution of the drug (rechal-metastases in breast cancer with adjuvant clodronate treatment.

lenge) should exacerbate the patient’s condition. N Engl J Med 1998; 339: 357-6318. Fietta P, Manganelli P, Lodigiani L. Clodronate induced uveitis.Knowledge of drugs with potential to produce ad-

Ann Rheum Dis 2003; 62: 378verse ocular reactions should be updated as knowl- 19. Thiebaud D, Sauty A, Burckhardt P, et al. An in vitro and in vivoedge of these adverse outcomes arises. study of cytokines in the acute-phase response associated with

bisphosphonates. Calcif Tissue Int 1997; 61: 386-9220. Mbekeani JN, Slamovits TL, Schwartz BH, et al. Ocular inflam-Acknowledgements

mation associated with alendronate therapy. Arch Ophthalmol1999; 117: 837-8This study was supported in part by an unrestricted grant

21. Malik AR, Campbell SH, Toma NM. Bilateral acute anteriorto Casey Eye Institute from Research to Prevent Blindness,uveitis after alendronate. Br J Ophthalmol 2002; 86: 1443

New York, NY and by the National Registry of Drug-Induced22. Siris ES. Bisphosphonates and iritis. Lancet 1993; 341: 436-7

Ocular Side Effects (www.eyedrugregistry.com). This report 23. Vinas G, Olive A, Holgado S, et al. Episcleritis secondary todid not require approval by the OHSU Institutional Review risedronate. Med Clin 2002 Apr 27; 118: 598-9Board. We would like to thank Genevieve Long for her 24. Watson PG. The nature and the treatment of scleral inflamma-

tion. Trans Ophthalmol Soc UK 1982; 102: 257-81assistance in the editing and modifications of this paper. The



25. Fraunfelder FW, Fraunfelder FT, Keates EU. Topiramate-asso- 47. Butler WH. The neuropathology of vigabatrin. Epilepsia 1989;ciated acute, bilateral, secondary angle-closure glaucoma. 30 Suppl. 3: S15-17Ophthalmology 2004 Jan; 111 (1): 109-11 48. Neal MJ, Cunningham JR, Shah MA, et al. Immunocytochemi-

26. Thambi L, Kapcala LP, Chambers W, et al. Topiramate-associ- cal evidence that vigabatrin in rats causes GABA accumulationated secondary angle-closure glaucoma: a case series. Arch in glial cells of the retina. Neurosci Lett 1989; 89: 29-32Ophthalmol 2002 Aug; 120 (8): 1108 49. Daneshvar H, Racette L, Coupland SG, et al. Symptomatic and

27. Banta JT, Hoffman K, Budenz DL, et al. Presumed topiramate- asymptomatic visual loss in patients taking vigabatrin. Oph-induced bilateral acute angle-closure glaucoma. Am J thalmology 1999; 106: 1792-8Ophthalmol 2001 Jul; 132: 112-4 50. Kaufman KR, Lepore FE, Keyser BJ. Visual fields and tia-

28. Bovino JA, Marcus FF. The mechanism of transient myopia gabine: a quandary. Seizure 2001; 10: 525-9induced by sulfonamide therapy. Am J Ophthalmol 1982 Jul; 51. Krauss G, Johnson M, Sheth S, et al. A controlled study compar-94: 99-101 ing visual function in patients treated with vigabatrin and

29. Chirls IA, Norris JW. Transient myopia associated with vaginal tiagabine. J Neurol Neurosurg Psychiatr 2003; 74: 339-43sulfanilamide suppositories. Am J Ophthalmol 1984 Jul 15; 52. Nousiainen I, Mantyjarvi M, Kalviainen R. Visual function in98: 120 patients treated with the GABAergic anticonvulsant drug tia-

30. Hook SR, Holladay JT, Prager TC, et al. Transient myopia gabine. Clin Drug Invest 2000; 20: 393-400induced by sulfonamides. Am J Ophthalmol 1986 Apr 15; 101: 53. Sorri I, Kalviainen R, Mantyjarvi M. Color vision and contrast495-6 sensitivity in epilepsy patients treated with initial tiagabine

31. Kimura R, Kasai M, Shoji K, et al. Swollen ciliary processes as monotherapy. Epilepsy Res 2005 Dec; 67 (3): 101-7an initial symptom in Vogt-Koyanagi-Harada syndrome. Am J 54. Fraunfelder FW, Fraunfelder FT, Corbett JJ. Isotretinoin-associ-Ophthalmol 1983 Mar; 95: 402 ated intracranial hypertension. Ophthalmology 2004 Jun; 111

32. Ben-Menachem E. Vigabatrin. Epilepsia 1995; 36 Suppl. 2: (6): 1248-50S95-104

55. Fraunfelder FT, Fraunfelder FW. Ocular side effects possibly33. Marson AG, Kadir ZA, Hutton JL, et al. The new antiepileptic associated with isotretinoin usage. Am J Ophthalmol 2001;

drugs: a systemic review of their efficacy and tolerability. 132 (3): 299-305Epilepsia 1997 Aug; 38: 859-80

56. Cchroeter T, Lanvers C, Herding H, et al. Pseudotumor cerebri34. Chiron C, Dulac O, Luna D, et al. Vigabatrin in infantile induced by all-trans-retinoic acid in a child treated for acute

spasms. Lancet 1990 Feb 10; 335: 363-4 promyelocytic leukemia. Med Pediatr Oncol 2000; 34: 284-635. Hancock E, Osborne JP. Vigabatrin in the treatment of infantile 57. Sano F, Tsuji K, Kunika N, et al. Pseudotumor cerebri in a

spasms in tuberous sclerosis: literature review. J Child Neurol patient with acute promyelocytic leukemia during treatment1999 Feb; 14: 71-4 with all-trans-retinoic acid. Intern Med 1998; 37 (6): 546-9

36. Meldrum BS. GABAergic mechanisms in the pathogenesis and58. Tiamkao S, Sirijirachai C. Pseudotumor cerebri caused by all-treatment of epilepsy. Br J Clin Pharmacol 1989; 27: 3-11S

trans-retinoic acid: a case report. J Med Assoc Thai 2000; 8337. Tanganelli P, Regesta G. Vigabatrin vs. carbamazepine (11): 1420-3

monotherapy in newly diagnosed focal epilepsy: a randomized59. Fraunfelder FW, Fraunfelder FT. Evidence for a probable causalresponse conditional crossover study. Epilepsy Res 1996; 25:

relationship between tretinoin, acitretin, and etretinate and257-62intracranial hypertension. J Neuroophthalmol 2004 Sep; 24

38. Elke T, Talbot JF, Lawden MC. Severe persistent visual field (3): 214-6constriction associated with vigabatrin. BMJ 1997; 314: 180-1

60. Guirgis MF, Lueder GT. Intracranial hypertension secondary to39. Wild JM, Martinez C, Reinshagen G, et al. Characteristics of a all-trans retinoic acid treatment for leukemia: diagnosis and

unique visual field defect attributed to vigabatrin. Epilepsia management. J AAPOS 2003; 7 (6): 432-41999; 40: 1784-94

61. Morrice G, Havener WH, Kapetansky F. Vitamin A intoxication40. Kalviainen R, Nousiainen I, Mantyjarvi M, et al. Vigabatrin, aas a cause of pseudotumor cerebri. JAMA 1960; 173: 1802-5GABAergic antiepileptic drug, causes concentric visual field

62. Jacobson DM, Berg R, Wall M, et al. Serum vitamin A concen-defects. Neurology 1999; 53: 922-66tration is elevated in idiopathic intracranial hypertension. Neu-41. Miller NR, Johnson MA, Paul SR, et al. Visual dysfunction inrology 1999; 53 (5): 1114-8patients receiving vigabatrin. Neurology 1999; 53: 2082-7

63. Visani G, Manfroi S, Tosi P, et al. All-trans-retinoic acid and42. Daneshvar H, Racette L, Coupland SG, et al. Symptomatic andpseudotumor cerebri. Leuk Lymphoma 1996; 23: 437-42asymptomatic visual loss in patients taking vigabatrin. Oph-

64. Roytman M, Frumkin A, Bohn TG. Pseudotumor cerebri causedthalmology 1999; 106: 1792-8by isotretinoin. Cutis 1988; 42: 399-40043. Harding GFA, Wild JM, Robertson KA, et al. Separating the

65. Lee AG. Pseudotumor cerebri after treatment with tetracyclineretinal electrophysiologic effects of vigabatrin: treatment ver-and isotretinoin for acne. Cutis 1995; 55 (3): 165-8sus field loss. Neurology 2000; 55: 347-52

66. Donahue SP. Recurrence of idiopathic intracranial hypertension44. Gross-Tsur V, Banin E, Shahar E, et al. Visual impairment inafter weight loss: the carrot craver. Am J Ophthalmol 2000;children with epilepsy treated with vigabatrin. Ann Neurol130 (6): 850-12000; 48: 60-4

67. Hoffner SE, Kallenius G, Beezer AE, et al. Studies on the45. Mustafa BA. Diversity of GABA receptors in the vertebratemechanisms of the synergistic effects of ethambutol and otherouter retina. Trends Neurosci 1995; 85: 118-20antibacterial drugs on Mycobacterium avium complex. Acta46. Pow DV, Baldridge W, Crook DK. Activity-dependent transportLeprol 1989; 7 Suppl. 1: 195-9of GABA analogues into specific cell types demonstrated at

high resolution using a novel immunocytochemical strategy. 68. Sadun AA. Metabolic optic neuropathies. Semin OphthalmolNeuroscience 1996; 73: 1129-43 2002; 17 (1): 29-32



69. Kaimbo WK, Bifuko ZA, Longo MB, et al. Color vision in 42 87. Tullio CJ. Ibuprofen-induced visual disturbance. Am J HospCongolese patients with tuberculosis receiving ethambutol Pharm 1981; 38: 1362treatment. Bull Soc Belge Ophtalmol 2002; 284: 57-61

88. Nicastro NJ. Visual disturbances associated with over-the-70. Karmon G. Bilateral optic neuropathy due to combined counter ibuprofen in three patients. Ann Ophthalmol 1989; 21:

ethambutol and isoniazid treatment. Ann Ophthalmol 1979;447-5011: 1013

89. Bovino JA, Marcus FF. The mechanism of transient myopia71. Kass I, Mandel W. Isoniazid as a cause of optic neuritis and

induced by sulfonamide therapy. Am J Ophthalmol 1982; 94:atrophy. JAMA 1957; 164: 1740-399-101

72. Kozak SF, Inderlied CB, Hsu HY, et al. The role of copper on90. Lutz EG. Allergic conjunctivitis due to diazepam. Am J Psychethambutol’s antimicrobial action and implications for

1975; 132 (5): 548ethambutol-induced optic neuropathy. Diagn Microbiol InfectDis 1998; 30: 83-7 91. Fraunfelder FW, Fraunfelder FT, Illingworth DR. Adverse ocu-

73. Sadun AA. Metabolic optic neuropathies. Semin Ophthalmol lar effects associated with niacin therapy. Br J Ophthalmol2002 17(1):29-32 1995; 79: 54-6

74. Feiner LA, Younge BR, Kazmier FJ, et al. Optic neuropathy and 92. Gass JDM. Nicotinic acid maculopathy. Am J Ophthalmolamiodarone therapy. Mayo Clin Proc 1987; 62: 702-17 1973; 76: 500-10

75. Macaluso DC, Shults WT, Fraunfelder FT. Features of ami- 93. Jampol LM. Niacin maculopathy. Ophthalmology 1988; 95:odarone-induced optic neuropathy. Am J Ophthalmol 1999; 1704-5127 (5): 610-2

94. Fraunfelder FW. Ocular side effects associated with dietary76. Mantyjarvi M, Tuppurainen K, Ikaheimo K. Ocular side effects

supplements and herbal medicines. Drugs Today 2005; 41 (8):of amiodarone. Surv Ophthalmol 1998; 42: 360-6537-45

77. Marmor MF, Carr RE, Easterbrook E, et al. Recommendations95. Physicians’ Desk Reference for herbal medicines. 2nd rev. ed.on screening for chloroquine and hydroxychloroquine reti-

Montvale (NJ): Thomson Healthcare, 2004nopathy: a report by the American Academy of Ophthalmolo-gy. Ophthalmology 2002; 109: 1377-82 96. Eisenberg DM, Davis RB, Ettner SL, et al. Trends in alternative

medicine use in the United States, 1990-1997: results of a78. Shroyer NF, Lewis RA, Lupski JR. Analysis of the ABCR(ABCA4) gene in 4 aminoquinoline retinopathy: is retinal follow-up national survey. JAMA 1998 Nov 11; 289 (18):toxicity by chloriquinine and hydroxychlorquine related to 1569-75Stargardt disease? Am J Ophthalmol 2001 Jun; 131 (6): 761-6

97. Fraunfelder FW. Ocular side effects from herbal medicines and79. Fraunfelder FW. Visual side effects associated with erectile

nutritional supplements. Am J Ophthalmol 2004 Oct; 138 (4):dysfunction agents. Am J Ophthalmol 2005 Oct; 140 (4):639-47723-4

98. Espaillat A. Canthaxanthine retinopathy. Arch Ophthalmol80. Reivich M, Holling HE, Roberts B, et al. Reversal of blood flow1999; 113: 412-3through the vertebral artery and its effect on cerebral circula-

tion. N Engl J Med 1961 Nov 2; 265: 878-85 99. Subiza J. Allergic conjunctivitis to chamomile tea. Ann Allergy

Asthma Immunol 1990; 65: 127-3281. Laties AM, Fraunfelder FT. Ocular safety of Viagra (sildenafilcitrate). Trans Am Ophthalmol Soc 1999; 97: 115-28 100. Blumenthal M, editor. The complete German Commission E

82. Gorin MB, Day R, Costantino JP, et al. Long-term tamoxifen monographs therapeutic guide to herbal medicines. Austincitrate use and potential ocular toxicity. Am J Ophthalmol (TX): American Botanical Council, 19981998; 125: 493-501

101. Grimm W, Muller H. A randomized controlled trial of the effect83. Gorin MB, Costantino JP, Kulacoglu DN, et al. Is tamoxifen a of fluid extract of echinacea purpurea on the incidence and

risk factor for retinal vaso-occlusive disease? Retina 2005; 25:severity of colds and respiratory infections. Am J Med 1999523-6Feb; 106: 138-43

84. Fraunfelder FW, Solomon J, Mehelas TJ. Ocular side effects102. Dobbins KR, Saul RF. Transient visual loss after licorice inges-associated with cyclooxygenase-2 inhibitors. Arch

tion. J Neurophthalmol 2000 Mar; 20: 38-41Ophthalmol 2006; 124: 277-279

85. Coulter DM, Clark DWJ, Savage RL. Celecoxib, rofecoxib, andacute temporary visual impairment. BMJ 2003; 327: 1214-5

Correspondence and offprints: Dr Frederick W. Fraunfelder,86. Haefliger IO, Meyer P, Flammer J, et al. The vascular endotheli-

Casey Eye Institute, 3375 SW Terwilliger Blvd, Portland,um as a regulator of the ocular circulation: a new concept inophthalmology. Surv Ophthalmol 1994; 39: 123-32 OR 97201, USA.


Drugs 2007; 67 (1): 95-106ADIS DRUG PROFILE 0012-6667/07/0001-0095/$49.95/0


Fixed-Dose CombinationLercanidipine/EnalaprilPhilip I. Hair, Lesley J. Scott and Caroline M. Perry

Wolters Kluwer Health | Adis, Auckland, New Zealand, an editorial office of Wolters KluwerHealth, Conshohocken, Pennsylvania, USA

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951. Pharmacodynamic Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962. Pharmacokinetic Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983. Therapeutic Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994. Tolerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025. Dosage and Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1046. Lercanidipine/Enalapril: Current Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Features and properties of fixed-dose combinationlercanidipine/enalapril (Zanipress, Zanitek)

IndicationAbstractHypertension not adequately controlled by monotherapy with

▲ Lercanidipine, a dihydropyridine calcium channel lercanidipine or enalaprilblocker, and enalapril, an ACE inhibitor, are estab- Mechanism of actionlished antihypertensive agents. A fixed-dose tablet Lercanidipine inhibits calcium entry through L-type channels into

vascular smooth muscle cells; enalaprilat, the active metaboliteformulation of lercanidipine/enalapril is approvedof enalapril, inhibits angiotensin converting enzyme, therebyin Germany for the treatment of hypertension inreducing levels of angiotensin IIpatients not responding to monotherapy.Dosage and administration

▲ Lercanidipine/enalapril 10mg/10mg once daily sig-Recommended dose Lercanidipine/enalapril 10mg/nificantly reduced sitting diastolic blood pressure 10mg in nonresponders to

and sitting systolic blood pressure, relative to ler- lercanidipine; lercanidipine/enalapril 10mg/20mg incanidipine 10mg once daily, in a 12-week, ran-nonresponders to enalaprildomised, double-blind trial in patients with mild to

Route of administration Oralmoderate hypertension who had previously not re-Frequency of administration Once daily, ≥15 min before asponded to 4 weeks’ treatment with lercanidipine.

meal▲ In a similarly designed trial, lercanidipine/enalapril Steady-state pharmacokinetics of S-lercanidipine (S-LER),

10mg/20mg once daily was significantly more ef- enalapril (EN) and enalaprilat (ENL) in patients (n = 20) withessential hypertension (lercanidipine/enalapril 10mg/20mgfective than enalapril 20mg once daily in hyperten-capsule once daily for 8 days)sive patients who had previously not responded toPeak plasma concentration 3.6; 91; 53 µg/Lenalapril monotherapy.(S-LER; EN; ENL)

▲ Fixed-dose lercanidipine/enalapril was generallyTime to peak plasma 1; 1; 4 hourswell tolerated, with a tolerability profile similar to concentration

that of either of the individual drugs alone or place- (S-LER; EN; ENL)bo. Cough was reported in ≤5.2% and peripheral Adverse eventsoedema in ≤1.5% of lercanidipine/enalapril recipi- Most frequent Cough, headache, dizziness,

peripheral oedema, flushing,ents.vertigo

96 Hair et al.

Hypertension is a significant factor in the devel- Lercanidipine[8,9] and enalapril[10,11] are well es-opment of cerebrovascular disorders, heart disease tablished therapies for hypertension, and have beenand renal failure, and is estimated to contribute to reviewed previously in Drugs. This review focuses≈5% of the global disease burden.[1] The conse- on the use of oral fixed-dose lercanidipine/enalaprilquences of hypertension represent a continuum of a (Zanipress, Zanitek)1 in the treatment of adultprogressive cardiovascular syndrome, rather than patients with essential hypertension.just elevated blood pressure (BP) alone, leading to

1. Pharmacodynamic Profilestructural and functional changes in the heart andvascular system.[2] The aim of treatment is to reduce

No published pharmacodynamic data are availa-the long-term risk of cardiovascular morbidity andble for the combination of lercanidipine andmortality, by targeting not only raised BP but alsoenalapril; consequently, a brief overview of therisk factors such as diabetes mellitus.[3]

pharmacodynamics of the individual drugs, basedReduction of systolic BP (SBP) to the desired largely on previous reviews,[8-11] is provided in this

target value (<140mm Hg, or <130mm Hg in pa- section.tients with diabetes mellitus or renal disease) is noteasily achieved with a single antihypertensive medi- Lercanidipinecation; a combination of drugs is required in a largeproportion of patients to achieve this goal.[3-5] A ● Lercanidipine inhibits calcium entry through L-combination regimen is more likely to achieve BP type calcium channels, for which the S-enantiomercontrol if the component drugs have different mech- has 100- to 200-fold higher affinity than the R-anisms of action, since elevated BP is the result of a enantiomer, in smooth muscle cells of the cardiovas-multifactorial process.[3-5] cular system, resulting in peripheral vasodilation

and reduced BP.[8,9,12] Antihypertensive effects haveCombining a calcium channel blocker (CCB)a gradual onset and long duration, relative to thosewith an ACE inhibitor is one of the combinationof other CCBs such as nifedipine and ni-regimens recommended in several antihypertensivetrendipine.[8,9]guidelines;[3-5] one such potential regimen is fixed-● In vitro data indicate that lercanidipine is highlydose lercanidipine/enalapril. Lercanidipine, a dihy-

selective for vascular smooth muscle, relative todropyridine CCB, induces vascular relaxation, withcardiac tissue,[8,9] possibly as a result of a greatera consequent decline in vascular resistance and arte-proportion of L-type Ca2+ channels existing in arial pressure.[5] However, this class of drugs is alsohigh-affinity inactivated state that is readily blockedlikely to have some undesirable effects (e.g. periph-by lercanidipine.[13]eral oedema), as a result of activation of both the

sympathetic nervous system and the renin-an- ● Lercanidipine may have antiatherogenic effectsgiotensin-aldosterone axis.[6] The concomitant ad- that are independent of its BP-lowering ef-ministration of an ACE inhibitor, such as enalapril, fects.[8,9,14-17] In hypertensive patients with coexist-may counteract these unwanted effects.[6] A further ing type 2 diabetes, lercanidipine 10 mg/day for 16advantage of ACE inhibitors is their potential to weeks reduced levels of low-density lipoprotein-reduce the incidence of new-onset diabetes.[7] Fur- cholesterol oxidation (measured as conjugated dienethermore, administering the agents as an individual formation) by 35% relative to baselinefixed-dose tablet or gelatin capsule is likely to pro- (p < 0.001).[16] The drug also delayed the develop-vide several potential advantages over separate ad- ment of fatty streaks and hyperplastic lesions in aministration, such as lower dose, improved patient rabbit model,[14] and inhibited smooth muscle cellcompliance and greater cost effectiveness.[6] proliferation and migration[15] and the release of

1 The use of trade names is for product identification purposes only and does not imply endorsement.


Fixed-Dose Combination Lercanidipine/Enalapril: Adis Drug Profile 97

metalloproteinases and esterification of cholesterol effect occurring 6–8 hours after administration;in macrophages[17] in vitro. higher doses prolong the duration of the BP-lower-

ing effect to a total of 24–36 hours.[10]● Lercanidipine appears to have no clinically rele-

vant effects on heart rate or ECG parameters in ● Enalapril 10 or 20 mg/day (n = 7) for 12 monthspatients with hypertension.[9] In addition, there was decreased both the media-to-lumen ratio (fromno significant change from baseline in left ventricu- 0.117 to 0.093; p < 0.01) in subcutaneous smalllar mass in 23 patients with hypertension who re- arteries and left ventricular mass index (from 94.8 toceived 6 months’ treatment with lercanidipine 78.7 g/m2; p < 0.05) in patients with hypertension(10mg once daily for the first 4 weeks and then 10 or and type 2 diabetes (n = 15).[24] Small artery20mg once daily).[18] remodelling is regarded as an early indicator of end-● Lercanidipine 10[19,20] or 20[20] mg/day also ap- organ damage as it not only increases peripheral

pears to have renoprotective effects; creatinine resistance but also impedes blood supply to organsclearance (CLCR) was increased by 10% (p = 0.019) such as heart, brain and kidney, thereby contributingin hypertensive patients with renal dysfunction to complications of hypertension.[25]

(CLCR <80 mL/min [<4.8 L/h]) over 6 months,[19]● Enalapril 10 or 20 mg/day changed carotid artery

and albumin excretion was reduced by 20% intima-media thickness (IMT) by –0.080mm (95%(p < 0.05) in hypertensive patients with microalbu- CI –0.097, –0.063) in hypertensive patients after 2minuria and type 2 diabetes over 12 months.[20] In years’ treatment (abstract presentation),[26] andspontaneously hypertensive rats, lercanidipine in- enalapril 10 mg/day (n = 48) halved the annualhibited glomerular hypertrophy (18%),[21] vasodilat- increase in IMT in patients with type 2 diabetesed afferent (23% reduction in wall-to-lumen ratio) relative to a control group (n = 50) that did notand efferent (19% reduction in wall-to-lumen ratio) receive enalapril (0.01 vs 0.02mm; p < 0.05).[27]

glomerular arterioles[22] and reduced albuminuria Increased IMT is associated with increased risk of(28%)[22] [all p < 0.05 vs baseline]. cardiovascular and cerebrovascular disease.[27]

● Lercanidipine 10–30mg once daily for 24–48● Enalapril improves some atherosclerotic risk fac-

weeks had neutral or favourable effects on lipids intors; an index of platelet activation (the ratio of β-

patients with hypertension, while doses of 10 or 20thromboglobulin to platelet factor 4) was signifi-

mg/day improved glucose tolerance and reducedcantly decreased (p < 0.05 vs baseline) at a dosage

fasting blood glucose, glycosylated haemoglobinof 20 mg/day for ≥4 weeks in 12 patients with

and serum fructosamine levels in hypertensive pa-essential hypertension,[28] and plasma levels of inter-

tients with type 2 diabetes after 8 weeks.[9]

cellular adhesion molecule-1 were significantly re-duced (p < 0.01 vs baseline) at a dosage of 10,[29] 20

Enalaprilor 40[30] mg/day for 24 weeks in 57 patients withessential hypertension and type 2 diabetes[29] and 21● Enalapril is an orally administered prodrug that ispatients with hypertension (abstract presenta-hydrolysed to form the active ACE inhibitortion).[30]enalaprilat (see section 2), reducing plasma levels of

the potent vasoconstrictor angiotensin II, with a ● Potentiation of the effects of bradykinin mayconsequent decrease in aldosterone secretion and an contribute to the apparent cardiovascular protectionincrease in plasma renin activity.[11] The reduction in associated with enalapril treatment.[31] In patientsangiotensin II levels results in peripheral vasodila- with atypical chest pain, enalapril 10mg once dailytion and reduced vascular resistance, leading to de- for 7 days (n = 25) significantly (p < 0.0001) in-creased BP.[10,11,23] Enalapril has little effect on heart creased bradykinin-induced release of tissue plasmi-rate or cardiac output.[11] nogen activator in the coronary circulation relative● Enalapril dose-dependently reduces BP in hyper- to that in a control group (n = 31) that did not receive

tensive patients at doses ≤10mg, with the maximum enalapril.[31] Enalapril may inhibit degradation of


98 Hair et al.

bradykinin, as angiotensin converting enzyme also es on the well established pharmacokinetic profilesfunctions as kininase II.[23] of the individual agents, lercanidipine[8,9] and

enalapril,[10,11] including data from the prescribing● ACE inhibitors have been reported to have re-information.[12,23]noprotective effects by reducing glomerular capilla-

ry pressure.[32] In a randomised, double-blind, 6- Lercanidipineyear trial in patients with type 2 diabetes and nor-

moalbuminuria at baseline, 5 of 77 (6.5%) enalaprilSince the pharmacological activity of racemic5–10 mg/day recipients and 15 of 79 (19.0%) place-

lercanidipine is mainly attributed to the S-enanti-bo recipients developed microalbuminuria (urinaryomer (section 1), and no in vivo interconversion ofalbumin excretion >30 mg/24h) [p = 0.001]; long-the S- and R-enantiomers has been demonstrated,[12]

acting CCBs (diltiazem or verapamil) and/ordiscussion in this section focuses on thehydrochlorothiazide were used for BP control whenpharmacokinetic profile of S-lercanidipine.SBP/diastolic BP (DBP) was consistently ≥145/● After oral administration of racemic ler-95mm Hg.[33]

canidipine, S-lercanidipine is completely absorbed● However, enalapril 10mg once daily did not af-

from the gastrointestinal tract and exhibits nonlinearfect the long-term development of nephropathy inpharmacokinetics.[9] Lercanidipine should be ad-18 patients with type 1 diabetes and albuminuria.[34]

ministered prior to meals, as the absorption of S-After 6 months, enalapril significantly (p < 0.05)lercanidipine is increased 4-fold when it is adminis-reduced the median albumin excretion rate (fromtered after a high-fat meal.[8,12]

148 to 90 µg/min), but this decrease was not main-● Lercanidipine exhibits a high level of plasmatained at 3 years.[34]

protein binding (>98%) and is lipophilic, aggregat-ing in lipid membranes of arterial wall cells.[8,9,12]

2. Pharmacokinetic ProfileThe magnitude of the apparent volume of distribu-tion (2–2.5 L/kg for intravenous infusion of ler-The pharmacokinetic profile of a fixed-dose gela-canidipine 2mg) has been ascribed to the high li-tin capsule of lercanidipine/enalapril 10mg/20mgpophilicity of the drug.[9] In addition, the absolute(used in a phase III trial;[35] see section 3) wasbioavailability of lercanidipine in fed hypertensivesimilar to that of the individual agents, based onpatients is about 10%.[12]absorption parameters, in a multiple-dose, ran-● Lercanidipine is metabolised by cytochromedomised, double-blind, crossover study in 20 pa-

P450 (CYP) 3A4, and undergoes substantial first-tients with essential hypertension (DBP ≥95mmpass metabolism; the resulting metabolites are main-Hg).[36] At steady state (on day 8), peak plasmaly inactive.[9] The drug is excreted as metabolites inconcentration (Cmax) values of S-lercanidipine,both urine and faeces (43.8% and 50.4% after ad-enalapril and enalaprilat were 3.6, 91 and 53 µg/L,ministration of lercanidipine 20mg in healthy volun-respectively, at 1, 1 and 4 hours after administrationteers).[38] After oral administration of lercanidipineof the combination.[36] In addition, the fixed-dose10mg in patients with mild to moderate hyperten-gelatin capsule was bioequivalent to a fixed-dosesion, the mean plasma terminal elimination half-lifetablet formulation of lercanidipine/enalapril 10mg/(t1/2β) was 8.0 hours.[9]20mg (intended for marketing) in a single-dose,

randomised, open-label, crossover study in 48 ● Coadministration of lercanidipine and strong in-healthy volunteers; the confidence intervals for the hibitors of CYP3A4 is not advised,[12] and caution istablet : capsule ratios of the area under the plasma recommended when coadministering lercanidipineconcentration-time curve (AUC) and Cmax for ler- and inducers or other substrates of CYP3A4. [12] Thecanidipine enantiomers, enalapril and enalaprilat drug should not be coadministered with ciclosporinwere within the bioequivalence acceptance range of or grapefruit juice.[12] Lercanidipine may increase0.80–1.25.[37] Thus, discussion in this section focus- Cmax of digoxin; patients receiving both drugs



should be monitored for digoxin toxicity.[12] Ler- Nevertheless, concomitant use of enalapril and lithi-canidipine does not interact significantly with war- um is not recommended; if coadministration of thesefarin, simvastatin, diuretics or ACE inhibitors.[12] agents is unavoidable, then serum lithium concen-

trations should be closely monitored.[23]● The use of lercanidipine in patients with severe

hepatic or renal impairment is not recommended3. Therapeutic Efficacybecause of the risk of drug accumulation.[8,9,12] Cau-

tion is advised when increasing the dose of ler-The efficacy of oral, once-daily, fixed-dose ler-canidipine to 20mg once daily in patients with mild

canidipine/enalapril gelatin capsules has been evalu-to moderate hepatic or renal impairment.[12]

ated in adult patients with mild to moderate essentialhypertension in a randomised, double-blind, multi- Enalaprilcentre, phase II, dose-finding trial (study CPL2-

● Enalapril is de-esterified in the liver to form 0008)[41] and two 12-week, multicentre, phase IIIenalaprilat, which does not undergo further metabo- trials (studies CPL1-0018[42] and CPL1-0019[35]),lism.[10,11] The Cmax of enalaprilat is linearly related with longer-term efficacy evaluated in noncompara-to the dose of enalapril across the 2.5–40mg dose tive 9-month extensions of the phase III trials.[35,42]

range;[10,11] however, the AUC for enalaprilat is not Data are currently available only as data onlinearly related to enalapril dosage because of the file.[35,41,42]

prolonged terminal elimination phase.[10,11] These data are supported by results from a● The absolute bioavailability of oral enalapril as randomised, double-blind trial, in which patients

enalaprilat is approximately 40%,[11] and no more with mild to moderate hypertension who had notthan 60% of circulating enalaprilat is bound to responded to 4 weeks of treatment with enalapril 20human plasma proteins.[23] Enalapril crosses the pla- mg/day received lercanidipine 10 mg/day (n = 56) orcental barrier; consequently, the drug is contraindi- hydrochlorothiazide 12.5 mg/day (n = 56) as add-oncated in pregnant patients.[11,23] therapy to enalapril for 20 weeks; lercanidipine was● Enalaprilat is eliminated predominantly by the non-inferior to hydrochlorothiazide for reduction of

renal route.[10] There is an initial elimination phase sitting DBP (sDBP) at trough levels (primaryfollowed by a prolonged terminal phase, with re- endpoint; reduction of 9.3 vs 7.4mm Hg [baselinespective t1/2 values of ≈5 hours and 30–35 hours.[11] 97mm Hg]).[43]

The latter phase represents tight binding to plasma All trials of fixed-dose lercanidipine/enalaprilangiotensin converting enzyme.[11] Following multi- had an initial washout period of 11–17 days[41] or ≥5ple doses of enalapril, the effective t1/2 of enalaprilat half-lives of any previously administered antihyper-is 11 hours.[23] tensive agents,[35,42] followed by a single-blind pla-● Dose reduction of enalapril is necessary in pa- cebo run-in phase of 2 weeks[35,42] or 25–31 days.[41]

tients with renal impairment (CLCR <80 mL/min), to Mild to moderate hypertension was defined in thereduce the risk of drug accumulation.[23] However, phase II trial as sDBP values in the rangedosage adjustment is not necessary in patients with 95–109mm Hg both before and after the washouthepatic dysfunction.[11] phase and also during and after the placebo run-in● A retrospective analysis of 20 hypertensive pa- phase,[41] and in phase III trials as sitting SBP

tients demonstrated the possibility of an increase in (sSBP) in the range 140–189mm Hg and sDBP inserum lithium concentrations when it was the range 95–114mm Hg both before and after thecoadministered with ACE inhibitors, including placebo run-in period.[35,42] Exclusion criteria in-enalapril.[39] This was particularly evident in pa- cluded secondary hypertension, cardiovascular dis-tients ≥50 years old (n = 12).[39] However, in a study ease (with the exception of stable angina pectoris,in healthy volunteers (n = 9), enalapril did not uncomplicated hypertensive cardiovascular disease,significantly affect serum lithium concentrations.[40] or uncomplicated myocardial infarction at least 6


100 Hair et al.

months previously), cardiovascular surgery within ents of this fixed-dose combination achieved a simi-the previous 6 months, a history of hypertensive lar reduction from baseline in sDBP to that observedencephalopathy or cerebrovascular accident, uncon- in recipients of lercanidipine 20mg once dailytrolled or complicated diabetes, hypertensive reti- (–9.9mm Hg; 95% CI –8.6, –11.2). Mean baselinenopathy or use of any medications (except NSAIDs) sDBP was 100mm Hg in all groups.[41]

that could affect BP.[35,41,42]● Analysis of covariance indicated that monother-

There were no between-group differences in the apy with once-daily lercanidipine 10 or 20mg orrandomised trials in terms of patient characteristics enalapril 10mg provided better antihypertensive ef-at baseline (the start of the double-blind ficacy than placebo (all p < 0.05); there were nophase),[35,41,42] with the exception of one trial (CPL1- between-group differences for 5mg doses of ler-0018)[42] where the mean age in the fixed-dose com- canidipine or enalapril versus placebo.[41] Notably,bination group was 53.5 versus 51.0 years in the all fixed-dose combinations were superior to place-monotherapy group (p = 0.026). bo in terms of antihypertensive efficacy (all

p < 0.05).[41]Analyses were based on the intent-to-treat (ITT)population, which was defined as all patients who ● Patient compliance (assessed by capsule count)had received at least one dose of treatment in the was ≥98.5% in all treatment groups.[41]

double-blind phase and had ≥1 BP measurement Phase III Trials18–48 hours after this point, with the last observa-

tion carried forward in the event of premature with-The two phase III trials evaluated lercanidipine/drawal.[35,41,42] Statistical analyses were based on an

enalapril fixed-dose combination therapy in hyper-analysis of covariance, with treatment and centretensive patients who were not responding to ler-as main effects and baseline value as a covari-canidipine (study CPL1-0018)[42] or enalapril (studyate.[35,41,42]

CPL1-0019)[35] monotherapy. During the single-blind monotherapy phase, patients received either Phase II Dose-Finding Triallercanidipine 10mg once daily for 4 weeks(CPL1-0018)[42] or once-daily enalapril 10mg for 2The appropriate once-daily fixed-dose combina-weeks, with the dosage increased to 20mg oncetion was determined using various combinations ofdaily for the next 4 weeks (CPL1-0019).[35]lercanidipine (5, 10 or 20mg), enalapril (5 or 10mg)Nonresponders were defined as patients who had aand placebo, in an ITT population of 653 patientstrough sDBP of 95–109mm Hg after the single-blind(n = 51–63/group).[41] The primary endpoint was themonotherapy phase, while maintaining a troughchange in trough sDBP (measured 22–26 hourssSBP of <180mm Hg.[35,42]post-dose) from baseline after 8 weeks of double-

blind treatment, assessed using a dose-response sur- Nonresponders to lercanidipine monotherapy re-face analysis (as recommended by the European ceived once-daily lercanidipine/enalapril 10mg/Agency for the Evaluation of Medicinal Prod- 10mg (n = 165) or once-daily lercanidipine 10mgucts[44]).[41] monotherapy (n = 172) for 12 weeks.[42] Mean base-

line sSBP/sDBP was 152/100mm Hg in both● Based on the additive response surface modelgroups.[42] Nonresponders to enalapril monotherapyanalysis of the primary endpoint, once-daily ler-received once-daily lercanidipine/enalapril 10mg/canidipine/enalapril 10mg/10mg was the optimal20mg (n = 162) or once-daily enalapril 20mgfixed-dose combination.[41] The estimated change inmonotherapy (n = 163) for 12 weeks; mean baselinetrough sDBP after 8 weeks was significantly greatersSBP/sDBP was 154/99mm Hg in both groups.[35]in recipients of lercanidipine/enalapril 10mg/10mg

once daily (–10.4mm Hg; 95% CI –11.4, –9.5) than In both trials, the primary efficacy endpoint wasin lercanidipine 10mg once daily recipients the change from baseline in mean trough sDBP after(–8.2mm Hg; 95% CI –9.2, –7.3). Moreover, recipi- 12 weeks of double-blind treatment.[35,42] Secondary



endpoints included the change from baseline in sDBP and from 4 weeks onwards for sSBP. Theremean trough sSBP after 12 weeks, the proportion of was a numerically higher percentage of combinationpatients with normalised SBP (sSBP <140mm Hg), therapy, versus monotherapy, recipients withnormalised DBP (sDBP <90mm Hg) or normalised normalised DBP (48% vs 37%), normalised SBPSBP/DBP (sSBP/sDBP <140/90mm Hg) after 12 (33% vs 28%), normalised SBP/DBP (24% vs 17%)weeks, and the proportion of patients classed as or responder rate (sDBP 53% vs 43%, sSBP 41% vsresponders after 12 weeks (normalised SBP or sSBP 33%), but these differences did not attain statisticaldecreased from baseline by ≥20mm Hg; normalised significance.[35]

DBP or sDBP decreased from baseline by ≥10mm ● A subgroup analysis of elderly nonresponders toHg).[35,42] enalapril showed that the mean change from base-

line in trough sDBP at 12 weeks was significantly● Lercanidipine/enalapril 10mg/10mg once daily(both p < 0.05) greater in combination therapy,showed better antihypertensive efficacy than ler-versus monotherapy, among recipients aged >60canidipine 10mg once daily after 12 weeks in ler-years (n = 164; –10.9 vs –7.9mm Hg) or ≥65 yearscanidipine nonresponders, according to the primary(n = 101; combination therapy vs monotherapyendpoint (mean change in sDBP from baseline –7.1treatment difference of –2.6mm Hg).[35] In thesevs –4.3mm Hg; p < 0.001).[42] The mean change insubgroups, the treatment differences for the changesSBP from baseline was also significantlyfrom baseline in trough sSBP did not reach statisti-(p < 0.001) greater in the combination therapy groupcal significance (combination therapy vs monother-(–7.7 vs –2.3mm Hg). Significantly (all p < 0.001)apy treatment difference of –3.2 and –2mm Hg ingreater reductions in sDBP and sSBP in the fixed-patients aged >60 and ≥65 years).[35]

dose combination group than in the monotherapygroup occurred after 2 weeks and were maintainedthroughout the 12 weeks of the study.[42]

Extension Phase● In lercanidipine nonresponders, after 12 weeks’

treatment, a significantly greater proportion of pa- Patients were eligible for inclusion in thetients receiving the fixed-dose combination had noncomparative extension phase of each of thenormalised SBP (39% vs 22%; p < 0.001), normal- phase III trials if they had completed the double-ised DBP (29% vs 19%; p = 0.023) and/or normal- blind treatment phase with an sSBP in the range ofised SBP/DBP (22% vs 12%; p = 0.012) than ler- 120–179mm Hg and an sDBP in the range ofcanidipine monotherapy recipients.[42] The respond- 80–109mm Hg.[35,42] This phase was primarily de-er rate was also higher in combination therapy signed to evaluate long-term tolerability (see sectionrecipients than in lercanidipine monotherapy recipi- 4).ents (sSBP 41% vs 24%, p < 0.001; sDBP 35% vs In the extension phase in nonresponders to ler-24%, p = 0.032).[42]

canidipine monotherapy, 221 patients (110 fixed-● Similarly, in enalapril nonresponders, the fixed- dose combination and 111 lercanidipine recipients)

dose combination was more effective than enalapril received lercanidipine/enalapril 10mg/10mg oncemonotherapy. After 12 weeks, the mean changes in daily, with a switch to lercanidipine/enalapril 10mg/trough sDBP from baseline were –9.2 and –7.5mm 20mg once daily permitted if sSBP/sDBP was >140/Hg in recipients of lercanidipine/enalapril 10mg/ 90mm Hg; 201 patients completed 9 months’ treat-20mg once daily and enalapril 20mg once daily ment in the extension phase.[42] In the extension(p = 0.015).[35] The corresponding mean changes in phase in nonresponders to enalapril monotherapy,sSBP from baseline were –9.8 and –6.7mm Hg 186 patients (94 fixed-dose combination and 92(p = 0.013). Significant (all p < 0.05) between-group enalapril recipients) received lercanidipine/enalaprildifferences in favour of fixed-dose combination 10mg/20mg once daily; 164 patients completed 9therapy were evident from 8 weeks onwards for months’ treatment.[35]


102 Hair et al.

● Among lercanidipine monotherapy nonrespon- patients receiving lercanidipine/enalapril 10mg/ders, 52% of extension-phase combination therapy 20mg per day in the pooled analysis.[45] Commonrecipients achieved a normalised SBP, 46% a treatment-emergent adverse events are summarisednormalised DBP and 37% a normalised SBP/DBP in figure 1.[45]

after 9 months.[42] Among patients switched from● Discontinuation of treatment because of an ad-

lercanidipine/enalapril 10mg/10mg to lercanidipine/ verse event occurred with a similar frequency in theenalapril 10mg/20mg during the extension phase combination therapy and monotherapy groups, with(n = 133; 60%), 35% achieved a normalised SBP,

the incidence in lercanidipine/enalapril 10mg/10mg33% a normalised DBP and 20% a normalised SBP/

and 10mg/20mg groups being 4.6% and 4.4%;DBP.[42]

cough was the most common cause of treatment● Similarly, among nonresponders to enalapril discontinuation (0.9% and 2.2%).[45]

monotherapy, 40% achieved a normalised SBP,● Cough and dizziness were two of the most com-60% a normalised DBP and 36% a normalised SBP/

mon treatment-emergent adverse events that oc-DBP in the extension phase.[35]

curred with both lercanidipine/enalapril dosages(figure 1).[45] Cough was attributed to the enalapril4. Tolerability

The fixed-dose formulation of lercanidipine/enalapril was generally well tolerated for ≤52 weeksby patients with mild to moderate hypertension inthe phase II and III trials[35,41,42] discussed in section3. This section focuses on a pooled safety analy-sis[45] of all patients who received at least one doseof lercanidipine/enalapril 10mg/10mg or 10mg/20mg in these trials[35,41,42] or in a pharmacokineticinteraction trial[36] discussed in section 2.● In the 8-week dose-finding trial (section 3), treat-

ment-emergent adverse events occurred with similarfrequencies in the fixed-dose combination groups tothose observed in the lercanidipine and enalaprilmonotherapy or placebo groups.[41] The proportionof patients experiencing any treatment-emergent ad-verse event was 25.0–42.2% in the treatment groupsand 36% in the placebo group. The most frequenttreatment-related adverse events were headache(1.7–7.4%), coughing (1.7–9.8% in enalapril recipi-ents and 0–1.9% in those not receiving enalapril)and dizziness (frequency not reported).[41] One ler-canidipine/enalapril 20mg/10mg recipient exper-ienced a serious adverse event (circulatory collapse)that was considered to be probably treatment-relat-ed.[41]

● Treatment-emergent adverse events consideredto be at least possibly related to therapy occurred in39 of 329 (11.9%) patients receiving lercanidipine/enalapril 10mg/10mg per day and 69 of 410 (16.8%)

0 21 43 5 6 7

1 3 5 7

8

Cough

Dizziness

Vertigo 1.2

1.2

1.5

2.4

2.7

4.4

1.5

5.2

a

b

0 2 4 6 8

Cough

Headache

Dizziness

Peripheral oedema

Flushing

Patients (%)

LER10/ENA10

LER10/ENA20

Fig. 1. Tolerability profile of once-daily, orally administered, fixed-dose lercanidipine/enalapril in patients with mild to moderate es-sential hypertension receiving (a) lercanidipine/enalapril 10mg/10mg per day (LER10/ENA10; n = 329) or (b) lercanidipine/enalapril 10mg/20mg per day (LER10/ENA20; n = 410) for ≤52weeks. Treatment-emergent adverse events considered to be atleast possibly related to therapy occurring in ≥1% of patients, usingpooled data[45] from an 8-day, double-blind, pharmacokinetic inter-action study,[36] an 8-week, double-blind, dose-finding study[41] andtwo 12-week, double-blind, phase III trials with 9-month, noncom-parative extension phases.[35,42]



Cho

lest

erol

(m

mol

/L)

Trig

lyce

rides

(m

mol

/L)

Glu

cose

(m

mol

/L)

Screening Baseline Endpoint Screening Baseline Endpoint

7

6

5

4

3

2

1

0

7

6

5

4

3

a

b

c

LER10/ENA10LER10

LER10/ENA20ENA20

Fig. 2. Effect on plasma lipids and glucose of once-daily, orally administered, fixed-dose lercanidipine/enalapril or component drug inpatients with mild to moderate essential hypertension. Mean values for (a) cholesterol, (b) triglycerides and (c) glucose in recipients oflercanidipine/enalapril 10mg/10mg per day (LER10/ENA10) [n = 167], lercanidipine 10 mg/day (LER10) [n = 175], lercanidipine/enalapril10mg/20mg per day (LER10/ENA20) [n = 163] or enalapril 20 mg/day (ENA20) [n = 164] at screening, baseline and endpoint (the lastmeasurement obtained during the 12-week double-blind treatment period). The normal ranges (fasting conditions) were <5.698 mmol/L(cholesterol), 0.56–2.15 mmol/L (triglycerides) and 3.33–5.55 mmol/L (glucose).[48] Data were derived from two randomised, double-blindtrials: CPL1-0018,[46] in which patients received LER10/ENA10 or LER10, and CPL1-0019,[47] in which patients received LER10/ENA20 orENA20.


104 Hair et al.

component, whereas peripheral oedema and flush- shown to effectively lower BP in hypertensive pa-ing were most likely due to lercanidipine-induced tients inadequately controlled by either componentperipheral vasodilation. Dizziness and vertigo were drug, and was generally well tolerated. Only a smallconsidered to be possibly linked to the BP-lowering number of patients discontinued treatment becauseeffect of the combination.[45] of an adverse event; simplification of the dosage

regimen to a single once-daily capsule resulted in● Serious adverse events occurred in 1.5% of ler-excellent compliance in the large phase II trial.canidipine/enalapril 10mg/10mg per day and 3.2%

of lercanidipine/enalapril 10mg/20mg per day recip- The absence of negative effects of the combina-ients in the pooled analysis, but all were considered tion on lipid and glucose metabolism (section 4) isto be either unrelated to treatment or unlikely to be also important, as glucose intolerance, diabetes andrelated to treatment.[45] hyperlipidaemia are additional risk factors that are

frequently present in the hypertensive population.● In the phase III trials (section 3), the proportionof patients with clinically significant abnormalities The tablet formulations intended for marketingin plasma levels of lipids or glucose was low (≤8%) have been shown to be bioequivalent to the gelatinat both baseline and endpoint (the last measurement capsule formulations used in pivotal clinical trials,obtained during the 12-week treatment period) with and offer the convenience of single-tablet adminis-lercanidipine/enalapril 10mg/10mg per day[42] or tration of a combination therapy, with potential10mg/20mg per day.[35] The mean values for plasma compliance benefits. These fixed-dose formulationslipids and glucose at screening, baseline and end- are now approved in Germany for the treatment ofpoint in the phase III trials are shown in figure hypertension in patients with BP not adequately2.[46,47] controlled with lercanidipine[49] or enalapril[50]

alone.5. Dosage and Administration

DisclosureFixed-dose combination lercanidipine/enalapril

is indicated for the treatment of hypertension in During the peer review process, the manufacturer of theagent under review was also offered an opportunity to com-patients aged ≥18 years;[49,50] the 10mg/10mg com-ment on this article; changes based on any comments receivedbination is used in patients in whom adequate BPwere made on the basis of scientific and editorial merit.

control has not been achieved with lercanidipinemonotherapy[49] and the 10mg/20mg combination is

Referencesused in patients not achieving BP control with1. Whitworth JA, World Health Organization, International Socie-enalapril monotherapy.[50] The fixed-dose combina-

ty of Hypertension Writing Group. 2003 World Health Organi-tion tablet is taken once daily at least 15 minutes zation (WHO)/International Society of Hypertension (ISH)

statement on management of hypertension. J Hypertens 2003;before a meal.[49,50]

21(11): 1983-92Lercanidipine/enalapril 10mg/10mg or 10mg/ 2. Giles TD, Berk BC, Black HR, et al. Expanding the definition

and classification of hypertension. J Clin Hypertens 2005 Sep;20mg tablets should not be administered during7 (9): 505-12pregnancy or lactation.[49,50] Local prescribing infor-

3. European Society of Hypertension-European Society of Cardi-mation should be consulted for detailed information, ology Guidelines Committee. 2003 European Society of Hy-

pertension-European Society of Cardiology guidelines for theincluding contraindications, precautions, drug inter-management of arterial hypertension. J Hypertens 2003; 21actions and use in special patient populations. (6): 1011-53

4. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Reportof the Joint National Committee on Prevention, Detection,6. Lercanidipine/Enalapril: Current StatusEvaluation, and Treatment of High Blood Pressure: the JNC 7report. JAMA 2003 May 21; 289 (19): 2560-72

In well controlled trials, once-daily administra- 5. Williams B, Poulter NR, Brown MJ, et al. Guidelines fortion of fixed-dose lercanidipine/enalapril has been management of hypertension: report of the fourth working



party of the British Hypertension Society, 2004-BHS IV. mary information on referral opinion pursuant to Article 30 ofJ Hum Hypertens 2004 Mar; 18 (3): 139-85 Council Directive 2001/83/EC for Renitec and associated

names (see Annex I) [online]. Available from URL: http://6. Sica DA. Rationale for fixed-dose combinations in the treatmentwww.emea.eu.int [Accessed 2005 Oct 5]of hypertension: the cycle repeats. Drugs 2002; 62 (3): 443-62

24. Rizzoni D, Porteri E, De Ciuceis C, et al. Effect of treatment7. Jandeleit-Dahm KAM, Tikellis C, Reid CM, et al. Why block-with candesartan or enalapril on subcutaneous small arteryade of the renin-angiotensin system reduces the incidence ofstructure in hypertensive patients with noninsulin-dependentnew-onset diabetes. J Hypertens 2005; 23 (3): 463-73diabetes mellitus. Hypertension 2005 Apr; 45 (Pt 2): 659-658. McClellan KJ, Jarvis B. Lercanidipine: a review of its use in

25. Park JB, Schiffrin EL. Small artery remodeling is the mosthypertension. Drugs 2000; 60 (5): 1123-40prevalent (earliest?) form of target organ damage in mild9. Bang LM, Chapman TM, Goa KL. Lercanidipine: a review of itsessential hypertension. J Hypertens 2001; 19 (5): 921-30efficacy in the management of hypertension. Drugs 2003; 63

26. Stumpe KO, Ludwig M, Fritschka E, et al. Differential regres-(22): 2449-72sive effects of amlodipine and enalapril on carotid artery10. Todd PA, Heel RC. Enalapril: a review of its pharmacodynamicintima-media thickening: outcome results of the ARES-trialand pharmacokinetic properties, and therapeutic use in hyper-[abstract no. OP 045]. J Hypertens 2004; 22 Suppl. 1: S14tension and congestive heart failure. Drugs 1986; 31 (3):

198-248 27. Hosomi N, Mizushige K, Ohyama H, et al. Angiotensin-con-verting enzyme inhibition with enalapril slows progressive11. Todd PA, Goa KL. Enalapril: a reappraisal of its pharmacologyintima-media thickening of the common carotid artery in pa-and therapeutic use in hypertension. Drugs 1992; 43 (3):tients with non-insulin-dependent diabetes mellitus. Stroke346-812001 Jul; 32 (7): 1539-4512. Napp Pharmaceuticals Ltd. Zanidip tablets: prescribing infor-

28. Leu HB, Charng MJ, Ding PYA. A double blind randomizedmation. In: Medicines Compendium 2005. Epsom: Datapharmtrial to compare the effects of eprosartan and enalapril onCommunications Ltd, 2005: 2676-8blood pressure, platelets, and endothelium function in patients13. Brixius K, Gross T, Tossios P, et al. Increased vascular selectiv-with essential hypertension. Jpn Heart J 2004 Jul; 45 (4):ity and prolonged pharmacological efficacy of the L-type Ca2+623-35channel antagonist lercanidipine in human cardiovascular tis-

29. Agabiti Rosei E, Rizzoni D, Muiesan ML, et al. Effects ofsue. Clin Exp Pharmacol Physiol 2005 Sep; 32 (9): 708-13candesartan cilexetil and enalapril on inflammatory markers of14. Soma MR, Natali M, Donetti E, et al. Effect of lercanidipine andatherosclerosis in hypertensive patients with non-insulin-de-its (R)-enantiomer on atherosclerotic lesions induced inpendent diabetes mellitus. CENTRO (CandEsartaN on aTh-hypercholesterolemic rabbits. Br J Pharmacol 1998 Dec; 125erosclerotic Risk factors) study investigators. J Hypertens(7): 1471-62005; 23 (2): 435-4415. Corsini A, Accomazzo MR, Canavesi M, et al. The new calcium

30. Moiseeva O, Villevalde S, Emelyanov I, et al. Comparison ofantagonist lercanidipine and its enantiomers affect majorenalapril and atenolol effects on nonhemodynamic factors inprocesses of atherogenesis in vitro: is calcium entry involved?cardiovascular remodeling in hypertensive patients [abstractBlood Press 1998; 7 Suppl. 2: 18-22no. P3.118]. J Hypertens 2005; 23 Suppl. 2: S31016. Rachmani R, Levi Z, Zadok B-S, et al. Losartan and ler-

31. Minai K, Matsumoto T, Horie H, et al. Bradykinin stimulatescanidipine attenuate low-density lipoprotein oxidation in pa-the release of tissue plasminogen activator in human coronarytients with hypertension and type 2 diabetes mellitus: a ran-circulation: effects of angiotensin-converting enzyme inhibi-domized, prospective crossover study. Clin Pharmacol Thertors. J Am Coll Cardiol 2001 May; 37 (6): 1565-702002 Sep; 72 (3): 302-7

32. Brown NJ, Vaughan DE. Angiotensin-converting enzyme inhib-17. Canavesi M, Baldini N, Leonardi A, et al. In vitro inhibitoryitors. Circulation 1998 Apr 14; 97 (14): 1411-20effect of lercanidipine on cholesterol accumulation and matrix

metalloproteinases secretion by macrophages. J Cardiovasc 33. Ravid M, Brosh D, Levi Z, et al. Use of enalapril to attenuatePharmacol 2004 Oct; 44 (4): 416-22 decline in renal function in normotensive, normoalbuminuric

patients with type 2 diabetes mellitus: a randomized, con-18. Cheung BMY, Man YB, Tse HF, et al. Advantages of bloodtrolled trial. Ann Intern Med 1998; 128 (12 Pt 1): 982-8pressure optimization. Adv Ther 2005; 22 (4): 285-96

19. Robles NR, Ocon J, Gomez CF, et al. Lercanidipine in patients 34. The European Study for the Prevention of Renal Disease inwith chronic renal failure: the ZAFRA study. Ren Fail 2005; Type 1 Diabetes (ESPRIT) Study Group. Effect of 3 years of27 (1): 73-80 antihypertensive therapy on renal structure in type 1 diabetic

patients with albuminuria: the European Study for the Preven-20. Dalla Vestra M, Pozza G, Mosca A, et al. Effect of lercanidipinetion of Renal Disease in Type 1 Diabetes (ESPRIT). Diabetescompared with ramipril on albumin excretion rate in hyperten-2001 Apr; 50 (4): 843-50sive type 2 diabetic patients with microalbuminuria: DIAL

study (Diabete, Ipertensione, Albuminuria, Lercanidipina). Di- 35. Recordati SpA. Efficacy and tolerability of a combination ofabetes Nutr Metab Clin Exp 2004; 17 (5): 259-66 lercanidipine and enalapril in patients with mild to moderate

essential hypertension not adequately controlled by enalapril21. Sabbatini M, Vitaioli L, Baldoni E, et al. Nephroprotectivetreatment (add-on to enalapril) [CPL1-0019]. Milan: Recordatieffect of treatment with calcium channel blockers in spontane-SpA, 2004 Mar 25. (Data on file)ously hypertensive rats. J Pharmacol Exp Ther 2000; 294 (3):

948-54 36. Recordati SpA. Lercanidipine-enalapril interaction (PK 0031).Milan: Recordati SpA, 2004 Nov 19. (Data on file)22. Sabbatini M, Leonardi A, Testa R, et al. Effect of calcium

antagonists on glomerular arterioles in spontaneously hyper- 37. Recordati SpA. Bioequivalence study of a fixed combinationtensive rats. Hypertension 2000 Mar; 35 (3): 775-9 versus a combination of marketed tablets of lercanidipine HCl

23. The European Agency for the Evaluation of Medicinal Products. (10 mg) and enalapril maleate (20 mg) [PK 0159]. Milan:Committee for Proprietary Medicinal Products (CPMP): sum- Recordati SpA, 2004 Nov 29. (Data on file)


106 Hair et al.

38. Barchielli M, Dolfini E, Farina P, et al. Clinical pharmacokinet- dose response information to support drug registration (CPMP/ics of lercanidipine. J Cardiovasc Pharmacol 1997; 29 Suppl. ICH/378/95) [online]. Available from URL: http://www.em2: S1-15 ea.eu.int/pdfs/human/ich/037895en.pdf [Accessed 2005 Oct 3]

39. Finley PR, O’Brien JG, Coleman RW. Lithium and angiotensin- 45. Recordati SpA. Pooled safety analysis of lercanidipine/enalaprilconverting enzyme inhibitors: evaluation of a potential interac- combination therapy. Milan: Recordati SpA, 2004 Nov 19.tion. J Clin Psychopharmacol 1996 Feb; 16 (1): 68-71 (Data on file)

40. DasGupta K, Jefferson JW, Kobak KA, et al. The effect of 46. Recordati SpA. Final clinical report: add-on to lercanidipine.enalapril on serum lithium levels in healthy men. J Clin Psy- Protocol no: REC 15/2375 - CPL1-0018. Milan: Recordatichiatry 1992 Nov; 53 (11): 398-400 SpA, 2003 Dec 10. (Data on file)

41. Recordati SpA. A multicentre, randomized, parallel group,47. Recordati SpA. Final clinical study report. Protocol no: REC 15/double-blind trial to determine the optimal dose combinations

2375 - CPL1-0019. Milan: Recordati SpA, 2003 Dec 18. (Dataof lercanidipine and enalapril in comparison to each compo-on file)nent administered alone and to evaluate the efficacy and safety

of the combinations in mild to moderate essential hypertension 48. Data on file, Recordati SpA, 2006 Oct 4(CPL2-008). Milan: Recordati SpA, 2000 Dec 21. (Data on

49. Recordati SpA. Gebrauchsinformation: ZANIPRESS 10mg/file)10mg filmtabletten. Milan: Recordati SpA, 2006. (Data on

42. Recordati SpA. Efficacy and tolerability of a combination of file)lercanidipine and enalapril in patients with mild to moderate

50. Recordati SpA. Gebrauchsinformation: ZANIPRESS 10mg/essential hypertension not adequately controlled by ler-20mg filmtabletten. Milan: Recordati SpA, 2006. (Data oncanidipine treatment (add-on to lercanidipine) [CPL1-0018].file)Milan: Recordati SpA, 2004 Mar 11. (Data on file)

43. Agrawal R, Marx A, Haller H. Efficacy and safety of ler-canidipine versus hydrochlorothiazide as add-on to enalapril

Correspondence: Philip I. Hair, Wolters Kluwer Health |in diabetic populations with uncontrolled hypertension.Adis, 41 Centorian Drive, Private Bag 65901, Mairangi Bay,J Hypertens 2006; 24 (1): 185-92Auckland 1311, New Zealand.44. The European Agency for the Evaluation of Medicinal Products

(Human Medicines Evaluation Unit). Note for guidance on E-mail: [email protected]


Drugs 2007; 67 (1): 107-108GUEST COMMENTARIES 0012-6667/07/0001-0107/$49.95/0


nation may assist treatment adherence by avoidingFixed-Dose Combinationmultiple daily pill intake.Lercanidipine/Enalapril

A Viewpoint by Nicolas R. Robles ACE inhibitor dosage is an important issue. Ac-cumulated evidence suggests that angiotensin axis-Hypertension Unit, Service of Nephrology,blocking drugs should be used at the highest doseHospital Infanta Cristina, Badajoz, Spaintolerated by the patient to reach full (not only antihy-pertensive) effects. It is likely that the maximum

A very strong recommendation, delivered by allrecommended lercanidipine dosage (20 mg/day)

official guidelines, is to bring blood pressure (BP)may be combined with a high dosage of enalapril

below 140/90mm Hg in every patient, and even(40 mg/day), if necessary, without risk of exacerba-lower in the presence of diabetes mellitus or renaltion of the well known clinical adverse effects ofdisease. Reaching these targets remains a difficultCCBs.task with currently available antihypertensive medi-

cations. Using monotherapies, one might indeed The lercanidipine/enalapril fixed-dose combina-expect to achieve the goal BP in no more than 40% tion actually offers an advantage over classical ACEof patients. In most patients, combination therapy is inhibitor/diuretic combinations in terms of effect onrequired in order to achieve optimal BP control, as carbohydrate metabolism, as it does not negativelyexemplified by the experience accumulated in large influence insulin resistance. Furthermore, the re-prospective interventional trials. ported effects of lercanidipine on proteinuria indi-

The coadministration of two agents acting by cate that this combination may have interestingdifferent mechanisms considerably improves the BP properties in proteinuric patients with renal dysfunc-control rate. Such combinations are not only effica- tion. It is likely to be most useful in patients withcious, but are also well tolerated, and some fixed mild to moderate chronic renal failure, where thelow-dose combinations even have a placebo-like low diuretic dose used in fixed-dose combinations istolerability. This is the case for the preparation

not effective. However, these advantages remaincontaining the ACE inhibitor enalapril (10 or 20mg)

speculative and further research is needed to docu-and the calcium channel blocker (CCB) ler-ment these potential beneficial effects. ▲canidipine (10mg). Moreover, a single-dose combi-



CCB and an ACE inhibitor has been recently report-Fixed-Dose Combinationed to be more effective than the ‘classic’ combina-Lercanidipine/Enalapril tion of a β-blocker and a diuretic.[1]

A Viewpoint by Roland AsmarThe advantages of such CCB/ACE inhibitor

The CardioVascular Institute, Paris, France combinations may remain speculative in the absenceof the medical practitioner’s involvement. Epidemi-ological studies have definitely shown that inertiaStudies have shown that the current rate of ade-and passivity of doctors in the choice of drug treat-quate blood pressure (BP) control is far from satis-ment of patients with uncontrolled hypertension is afactory and does not reach the level of 30% ofsignificant cause of high BP. Therefore, practition-treated hypertensive patients across almost all West-ers have to increase the use of combination therapyern countries. Insufficient use of drug combinationsin order to improve the BP control rate.has been identified as a major reason for the poor

rate of BP control. Therefore, guidelines on hyper- Further research is needed to document the effi-tension management now support the crucial role of cacy : tolerability comparison between the variouscombination therapy in increasing the extent of BP fixed-dose antihypertensive combinations availablecontrol. on the market. Moreover, clinical trials aiming to

define the suitable combination to be used, accord-A combination of two antihypertensive drugs in aing to patient characteristics, will be very helpful insingle pill offers advantages that include high effica-further refining the guidelines for hypertensioncy, a low incidence of adverse effects, good treat-management. ▲ment compliance and relatively low cost. The com-

bination of a calcium channel blocker (CCB) withan ACE inhibitor is one of the attractive combina-tion regimens. In fact, studies have shown that this Referencetype of combination presents a high effica- 1. Poulter NR, Wedel H, Dahlof B, et al. Role of blood pressure

and other variables in the differential cardiovascular eventcy : tolerability ratio, reverses organ damage andrates noted in the Anglo-Scandinavian Cardiac Outcomes Tri-improves cardiovascular prognosis in terms of mor- al-Blood Pressure Lowering Arm (ASCOT-BPLA). ASCOTinvestigators. Lancet 2005 Sep 10; 366 (9489): 907-13bidity and mortality. Indeed, the combination of a

Drugs 2007; 67 (1): 109-118ADIS DRUG PROFILE 0012-6667/07/0001-0109/$49.95/0


LimaprostTracy Swainston Harrison and Greg L. Plosker


ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091. Pharmacodynamic Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1102. Pharmacokinetic Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1113. Therapeutic Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124. Tolerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165. Dosage and Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1176. Limaprost: Current Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Features and properties of limaprost (Opalmon)Abstract▲ Limaprost, an alprostadil (prostaglandin E1) ana- Featured indications

logue, is a vasodilator that increases blood flow andThromboangiitis obliterans (TAO) [Buerger’s disease]inhibits platelet aggregation.

▲ The efficacy of oral limaprost was evaluated in Lumbar spinal canal stenosis (LSCS) [in patients with bilateralintermittent claudication and a normal straight leg raise testadult Japanese patients in three randomised,result]double-blind, 6-week trials. One study included pa-

tients with thromboangiitis obliterans and two trials Mechanism of actionincluded patients with lumbar spinal canal stenosis.

Vasodilation, increased blood flow and inhibition of plateletLimaprost was generally well tolerated and seriousaggregationadverse events were uncommon.

▲ Thromboangiitis Obliterans: In a randomised, Dosage and administration in adultsdouble-blind trial in Japanese patients primarily

Route of administration Oralwith thromboangiitis obliterans (n = 136), there wasno significant difference between patients receiving

Frequency of administration Daily in three divided doseslimaprost 30 µg/day and those receiving oralticlopidine 500 µg/day in the improvement of is- Dose TAO 30 µg/daychaemic symptoms.

Dose LSCS 15 µg/day▲ Lumbar Spinal Canal Stenosis: Limaprost 15 µg/day was superior to limaprost 3 µg/day for overall Single-dose pharmacokinetic properties of oral limaprost indrug usefulness and overall improvement from adult healthy volunteers (n = 12)baseline to study end in a phase III trial in 146

Mean maximum plasma 5µg: 1.18 pg/mLpatients with lumbar spinal canal stenosis. Assess-concentration 10µg: 2.06 pg/mLment of overall improvement considered various

objective symptoms (e.g. muscle strength, walking Median time to maximum plasma 5µg: 0.75hconcentration 10µg: 0.5hability) and subjective symptoms (e.g. pain or

numbness in extremities), while overall usefulnessMean area under the 5µg: 1.97 pg • h/mLalso considered safety issues. concentration-time curve from 10µg: 3.43 pg • h/mLtime 0 to infinity▲ The efficacy of limaprost 15 µg/day was not signifi-

cantly different from that of 30 µg/day, but tendedMean elimination half-life 5µg: ≈1hto be better than that of 6 µg/day in a phase II trial in 10µg: ≈1h

patients with lumbar spinal canal stenosis and nor-Adverse eventsmal straight leg raise test results. The optimal dos-

age of limaprost for this indication was thereforeGastrointestinal-related, rash, hot flushes/flushing, headachesdeemed to be 15 µg/day. and anaemia

110 Swainston Harrison & Plosker

the sacrum (L5/S1),[5] which is usually caused bybone and ligamentous hypertrophy or intervertebraldisc degeneration.[5] This constriction leads to en-trapment of the cauda equina (a collection of spinalnerve roots), causing ischaemia, which is a contrib-uting factor in the development of the intermittent

HOH

HO

HCH3

CH3H H

COOH

OH

Limaprost neurogenic claudication.[6] Patients also experiencepain in the back, buttocks, thighs and legs, a feelingof weakness and numbness in the legs, gait distur-Limaprost is an alprostadil (prostaglandin E1)bance[5,6] and, in severe cases, bowel or bladderanalogue that was developed in Japan to treat nu-disturbances.[7] The middle-aged and elderly aremerous ischaemic symptoms of thromboangiitis ob-commonly affected; disease progression is slow andliterans (TAO) and lumbar spinal canal stenosisthe lower back and leg pain may be incapacitating.[6]

(LSCS)[1] because of the well known vasodilatory,Treatment of the symptoms of LSCS depends onantiplatelet and cytoprotective properties of pros-disease severity (severity is based on level of legtaglandins.[2]

pain and related disability); discussion of surgicalTAO (Buerger’s disease) is a nonatherosclerotic, management and conservative treatment options

occlusive, inflammatory disease of distal, small- and such as bed rest, physical therapy and lumbar brac-medium-size arteries and veins of the arms and ing are beyond the scope of this review.[8] Pharma-legs.[2,3] Signs and symptoms are pain at rest, is- cological treatment options include epidural cortico-chaemic ulcers (in the arms, legs or both), thrombo- steroids, anaesthetic nerve block or narcoticphlebitis, Raynaud’s phenomenon, an abnormal Al- analgesics, which are used to relieve pain.[8] In Ja-len-test result, sensory findings and intermittent pan, limaprost is approved not only for pain relief,claudication.[2,3] This disease is more prevalent in but also for relief of other ischaemic symptoms,Japan and Korea than in Western Europe (16–66% such as numbness, and also of walking distur-vs 0.5–5.6% of patients with peripheral arterial dis- bances.[1]

ease) and occurs mainly in young, male smokers.[3]The pharmacological properties, efficacy and tol-

Patients experience alternating periods of acute ex- erability of limaprost (Opalmon)1 in patients withacerbation and remission.[2] The only proven treat- TAO or LSCS are the focus of this review.ment of TAO is cessation of tobacco use; diseaseprogression is prevented and amputations may be 1. Pharmacodynamic Profilesuccessfully avoided.[3] Nevertheless, alleviation ofischaemic pain, improved microcirculation and ● Limaprost has vasodilatory properties and in-healing of ulcers may be achieved via pharmacolog- creases blood flow.[9] In an animal model of periph-ical treatment with prostaglandins or aspirin; how- eral circulatory disorder, oral administration ofever, the latter agent appears less effective than limaprost inhibited the development of ischaemicprostaglandins.[2] Iloprost (an analogue of epopros- lesions in limbs and peripheral extremities.[10] Thetenol[4]) and alprostadil are available in numerous effect of limaprost on ischaemic lesions in patientsEuropean countries for the treatment of TAO[2,3] and with TAO or LSCS is reviewed in section 3.limaprost is approved only in Japan and South Ko-

● In animal studies with relevance for TAO,rea in this indication.[1]

limaprost decreased coronary vessel resistance,[11]

LSCS is a constriction of the spinal canal at the increased coronary blood flow[11] and increasedthird to fourth (L3/L4) or fourth to fifth (L4/L5) femoral arterial and cutaneous blood flow in thelumbar vertebrae, or at the fifth lumbar vertebra to hind limbs of dogs.[12]

1 The use of trade names is for product identification purposes only and does not imply endorsement.


Limaprost: Adis Drug Profile 111

● The inhibitory effect of limaprost on platelet also increased the diameter of the cauda equinaaggregation and platelet adhesiveness was 10- to blood capillary (at the seventh lumbar vertebra).[20]

16-fold and 20-fold more potent than with epopros- ● Additionally, animal models of LSCS havetenol (prostacyclin or prostaglandin I2) in an in vitro shown that limaprost may improve nerve func-study using guinea-pig platelets.[13] This inhibitory tion[18,21,22] and walking ability.[19,23] For example,effect on platelet function may result in inhibition of intravenous limaprost inhibited an induced reduc-thrombus formation; in an in vivo guinea-pig model tion in nerve conduction velocity in the caudaof electrically induced thrombosis, limaprost signif- equina nerve at the seventh lumbar vertebraicantly (p < 0.05) inhibited thrombus formation as (p < 0.05 vs control), but had no significant effect onassessed by the change of the threshold voltage for another measure of nerve function (the reduction offormation of a thrombus.[14] muscle action potential area).[21]

● Sciatic nerve function may be affected in LSCS● By contrast, limaprost inhibited platelet aggrega-since the stenosis may occur at the fourth and fifthtion with similar potency to that of epoprostenol inlumbar vertebrae.[5] Sciatic nerve ligation in ratsan in vitro study with human platelets.[15]

caused prolongation of heat-stimulated myogenic● Oral limaprost 20–40µg produced no significant

nerve discharges in the femoral muscles ipsilateralchange in platelet function (either bleeding time orto the nerve, but with oral administration ofplatelet adhesiveness) in a single-dose study inlimaprost, this prolongation was inhibited (p < 0.05healthy volunteers.[16] However, in patients withvs control).[22] In another in vivo rat model of dis-thromboembolic disorders (n = 6), a single 30 orturbed sciatic nerve function, chronic administration40µg dose inhibited platelet aggregation by ≈3–35%of oral limaprost reduced hyperalgesia.[18]

and adhesiveness by ≈13% (30µg) or ≈35% (40µg)● Oral administration of limaprost in rats signifi-in a dose-dependent manner up to 2 hours after

cantly (p < 0.05) increased walking[23] or running[19]administration (values estimated from a graph).[15]

distance compared with a sham control in in vivoMoreover, oral limaprost increased platelet cyclicmodels of walking dysfunction.[19,23] By contrast,AMP levels by 10–20% after a single 30 or 40µgberaprost did not improve walking distance in onedose in patients with thromboembolic disorders.[15]

trial.[23] Section 3 reviews the efficacy of limaprost● Oral limaprost produced transient decreases in

in improving walking ability in patients with LSCS.blood pressure in healthy volunteers in a single-dose

● There is potential for a pharmacodynamic drugstudy (n = 12).[16] These decreases were not statisti-interaction (an increased tendency to bleed) betweencally significant with a 20µg dose. Significant re-limaprost and the following agents if coadminis-ductions from baseline were reported in systolictered: antiplatelets (aspirin, ticlopidine, cilostazole),blood pressure only at 2 hours post-dose with a 40µgthrombolytics (urokinase) and anticoagulants (hepa-dose, and in both diastolic and mean blood pressurerin, warfarin).[9]

up to 2 hours after a 30µg dose and 1–3 hours after a40µg dose (all p < 0.05). All blood pressure parame-

2. Pharmacokinetic Profileters returned to baseline values after 3–6 hours.[16]

● In animal models of ischaemia with relevance for The pharmacokinetic properties of oral limaprostLSCS, limaprost improved blood flow to the cauda 5 and 10µg have been evaluated in a single-dose,equina (p < 0.05 vs baseline),[17] sciatic nerve[18] and crossover study in Japanese adult healthy volunteersnerve tissue in the lumbar vertebral canal.[19] Blood (n = 12).[24] Similar results to those reviewed hereflow in the tissue between two ligations around the have been reported in a study of parallel-groupright sciatic nerve[18] and in the nerve tissue of the design (n = 24), which is not discussed further.[25]

fifth lumbar vertebra[19] increased significantly after Additional data are available from animal[26,27] andmultiple-dose oral administration of limaprost in vitro[28] studies. The pharmacokinetics of(p < 0.001[18] and p < 0.01[19] vs control). Limaprost limaprost in the elderly and in paediatric patient



populations, or in patients with renal or hepatic 3. Therapeutic Efficacyimpairment have not been investigated.

This section reviews data from fully published,randomised, double-blind trials with more than 50

Absorption and Distribution patients. All studies are in adult Japanese patients.

● Limaprost is rapidly absorbed following oral ad- Thromboangiitis Obliteransministration; the median time to peak plasma con-centration (tmax) was under 1 hour after a single dose The efficacy of oral limaprost 30 µg/day in theof 5 or 10µg (0.75 and 0.5 hours).[24] Oral limaprost treatment of ischaemic symptoms (ulcer, feeling oftablets were taken with water after a meal.[24] coldness, pain) associated with TAO was evaluated

versus oral ticlopidine 500 µg/day in a randomised,● The mean maximum plasma concentrationdouble-blind, multicentre 6-week trial.[29] Patients(Cmax) of limaprost after a single 5 and 10µg dosewith occlusive arteriosclerosis were enrolled in thiswas 1.18 and 2.06 pg/mL.[24] Respective mean areatrial (n = 49); however, since limaprost is not ap-under the concentration-time curve from time zeroproved for use in this indication, these data are notto infinity (AUC∞) values were 1.97 and 3.43 pg • h/reviewed here. A small number of patients withmL.chronic arterial occlusion or diabetic vascular dis-● Absorption of limaprost appears to be dose-pro-ease (n = 13) were enrolled in the trial and wereportional; the ratios for the dose-adjusted meangrouped together with patients with TAO (n = 136)Cmax and AUC∞ values of a single 10µg dose versus[‘TAO group’]. Patients were required to have ul-a single 5µg dose were 0.971 and 0.935, which wascers on their extremities and were excluded if theywithin the 0.90–1.11 range considered by investiga-had any of the following: prior surgical revascu-tors to indicate linear pharmacokinetics.[24]

larisation or excision of sympathetic ganglia; a ten-● In an in vitro study,[28] limaprost was highlydency for haemorrhage; granulocytopenia or a histo-protein bound in human plasma (95.8%).ry thereof; and severe hepatic or renal impairment.Baseline characteristics were not significantly dif-

Metabolism and Elimination ferent between treatment groups.[29]

Study drugs were taken three times daily in divid-● According to a study in rats,[26] ≈70% of an orally ed doses after meals for 6 weeks following a 1-week

administered radiolabelled dose of limaprost is ex- placebo run-in period.[29] Concomitant use of an-creted in the faeces and ≈30% is eliminated in the tibacterial agents or analgesics was permissible, buturine over a 96-hour post-dose period. Two major any agents that may have affected trial results (e.g.metabolites have been identified in rats; both metab- antiplatelets, anticoagulants, vasodilators and barbi-olites are excreted primarily in the faeces and bile turates) were not allowed.(5–18% of the administered dose).[28] The extent or Efficacy endpoints were the following: generalsite of metabolism, and route of excretion of usefulness, which considered both efficacy and tol-limaprost has not yet been investigated in humans. erability and was assessed via a 100-point visual● Limaprost is rapidly eliminated from the body analogue scale (VAS) [a score of ≥70 of 100 was

after a single 5 or 10µg dose. The mean elimination considered ‘useful’]; improvement in ulcer size (anhalf-life was approximately 1 hour with either objective symptom); improvement in rest pain (adose.[24] Thus, oral limaprost is taken daily in three subjective symptom); and overall improvement,divided doses (section 5). which considered all clinical symptoms taken to-● Similarly, the apparent clearance of oral gether and was assessed using a 5-point study-de-

limaprost from the plasma was not dose-proportion- fined categorical scale.[29] All endpoints were inves-al, and was 3110 L/h for both 5 and 10µg doses.[24] tigator-rated except for an evaluation of drug effec-



56

53

53

38

57

53

52

54

45

56

0 10 20 30 40 50 60 70 80 90 100

Usefulness rate (expert)

Usefulness rate (investigator)

Overall improvement rate

Rest pain improvement rate

Ulcer size improvement rate

Patients (%)

LIM 30 µg/dayTIC 500 µg/day

Fig. 1. Efficacy of limaprost (LIM) vs ticlopidine (TIC) in patients (pts) with thromboangiitis obliterans (n = 136), or chronic arterial occlusionor diabetic vascular disease (n = 13).[29] In this randomised, double-blind, multicentre, 6-week trial, pts received LIM 30 µg/day (n = 53–60)or TIC 500 µg/day (n = 64–76) in three divided doses after meals; evaluable patient numbers differed for each endpoint. Efficacy endpointswere assessed via investigator-rated 5-point categorical or 100-point visual analogue scales. Improvement or usefulness rates were theratio of the number of pts in the two highest ranking categories of improvement/usefulness to the total number of evaluable pts.

tiveness, which was based on changes in ulcer size Lumbar Spinal Canal Stenosis

and reported using a VAS by an expert committee (aPatients with LSCS and symptoms of degenera-score of ≥70 of 100 indicated that the drug was

tive spondylolisthesis and bilateral intermittent clau-considered ‘useful’). Analyses were per-protocol,dication,[30] or degenerative, combined, spondylolis-unless specified otherwise.thetic, or spondolytic symptoms[31] were enrolled in

● The efficacy of limaprost was not significantly two randomised, double-blind, multicentre, 6-weekdifferent from that of ticlopidine in the TAO trials.[30,31] Patients were excluded from the studiesgroup.[29] Thus, a similar proportion of patients in if they met any of the following criteria: had prior

surgery or were suitable candidates for surgery;[30,31]the two treatment groups had improvement ratingshad chronic occlusive arterial disease;[30] had con-in the top two categories for ulcer size (healed orcomitant internal organ disease;[30,31] had severe im-reduced), rest pain (markedly improved or im-pediment to lower extremity movement;[30] or wereproved), overall improvement (markedly improvedpregnant women or nursing mothers.[30] Patients en-

or improved), investigator-rated usefulness (scores rolled in the trials were elderly (mean patient age ofof 90–100 or 70–89) and usefulness rated by an 66–70 years), the majority of whom had degenera-expert panel (90–100 or 70–89) [figure 1]. tive spinal canal stenosis (75%[30] and 72%[31] of

patients) of mild or moderate severity (84%[30] and● In addition, the reduction in ulcer size from base-90%[31]) and of less than 2 years’ durationline to study end was significant in both the(66–79%).[30,31]

limaprost (n = 57) and ticlopidine (n = 99) treatmentNerve blocking treatment was not allowed duringgroups (12–8 mm2 and 13–8 mm2) [p-value not

the trial; physical therapies (e.g. physiotherapy orreported].[29] This analysis included patients present-

corsets) were permissible only where patients hading with occlusive arteriosclerosis as well as those in been receiving these treatments prior to the tri-the TAO group. al.[30,31] Concomitant pharmacological treatment



with vasodilatory, muscle-relaxing or bone metabo- actions’),[31] ‘patient complaints’ (at rest)[30] and thedistance to onset of intermittent claudication[30] werelism agents and cyanocobalamin preparations wasadditional endpoints. With the exception of the SLRgenerally excluded.[30,31] In one trial, treatment oftest and the distance to onset of intermittent claudi-concomitant illnesses was permitted[30] and, in thecation, most assessments were categorical, and val-other trial,[31] analgesics and anti-inflammatoryues at the end of study were compared with baselineagents were allowed but, in either case, only wherevalues to assess the degree of improvement shown.patients had been receiving these treatments prior toThe scales used to rate subjective and objectivethe trial.symptoms were different between the two trials,Patients received oral limaprost 3[30] or 6[31] µg/thus results are reported separately. Statistical analy-day (i.e. control treatment groups; n = 77[30] andses were generally per-protocol, unless otherwise38[31]), or 15 µg/day (n = 69[30] and 47[31]) or 30 µg/specified.day (n = 44)[31] in three divided doses after meals.● Limaprost 15 µg/day was considered by investi-Patient baseline demographic and disease character-

gators to provide greater overall improvement and toistics were not significantly different between treat-be a more useful treatment than the control dosagement groups in one trial,[30] although in the otherof 3 µg/day (primary endpoints) in the phase III trialstudy[31] there were more women in the 30 µg/dayin 146 patients with LSCS.[30] When comparingthan in the 6 or 15 µg/day groups (n = 25 vs 13 andassessment rates, there were more patients in the16) and more patients with degenerativehigher dosage group than in the control dosagespondylolisthetic, combined or spondolytic spinalgroup rated as ‘improved’ or ‘markedly improved’canal stenosis in the 6 and 15 µg/day than in the 30(overall improvement rate 50.7% vs 27.3%; p <µg/day group (n = 15 and 16 vs 5). Where stated,0.01) at the end of treatment or for whom limaprostapproximately two-thirds of patients showed onsetwas rated as ‘useful’ or ‘extremely useful’ (useful-of intermittent claudication within a walking dis-ness rate 50.7% vs 27.3%; p < 0.01) [figure 2].[30]

tance of 500m.[30]

Assessment of overall improvement considered va-Primary efficacy endpoints, which were investi-

gator-rated measurements, were specified only inone trial and were the overall improvement in symp-toms at the end of treatment (both subjective andobjective symptoms were assessed) and drug useful-ness.[30] The third primary endpoint was a tolerabili-ty endpoint (see section 4).[30] The straight leg raise(SLR) test score, sensation, muscle strength and, inone trial, walking ability,[30] were assessed objec-tively, whereas lower extremity or lumbar pain andlower extremity numbness, and walking ability inthe other trial,[31] were assessed subjectively. In onetrial, baseline assessment of subjective symptomswas made at the time of onset of intermittent claudi-cation; however, further on-treatment assessmentsof these symptoms were conducted after patientshad walked the same distance that they had walkedat baseline before intermittent claudication be-gan.[30] By contrast, at least one subjective symptomwas assessed at rest in the other trial (lumbarpain).[31] Activities of daily living (‘everyday life

LIM 15 µg/day (n = 69)LIM 3 µg/day (n = 77)

50.7 50.7

27.3 27.3

60

50

40

30

20

10

0Overall improvement

rateUsefulness rate

Pat

ient

s (%

)

* *

Fig. 2. Efficacy of oral limaprost (LIM) 15 vs 3 µg/day in adultpatients (pts) with lumbar spinal canal stenosis in a randomised,double-blind, multicentre, 6-week trial.[30] Investigator-rated cate-gorical global measures of efficacy were the overall improvement atend of treatment and usefulness. Overall improvement rate andusefulness rate were the ratio of the number of pts in the twohighest ranking categories of improvement/usefulness to the totalnumber of evaluable pts. * p < 0.01 vs comparator.



LIM 6 µg/dayLIM 15 µg/dayLIM 30 µg/day

43.5

20

26.1

34.8

59.4

41.7

50

59.4

55.2

21.1

48.351.7

70

60

50

40

30

20

10

0Subjectivesymptoms

Objectivesymptoms

Everyday lifeactions

Overallimprovement

Impr

ovem

ent r

ate

(% o

f pts

)

Fig. 3. Improvement rates with oral limaprost (LIM) in patients (pts) with lumbar spinal canal stenosis and normal straight leg raise testresults. In this randomised, double-blind, multicentre, 6-week trial, patients received oral LIM in three daily divided doses after meals.[31]

Improvement rates are shown for subjective symptoms, objective symptoms, everyday life action and overall improvement, and representthe ratio of the number of pts in the two highest ranking categories (e.g. markedly improved or improved subjective symptoms) to the totalnumber of evaluable pts. The total number of pts varied for each assessment: LIM 6 µg/day (n = 20–23), LIM 15 µg/day (n = 24–32) and LIM30 µg/day (n = 19–29). There were no statistically significant differences between treatment groups.

rious objective symptoms (e.g. muscle strength, ly improved subjective symptoms (42% vs 26%),walking ability) and subjective symptoms (e.g. pain patient impressions of ‘got better’ (39% vs 25%), oror numbness in extremities), while assessment of with improved or markedly improved overall gener-usefulness also considered safety issues. al improvement (55% vs 35%) was greater in the

higher than in the control dosage group (secondary● Between-group differences in improvement inefficacy endpoints) [all p < 0.05].most objective symptoms at study end were not

significant except for muscular strength and dis- ● The optimal dosage of limaprost in LSCS wastance walked to onset of intermittent claudication determined in a phase II trial comparing limaprost 6,according to Japanese Orthopaedic Association cri- 15 and 30 µg/day in a total of 129 patients.[31] Ateria for walking ability in the same trial (secondary post-hoc analysis was conducted in 84 patients withefficacy endpoints).[30] More patients in the 15 µg/ normal SLR test results; patients with abnormalday than the 3 µg/day dosage group experienced an SLR results were excluded because they are typical-improvement of at least one grade in muscular ly good surgical candidates, often having concurrentstrength (59% vs 15% of patients; p = 0.0025) or of disc hernia or lateral stenosis of the lumbar canal.at least one scale point in the distance walked to The efficacy of limaprost 15 µg/day was not signifi-onset of intermittent claudication (40% vs 25%; cantly different from that of 30 µg/day but tended top = 0.0187).[30]

be better than that of 6 µg/day when patients withabnormal SLR were excluded (figure 3). For exam-● The efficacy of the 15 µg/day dosage was gener-ple, improvement rates for subjective symptoms,ally superior to that of the 3 µg/day group accordingobjective symptoms and overall improvement wereto subjective measures of efficacy.[30] At study end,approximately 15–20% higher with limaprost 15 µg/the proportion of patients with improved or marked-



day than with limaprost 6 µg/day, and results with 15 µg/day and 16%[31] of those receiving 30 µg/day.limaprost 30 µg/day were generally similar to or However, between-group differences in the inci-slightly lower than those with limaprost 15 µg/day. dence of these adverse events in one trial,[30] or in

the number of patients without these adverse events● Efficacy data from a postmarketing surveillancein the other trial,[31] were not significant.study are available from 1800 patients with LSCS

who received limaprost (mean dosage 15.1 µg/day) ● The most frequent treatment-related adverse ef-for a mean duration of 116 days.[32] The overall fects included gastrointestinal (GI)-related effects,improvement rate was 49.9%, i.e. approximately rash, hot flushes/flushing, headaches and anaemia.[9]

half of the patients improved or markedly improved In patients with TAO, diarrhoea, nausea/vomiting/with limaprost therapy. Lower extremity pain, both retching, hot flushes/flushing, rash and abdominalat rest and after walking, was significantly reduced or epigastric discomfort were reported in 1.1%,by limaprost and walking ability was significantly 0.5%, 0.5%, 0.4% and 0.4% of patients (n = 4582;improved. For example, 55.7% of patients could not pooled analysis);[9] in patients with LSCS, therewalk 500m before administration of limaprost, but were 25 reports of adverse events affecting the GIthis decreased to 27.4% of patients following treat- tract (e.g. stomach discomfort, diarrhoea, abdominalment (p < 0.0001). Mean walking time almost discomfort) among 397 patients (pooled analy-doubled from 12.98 minutes at baseline to 21.23 sis).[32]

minutes with limaprost (p < 0.0001). ● Other treatment-related adverse events occurring● A total of 363 patients who were evaluated for in 0.2–0.3% of patients with TAO were abdominal

efficacy in the postmarketing surveillance study re- or gastric pain, headache, hepatic function abnor-ceived limaprost treatment for >6 months (including malities (e.g. increased AST or ALT levels) and140 who received limaprost for >1 year).[32] The anorexia.[9]

overall improvement rate was 52.9% among patients● Serious treatment-emergent adverse events de-

treated for >6 months and 60.7% among those treat- scribed by investigators as occurring in patients fored for >1 year. A comparison of improvement rates whom treatment was ‘not safe’ were reported onlyamong patients treated for >1 year versus those in one trial, in three patients receiving limaprosttreated for <1 year showed a significantly higher 30 µg/day.[29] These adverse events were GI-relatedrate in the longer-term treatment group (60.7% vs in two patients and increased AST or ALT levels in47.9%; p-value not stated). the third patient; all three patients discontinued the

study drug.4. Tolerability

● Four of 92 patients receiving limaprost and oneof 106 patients receiving ticlopidine discontinuedData in this section are primarily from pooledtreatment because of an adverse event in a 6-weektolerability analyses available in the manufacturer’strial (see section 3 for trial design details).[29] Rea-Japanese prescribing information,[9] postmarketingsons for discontinuation were not reported for allsurveillance data[32] and the 6-week trials discussedpatients; however, two patients stopped treatmentin section 3.[29-31] The duration of drug exposure inbecause of GI-related adverse events.[29]patients included in the pooled analyses was not

reported. ● In this trial versus ticlopidine,[29] adverse events● Oral limaprost 3–30 µg/day was generally well were not reported separately for the two indication

tolerated in adult patients with TAO[9,29] or groups (TAO and occlusive arteriosclerosis). Treat-LSCS.[9,30-32] The overall incidence of treatment- ment-emergent adverse events occurred in 12% ofemergent adverse events with limaprost in the latter patients receiving limaprost versus 8% of patientsindication appeared to be dose-related, occurring in receiving ticlopidine. Nevertheless, the investigator-3% or 4% of patients receiving 3[30] or 6[31] µg/day rated safety of limaprost 3–30 µg/day was not sig-compared with 9%[30] and 6%[31] of those receiving nificantly different from that of ticlopidine in the



TAO group (safety rate of 88% [n = 69] vs 95% 15µg in patients with LSCS, taken in three divided[n = 80] of patients).[29] doses.[9]

● In similar investigator-rated assessments of over- Currently no dosage recommendations for spe-all safety in the trials in patients with LSCS, safety cial patient populations exist, although the drugrates were 94–97% of 38–47 evaluable patients re- should be administered with caution in patients withceiving limaprost 3–15 µg/day,[30,31] and 84% of 44 a tendency to bleed or those receiving treatmentpatients receiving 30 µg/day, i.e. twice the recom- with thrombolytic, antiplatelet or anticoagulantmended dosage (section 5).[31] Between-group dif- agents. Local prescribing information should beferences were not significant.[30,31] The safety rate consulted for information on precautions, contrain-was the ratio of the number of patients for whom the dications and drug interactions.[9]

drug was rated by the investigator as ‘fairly safe’ or‘safe’ to the number of all evaluable patients. 6. Limaprost: Current Status● In addition to elevated AST and ALT

levels,[9,29-31] limaprost has been associated with In Japan, oral limaprost is approved for the treat-thrombocytopenia[30] and increased blood urea ni- ment of various ischaemic symptoms, such as pain,trogen levels[31] in a small number of patients. The ulcers and a sensation of coldness in the peripheraldrug may also be associated with hepatic impair- extremities, in patients with TAO and of subjectivement or jaundice.[9] symptoms, such as pain and numbness in the lower

legs and reduced walking ability in patients with● Tolerability data from a postmarketing surveil-acquired LSCS. In a randomised, double-blind trial,lance study are available from 1930 patients withthe efficacy of limaprost 30 µg/day was not signifi-LSCS who received limaprost (mean dosage 15.1cantly different from that of ticlopidine 500 µg/dayµg/day) for a mean duration of approximately 4in patients with TAO. In patients with LSCS whomonths.[32] Adverse events occurred in 5.2% of pa-were deemed to be good candidates for medicaltients. A total of 123 events were reported, almostrather than surgical intervention on the basis of SLRhalf of which involved GI disorders, such as stom-test results, the optimal dose of limaprost was foundach discomfort or diarrhoea. One serious adverseto be 15 µg/day. Limaprost was generally well toler-event was reported (bleeding duodenal ulcer).ated in clinical trials.Among 397 patients treated for >6 months (includ-

ing 141 who received limaprost for >1 year), 9.1%experienced an adverse event, in most cases a GI Disclosuredisorder.[32] The incidence of adverse events was

During the peer review process, the manufacturer of thesignificantly lower among patients who receivedagent under review was offered an opportunity to commenttreatment for >1 year than among those who re-on this article; changes based on any comments received wereceived treatment for <1 year (2.84% vs 12.5%; p =made on the basis of scientific and editorial merit.

0.0025).

References5. Dosage and Administration1. Ono Pharmaceutical Co Ltd. Additional indication approved for

limaprost, oral prostaglandin E1 derivative (media release)Oral limaprost is approved in Japan for the treat- [online]. Available from URL: http://www.ono.co.jp [Ac-ment of ischaemic symptoms (ulcer, feeling of cold- cessed 2006 Apr 12]

2. Szuba A, Cooke JP. Thromboangiitis obliterans: an update onness, pain) in adult patients with TAO and for theBuerger’s disease. West J Med 1998 Apr; 168 (4): 255-60

treatment of pain and numbness in the lower legs 3. Olin JW. Thromboangiitis obliterans (Buerger’s disease).and of abnormal gait in adult patients with acquired N Engl J Med 2000 Sep 21; 343 (12): 864-9

4. Sweetman SC. Martindale: the complete drug reference. 34thLSCS who have bilateral intermittent claudicationed. London: Pharmaceutical Press, 2005

and a normal result on the SLR test.[2] The recom- 5. Arbit E, Pannullo S. Lumbar stenosis: a clinical review. Clinmended daily dose is 30µg in patients with TAO and Orthop Relat Res 2001 Mar; (384): 137-43



6. Alvarez JA, Hardy Jr RH. Lumbar spine stenosis: a common mental acute cauda equine compression: a study using a videocause of back and leg pain. Am Fam Physician 1998 Apr 15; recording system (digital hi-scope) [in Japanese]. Progr ad57 (8): 1825-39 Med 2002; 22 (2): 443-5

7. Chosa E, Sekimoto T, Kubo S, et al. Evaluation of circulatory 21. Kayama S, Konno S, Kikuchi S, et al. Effect of OP-1206 α-CDcompromise in the leg in lumbar spinal canal stenosis. Clin (prostaglandin E1 derivative) on nerve conduction velocity inOrthop Relat Res 2005 Feb; (431): 129-33 the dog cauda equina subjected to acute experimental compres-

sion [in Japanese]. Kiso to Rinsho 1996; 30 (2): 229-368. Snyder DL, Doggett D, Turkelson C. Treatment of degenerativelumbar spinal stenosis. Am Fam Physician 2004 Aug 1; 70 (3): 22. Fujitani B, Kii Y, Tashibu H, et al. Effect of OP-1206 • α-CD517-20 on electromyograms in sciatic nerve-ligated rate [in Japanese].

Kiso to Rinsho 1996; 30 (2): 245-509. Ono Pharmaceutical Co Ltd. Opalmon tablets 5 µg prescribinginformation [online]. Available from URL: http://www.e- 23. Liu Y, Noguchi K, Takenobu Y, et al. Comparison the effect ofsearch.ne.jp/~jpr/PDF/ONO06.PDF [Accessed 2006 Apr 21] beraprost sodium with that of limaprost alfadex in rat neuro-

pathic intermittent claudication model [in Japanese]. Jpn10. Kitagawa T, Wakitani K, Ogaki Y, et al. Effects of a pros-Pharmacol Ther 2002; 30 (10): 875-80taglandin E1 analogue, OP-1206•α-cyclodextrin clathrate

(OP-1206•α-CD) in a model of experimental peripheral circu- 24. Mikami H, Hishita N, Itoh T. Clinical pharmacokinetic study oflation disorder in rats [in Japanese]. Gendai Iryo 1986; 18 a single oral administration of limaprost alfadex (Prorenal)Suppl. II: 1-11 [in Japanese]. Rinsho Iyaku 2005; 21 (3): 361-6

11. Tsuboi T, Hatano K, Nakatsuji B, et al. Pharmacological evalua- 25. Komaba J, Masuda Y, Nako S, et al. Clinical pharmacokinetiction of OP 1206, a prostaglandin E1 derivative, as an an- study of Opalmon (limaprost alfadex) tablet in healthy Japa-tianginal agent. Arch Int Pharmacodyn Ther 1980; 247 (1):

nese male volunteers [in Japanese]. Igaku to Yakugaku 2005;89-102

53 (2): 265-7112. Kitagawa T, Sakaguchi N, Kira H, et al. Effects of a pros-

26. Miyamoto S, Taniguchi K, Kajiwara I, et al. Pharmacokineticstaglandin E1 analogue, OP-1206•α-cyclodextrin clathrateof OP-1206•α-cyclodextrin clathrate (OP-1206•α-CD): first(OP-1206•α-CD) on the femoral arterial blood flow, hindlimbreport: absorption and excretion after oral and intracaudalcutaneous blood flow and hindlimb cutaneous temperature [inadministration in rats [in Japanese]. Gendai Iryo 1986; 18Japanese]. Gendai Iryo 1986; 18 Suppl. II: 12-20Suppl. II: 56-69

13. Tsuboi T, Fujitani B, Maeda J, et al. Effect of OP 1206, a27. Miyamoto S, Kida J, Okada K, et al. Pharmacokinetics ofprostaglandin E1 derivative, on guinea-pig platelet functions.

OP-1206-cyclodextrin clathrate (OP-1206-CD): second report:Thromb Res 1980; 20 (5-6): 573-80distribution after oral administration in rats [in Japanese].

14. Fujitani B, Watanabe M, Kuwashima J, et al. Effect of a Gendai Iryo 1986; 18 Suppl. II: 70-9prostaglandin E1 derivative (OP-1206) and acetylsalicyclic

28. Miyamoto S, Ishido M, Sawada M, et al. Pharmacokinetics ofacid on electrically induced thrombosis in guinea-pig mesen-OP-1206•α-cyclodextrin clathrate (OP-1206•α-CD): third re-teric artery and its modification by an inhibitor of pros-port: metabolism in rats [in Japanese]. Gendai Iryo 1986; 18taglandin I2 synthetase, tranylcypromine. Jap J PharmacolSuppl. II: 80-1031986; 40: 31-5

29. Kusaba A, Tanabe T, Mishima Y, et al. Therapeutic effect of15. Maeda Y, Kanayama S, Okajima Y, et al. Effect of PGE1OP-1206•α-CD in patients with chronic arterial occlusiveanalogue (ONO-1206) on the platelet functions [in Japanese].disease of the extremities: double-blind comparative trial withKetsueki to Myakkan 1982; 13 (1): 142-5ticlopidine [in Japanese]. J Clin Exp Med 1986; 138: 217-26

16. Yamamoto T, Hiromoto J. Phase I clinical trial of 17(S)-methyl-30. Kurihara A, Kataoka O, Sugawara S, et al. Clinical benefit ofomega-homo-trans-delta2-prostaglandinE1•α-cyclodextrin

(OP-1206•α-CD): part 1. Single dose study [in Japanese]. OP-1206•α-CD on lumbar spinal canal stenosis: multi-centerYakuri To Chiryo 1981; 9 (4): 1463-76 comparative double-blind clinical study [in Japanese]. Rinsho

Iyaku 1996; 12 (3): 511-2917. Ito K, Nagashima K, Takenobu Y, et al. Effect of OP-1206 • α-

CD on cauda equina blood flow in dog cauda equina with 31. Uratsuji M, Kurihara A, Iguchi T, et al. The optimal dose forexperimental compression [in Japanese]. Kiso to Rinsho 1995; OP-1206 • α-CD on lumbar spinal canal stenosis: multi-center29 (10): 2577-85 comparative double-blind clinical study [in Japanese]. Rinsho

Iyaku 1996; 12 (3): 489-50918. Sawaragi H, Takenobu Y, Nonaka S, et al. Effect of OP-1206 α

CD on the thermal hyperalgesia induced by constriction injury 32. Ono Pharmaceutical Co Ltd. Postmarketing surveillance: Aprilto the sciatic nerve in the rat [in Japanese]. Kiso to Rinsho 2001-March 2004: Prospective Central Registration System1996; 30 (2): 237-44 Trial [data on file]. Japan: Ono Pharmaceutical Co, Ltd, 2006

19. Takenobu Y, Katsube N, Nakai H. Effect of OP-1206• α-CD onthe walking dysfunction in the rat cauda equina nerve com-

Correspondence: Greg L. Plosker, Wolters Kluwer Health |pressed model [in Japanese]. Kiso to Rinsho 1996; 30 (2):221-7 Adis, 41 Centorian Drive, Private Bag 65901, Mairangi Bay,

Auckland 1311, New Zealand.20. Sekiguchi M, Kikuchi S, Konno S. Effects of liimaprost, anE-mail: [email protected] prostaglandin E1 analogue, in a model of experi-




equina, leading to a lack of nutrition in that area, andLimaprost eventually results in neurological impairment. ThisA Viewpoint by Akira Dezawacauses pain and numbness in the lower limbs, which

Department of Orthopedic Surgery, Teikyo consequently lead to difficulty in walking (intermit-University Mizonokuchi Hospital,

tent claudication). Surgical intervention cannot beKawasaki, Japan

recommended for all patients and LSCS is resistantto drug therapy, such as analgesics and anti-inflam-

OP-1206 α-CD (limaprost alfadex) is an orally matory drugs, as well as orthosis therapy and phys-active prostaglandin E1 (PGE1) analogue that was iotherapy. Therefore, a multicentre, comparative,approved in 1988 for the treatment of various is- double-blind clinical study was performed for thechaemic symptoms, such as ulcer, pain and sensa- purpose of investigating the efficacy, safety andtion of coldness of the hands and feet, associated usefulness of limaprost (15 µg/day for 6 weeks) inwith thromboangiitis obliterans (TAO). In 2001, the LSCS.[1] Indeed, intravenous PGE1 infusions in in-indication of limaprost for improvement in pain and termittent claudication are effective and well tolerat-numbness in lower limbs and walking ability was ed, but most patients would prefer to avoid an infu-expanded to include lumbar spinal canal stenosis sion as a first-line therapy. It is therefore expected(LSCS). Limaprost is a potent inhibitor of platelet that limaprost will be useful in improving the qualityaggregation as well as a strong vasodilator. of life of patients suffering from ischaemic involve-Limaprost 30 µg/day for TAO has a beneficial effect ment by inducing remission of various symp-especially on relieving pain and healing ulcers. Of toms. ▲course, physical training in combination with drugtreatment is a useful therapeutic option.

LSCS most commonly affects the middle-aged Reference1. Kurihara A, Kataoka O, Sugawara S, et al. Clinical benefit ofand elderly population and its main symptom is

OP-1206•α-CD on lumbar spinal canal stenosis: multi-centerintermittent claudication. Such a narrowed spinal comparative double-blind clinical study [in Japanese]. Rinshocanal results in poor blood circulation of the cauda Iyaku 1996; 12 (3): 511-29



effect of improving vascular endothelial functionLimaprostand a vasodilating effect, so it is considered to beA Viewpoint by Shin-Ichi Konnoefficacious in LSCS.

Department of Orthopaedic Surgery, FukushimaBecause PGE1 is rapidly inactivated in the lungs,Medical University, School of Medicine,

large doses are needed; this can lead to hypotension,Fukushima, Japan

diarrhoea and local irritation. To avoid this, lipo-Prostaglandin E1 (PGE1) clathrate compounds PGE1 was developed. Clinical studies have demon-

were developed as antiplatelet drugs: alprostadil strated that lipo-PGE1 is more effective and saferalfadex for injection (1979) and limaprost alfadex than conventional free PGE1 in the treatment offor oral use (1988). PGE1 is both a vasodilator and peripheral vascular disease. However, oral limaprostan inhibitor of platelet aggregation, and has been overcomes the safety concerns of an intravenouswidely used in the treatment of peripheral vascular infusion. Most patients would prefer to avoid andisease in Japan. Intermittent claudication due to

infusion as first-line therapy. Oral limaprost 3–30lumbar spinal canal stenosis (LSCS) is very likely to µg/day was generally well tolerated in clinical trials.account in part for the development of symptoms of

Larger studies that use a double-blind, randomised,reversible functional impairment caused by a rela-

controlled design are needed to evaluate the efficacytive ischaemic state of the cauda equina. Based on

of limaprost for treatment of LSCS and peripheralthese findings, it would be reasonable to considervascular disease, with assessments of patient-basedthat reversal of compression-induced decreasedoutcomes using instruments such as the Roland-blood flow in the cauda equina may be effective inMorris disability questionnaire and the Short-Formimproving symptoms due to LSCS. Limaprost has36. ▲been proven to have an antithrombocytic effect, an

Drugs 2007; 67 (1): 121-153ADIS DRUG EVALUATION 0012-6667/07/0001-0121/$49.95/0


FenofibrateA Review of its Use in Primary Dyslipidaemia, theMetabolic Syndrome and Type 2 Diabetes Mellitus

Gillian M. Keating and Katherine F. Croom


Various sections of the manuscript reviewed by:M. Farnier, Point Medical, Rond Point de la Nation, Dijon, France; P. Gervois, Laboratoire de Biochimie,Faculte des Sciences Pharmaceutiques et Biologiques 3, Lille, France; P.H. Jones, Methodist DeBakey HeartCenter, Baylor College of Medicine, Houston, Texas, USA; D.N. Kiortsis, Medical School, University ofIoannina, Ioannina, Greece; V. Melenovsky, Department of Cardiology, IKEM, Prague, Czech Republic; K.G.Parhofer, University of Munich, Klinikum Grosshadern, Munich, Germany; G.F. Watts, School of Medicineand Pharmacology, University of Western Australia, Royal Perth Hospital, Perth, Western Australia,Australia.

Data SelectionSources: Medical literature published in any language since 1980 on ‘fenofibrate’, identified using MEDLINE and EMBASE, supplementedby AdisBase (a proprietary database of Adis International). Additional references were identified from the reference lists of publishedarticles. Bibliographical information, including contributory unpublished data, was also requested from the company developing the drug.Search strategy: MEDLINE, EMBASE and AdisBase search terms were ‘fenofibrate’ or ‘fibric acid derivatives’. Searches were last updated18 December 2006.Selection: Studies in patients with primary dyslipidaemia, the metabolic syndrome or type 2 diabetes mellitus who received fenofibrate.Inclusion of studies was based mainly on the methods section of the trials. When available, large, well controlled trials with appropriatestatistical methodology were preferred. Relevant pharmacodynamic and pharmacokinetic data are also included.Index terms: Fenofibrate, dyslipidaemia, metabolic syndrome, type 2 diabetes mellitus, pharmacodynamics, pharmacokinetics, therapeuticuse.

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1221. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1232. Pharmacodynamic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

2.1 Effects on Lipids and Apolipoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1252.2 Effects on Lipid Transporters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1252.3 Other Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

3. Pharmacokinetics Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1273.1 Special Patient Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1273.2 Potential Drug Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

4. Clinical Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1284.1 In Primary Dyslipidaemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

4.1.1 Placebo-Controlled Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1294.1.2 Comparisons with HMG-CoA Reductase Inhibitors (Statins) . . . . . . . . . . . . . . . . . . . . . . . . . 1294.1.3 Comparisons with Other Fibrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1294.1.4 In Combination with Statins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

122 Keating & Croom

4.1.5 In Combination with a Cholesterol Absorption Inhibitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1314.1.6 In Combination with a Bile Acid Sequestrant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

4.2 In the Metabolic Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1324.2.1 Comparisons with Placebo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1334.2.2 Comparison with a Statin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1334.2.3 In Combination with Statins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1334.2.4 In Combination with Metformin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1344.2.5 In Combination with Orlistat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

4.3 In Type 2 Diabetes Mellitus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1354.3.1 Effects on Lipid Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374.3.2 Effects on Macrovascular Endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384.3.3 Effects on Microvascular Endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384.3.4 Effects on Glycaemic Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

5. Tolerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396. Dosage and Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1417. Place of Fenofibrate in the Management of Primary Dyslipidaemia, the Metabolic Syndrome

and Type 2 Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

Fenofibrate is a fibric acid derivative indicated for use in the treatment of primaryAbstracthypercholesterolaemia, mixed dyslipidaemia and hypertriglyceridaemia in adultswho have not responded to nonpharmacological measures. Its lipid-modifyingeffects are mediated by activation of peroxisome proliferator-activatedreceptor-α. Fenofibrate also has nonlipid (i.e. pleiotropic) effects (e.g. it reducesfibrinogen, C-reactive protein and uric acid levels and improves flow-mediateddilatation).

Fenofibrate improves lipid levels (in particular triglyceride [TG] and high-den-sity lipoprotein-cholesterol [HDL-C] levels) in patients with primary dys-lipidaemia. Its lipid-lowering profile means that fenofibrate is particularly wellsuited for use in atherogenic dyslipidaemia (characterised by high TG levels, lowHDL-C levels and small, dense low-density lipoprotein [LDL] particles), which iscommonly seen in patients with the metabolic syndrome and type 2 diabetesmellitus. Indeed, fenofibrate improves the components of atherogenic dys-lipidaemia in patients with these conditions, including a shift from small, denseLDL particles to larger, more buoyant LDL particles. Greater improvements inlipid levels are seen when fenofibrate is administered in combination with anHMG-CoA reductase inhibitor (statin) or in combination with ezetimibe, com-pared with monotherapy with these agents. In the DAIS study, fenofibratesignificantly slowed the angiographic progression of focal coronary atherosclero-sis in patients with type 2 diabetes. In terms of clinical outcomes, although nosignificant reduction in the risk of coronary events was seen with fenofibrate inthe FIELD trial in patients with type 2 diabetes, treatment was associated with asignificantly reduced risk of total cardiovascular disease (CVD) events, primarilythrough the prevention of non-fatal myocardial infarction and coronary revascu-larisation. Subgroup analyses revealed significant reductions in total CVD eventsand coronary heart disease events in patients with no previous CVD, suggesting apotential role for primary prevention with fenofibrate in patients with early type 2diabetes. Improvements were also seen in microvascular outcomes withfenofibrate in the FIELD trial. Fenofibrate is generally well tolerated, both as


Fenofibrate: A Review 123

monotherapy and when administered in combination with a statin. Combinationtherapy with fenofibrate plus a statin appears to be associated with a low risk ofrhabdomyolysis; no cases of rhabdomyolysis were reported in patients receivingsuch therapy in the FIELD trial. Thus, fenofibrate is a valuable lipid-loweringagent, particularly in patients with atherogenic dyslipidaemia.

1. Introduction

There are three major classes of lipoprotein: low-density lipoproteins (LDL), the major cholesterol-containing lipoproteins; high-density lipoproteins(HDL), which are integral to the transfer of choles-terol from the peripheral tissues to the liver forbiliary elimination (i.e. reverse cholesterol trans-port); and very-low-density lipoproteins (VLDL),triglyceride (TG)-rich lipoproteins that are, alongwith intermediate-density lipoproteins, precursorsof LDL.[1-3] There is also a fourth class of lipoprote-in, the TG-rich chylomicrons, which are formed inthe intestine from dietary fat.[1]

Risk factors for coronary heart disease (CHD)include elevated LDL-cholesterol (LDL-C) levels,low HDL-cholesterol (HDL-C) levels and elevatedTG levels (especially VLDL remnants).[1] In addi-tion, emerging risk factors include elevated li-poprotein(a) levels, elevated apolipoprotein (Apo) B

Table II. Lipid goals (mmol/L) based on CHD event risk, accordingto the National Cholesterol Education Program Adult TreatmentPanel III guidelines[1,4]

Risk Primary Secondary non-LDL-C HDL-C targettarget (if TG ≥2.3)

Low risk (0–1 risk factors) <4.1 <4.9

Moderate risk (≥2 risk factors; <3.4 <4.110-year risk <10%)

Moderate high risk (≥2 risk <3.4a <4.1factors; 10-year risk 10–20%)

High risk (CHD or CHD risk <2.6 <3.4equivalent [e.g. diabetesmellitus])b

Very high risk (e.g. CHD plus <1.8c <2.6other risk factors [e.g. diabetes ormultiple risk factors of themetabolic syndrome])

a An LDL-C goal of <2.6 mmol/L is an option in these patients.

b Includes patients with ≥2 risk factors and a 10-year risk of>20% (considered a CHD risk equivalent).

c Optional target in patients at very high risk.

CHD = coronary heart disease; HDL-C = high-density lipoprotein-cholesterol; LDL-C = low-density lipoprotein-cholesterol;TG = triglyceride.

levels, low ApoAI levels, elevated homocysteinelevels, elevated fibrinogen levels and elevated C-

event (table II).[1] For example, patients with pre-reactive protein (CRP) levels.[1]

existing CHD and the metabolic syndrome (a pre-A system commonly used to classify the dys-cursor of diabetes mellitus) or diabetes are consid-lipidaemias is shown in table I. LDL-C is the prima-ered to be at very high risk of a CHD event (tablery target of lipid-lowering therapy, and the LDL-CII).[4]

treatment goal varies according to the risk of a CHDThere has been a substantial increase in the prev-

alence of obesity, the metabolic syndrome and dia-betes in recent times.[5] Indeed, it was recently esti-mated that almost one-quarter of adults in the UShave the metabolic syndrome;[6] the increase in thenumber of people with the metabolic syndrome inthe US between 1990 and 2000 (from 50 to 64million) suggests that its prevalence will continue toincrease.[7] It has been suggested that the growingnumber of people affected by this disorder poses athreat to the reduction in CHD risk achieved in theUS population over the past 3 decades (this reduc-

Table I. Fredrickson classification of dyslipidaemias[3]

Type Elevated particles TC TG

I Chylomicrons ↔ ↑↑IIa LDL ↑↑ ↔IIb LDL, VLDL ↑↑ ↑III IDL ↑ ↑IV VLDL ↔↑ ↑↑V Chylomicrons, VLDL ↑ ↑↑IDL = intermediate-density lipoprotein; LDL = low-densitylipoprotein; TC = total cholesterol; TG = triglyceride; VLDL = very-low-density lipoprotein; ↑ indicates increased; ↑↑ indicates greatlyincreased; ↔ indicates normal; ↔↑ indicates normal or increased.


124 Keating & Croom

tion has been achieved via declining LDL-C diabetes have retinopathy at the time of diagno-levels).[1] Indeed, the metabolic syndrome poses a sis.[14] In addition, approximately 20–40% of pa-substantial burden in terms of increased morbidity, tients with diabetes will develop nephropathy, themortality and healthcare costs. The metabolic syn- most common cause of end-stage renal disease.[15]

drome is characterised by the presence of atherogen- Neuropathy affects 40–50% of patients with diabe-ic dyslipidaemia, abdominal obesity, raised blood tes[14] and commonly leads to foot ulceration andpressure (BP), prothrombotic and proinflammatory amputation, major sources of morbidity and disabili-states and insulin resistance.[1] Atherogenic dys- ty in this patient group.[15]

lipidaemia comprises the triad of high TG levels, Lipid-lowering therapy plays an important role inlow HDL-C levels and the presence of small, dense improving the lipid profile of patients with dys-LDL particles (i.e. a pattern B LDL phenotype).[1] lipidaemia, thereby reducing CHD risk. Lipid-low-According to the National Cholesterol Education ering agents include the HMG-CoA reductase inhib-Program (NCEP) Adult Treatment Panel III guide- itors (statins), the fibric acid derivatives (fibrates),lines, the metabolic syndrome is diagnosed when at the bile acid sequestrants, nicotinic acid (niacin) andleast three of the following criteria are present: the cholesterol absorption inhibitors.[1,16]

abdominal obesity (a waist circumference of The fibrate fenofibrate1 has been available since>102cm in men and >88cm in women); a TG level 1975,[17] although the original formulation had poor≥1.7 mmol/L; an HDL-C level <1.0 mmol/L in men solubility resulting in low bioavailability.[18] How-and <1.3 mmol/L in women; BP ≥130/85mm Hg; ever, development of newer formulations of theand a fasting glucose level ≥110 mg/dL.[1] Alterna- drug utilising smaller fenofibrate particle sizes hastive diagnostic criteria have been proposed by the improved its bioavailability. A capsule formulationWHO and other organisations.[8] Patients with the of micronised fenofibrate is available in 67, 200 andmetabolic syndrome have an ≈2-fold increase in the 267mg doses[19-21] and a microcoated tablet formula-risk of developing cardiovascular disease (CVD) tion of micronised fenofibrate is available in 54, 160and an ≈5-fold increase in the risk of developing and 215mg doses.[22,23] A new nanoparticle tablettype 2 diabetes.[9] formulation of fenofibrate utilising NanoCrystal

technology has recently become available in the US,The prevalence of type 2 diabetes is also increas-France and Canada, with launches in other countriesing; it is estimated that 366 million persons will beanticipated.[18,23] This new nanoparticle formulationaffected by diabetes worldwide by 2030.[10] Diabe-can be administered without regard to food (i.e. it istes in itself is a strong independent risk factor fora non-food effect [NFE] formulation) and is availa-CHD,[1] with 2- to 4-fold higher rates of coronaryble in 48 and 145mg doses (section 3).[23,24] Aartery disease seen in patients with diabetes than in200mg dose of the capsule formulation, a 160mgthose without.[11] Atherogenic dyslipidaemia and thedose of the microcoated tablet formulation and ametabolic syndrome are common features of type 2145mg dose of the new NFE formulation are consid-diabetes.[1] Patients with diabetes are at risk of bothered bioequivalent.[18]macrovascular (e.g. CHD, peripheral vascular dis-

ease, cerebrovascular disease) and microvascular This article discusses the pharmacological(retinopathy, nephropathy, neuropathy) complica- properties of fenofibrate, and examines its efficacytions.[12] In terms of microvascular complications, and tolerability in primary dyslipidaemia, the meta-retinopathy is the most common cause of new cases bolic syndrome and type 2 diabetes. Throughout thisof blindness in adults aged 20–74 years.[13] Approxi- article, lipid levels are reported in mmol/L; to con-mately 20% of patients with newly diagnosed type 2 vert to mg/dL, total cholesterol (TC), LDL-C and

1 Trade names for the various fenofibrate formulations include Catalip, Fulcro, Lipanthyl, Lipantil, Lipidil,Lipcor, Secalip, Supralip and Tricor. The use of trade names is for product identification purposes only and doesnot imply endorsement.



HDL-C values should be multiplied by 38.7 and TG sis.[25,32,33] De novo fatty acid synthesis is also inhib-values should be multiplied by 88.6. ited with fenofibrate, through reductions in acetyl-

CoA carboxylase and fatty acid synthase activity;2. Pharmacodynamic Properties this also reduces the availability of fatty acids for

TG synthesis.[32,33] ApoB and VLDL production andsecretion is also reduced with fenofibrate.[32,33]

2.1 Effects on Lipids and ApolipoproteinsFenofibrate increases LDL clearance and reduces

Fenofibrate is a prodrug that is converted into the small dense LDL.[34-38] Larger, less dense LDL par-pharmacologically active metabolite fenofibric acid ticles have a high binding affinity for cellular LDL(section 3).[18] The lipid-modifying effects of fe- receptors, and are less susceptible to oxida-nofibrate are mediated by its activation of the nucle- tion.[28,34-36]

ar transcription factor peroxisome proliferator-acti- PPARα activation also increases ApoAI andvated receptor-α (PPARα) [figure 1].[25] Activated ApoAII synthesis (the major proteins inPPARα forms a heterodimer with another nuclear HDL).[32,33,39,40] Fenofibrate reduces cholesteryl es-receptor, retinoid X receptor, which then binds spe- ter transfer protein (CETP) activity;[41,42] the reduc-cific peroxisome proliferator response elements, tion in CETP-mediated transfer of lipid from HDLthereby modulating the expression of genes regulat- to VLDL may also contribute to the observed in-ing lipid metabolism.[25,26] It has been suggested that crease in HDL-C levels.[34,43]

PPARα expression levels may determine patient In keeping with the above-mentioned mecha-response to fibrates.[27] Whereas fenofibrate appears nisms, the following significant (p < 0.05 vs base-to act as a full agonist at PPARα, gemfibrozil seems line or placebo) changes were seen with fenofibrateto act as a selective PPARα modulator.[27] in patients with primary dyslipidaemia,[37,44-62] the

Activation of PPARα results in increased lipoly- metabolic syndrome[63,64] or type 2 diabetes:[65-68]

sis and plasma clearance of atherogenic TG-rich • reduced ApoCIII levels;[37,44,46,56,60,63,64]

lipoproteins via the activation of lipoprotein lipase • increased ApoAI levels;[37,44-46,49,51-53,57-63,65,68]

and ApoAV and reduced production of the lipopro- • increased ApoAII levels;[37,44,60,63]

tein lipase inhibitor ApoCIII.[2,28-32] Fenofibrate also • reduced ApoB levels;[37,44-49,51-56,58-67]

promotes the β-oxidation of fatty acids, thus reduc- • reduced lipoprotein(a) levels in some stud-ing the availability of free fatty acids for TG synthe- ies.[46,49,50]

Clinical trials in patients with primary dys-lipidaemia, the metabolic syndrome or type 2 diabe-tes revealed that fenofibrate monotherapy was asso-ciated with changes from baseline of –6% to –27%in TC levels, –33% to +1% in LDL-C levels, +1% to+34% in HDL-C levels and –25% to –59% in TGlevels (see section 4 for detailed results). It shouldbe noted that these studies included patients withvarious types of dyslipidaemia (e.g. primaryhypercholesterolaemia, mixed dyslipidaemia,hypertriglyceridaemia).

2.2 Effects on Lipid Transporters

ATP-binding cassette transporter A1 (ABCA1) isinvolved in the control of ApoAI-mediated choles-terol efflux. Fenofibric acid increased transcription

↓ Triglycerides

↓ Inflammation↑ HDL

synthesis

↑ LDLparticle size

Fibrate

ActivatedPPARα/RXR

↑ Reversecholesterol transport

PPARα

PPRE/targetgenes

Fig. 1. Mechanism of action of fenofibrate.[25,28] HDL = high-densitylipoprotein; LDL = low-density lipoprotein; PPARα = peroxisomeproliferator-activated receptor-α; PPRE = peroxisome proliferatorresponse elements; RXR = retinoid X receptor; ↑ indicates in-crease; ↓ indicates decrease.


126 Keating & Croom

of ABCA1 in macrophages, as well as increasing ry cytokines interleukin (IL)-6,[87,91,92,95]

ApoAI-mediated lipid release.[26] The effect of fe- IL-1β,[73,87] tumour necrosis factor-αnofibric acid on ABCA1 was dependent on activa- (TNFα)[51,94,95] and monocyte chemoattractanttion of the nuclear receptor liver X receptor protein-1 (MCP-1);[81,82,86,87,92]

(LXR).[26] Significant (p < 0.05) increases from • reduced intercellular adhesion molecule-1baseline in ABCA1 and LXR-α mRNA levels were (ICAM-1) levels were seen in some[82] but notseen following fenofibrate therapy in patients with all[92] studies;diabetic dyslipidaemia.[69]

• increased serum PON1 activity[58,74,83] and adi-Scavenger receptors play a vital role in the ac- ponectin levels;[52,53]

cumulation of lipids by macrophages.[2] Scavenger • improvement in flow-mediated dilata-receptor class B type I (SR-BI) is thought to be tion,[51-54,73,94] but not the dilator response to ni-involved in cholesterol efflux.[2] Fenofibrate was troglycerin;[51-54,73]

shown to stimulate the expression of SR-BI/CD36- • improved forearm blood flow in response to ace-and LIMPII-analogous 1 in vitro;[70] however, no tylcholine, nitroprusside and verapamil;[76]

increase in SR-BI mRNA levels was seen in patients• reduced uric acid levels;[44,45,48,55,57,60,72,80,84,95]

with diabetic dyslipidaemia who received fenofi-• increased homocysteine levels.[57,72,75,78,79,93]

brate.[69]

Given that the metabolic syndrome is character-ised by a prothrombotic/proinflammatory state (sec-2.3 Other Effectstion 1), the effect of fenofibrate therapy is of particu-

As well as regulating lipid metabolism, PPARα lar interest in patients with this condition. For exam-modulates haemostasis and the inflammatory re- ple, fenofibrate 200 mg/day for 12 weekssponse.[71] Increased levels of fibrinogen, plasmi- significantly reduced fibrinogen levels from base-nogen activator inhibitor-1 (PAI-1) and the inflam- line by 19% (p < 0.01) in a noncomparative study inmatory marker CRP have been linked with an inpatients with the metabolic syndrome (n = 37).[97]

creased risk of atherosclerosis.[71] Raised uric acid Moreover, significantly (p < 0.05) greater reduc-and homocysteine levels have also been linked to tions from baseline were seen with fenofibrate 160atherosclerosis,[72] and impaired vasoreactivity pre- mg/day than with placebo in IL-1β (–3% vs –2%),cedes the development of atherosclerosis.[51,73] By MCP-1 (–5% vs –1%), vascular cell adhesion mole-contrast, paraoxonase (PON1), an enzyme present cule-1 (VCAM-1; –10% vs –3%) and ICAM-1on HDL, is thought to have an antioxidant effect,[74] (–15% vs –2%) levels in a well designed trial inand increased adiponectin levels are thought to im- patients with the metabolic syndrome (n = 55)prove insulin sensitivity.[52] [IL-1β and MCP-1 levels measured in the postpran-

The effect of fenofibrate on these various factors dial state and VCAM-1 and ICAM-1 levels mea-has been examined in studies in patients with dys- sured in the fasting state];[98] in addition, adiponec-lipidaemia,[37,44-49,51-54,57,58,61,72-95] the metabolic syn- tin levels increased from baseline to a significantlydrome,[96-98] impaired glucose tolerance[99] or type 2 greater extent with fenofibrate than with placebodiabetes.[65,68,100] Some studies are available only as (+0.34 vs +0.07 µg/mL; p = 0.0003).[101] Similarly,abstracts.[83,96,98,101] fenofibrate 200 mg/day was associated with signifi-

cant (p < 0.05) improvements in CRP (–34%), IL-6In patients with dyslipidaemia, fenofibrate ther-(–27%) and ICAM (–5%) levels, but not VCAMapy was associated with the following significantlevels, in another placebo-controlled study in 25(p < 0.05 vs baseline or placebo) changes:patients with the metabolic syndrome.[96]• reduced plasma fibrinogen[45-49,51-54,61,74,85,88-91]

and PAI-1[51,85] levels; Patients with impaired glucose tolerance who

• reduced CRP levels,[37,52,53,57,61,73,74,77,91,92,95] received fenofibrate 267 mg/day for 1 monthalong with reduced levels of the pro-inflammato- (n = 31) had significant (p < 0.01) reductions from



baseline in plasma levels of fibrinogen (17%), quivalence in terms of Cmax and area under thePAI-1 (26%) and CRP (25%), plasma factor VII plasma concentration-time curve (AUC) valuesactivity (19%) and release of TNFα (32%), IL-1β when 145mg of the NFE formulation was adminis-(35%), IL-6 (22%) and MCP-1 (17%).[99] tered in the fasting state or with high- or low-fat

meals.[18] In the fasting state, a single dose of fe-Significant reductions in fibrinogen (–21%[65]

nofibrate 145mg was associated with a Cmax of 7.9and –14%[68]) and uric acid (–25%)[68] levels wereµg/mL, a tmax of 2.3 hours and an AUC from timeseen with fenofibrate 200 mg/day in studies in pa-zero to infinity of 123.8 µg • h/mL.[18] Thus, thetients with type 2 diabetes and dyslipidaemiaNFE formulation is not prone to food-related(n = 40[68] and 120[65]) [p < 0.05 vs baseline orchanges in bioavailability, resulting in predictableplacebo].[65,68] Moreover, fenofibrate 200 mg/dayefficacy.was associated with a significantly greater improve-

Fenofibrate does not accumulate with repeat ad-ment than placebo in brachial artery flow-mediatedministration,[22] and plasma concentrations ofdilatation (+44% vs –9%; p = 0.01) in patients withfenofibric acid attained steady state within 5 days.[24]diabetic dyslipidaemia.[68] Fenofibrate 200 mg/dayFenofibric acid is >99% bound to plasma albu-was also associated with a significant increase frommin.[22]baseline in homocysteine levels (from 11.0 to 16.5

µmol/L; p < 0.001) in patients with type 2 diabetes No unchanged fenofibrate was detectable in plas-(n = 418), according to an analysis of the DAIS ma following oral administration of the drug and itstrial.[100] However, this increase did not seem to metabolism to fenofibric acid.[24] The drug wasattenuate the improvement in the progression of mainly excreted in the urine as fenofibric acid andfocal coronary atherosclerosis seen with fenofibrate its glucuronide conjugate.[22] Following administra-(section 4.3).[102] tion of radiolabelled fenofibrate, ≈60% of the radio-

activity was detected in the urine and ≈25% in thefaeces.[24] Fenofibric acid has an elimination half-3. Pharmacokinetics Propertieslife of ≈20 hours.[18]

Following oral administration, fenofibrate is rap-idly hydrolysed by esterases to its active metabolite 3.1 Special Patient Populationsfenofibric acid.[18] Repeat administration of themicrocoated tablet formulation of fenofibrate 160 In elderly volunteers aged 77–87 years, the oralmg/day resulted in a steady-state mean maximum clearance of fenofibric acid was 1.2 L/h, comparedplasma concentration (Cmax) of fenofibric acid of with 1.1 L/h in young adults.[24] No dosage adjust-12.2 µg/mL, reached in a time (tmax) of 3.5 hours.[18] ment is recommended in the elderly for theThe mean trough plasma concentration of fenofibric microcoated tablet formulation of fenofibrate,[22] al-acid was 4.1 µg/mL.[18] The mean plasma concentra- though a reduced initial dosage is recommended fortion of fenofibric acid following administration of the NFE formulation of the drug, according to the200 mg/day of the capsule formulation of US prescribing information.[24] The pharmacokinet-micronised fenofibrate was 15 µg/mL.[20] The ab- ics of fenofibrate were similar in men and wo-sorption of fenofibrate from these formulations is men.[24] Fenofibrate has not been studied in paedia-less efficient when administered in a fasting state, tric patients or in patients with hepatic impairment.meaning that they must be administered with The clearance of fenofibric acid is reduced infood.[20,22] For example, the extent of absorption of patients with severe renal impairment, with accumu-the microcoated tablet formulation is increased by lation seen with repeat administration.[24] Dose re-≈35% in the fed versus fasting state.[18] duction is recommended in patients with renal im-

By contrast, the nanoparticle NFE formulation of pairment (creatinine clearance [CLCR] <50 or <60fenofibrate may be administered without regard to mL/min). The initial recommended dose in patientsfood.[24] A study in healthy volunteers revealed bioe- with renal impairment (CLCR <50 mL/min) is 48mg


128 Keating & Croom

of the NFE formulation, according to US prescrib- nine levels were significantly (p = 0.003) in-ing information,[24] and one (CLCR <20 mL/min) or creased.[113] Thus, coadministration of the twotwo (CLCR 20–60 mL/min) 67mg capsules of agents may increase the inherent nephrotoxicity ofmicronised fenofibrate, according to UK prescribing ciclosporin.information.[19]

4. Clinical Efficacy3.2 Potential Drug Interactions Most studies included in this section had a drug-

free or placebo run-in period, in addition to dietaryFenofibrate has a low potential for drug interac-control, prior to starting active treatment. Dietarytions. In vivo studies showed that neither fenofibraterestrictions were generally maintained for the dura-nor fenofibric acid undergo significant oxidativetion of the study. Unless stated otherwise, patientsmetabolism.[24] Fenofibrate and fenofibric acid didreceived once-daily administration of the capsulenot inhibit cytochrome P450 (CYP) isozymesformulation of micronised fenofibrate 200mg or theCYP3A4, CYP2D6, CYP2E1 or CYP1A2, accord-microcoated tablet formulation of micronised fe-ing to the results of in vitro studies in human livernofibrate 160mg. Definitions for the trial abbrevia-microsomes.[24] They showed mild to moderate inhi-tions and acronyms used in this section are found inbition of CYP2C9 and weak inhibition of CYP2C19table III.and CYP2A6.[24]

Concomitant administration of fenofibrate had4.1 In Primary Dyslipidaemia

no clinically significant effect on the pharmacoki-netics of the statins simvastatin,[103] rosuvastatin,[104] The potential of monotherapy with oral fenofi-fluvastatin[105] or atorvastatin,[106] and only modest brate 160[89] or 200[59,88,114-118] mg/day in the treat-effects on the exposure of pravastatin and its activemetabolite 3α-hydroxy-iso-pravastatin.[107] Thepharmacokinetics of fenofibrate were not altered toa significant extent by the concomitant administra-tion of simvastatin,[103] rosuvastatin,[104] atorvasta-tin[24] or pravastatin.[108]

In terms of lipid-lowering agents other than sta-tins, concomitant ezetimibe did not have a signifi-cant effect on the pharmacokinetics of fenofibrate inhealthy volunteers[109] or patients with primaryhypercholesterolaemia.[110] Although fenofibrate in-creased ezetimibe exposure, this effect was not con-sidered of clinical significance.[109,110] Concomitantadministration of colesevelam did not have a signifi-cant effect on the bioavailability of fenofibrate.[111]

Fenofibrate did not have a significant effect onthe pharmacokinetics of the meglitinide analoguerepaglinide.[112] Fenofibrate has been shown topotentiate the effect of coumarin-type anticoagu-lants, with prolongation of the prothrombin time orinternational normalised ratio.[24] Whole-blood con-centrations of ciclosporin were significantly(p = 0.03) reduced in heart transplant patients re-ceiving concomitant fenofibrate, and serum creati-

Table III. Trial abbreviations/acronyms and definitions

Abbreviation/ Definitionacronym

4D Deutsche Diabetes Dialyse Studie

4S Scandinavian Simvastatin Survival Study

ACCORD Action to Control Cardiovascular Risk inDiabetes

ASCOT Anglo-Scandinavian Cardiac Outcomes Trial

ASPEN Atorvastatin Study for Prevention of CoronaryHeart Disease Endpoints in Non-InsulinDependent Diabetes Mellitus

BIP Bezafibrate Infarction Prevention

CARDS Collaborative Atorvastatin Diabetes Study

CARE Cholesterol and Recurrent Events

DAIS Diabetes Atherosclerosis Intervention Study

DIACOR Diabetes and Combined Lipid Therapy Regimen

FIELD Fenofibrate Intervention and Event Lowering inDiabetes

HHS Helsinki Heart Study

HPS Heart Protection Study

LIPID Long-term Intervention with Pravastatin inIschaemic Disease

SAFARI Study of simvastatin plus fenofibrate forcombined hyperlipidemia

UKPDS UK Prospective Diabetes Study

VA-HIT Veterans Affairs HDL-C Intervention Trial



4.1.2 Comparisons with HMG-CoA Reductasement of adults with primary dyslipidaemia wasInhibitors (Statins)shown in noncomparative trials. However, theseStudies comparing the efficacy of fenofibratetrials are not discussed further, given the availability

with statins in patients with primary dyslipidaemiaof numerous well designed, comparative trials.have been reviewed previously.[130] Thus, this sec-Comparative trials included patients with prima-tion provides only a brief overview, focusing onry hypercholesterolaemia[48,55,119-122] and/or mixedfully published studies of best design (see table IVdyslipidaemia;[48,52,55,57,60,119-121,123-127] mean patientfor study design details and detailed results).age was 45–60 years.[46,48,52,55,57,60,119-126,128,129]

In summary, fenofibrate 200 mg/day was consist-Where specified, primary endpoints included ently associated with significantly greater improve-

the change from baseline in LDL-C ments than statins in TG levels,[57,119-121,126] and inlevels,[55,119,122,124,125,127] HDL-C levels[126] or TG HDL-C levels in most studies,[57,119,121,126] in patientslevels.[119,123]

with primary hypercholesterolaemia[119-121] or mixedWhere specified, analyses were generally con- dyslipidaemia[57,119-121,126] (table IV). By contrast,

ducted in the intention-to-treat (ITT) popula- statins were generally associated with significantlytion[55,119-121,123-126,129] (using last observation carried greater improvements than fenofibrate in TC andforward [LOCF] analysis)[119,123] or efficacy evalu- LDL-C levels (table IV).[57,121,126]

able population.[122]

4.1.3 Comparisons with Other Fibrates4.1.1 Placebo-Controlled Trials Fenofibrate 200 mg/day was associated with sig-The efficacy of fenofibrate in placebo-controlled nificantly greater reductions from baseline in TC

trials in patients with primary dyslipidaemia and LDL-C levels than sustained-release gemfi-(n = 20–340) has been reviewed previously[130] and brozil 900 mg/day among 21 patients with type IIawill only be discussed briefly in this section. Trials or IIb dyslipidaemia, according to the results of awere of randomised, double-blind, placebo-con- randomised, double-blind, crossover study.[48] Aftertrolled design[46,55,60,129] and, where specified, in- 6 weeks’ treatment, TC levels were reduced fromcluded patients with types IIa,[55] IIb,[55,60] IV[60] or baseline by 22% with fenofibrate and 15% withV[60] dyslipidaemia. Other trials included patients gemfibrozil (p = 0.0148) and LDL-C levels werewith TG levels of 2.3–6 mmol/L and HDL-C levels reduced by 27% and 16% in the correspondingof <0.9 mmol/L,[46] or TG levels ≥2.3 mmol/L and treatment groups (p = 0.0117). There was no signifi-an LDL-C : HDL-C ratio of ≥5.[129] Patients re- cant difference between fenofibrate and gemfibrozilceived micronised fenofibrate 200 mg/day,[46] recipients in the change from baseline in HDL-C200–400 mg/day (only results pertaining to the 200 (+9% vs +9%) or TG (–54% vs –47%) levels.mg/day dosage will be discussed),[55] or 67–201 mg/ Similarly, fenofibrate improved the lipid profileday;[129] in one study, patients received standard to a significantly greater extent than gemfibrozil in afenofibrate 300 mg/day (equivalent to 200 mg/day nonblind switch study in 80 patients with dys-of micronised fenofibrate).[60] lipidaemia (LDL-C >3.4 mmol/L or TG >2.3 mmol/

In all trials, significantly (p < 0.05) greater reduc- L) and CHD.[128] Almost half of the patients (n = 39)tions from baseline in TG levels were seen with were receiving treatment with a statin at baselinefenofibrate than with placebo (–27% to –56% vs and continued to do so during the study. After–12% to +14%).[46,55,60,129] Significantly (p < 0.05) receiving gemfibrozil 600mg twice daily for ≥3greater improvements in TC (–6% to –25% vs +1% months, they switched to fenofibrate 201 mg/dayto +2%),[46,55,60] LDL-C (–32% vs +1%)[55] and (three 67mg capsules) for 12 weeks. After 12 weeks’HDL-C (+11% to +34% vs 0% to +9%)[46,60] levels therapy, improvements from baseline in lipid levelswere also seen with fenofibrate than with placebo in were significantly (p < 0.001) greater with fenofi-some trials. brate than gemfibrozil both among patients who


130 Keating & Croom

Table IV. Efficacy of fenofibrate (FEN) [micronised FEN capsule formulation] compared with HMG-CoA reductase inhibitors (statins) inpatients (pts) with primary dyslipidaemia

Study (design details) Lipid entry criteria Treatment (mg/day) No. of pts Mean change from baseline in lipid level (%)(mmol/L) [treatment duration] [mean baseline level; mmol/L]

TC LDL-C HDL-C TG

Comparisons with atorvastatin (ATO)

Despres et al.[126] LDL-C >3.2, HDL-C FEN 200 [12wk] 79 –15 [6.3] –16 [4.2] +13†††a [0.9] –30†† [2.5](r, nb, pg, mc) <1.2 (women) or <1.1 ATO 10 [12wk] 86 –28*** [6.1] –39*** [4.2] +5a [0.9] –15 [2.3]

(men), TG <4.5

Malik et al.[57] TC >6.2, TG >1.5 FEN 200 [10wk] 29 –12 [7.6] –8 [4.4] +13†† [1.3] –50†† [5.4](r, sb, co) ATO 10 [10wk] 29 –28*** [7.6] –34*** [4.4] –1 [1.3] –32 [5.4]

Comparison with pravastatin (PRA)

Ducobu et al.[119] IIa: TC ≥6.5, TG <2.3; FEN 200 [3mo] 75 –18 [7.6] –18a [5.2] +13†† [1.1] –39†††b [2.0](r, db, pg, mc) or IIb: TC >6.5, TG PRA 20 [3mo] 76 –15 [7.6] –17a [5.3] +6 [1.0] –12b [2.2]

2.3–4.5

Comparisons with simvastatin (SIM)

Farnier et al.[120]c IIa: TC ≥6.2, LDL-C FEN 200 [3mo] 15 –27 [8.3] –33 [6.4] +1.3 [1.5] –36††† [1.0](r, db, co) ≥4.1, TG <1.5 SIM 20 [3mo] 16 –28 [8.8] –36 [6.9] +1.4 [1.4] –1 [1.0]

IIb: TC ≥6.2, TG >1.5 FEN 200 [3mo] 11 –23 [7.8] –25 [5.6] +25 [1.1] –53††† [2.6]but <5.2 SIM 20 [3mo] 10 –21 [7.6] –29 [5.1] +15 [1.1] +25 [3.1]

Steinmetz et al.[121] IIa: TC >6.5; or IIb: TC FEN 200 [12wk] 66 –19 [7.6] –21 [5.5] +18† [1.0] –41††† [2.4](r, db, pg, mc) >6.5, TG 2.9–5.8 SIM 20 [12wk] 64 –25* [7.6] –35** [5.4] +15 [0.9] –17 [2.9]

a Primary endpoint.

b Primary endpoint in the subgroup of pts with type IIb dyslipidaemia.

c Pts were randomised to FEN for 3mo followed by SIM for 3mo or vice versa. Because of the absence of a washout period betweenthe first and second treatment periods, the changes in lipid levels are only reported for the first 3mo treatment period.

IIa/b = Fredrickson classification type IIa/b; co = crossover; db = double-blind; HDL-C = high-density lipoprotein-cholesterol; LDL-C = low-density lipoprotein-cholesterol; mc = multicentre; nb = nonblind; pg = parallel-group; r = randomised; sb = single-blind; TC = totalcholesterol; TG = triglyceride; * p < 0.05, ** p < 0.001, *** p < 0.0001 vs FEN; † p < 0.05, †† p < 0.01, ††† p < 0.001 vs statin.

were receiving a statin (TC: –36% vs –32%; LDL- Noncomparative[80,131,132] and nonrandomised,C: –44% vs –40%; HDL-C: +31% vs +26%; TG: nonblind[49,133] trials examining the use of combina-–53% vs –48%) and those who were not (TC: –15% tion therapy are not discussed.vs –11%; LDL-C: –18% vs –13%; HDL-C: +21% vs In general, combination therapy with fenofibrate+14%; TG: –47% vs –42%). plus a statin improved lipid levels to a greater extent

A significantly greater increase from baseline in than fenofibrate monotherapy and/or statinApoAI levels occurred with fenofibrate than with monotherapy in patients with primary dyslipi-gemfibrozil (+9% vs +2%; p < 0.001) in patients daemia.[52,122,123] For example, in the SAFARI trial,with mixed dyslipidaemia.[27] Patients in this

fenofibrate 160 mg/day plus simvastatin 20 mg/dayrandomised, double-blind, multicentre study re-

improved TC, LDL-C, HDL-C and TG levels to aceived fenofibrate 200 mg/day (n = 116) or

significantly greater extent than simvastatin 20 mg/gemfibrozil 1200 mg/day (n = 118) for 24 weeks.day alone in patients with mixed dyslipidaemia (ta-There was no significant difference between fe-ble V).[123] Similarly, fenofibrate, administerednofibrate and gemfibrozil recipients in the changealone or in combination with atorvastatin, was asso-from baseline in TG (–39% vs –41%) or HDL-Cciated with significantly greater improvements in(+16% vs +12%) levels.HDL-C and TG levels, compared with atorvastatinalone, in patients with mixed dyslipidaemia (table4.1.4 In Combination with StatinsV).[52] In addition, TC and LDL-C levels decreasedThis section focuses on the results of well de-to a significantly greater extent in patients receivingsigned, randomised, double-blind trials.[52,122,123]



atorvastatin alone or combination therapy than in ents, 22% of ezetimibe monotherapy recipients,those receiving fenofibrate alone.[52] Finally, im- 64% of fenofibrate plus ezetimibe recipients andprovements in TC, LDL-C, HDL-C and TG levels 12% of placebo recipients.were significantly greater with fenofibrate plus At the end of this study, 576 patients entered afluvastatin, than with fenofibrate alone, in patients 48-week, double-blind, extension study duringwho mainly had primary hypercholesterolaemia (ta- which they received fenofibrate (n = 236) or fe-ble V).[122] nofibrate plus ezetimibe (n = 340).[134] After an

additional 48 weeks of treatment, improvements4.1.5 In Combination with a Cholesterol from baseline (i.e. the start of the initial study) in TCAbsorption Inhibitor (–23% vs –14%; p < 0.001), LDL-C (–22% vs –9%;Combination therapy with fenofibrate plus the p < 0.001), HDL-C (+21% vs +18%; p = 0.02) and

cholesterol absorption inhibitor ezetimibe was asso- TG (–46% vs –42%; p = 0.002) levels were signifi-ciated with significantly greater reductions from cantly greater with combination therapy than withbaseline in TC, LDL-C and TG levels than fenofibrate alone.monotherapy with either agent alone in patients with Triple therapy with fenofibrate 160 mg/day plusmixed dyslipidaemia (n = 588) [figure 2], according a fixed combination of ezetimibe/simvastatin 10mg/to the results of a well designed study.[124] Fe- 20mg per day for 12 weeks improved the lipidnofibrate, alone or in combination with ezetimibe, profile to a significantly greater extent than fe-was also associated with significantly greater in- nofibrate 160 mg/day alone or ezetimibe/simvasta-creases from baseline in HDL-C levels, compared tin 10mg/20mg per day in patients with mixed dys-with ezetimibe alone (figure 2). lipidaemia (n = 611).[127] In this randomised, double-

At baseline, ≥70% of patients in each treatment blind, placebo-controlled, multicentre study (availa-group had a pattern B LDL phenotype.[124] Follow- ble as an abstract), triple therapy reduced LDL-Cing treatment, a shift to a larger LDL size pattern had levels from baseline to a significantly (p < 0.05)occurred in 62% of fenofibrate monotherapy recipi- greater extent than fenofibrate alone or placebo

Table V. Efficacy of combination therapy with fenofibrate (FEN) [micronised FEN capsule[52,122] or microcoated tablet[123] formulation] and anHMG-CoA reductase inhibitor (statin) in patients (pts) with primary hypercholesterolaemia[122] or mixed dyslipidaemia[52,123]

Study (trial design) Lipid entry criteria Treatment (mg/day) No. of pts Mean change from baseline in lipid level (%)(mmol/L) [treatment duration] [mean baseline level; mmol/L]

TC LDL-C HDL-C TG

Farnier et al.[122] LDL-C ≥4.9, TG ≤3.9 FEN 200 [16wk] 32 –19** [9.1] –21**a [7.1] +4* [1.2] –29* [1.8]

(r, db, pg, mc) FEN 200 + FLU 20 33 –27 [9.7] –32a [7.6] +14 [1.3] –39 [1.8][16wk]

FEN 200 + FLU 40 31 –35 [9.2] –41a [7.1] +3 [1.4] –40 [1.6][16wk]

Grundy et al.[123] LDL-C >3.4, TG SIM 20 [12wk] 207 –20 [6.6] –26 [4.2] +10 [1.1] –20a,b [2.6]

(r, db, pg, mc) 1.7–5.6 FEN 160 + SIM 20 411 –26‡‡ [6.6] –31‡‡ [4.2] +19‡‡ [1.1] –43‡‡a,b [2.6][SAFARI trial] [12wk]

Koh et al.[52] TC ≥5.2, TG 2.3–9.0 FEN 200 [2mo] 56 –13 [6.0] –6 [3.4] +23‡ [1.1] –55‡ [3.8]

(r, db, co) ATO 10 [2mo] 56 –29† [6.3] –40† [3.5] 0 [1.2] –25 [3.4]

FEN 200 + ATO 10 56 –29† [6.2] –30† [3.3] +15‡ [1.2] –57‡ [3.6][2mo]

a Primary endpoint.

b Median value.

ATO = atorvastatin; co = crossover; db = double-blind; FLU = fluvastatin; HDL-C = high-density lipoprotein-cholesterol; LDL-C = low-density lipoprotein-cholesterol; mc = multicentre; pg = parallel-group; r = randomised; SIM = simvastatin; TC = total cholesterol;TG = triglyceride; * p < 0.05, ** p < 0.001 (analysis of variance F test for the comparison of the three treatment groups); † p < 0.05 vs FENmonotherapy; ‡ p < 0.05, ‡‡ p < 0.001 vs statin monotherapy.


132 Keating & Croom

4.1.6 In Combination with a Bile Acid SequestrantFollowing 8 weeks’ treatment with fenofibrate

160 mg/day, the addition of the bile acid sequestrantcolesevelam 3.75 g/day for a further 6 weeks result-ed in significantly (p < 0.0001) greater reductions inTC and LDL-C levels, compared with continuationof fenofibrate alone, in patients with mixed dys-lipidaemia.[125] However, the addition of coleseve-lam did not result in additional significant improve-ment in HDL-C or TG levels, compared with fe-nofibrate alone. Lipid levels with fenofibrate pluscolesevelam versus fenofibrate alone at baseline, theend of the fenofibrate run-in period and study endwere 6.6 versus 6.5 mmol/L, 6.0 versus 5.8 mmol/Land 5.6 versus 5.9 mmol/L for TC (mean values);4.1 versus 4.1 mmol/L, 3.8 versus 3.7 mmol/L and3.4 versus 3.8 mmol/L for LDL-C (mean values);1.2 versus 1.2 mmol/L, 1.3 versus 1.3 mmol/L and1.3 versus 1.3 mmol/L for HDL-C (mean values);and 2.6 versus 2.5 mmol/L, 1.8 versus 1.8 mmol/Land 1.9 versus 1.6 mmol/L for TG (median values).This randomised, double-blind, parallel-group, mul-ticentre study included 122 patients.[125]

4.2 In the Metabolic Syndrome

The potential of fenofibrate in the treatment ofpatients with the metabolic syndrome was shown innoncomparative trials.[97,135] However, these studiesare not discussed further, given the availability of

Per

cent

age

chan

ge fr

om b

asel

ine

in li

pid

leve

ls

−50

−45

−40

−35

−30

−25

−20

−15

−10

−5

0

5

10

15

20

25

FEN (n = 179)EZE (n = 173)FEN + EZE (n = 175)PL (n = 61)

TC

†

‡ ‡

‡**

HDL-C

‡ ‡††

TG

† †*‡ ‡

LDL-C

** †‡

**

Fig. 2. Efficacy of combination therapy with fenofibrate (FEN) plusezetimibe (EZE) in patients with mixed dyslipidaemia. Results of arandomised, double-blind, parallel-group, multicentre study inwhich patients received FEN 160 mg/day (micronised FENmicrocoated tablet formulation), EZE 10 mg/day, FEN 160 mg/dayplus EZE 10 mg/day or placebo (PL) for 12wk.[124] The primaryendpoint was the change from baseline in LDL-C levels. Meanvalues are reported except for TG levels (median values). Baselinelipid levels (mmol/L) across treatment groups were as follows: TC =6.7–6.9; LDL-C = 4.1–4.3; HDL-C = 1.1; TG = 2.9–3.2. HDL-C =high-density lipoprotein-cholesterol; LDL-C = low-density lipopro-tein-cholesterol; TC = total cholesterol; TG = triglyceride; * p < 0.05,** p < 0.001 vs FEN monotherapy; † p < 0.001 vs EZE monother-apy; ‡ p < 0.001 vs PL.

well designed, comparative trials.(–46% vs –16% and –4%); LDL-C levels were Several comparative trials have examined the

efficacy of fenofibrate in patients with the metabolicreduced in patients receiving ezetimibe/simvastatinsyndrome.[63,96,136-140] Where specified, randomised

by 47%. In addition, HDL-C levels increased by astudies were of double-blind[63,96,136-138] or non-

significantly (p < 0.05) greater extent with triple blind[139] design. Studies[63,136-139] generally used theNCEP definition[1] of the metabolic syndrome (sec-therapy than with ezetimibe/simvastatin or placebotion 1). Patients included in these studies did not(+19% vs +9% and +1%); HDL-C levels increasedgenerally have diabetes.[63,96,136-139] Where specified,

in patients receiving fenofibrate alone by 18%. Both mean patient age was 46–63 years.[63,96,136,137,139,140]

TG (–50% vs –41%, –29% and –3%) and non-HDL- The main endpoint discussed in this section is theeffect of treatment on atherogenic dyslipidaemia.C (–51% vs –21%, –45% and –2%) levels wereThe effect of treatment on other relevant endpointsreduced to a significantly (p < 0.05) greater extent(e.g. prothrombotic and proinflammatory markers)

with triple therapy than with fenofibrate alone, eze- is discussed in section 2. Where specified, primarytimibe/simvastatin or placebo.[127] endpoints included the change from baseline in TG



levels,[137] reversal of the metabolic syndrome diag- statin 10 mg/day alone in patients with the metabolicnosis[139] and the proportion of ‘normalised’ patients syndrome.[137] For example, after 3 months’ therapy(i.e. patients with no biochemical abnormalities inin this crossover study (n = 20), TG and HDL-Cdicative of the metabolic syndrome).[136] levels were improved to a significantly greater ex-

tent with fenofibrate plus simvastatin than withSome of these studies are available only as ab-simvastatin alone (figure 3). The ratio ofstracts.[96,138,140]

large : small LDL particles was significantly (p <4.2.1 Comparisons with Placebo 0.05) larger with fenofibrate plus simvastatin thanFenofibrate improved aspects of atherogenic dys- with simvastatin alone or placebo (2.4 vs 1.1 and

lipidaemia in patients with the metabolic syndrome, 0.9).according to the results of two 12-week, placebo-

The addition of fenofibrate 160 mg/day to simva-controlled trials (n = 25[96] and 45[138]). Comparedstatin significantly reduced mean TG levels (fromwith placebo, fenofibrate 160[138] or 200[96] mg/day3.4 to 2.1 mmol/L; p < 0.02) and increased meansignificantly reduced TG levels by 43% (from 3.2 topeak LDL particle size (from 251 to 259A;1.9 mmol/L; p = 0.0003)[138] and 59% (from 5.6 top < 0.0002), in 34 patients with the metabolic syn-2.3 mmol/L; p < 0.001),[96] reduced LDL-C levels bydrome.[140] After 4 months’ treatment, 11 of 346% (from 3.5 to 3.3 mmol/L; p = 0.006)[138] andpatients (32%) had converted from a pattern B LDLincreased LDL particle size (p < 0.0001).[138] Thephenotype (characterised by small, dense LDL parti-change from baseline in HDL-C levels in fenofibratecles) to a pattern A LDL phenotype (characterisedrecipients did not significantly differ from that inby larger, more buoyant LDL particles). The re-placebo recipients in either trial.[96,138] Fenofibratemaining patients were then randomised to receivetherapy was not associated with any significant

changes in glucose metabolism.[96]

In a crossover study in patients with the metabol-ic syndrome (n = 11), TG levels were significantlylower (1.7 vs 2.4 mmol/L; p < 0.001), and HDL-Clevels were significantly higher (1.0 vs 0.9 mmol/L;p < 0.001), after 5 weeks’ treatment with fenofibrate200 mg/day versus placebo.[63] Results pertaining tothe atorvastatin arm of this trial are presented insection 4.2.2.

4.2.2 Comparison with a StatinFollowing 5 weeks’ treatment, HDL-C levels

were significantly higher with fenofibrate 200 mg/day than with atorvastatin 40 mg/day (1.0 vs 0.9mmol/L; p = 0.001), according to the results of acrossover trial in 11 patients with the metabolicsyndrome.[63] By contrast, TC (3.5 vs 5.6 mmol/L;p < 0.001) and LDL-C (1.9 vs 3.7 mmol/L;p < 0.001) levels were significantly lower withatorvastatin than with fenofibrate.

4.2.3 In Combination with StatinsCombination therapy with fenofibrate 200 mg/

day plus simvastatin 10 mg/day improved ather-ogenic dyslipidaemia to a greater extent than simva-

Mea

n lip

id le

vel (

mm

ol/L

)

0

1

2

3

4

5

6

7

8 SIMFEN + SIMPL

TC

**

Non-HDL-C

**

LDL-C

**

HDL-C

†

TG

*†

Fig. 3. Effect of fenofibrate (FEN) plus simvastatin (SIM) on ather-ogenic dyslipidaemia in patients with the metabolic syndrome. Re-sults of a randomised, double-blind, crossover study in which 20patients received SIM 10 mg/day, FEN 200 mg/day (micronisedFEN capsule formulation) plus SIM 10 mg/day or placebo (PL) for3mo each.[137] Patients had TG levels of 2.3–9.0 mmol/L and LDL-Clevels of 4.1 mmol/L at baseline. HDL-C = high-density lipoprotein-cholesterol; LDL-C = low-density lipoprotein-cholesterol; TC = totalcholesterol; TG = triglyceride; * p < 0.05 vs PL; † p < 0.05 vs SIMalone.


134 Keating & Croom

the thiazolidinedione (and PPARγ agonist) rosig-litazone 8 mg/day (n = 12) or nicotinic acid 2 g/day(n = 11) for a further 6 months. Further significant(p < 0.02 vs simvastatin plus fenofibrate) improve-ments were seen in TG levels and peak LDL particlesize with the addition of rosiglitazone, but not withnicotinic acid. A pattern A LDL phenotype was seenin 9 of 12 (75%) patients receiving simvastatin plusfenofibrate and rosiglitazone and in 7 of 11 (64%)patients receiving simvastatin plus fenofibrate andnicotinic acid.

4.2.4 In Combination with MetforminCombination therapy with fenofibrate plus the

oral antidiabetic agent metformin was more effec-tive than monotherapy with either agent alone interms of normalising biochemical abnormalities inpatients with the metabolic syndrome.[136] Patients(n = 681) in this 3-month study were randomised toreceive twice-daily treatment with one of the follow-ing regimens: fenofibrate 80mg plus metformin850mg; fenofibrate 80mg plus metformin 500mg;fenofibrate 40mg plus metformin 850mg; fenofi-brate 40mg plus metformin 500mg; fenofibrate80mg; metformin 850mg; or placebo.

The normalisation rate (i.e. the proportion ofpatients with no biochemical abnormalities at 3months) was significantly higher in patients receiv-ing fenofibrate 160 mg/day plus metformin 1700mg/day than in those receiving monotherapy withfenofibrate 160 mg/day or metformin 1700 mg/day(figure 4).[136] In terms of individual parameters,normalisation rates in patients receiving fenofibrateplus metformin, fenofibrate monotherapy, met-formin monotherapy and placebo were 33.3–39.4%,19.6%, 30.0% and 20.0% for fasting glucose levels;39.8–55.0%, 50.0%, 18.5% and 15.4% for TGlevels; and 31.3–35.0%, 28.8%, 18.8% and 16.7%for HDL-C levels. Significant differences were seenbetween patients receiving fenofibrate 160 mg/dayplus metformin 1700 mg/day and fenofibratemonotherapy for fasting glucose levels (p = 0.002)and metformin monotherapy for TG levels(p < 0.001). Percentage changes from baseline infasting glucose, TG and HDL-C levels are alsoshown in figure 4.

Per

cent

age

of p

atie

nts

0

2

4

6

8

10

12

14

16

18

20

FEN 160

+ M

ET 170

0

FEN 160

+ M

ET 100

0

FEN 80

+ M

ET 170

0

FEN 80

+ M

ET 100

0

FEN 160

MET 1

700 PL

Leas

t squ

ares

mea

n pe

rcen

tage

cha

nge

from

bas

elin

e

−40

−35

−30

−25

−20

−15

−10

−5

0

5

10

Fasting glucose

**††

HDL-C

*‡

TG

**‡‡

a

b

‡†

FEN 160 + MET 1700FEN 160 + MET 1000FEN 80 + MET 1700FEN 80 + MET 1000FEN 160MET 1700PL

Fig. 4. Efficacy of fenofibrate (FEN) [the micronised FENmicrocoated tablet formulation] + metformin (MET) in the metabolicsyndrome. Results of a randomised, double-blind, multicentre studyin patients receiving FEN 160 mg/day + MET 1700 mg/day (n =109), FEN 160 mg/day + MET 1000 mg/day (n = 104), FEN 80 mg/day + MET 1700 mg/day (n = 106), FEN 80 mg/day + MET 1000mg/day (n = 104), FEN 160 mg/day (n = 103), MET 1700 mg/day (n= 100) or placebo (PL; n = 55).[136] Shown are (a) the proportion ofpatients with normalisation of metabolic syndrome biochemical ab-normalities (primary endpoint) and (b) the change from baseline infasting glucose, TG and HDL-C levels. p-Values were calculatedagainst the best treatment in figure b. TG = triglyceride; HDL-C =high-density lipoprotein-cholesterol; * p = 0.001, ** p < 0.0001 vsPL; † p = 0.009, †† p < 0.0001 vs FEN monotherapy; ‡ p < 0.01, ‡‡p < 0.0001 vs MET monotherapy.



4.2.5 In Combination with Orlistat although the improvements were significantlyFenofibrate, alone or in combination with the (p < 0.05) greater with orlistat-containing regimens

lipase inhibitor orlistat, improved metabolic param- than with fenofibrate monotherapy.eters in patients with the metabolic syndrome.[139]

The Homeostasis Model Assessment Index wasAfter 3 months’ therapy, 48% of patients receiving significantly (p < 0.01) improved from baseline inmonotherapy with fenofibrate 200 mg/day (n = 29), patients receiving fenofibrate alone (from 3.6 to44% of patients receiving monotherapy with orlistat 3.0), orlistat alone (from 3.9 to 2.6) and fenofibrate120mg three times daily (n = 29) and 50% of pa- plus orlistat (from 3.9 to 3.0).[139] The improvementtients receiving fenofibrate 200 mg/day plus orlistat in recipients of orlistat monotherapy was signifi-120mg three times daily (n = 28) no longer fulfilled cantly (p < 0.05) greater than that in recipients ofthe criteria for the metabolic syndrome (all fenofibrate monotherapy.p < 0.0001 vs baseline); there were no significantbetween-group differences.

4.3 In Type 2 Diabetes MellitusPatients receiving fenofibrate monotherapy, orli-stat monotherapy and fenofibrate plus orlistat all had

This section examines the efficacy of fenofibratesignificant (p < 0.05) improvements from baselinein patients with type 2 diabetes. Small (n = 32–40)in serum levels of TC (–13%, –15% and –26%),trials of placebo-controlled[66-68] or noncompara-LDL-C (–12%, –18% and –30%), non-HDL-Ctive[142] design in patients with type 2 diabetes(–17%, –17% and –26%) and TG (–35%, –14% anddemonstrated that fenofibrate 200 mg/day had a–37%) after 3 months’ therapy.[139] Moreover, re-beneficial effect on the lipid profile in this patientductions from baseline in TC and LDL-C werepopulation; however, these trials are not discussedsignificantly (p < 0.05) greater with combinationfurther.therapy than with either monotherapy. In addition,

HDL-C levels increased to a significantly (p < 0.05) This section focuses on the results of two large,greater extent with fenofibrate monotherapy than long-term, randomised, double-blind, placebo-con-with orlistat alone or fenofibrate plus orlistat (+3% trolled, multicentre trials in patients with type 2vs –9% and –10%), and TG levels decreased to a diabetes: the DAIS (n = 418)[102] and FIELDsignificantly (p < 0.05) greater extent with fenofi- (n = 9795)[143] trials. Study design details, includingbrate-containing regimens than with orlistat alone. lipid entry criteria, are presented in table VI. Pa-At baseline, mean lipid levels across treatment tients in the DAIS study had mild lipoprotein abnor-groups were as follows: TC 7.6–7.9 mmol/L; LDL- malities.[102] Patients in the FIELD study were con-C 3.9–4.4 mmol/L; HDL-C 1.3–1.4 mmol/L; non- sidered at increased risk of CHD, although theyHDL-C 6.3–6.6 mmol/L; and TG 2.0–2.5 mmol/L. generally had no clear indication for lipid-loweringAt 6 months, reductions in small dense LDL-C therapy; ≈40% of patients had dyslipidaemia atlevels (–63% and –77% vs –35%) and increases in baseline (dyslipidaemia was defined as a TG level ofmean LDL diameter (+2.3 and +2.7 vs +0.7nm) >1.7 mmol/L and an HDL-C level of <1.03 mmol/Lwere significantly (p < 0.05) greater with fenofibrate for men and <1.29 mmol/L for women).[143] More-monotherapy or fenofibrate plus orlistat, than with over, 2131 patients had previous CVD and 7664 didorlistat alone.[141] not.[143] In the FIELD study, the median duration of

At 3 months, significant (p < 0.05) reductions type 2 diabetes was 5 years and the baselinefrom baseline in BP were also seen in all three glycosylated haemoglobin (HbA1c) level was 6.9%,treatment groups, with no significant between-group indicating good glycaemic control.[143] Patientsdifferences.[139] Similarly, anthropometric measure- (mean age 56–62 years) were followed for ≥3 yearsments such as body mass index, waist circumference in the DAIS trial[102] and for ≈5 years in the FIELDand bodyweight were significantly (p < 0.05) im- trial.[143] The drop-out rate in the FIELD study,proved from baseline in all three treatment groups, averaged over 5 years, was 11% in fenofibrate recip-


136K

eating & C

room

2007 A

dis D

ata

Info

rma

tion

BV. A

ll righ

ts rese

rved

.D

rug

s 2007; 67 (1)

Table VI. Effect of fenofibrate (FEN) [micronised FEN capsule[65,102,143-146,148] or microcoated tablet[147] formulation] on lipid levels in patients (pts) with type 2 diabetes mellitus

Study Lipid entry criteria Treatment (mg/day) No. of Mean change from baseline in lipid level (%) [baseline level; mmol/L]

(trial design) (mmol/L) [treatment duration] pts TC LDL-C HDL-C TGCompared with placebo (PL)

DAIS investigators[102]a TC : HDL-C ratio ≥4 plus FEN 200 [≥3y] 207 –9.6** [5.6] –5.9** [3.4] +7.4** [1.0] –28.5** [2.6]

(r, db, pg, mc) either LDL-C 3.5–4.5 and PL [≥3y] 211 +0.4 [5.6] +0.4 [3.4] +1.5 [1.1] +1.1 [2.4]TG ≤5.2, or TG 1.7–5.2 andLDL-C ≤4.5

FIELD study TC 3.0–6.5 plus TC : HDL-C FEN 200 [≈5y] 4895 –16* [5.0] –21* [3.1] +3* [1.1] –25* [2.0]

investigators[143] ratio ≥4 or TG 1.0–5.0 PL [≈5y] 4900 –9 [5.0] –15 [3.1] +2 [1.1] –3 [1.9](r, db, pg, mc)

Compared with, or in combination with, HMG-CoA reductase inhibitors (statins)

Athyros et al.[65] TC >5.7, LDL-C >3.4, TG FEN 200 [24wk] 40 –16† [6.5] –15† [4.2] +16† [0.9] –41† [3.2]

(r, nb, pg) 2.3–4.5, HDL-C <1.04 ATO 20 [24wk] 40 –31† [6.5] –40† [4.2] +9† [0.9] –30† [3.1]

FEN 200 + ATO 20 [24wk] 40 –37†‡§ [6.6] –46†‡§ [4.2] +22†‡ [0.9] –50†‡§ [3.1]

Derosa et al.[148] TC ≥5.2, LDL-C ≥2.6, TG FLU ER 80 [12mo] 23 –20†b [6.7] –25†b [4.8] +14†b [1.1] –17b [1.7]

(r, db, pg) ≥1.7 FEN 200 + FLU ER 80 25 –26†b [6.8] –35†‡b [4.9] +34†‡b [1.1] –32†‡b [1.8][12mo]

Durrington et al.[146]c TC ≥5.2, TG 2.3 to <9.0 PL → FEN 67–201 [24wk] 49 –8 [6.3] +1 [3.7] +9 [1.0] –34b [4.2]

(r, pg, mc) PL → ROS 10–40 [24wk] 51 –37§ [6.2] –47§θ [3.7] +6 [1.0] –30b [3.6]

ROS 5 → ROS 5 + FEN 60 –31 [6.5] –34 [3.9] +11 [1.1] –41 [3.5]67–201 [24wk]

ROS 10 → ROS 10 + FEN 53 –36 [6.4] –42 [3.9] +12 [1.0] –47‡b [3.5]67–201 [24wk]

Frost et al.[144] LDL-C >3.5, TG >2.3 FEN 200 [6wk] 11 –16† [7.0] –11 [4.4] +11 [1.3] –39† [2.9]

(r, nb, co) ATO 10 [6wk] 11 –24† [6.7] –29† [4.1] +10† [1.2] –4 [2.9]

Muhlestein et al.[147] LDL-C ≥2.6, HDL-C <1.0, FEN 160 [12wk] 100 –13† [5.8] –10† [3.5] +14† [0.9] –38†‡d [3.1]

(DIACOR; r, db, pg) TG ≥2.3 SIM 20 [12wk] 100 –26†§ [5.9] –34†§ [3.7] +7† [1.0] –25†d [2.6]

FEN 160 + SIM 20 [12wk] 100 –27†§ [6.0] –29†§ [3.5] +13† [0.9] –49†‡d [3.2]

Sarano et al.[145]e LDL-C >4.1, TG 2.4–4.5 FEN 200 [4wk] 16 –22† [7.9] –26† [5.3] +16† [1.1] –42† [3.2]

(r, pg) FLU 40 [4wk] 16 –22† [7.9] –28† [5.3] +9 [1.1] –23 [3.2]

FEN 200 + FLU 40 [12wk]f 30 –31† [7.9] –41† [5.3] +25† [1.1] –45† [3.2]

a Percentage changes were estimated from a graph.

b Primary endpoint.

c Pts received 6 weeks of double-blind treatment with PL or ROS 5 or 10 mg/day, after which they entered an 18-week nonblind dose-titration phase. The dosage of FEN orROS was titrated upwards as long as the LDL-C level remained >1.3 mmol/L.

d Median.

e Abstract. Baseline values were reported for the entire treatment group, rather than individual treatment arms.

f After the initial 4-week treatment period, pts with an LDL-C level >2.6 mmol/L received combination therapy.

ATO = atorvastatin; co = crossover; db = double-blind; ER = extended release; FLU = fluvastatin; HDL-C = high-density lipoprotein-cholesterol; LDL-C = low-density lipoprotein-cholesterol; mc = multicentre; nb = nonblind; pg = parallel-group; r = randomised; ROS = rosuvastatin; SIM = simvastatin; TC = total cholesterol; TG = triglyceride; * p < 0.05, ** p <0.001 vs PL; † p < 0.05 vs baseline; ‡ p < 0.05 vs statin monotherapy; § p < 0.05 vs FEN monotherapy; θ p < 0.05 vs ROS 5 + FEN 67–201.


ients and 10% in placebo recipients (hazard ratio tin monotherapy were generally not performed in[HR] 1.01; 95% CI 0.93, 1.11).[143] these studies, although in two studies,[146,147] statin

monotherapy was associated with significantlyThe primary endpoint in the DAIS trial was thegreater reductions from baseline in TC and LDL-Cmean segment diameter, reflecting the angiographiclevels than fenofibrate monotherapy, and in oneprogression of diffuse coronary atherosclerosis.[102]

study,[147] fenofibrate monotherapy was associatedIn the FIELD trial, the initial primary endpoint waswith significantly greater reductions from baselinedeath from CHD; however, this was subsequentlyin TG levels than statin monotherapy (table VI). Aamended to CHD death or non-fatal myocardialsignificant (p < 0.05) reduction from baseline ininfarction (MI) [i.e. total CHD events].[143]

small, dense LDL particles was also reported withAlso discussed are the results of smallerfenofibrate in one study.[144](n = 11–300) studies comparing the use of fe-

nofibrate with a statin[65,144-147] and/or examining theIn Combination with Statinseffect of combination therapy with fenofibrate plusCombination therapy with fenofibrate plus a sta-a statin[65,145-148] in patients with type 2 diabetes.

tin was generally associated with significantlyPatients generally had mixed dys-greater improvements in the lipid profile than treat-lipidaemia;[65,144,146-148] patients in one study alsoment with fenofibrate alone and/or a statin alone inhad CHD (i.e. they were secondary prevention pa-patients with type 2 diabetes and dyslipi-tients).[148] Mean/median patient age was 54–61daemia.[65,145-148]years.[65,144-148] Study design details are presented in

For example, fenofibrate plus atorvastatin gener-table VI. Where specified, primary efficacyally improved the lipid profile to a significantlyendpoints included the change from baseline ingreater extent than monotherapy with either agentTC,[148] LDL-C,[148] HDL-C[148] and TG[146,148]

alone (table VI).[65] In addition, combination therapylevels, and the proportion of patients achieving allwith fenofibrate plus fluvastatin resulted in furtherthree NCEP goals (i.e. LDL-C level <2.6 mmol/L,improvements in the lipid profile among patientsHDL-C level ≥1.0 mmol/L and TG level <1.7 mmol/with type 2 diabetes who had an LDL-C level ofL).[149]

>2.6 mmol/L, despite having received treatmentWhere specified, efficacy analyses were conduct-with either agent alone (table VI).[145] Moreover,ed on the ITT population[102,143,146] using LOCFcombination therapy with fenofibrate plus fluvasta-analysis.[146] Three analyses are only available astin,[148] rosuvastatin[146] or simvastatin[147] improvedabstracts.[145,149,150]

LDL-C,[148] HDL-C[148] and/or TG[146-148] levels to a4.3.1 Effects on Lipid Levels significantly greater extent than treatment with a

statin alone (table VI).Comparisons with PlaceboPreliminary results of an add-on substudy[149] ofResults of the DAIS[102] and FIELD[143] trials in

the main DIACOR trial[147] (see table VI) indicatepatients with type 2 diabetes revealed significantlythat triple therapy with fenofibrate 160 mg/day plusgreater improvements in the lipid profile with fe-simvastatin 20 mg/day and ezetimibe 10 mg/daynofibrate 200 mg/day than with placebo (table VI).was associated with a significantly greater propor-It should be noted that patients did not have substan-tion of patients achieving all three NCEP goals,tial dyslipidaemia at baseline in either trial.compared with dual therapy with fenofibrate 160

Comparisons with Statins mg/day plus simvastatin 20 mg/day (23.5% vs 0%;As seen in patients with primary dyslipidaemia p = 0.006).[149] In the DIACOR study, patients re-

(section 4.1.2), fenofibrate and statins tended to ceived fenofibrate, simvastatin or fenofibrate plusimprove different components of the lipid profile in simvastatin for 12 weeks, followed by 9 months’patients with type 2 diabetes (table VI).[65,144-147] dual combination therapy for all patients. PatientsStatistical analyses comparing fenofibrate with sta- (n = 37) who still had an LDL-C level ≥2.6 mmol/L


138 Keating & Croom

or a TG level ≥1.7 mmol/L were then randomised in group difference in the total incidence of CHDevents in the FIELD trial.[143]a double-blind manner to receive fenofibrate plus

simvastatin in combination with either ezetimibe or A prespecified subgroup analysis of the FIELDplacebo. These results should be interpreted with study data revealed a greater treatment effect incaution given the small number of patients receiving patients with no previous CVD (primary prevention)triple therapy. than in patients with prior CVD (secondary preven-

tion).[143] A significant (p = 0.004) 19% reduction intotal CVD events occurred in patients with no previ-4.3.2 Effects on Macrovascular Endpointsous CVD (HR 0.81; 95% CI 0.70, 0.94), but not inIn the FIELD trial, there was no significant dif-patients with previous CVD (HR 1.02; 95% CI 0.86,ference between fenofibrate and placebo recipients1.20) [interaction p = 0.05]. Similarly, a post hocin the incidence of total CHD events (i.e. CHDsubgroup analysis showed a significant (p = 0.014)mortality or non-fatal MI) [5.2% vs 5.9%, respec-25% reduction in CHD events in patients withouttively] (HR 0.89; 95% CI 0.75, 1.05) or CHD mor-prior CVD (HR 0.75; 95% CI 0.59, 0.94), but not intality (2.2% vs 1.9%) [HR 1.19; 95% CI 0.90, 1.57].those with prior CVD (HR 1.08; 95% CI 0.84, 1.38)However, fenofibrate recipients were significantly[interaction p = 0.03].less likely than placebo recipients to experience

Compared with placebo recipients, fenofibratenon-fatal MI (3.2% vs 4.2% of patients; p = 0.01),recipients also had a significant 38% reduction incorresponding to a 24% relative reduction (HR 0.76;the number of non-traumatic amputations95% CI 0.62, 0.94).[143]

(p = 0.011).[150]

In terms of prespecified secondary endpoints, theIn the DAIS study, there was no significant dif-incidence of total CVD events was significantly

ference between fenofibrate and placebo recipientslower with fenofibrate than with placebo (12.5% vsin the change from baseline in mean segment diame-13.9%; p = 0.035) [HR 0.89; 95% CI 0.80, 0.99],ter (–0.06 vs –0.08mm); this measure reflects dif-and fenofibrate recipients were significantly lessfuse disease.[102] However, fenofibrate significantlylikely than placebo recipients to undergo coronaryslowed the angiographic progression of focal coro-revascularisation (5.9% vs 7.4%; p = 0.003) [HRnary atherosclerosis. The reduction in average mini-0.79; 95% CI 0.68, 0.93] or any revascularisationmum lumen diameter was 40% less with fenofibrate(i.e. coronary, carotid or peripheral; 7.8% vs 9.6%;than with placebo (–0.06 vs –0.10mm; p = 0.029),p = 0.001) [HR 0.80; 95% CI 0.70, 0.92].[143] Thereand progression in the percentage diameter stenosiswere no significant between-group differences inwas reduced by 42% (+2.11% vs +3.65%;the incidence of CVD mortality (2.9% vs 2.6%) [HRp = 0.02).[102]

1.11; 95% CI 0.87, 1.41], total mortality (7.3% vs6.6%) [HR 1.11; 95% CI 0.95, 1.29] or total stroke 4.3.3 Effects on Microvascular Endpoints(3.2% vs 3.6%) [HR 0.90; 95% CI 0.73, 1.12]. In terms of the effect of fenofibrate on microvas-

cular disease, patients receiving fenofibrate wereIt should be noted that in the FIELD trial, signifi-significantly (p = 0.0003) less likely than thosecantly more placebo than fenofibrate recipients re-receiving placebo to need laser treatment for reti-ceived additional lipid-lowering therapy during thenopathy in the FIELD trial (HR 0.70; 95% CI 0.58,study (17% vs 8%; p < 0.0001) [averaged over the0.85).[143]5-year study period].[143] In 93% of placebo recipi-

ents and 94% of fenofibrate recipients this addition- Fenofibrate also reduced the progression of albu-al drug was a statin. While it was expected that some minuria, according to the results of the FIELD andpatients would start other lipid-lowering therapies DAIS trials.[102,143] In the FIELD trial, fenofibrateduring the trial, it has been suggested that this higher recipients were significantly (p = 0.002) more likelyrate of statin use among placebo recipients may than placebo recipients to regress from, or not pro-partially explain the lack of a significant between- gress to, albuminuria.[143] With fenofibrate and pla-



cebo, 10% and 11% of patients progressed (from concerning the use of fenofibrate in combinationnormo- to microalbuminuria or from micro- to with ezetimibe, metformin or statins were also ob-macroalbuminuria) and 9% and 8% regressed. A tained from trials discussed in section 4.similar significant (p = 0.031) effect on progression Fenofibrate was generally well tolerated in theto albuminuria was seen in a subgroup analysis treatment of dyslipidaemia.[24] The pooled analysis(n = 314)[151] of the DAIS trial data.[102] Progression revealed that the most commonly occurring adverseof albuminuria occurred in 8% of fenofibrate recipi- events (regardless of causality) in fenofibrate recipi-ents and 18% of placebo recipients, regression oc- ents included liver function test (LFT) abnormali-curred in 13% and 11%, and no change occurred in ties, respiratory disorder, abdominal pain, back pain,79% and 71%.[151]

increased AST levels, headache, increased ALTThese results are of particular interest given that levels, increased creatine phosphokinase (CPK)

patients in the FIELD trial had relatively early type 2 levels, diarrhoea, nausea, rhinitis, asthenia, flu syn-diabetes.[143]

drome and constipation (figure 5).[24] Only LFTabnormalities and increased AST levels occurred in4.3.4 Effects on Glycaemic Controlsignificantly more fenofibrate than placebo recipi-Glycaemic control was generally not altered to aents (figure 5). Increases in creatinine levels haveclinically significant extent in patients receiving fe-also been reported in patients receiving fibrates,nofibrate. In the FIELD trial, median HbA1c wasincluding fenofibrate.[105,152] The mechanism behind6.9% at baseline and 7.0% at study end in fe-this increase in creatinine levels is unclear,[105] al-nofibrate recipients; corresponding values in place-though it has been suggested that it reflects anbo recipients were 6.9% and 6.9%.[143] Small in-increase in the metabolic production of creatinine,creases from baseline in mean HbA1c levels oc-rather than impaired renal function.[153]curred in recipients of both fenofibrate (from 7.50%

to 7.97%) and placebo (from 7.56% to 7.80%) in the In patients with type 2 diabetes who participatedDAIS study; these increases were too small to be in the FIELD trial, serious adverse effects that wereconsidered clinically relevant.[102] Moreover, HbA1c deemed possibly related to treatment occurred inlevels were unchanged in patients receiving fe- 0.8% of fenofibrate 200 mg/day recipients and 0.5%nofibrate, atorvastatin, or fenofibrate plus atorvasta- of placebo recipients.[143] No cases of rhabdomyol-tin in another trial.[65] A significant 12% relative ysis occurred in patients receiving combination ther-reduction from baseline in mean HbA1c level oc- apy with fenofibrate plus a statin. Overall, rhabdo-curred in patients receiving combination therapy myolysis occurred in three fenofibrate recipientswith fenofibrate plus fluvastatin in one study.[148]

(0.06%) and one placebo recipient (0.02%); all casesfully resolved. Significantly more fenofibrate than5. Tolerabilityplacebo recipients developed pulmonary embolism(1.1% vs 0.7%; p = 0.022) or pancreatitis (0.8% vsInformation concerning the general tolerability0.5%; p = 0.031), although it should be noted thatof fenofibrate were obtained from a pooled analysisthe incidence of these events was low. Other clini-of patient data (804 patients) from randomised,cally important adverse events occurring in <2% ofdouble-blind, placebo-controlled trials reported inpatients included deep vein thrombosis (1.4% ofthe manufacturer’s prescribing information.[24] Ad-fenofibrate recipients vs 1.0% of placebo recipi-verse event data were also obtained from the wellents), myositis (0.04% vs 0.02%) and renal diseasedesigned FIELD trial (n = 9795) [see section 4.3 forneeding dialysis (0.3% vs 0.4%). Less than 1% ofstudy design details].[143] The FIELD trial includedpatients in either treatment arm developed ALTpatients with type 2 diabetes who were followed forlevels 3–5 times or >5 times the upper limit of≈5 years. At study end, 36% of placebo recipientsnormal, or CPK levels 5–10 times or >10 times theand 19% of fenofibrate recipients were receivingupper limit of normal. At study end, median creati-non-study lipid-lowering drugs. Tolerability data


140 Keating & Croom

Inci

denc

e (%

of p

atie

nts)

0

1

2

3

4

5

6

7

8 FENPL*

*

LFT a

bnor

mali

ty

Respir

ator

y diso

rder

Abdom

inal p

ain

Back p

ain

Incr

ease

d AST le

vels

Heada

che

Incr

ease

d ALT

leve

ls

Incr

ease

d CPK le

vels

Diarrh

oea

Nause

a

Rhinitis

Asthen

ia

Flu sy

ndro

me

Consti

patio

n

Fig. 5. Tolerability of fenofibrate (FEN) in patients with dyslipidaemia. Results of a pooled analysis of data from randomised, double-blind,placebo (PL)-controlled trials.[24] Patients received FEN (n = 439) or PL (n = 365); the dosage of FEN was equivalent to 200mg of thecapsule formulation. Shown are adverse events occurring with an incidence of >2%, regardless of causality. ALT = alanine aminotransfer-ase; AST = aspartate aminotransferase; CPK = creatine phosphokinase; LFT = liver function test; * significantly different vs PL (p-value notstated).

nine levels were significantly higher in fenofibrate Adverse events (most commonly related to thegastrointestinal system) were reported inthan in placebo recipients (91 vs 80 µmol/L; p <44.0–52.8% of patients with the metabolic syn-0.01); creatinine levels >200 µmol/L occurred in 2%drome receiving fenofibrate 80 or 160 mg/day inof fenofibrate recipients and 1% of placebo recipi-combination with metformin 1000 or 1700 mg/day;ents. The increase in creatinine levels seen in fenofi-no dose relationship was evident.[136] Adversebrate recipients was thought unlikely to be ofevents were reported in 34.0% of patients receivingclinical significance and reversed following discon-fenofibrate alone, 44.0% of patients receiving

tinuation of treatment.[143]

metformin alone and 38.2% of placebo recipients.Fenofibrate was generally well tolerated when Serious adverse events were reported in <4% of

administered in combination with ezetimibe in pa- patients in any treatment arm.tients with mixed dyslipidaemia.[124] Treatment-re-

The adverse event profile in patients with prima-lated adverse effects were reported in 11.4% of

ry dyslipidaemia receiving combination therapypatients receiving fenofibrate 160 mg/day plus eze- with fenofibrate plus fluvastatin was similar to thattimibe 10 mg/day, compared with 14.3% of patients in patients receiving fenofibrate alone.[122] The mostreceiving fenofibrate alone, 6.4% of patients receiv- commonly reported adverse events included gastro-ing ezetimibe alone and 7.8% of placebo recipients. intestinal adverse events (26% of fenofibrate 200There were no reports of myopathy, rhabdomyolysis mg/day recipients vs 6% of fenofibrate 200 mg/dayor pancreatitis, and no patient experienced an in- plus fluvastatin 20 mg/day recipients and 24% ofcrease from baseline in CPK level to >10-fold the fenofibrate 200 mg/day plus fluvastatin 40 mg/dayupper limit of normal. recipients), musculoskeletal adverse events (e.g.



myalgia) [24% vs 17% and 15%] and respiratory tions with fenofibrate, yielding an odds ratio (OR) of1.78 (95% CI 1.43, 2.22). The rate of rhabdomyoly-adverse events (15% vs 11% and 24%). There weresis was also significantly higher with gemfibrozilno significant between-group differences in terms ofthan with fenofibrate (59.6 vs 5.5 per million pre-the changes from baseline in creatinine, AST, ALT,scriptions) [OR 10.84; 95% CI 8.44, 13.95]. Therealkaline phosphatase, CPK or myoglobin levels.[122]

were 68 cases of rhabdomyolysis in fenofibrate re-Similarly, changes in LFTs and muscle enzymescipients and 1304 cases in gemfibrozil recipients (inwere generally transient and resolved followingtotal, 3161 adverse events were filed that listedtreatment discontinuation in 56 patients receivingfenofibrate or gemfibrozil as a suspect agent). Incombination therapy with fenofibrate 200 mg/dayfenofibrate recipients, 24% of patients experiencingplus atorvastatin 10 mg/day.[52] In addition, no casesrhabdomyolysis were receiving concomitant ceriva-of rhabdomyolysis or clinical myopathy were re-statin (no longer on the market), 12% were receivingported among 403 patients receiving fenofibrateanother statin and 64% were receiving no statin.160 mg/day plus simvastatin 20 mg/day in anotherCorresponding percentages among gemfibrozil re-study.[123]

cipients were 89%, 10% and 1%.[161]In patients with the metabolic syndrome[137] or

Similarly, the use of fenofibrate plus a statin wastype 2 diabetes,[65,146,148] combination therapy withassociated with fewer reports of rhabdomyolysisfenofibrate plus simvastatin, fluvastatin, atorvastatinthan the use of gemfibrozil plus a statin, accordingor rosuvastatin was not associated with clinicallyto the results of another retrospective analysis of onerelevant changes in CPK levels. Myalgia was report-of the FDA adverse event databases referred toed in 1.8–3.3% of patients receiving fenofibrate plusabove (January 1998 to March 2002).[162] Thererosuvastatin[146] and in no patients receiving fenofi-were 4.5 cases of rhabdomyolysis per million pre-brate plus atorvastatin (n = 40).[65]

scriptions reported with fenofibrate plus a statinConcern has been raised over the possibility of compared with 87 cases per million prescriptions

rhabdomyolysis in fibrate recipients, particularly in with gemfibrozil plus a statin. Notably, of the totalpatients receiving a fibrate in combination with a 606 reports of rhabdomyolysis, 14 (2%) were asso-statin. Indeed, a retrospective analysis using claims ciated with fenofibrate plus cerivastatin and 533data from US managed healthcare plans revealed a (88%) were associated with gemfibrozil plus ceriva-2-fold increase in the risk of rhabdomyolysis with statin. The mechanisms possibly underlying thethe use of fibrate plus statin combination therapy, higher rate of rhabdomyolysis seen with gemfibrozilcompared with the use of a fibrate alone,[154] and plus statin combination therapy are discussed fur-there have been rare case reports of rhabdomyolysis ther in section 7.in patients receiving both monotherapy with fenofi-brate[155] or combination therapy with fenofibrate 6. Dosage and Administrationand a statin.[156-158] Myopathy has also been reportedrarely in patients receiving monotherapy with fe- As previously mentioned (section 1), there arenofibrate[159] or combination therapy with fe- several formulations of fenofibrate available, in-nofibrate plus rosiglitazone.[160]

cluding capsule and microcoated tablet formulationsof micronised fenofibrate (available in the EU), andHowever, fenofibrate is associated with a lowera nanoparticle NFE formulation of fenofibrate (cur-risk of rhabdomyolysis than gemfibrozil, accordingrently available in the US, France and Canada, withto the results of a retrospective analysis of adversefurther launches anticipated).event data from two US FDA databases (January

1999 to December 2002).[161] The rate of muscle- In the EU, fenofibrate is approved for use in therelated adverse events (excluding frank rhabdomy- treatment of hypercholesterolaemia and hypertrig-olysis) was 15.7 per million prescriptions with lyceridaemia, either alone or combined (i.e. Fred-gemfibrozil compared with 8.8 per million prescrip- rickson type IIa, IIb, III, IV or V dyslipidaemia; see


142 Keating & Croom

table I), in adults who have not responded to dietary with diabetes had an LDL-C level of <2.6 mmol/and other nonpharmacological measures.[22] Fenofi- L[164] (the NCEP target).brate is also approved for use in patients with secon- Treatment options for dyslipidaemia include thedary hyperlipoproteinaemia that has not resolved statins, the fibrates, the bile acid sequestrants anddespite treatment of the underlying disease (e.g. in nicotinic acid,[1] as well as newer agents such as thepatients with diabetic dyslipidaemia). The recom- cholesterol absorption inhibitor ezetimibe (tablemended dosage of the microcoated tablet formula- VII). The benefit of statins in terms of reducingtion of micronised fenofibrate is 160mg once daily, cardiovascular morbidity and mortality has beentaken with food.[22] For the capsule formulation of clearly demonstrated in clinical trials,[165] and thesemicronised fenofibrate, the initial recommended agents are generally the first-line option for lower-dosage is three 67mg capsules per day (in divided ing LDL-C levels.[1] However, statins have moredoses)[19] or one 200mg capsule per day;[20] an in- modest effects on TG and HDL-C levels (table VII),creased dosage of one 267mg capsule per day is meaning that even with optimal statin therapy, someappropriate in patients with severe dyslipidaemia.[21]

patients (e.g. those with mixed dyslipidaemia) mayThe capsule formulation should also be adminis- have residual CHD risk that is amenable to furthertered with food. In addition, the 48 and 145mg NFE reduction.[165,166] The presence of conditions such asformulation is available in France and can be admin- diabetes and the metabolic syndrome predicts a highistered without regard to meals. residual risk.[12,165] For example, subgroup analysis

In the US, fenofibrate is approved for use in in a recent meta-analysis of 14 trials found that statinadults with Fredrickson type IIa, IIb, IV or V dys- therapy reduced the risk of major coronary events bylipidaemia who have not responded to nonpharma- only 22% in patients with a history of diabetes.[167]

cological measures.[24] The recommended initial Moreover, when patients in some of the major statindosage of the NFE formulation is 145 mg/day in trials (e.g. 4S) were classified according to whetherprimary hypercholesterolaemia and mixed dys- they had lower or higher HDL-C levels, statins onlylipidaemia and 48–145 mg/day in hypertriglycer- reduced the risk of coronary events in patients withidaemia; as mentioned previously, this formulation lower HDL-C levels to a similar extent to that seencan be administered without regard to meals. in patients with higher HDL-C levels who were

Local prescribing information should be con- receiving placebo;[168] this finding is of particularsulted for contraindications, warnings and precau- relevance to patients with atherogenic dyslipidaemiations, and for information concerning drug interac- and low HDL-C levels, such as those with type 2tions and dosage adjustments in special patient diabetes.populations.

Fibrates activate PPARα. As previously men-tioned, fenofibrate appears to act as a full PPARα7. Place of Fenofibrate in theagonist, whereas gemfibrozil appears to be a partialManagement of Primary Dyslipidaemia,PPARα agonist (section 2.1).[27] Bezafibrate is anthe Metabolic Syndrome and Typeagonist at all three PPAR isoforms, including2 DiabetesPPARγ.[5] Newer formulations of fenofibrate haveimproved bioavailability. In particular, the new na-Lipid-lowering therapy is underutilised in pa-noparticle NFE formulation can be administeredtients with dyslipidaemia. For example, a recent USwithout regard to food (section 3).study in >6500 individuals without clinical CVD

Fenofibrate is particularly associated with im-found that ≈30% of the population had dys-provements in TG and HDL-C levels in patientslipidaemia, and that only about half of these patientswith primary dyslipidaemia, as shown by the resultswere receiving lipid-lowering therapy.[163] More-of studies demonstrating the efficacy of fenofibrateover, another recent US study (n >7000) indicatedmonotherapy in such patients (section 4.1).that in a managed care setting, <25% of patients



Table VII. Mechanism of action and main lipid effects of lipid-lowering drugs[1,16,169]

Drug class Mechanism of action Main lipid effects

Fibric acid derivatives (e.g. Agonist at the peroxisome proliferator-activated receptor-α, resulting LDL-C: ↓5–20%; HDL-C:fenofibrate, gemfibrozil, in increased lipolysis and plasma clearance of atherogenic TG-rich ↑10–35%; TG: ↓20–50%bezafibrate) lipoproteins, reduced availability of fatty acids for TG synthesis,

increased ApoAI and ApoAII synthesis (the major proteins in HDL),increased reverse cholesterol transport and a shift to larger, morebuoyant LDL particles (see also section 2)

HMG-CoA reductase inhibitors Inhibit HMG-CoA reductase leading to reduced cholesterol synthesis, LDL-C: ↓18–55%; HDL-C(e.g. lovastatin, pravastatin, decreased hepatic cholesterol content, up-regulation of hepatic LDL ↑5–15%; TG ↓7–30%simvastatin, fluvastatin, receptors and decreased serum LDL-C levelsatorvastatin, rosuvastatin)

Cholesterol absorption inhibitors Inhibits intestinal absorption of biliary and dietary cholesterol LDL-C: ↓18%; HDL-C: ↑4%;(e.g. ezetimibe) TG: ↓5%

Bile acid sequestrants (e.g. Bind bile acids in the intestine, resulting in conversion of cholesterol LDL-C: ↓15–30%; HDL-C:colesevelam, colestipol, to bile acids in the liver. Decreased hepatic cholesterol content ↑3–5%cholestyramine) increases LDL-receptor expression, resulting in decreased serum

LDL-C levels

Nicotinic acid Inhibits lipoprotein synthesis, decreases hepatic synthesis of VLDL LDL-C: ↓5–25%; HDL-C:particles, and inhibits peripheral mobilisation of free fatty acids. Shifts ↑15–35%; TG: ↓20–50%LDL composition from small, dense particles to larger, more buoyantparticles

Apo = apolipoprotein; HDL = high-density lipoprotein; HDL-C = HDL-cholesterol; LDL = low-density lipoprotein; LDL-C = LDL-cholesterol;TG = triglyceride; VLDL = very-low-density lipoprotein.

Fibrates and statins regulate serum lipids by dif- well designed trials. Combination therapy with fe-nofibrate plus either ezetimibe or colesevelam alsoferent mechanisms (table VII) and, as expected,appeared beneficial in patients with primary dys-fenofibrate tended to improve different componentslipidaemia (sections 4.1.5 and 4.1.6).of the lipid profile as compared with statins in

clinical trials (section 4.1.2). Given this, combina- Fenofibrate also demonstrated beneficial effectstion therapy may be a useful option, especially in on various emerging risk factors (e.g. ApoB, ApoAI,patients with mixed dyslipidaemia.[123] Indeed, com- fibrinogen and CRP levels) in patients with dys-bination therapy with fenofibrate plus a statin gener- lipidaemia (sections 2.1 and 2.3). It appears thatally improved lipid levels to a greater extent than lowering lipids does not fully explain the beneficialmonotherapy with either agent alone in patients with effect of fibrates seen in clinical trials (e.g. in theprimary dyslipidaemia, including patients with mix- VA-HIT study, changes in lipid levels accounted fored dyslipidaemia (section 4.1.4). It remains to be only 23% of the beneficial effect of gemfibrozil

therapy).[170] Thus, the nonlipid pleiotropic effectsseen if combination therapy with fenofibrate plus aassociated with PPARα agonism may contribute tostatin will have a positive impact on clinical out-the clinical efficacy of fibrates.[171] Homocysteinecomes in such patients.[123] Fixed combinations suchlevels (another emerging risk factor) were increasedas fenofibrate/simvastatin are currently under devel-with fenofibrate in patients with dyslipidaemia (sec-opment for the treatment of dyslipidaemia. Suchtion 2.3). Interestingly, results from the DAIS trialcombinations have the advantage of reducing thesuggest that the increase in homocysteine levelspill burden, with the potential for improved patientseen with fenofibrate did not attenuate its beneficialadherence. Although combination therapy may beeffect on the progression of focal coronary athero-associated with increased drug acquisition costs, itsclerosis (section 2.3).has been suggested that more effective lowering of

lipid levels may decrease cardiovascular events, ul- As previously mentioned (section 1), the pres-timately resulting in reduced hospitalisation ence of the metabolic syndrome or diabetes in-costs;[148] this hypothesis needs to be confirmed in creases the risk of a CHD event.[1] Although lifestyle


144 Keating & Croom

modification is important in managing both condi- Clinical trials demonstrate that fenofibratetions, pharmacological therapy also has a vital role monotherapy has a beneficial effect on atherogenicto play. In particular, lipid-lowering therapy to ad- dyslipidaemia in patients with the metabolic syn-dress the atherogenic dyslipidaemia associated with drome, reducing TG levels, tending to increaseboth conditions will help reduce the CHD risk.[1] HDL-C levels and promoting a shift to larger, more

buoyant LDL particles (section 4.2.1). FenofibrateMixed results have been seen with statin therapyalso had beneficial effects on prothrombotic andin patients with diabetes in subgroup analyses ofproinflammatory factors in these patients (sectionlarge primary and/or secondary prevention trials.2.3). Moreover, combination therapy with fenofi-Statin therapy significantly reduced the risk of CHDbrate and a statin generally improved the lipid pro-events in some analyses (e.g. 4S,[172,173] CARE,[174]

file to a greater extent than statin therapy aloneHPS[175]), but not in others (e.g. LIPID,[176] AS-(section 4.2.3). Thus, combination therapy is anCOT[177]). The differences in outcomes between tri-attractive option for achieving optimal improvementals can be explained in part by differences betweenof atherogenic dyslipidaemia in patients with thethe patient populations in baseline characteristicsmetabolic syndrome.(e.g. baseline level of CHD risk). In addition, the

residual risk concept has been proposed as a possi- The oral antidiabetic agent metformin is widelyble explanation for these mixed results (i.e. patients used for the treatment of type 2 diabetes.[9] Indeed,with diabetes may still have high TG and low HDL- intensive glucose control with metformin appearedC levels despite statin therapy).[166] Atorvastatin re- to reduce the risk of diabetes-related endpoints induced the risk of CVD events in a primary preven- patients with type 2 diabetes in the UKPDS trial.[185]

tion trial conducted specifically in patients with type Results from a study examining the use of fe-2 diabetes (the CARDS study[178]); however, no

nofibrate plus metformin in patients with the meta-significant reduction in CVD events occurred with

bolic syndrome indicate that patients receiving com-atorvastatin in ASPEN (a trial conducted in patients

bination therapy were more likely to experiencewith type 2 diabetes and LDL-C levels below guide-

normalisation of biochemical abnormalities thanline targets)[179] or in the 4D study (conducted inthose receiving either agent alone (section 4.2.4). Apatients with type 2 diabetes who were receivingfixed combination of fenofibrate/metformin is cur-maintenance haemodialysis).[180]

rently under development for the treatment of meta-The potential of fibrate therapy in patients with bolic disorders; it will be available for twice daily

the metabolic syndrome or diabetes was shown in (fenofibrate/metformin 80mg/500mg, 80mg/850mgsubgroup analyses of the primary prevention HHS or 80mg/1000mg) or three times daily (54mg/study[181] and the secondary prevention BIP[182,183]

850mg) administration.[23] Such a formulation mayand VA-HIT[184] trials. Eighteen-year follow-up data be of particular benefit in patients with the metabol-from the HHS trial revealed that patients with risk ic syndrome, given that these patients often requirefactors for the metabolic syndrome (e.g. high body multiple agents to treat different aspects of theirmass index, high TG levels and low HDL-C levels) condition (e.g. lipid-lowering agents, antihyperten-received the most benefit from early gemfibrozil sives, antidiabetic agents),[9] as well as in patientstherapy.[181] In the BIP trial, bezafibrate reduced the with type 2 diabetes.incidence of MI in patients with the metabolic syn-

According to the NCEP guidelines, type 2 diabe-drome[183] and attenuated the progression of insulintes constitutes a CHD risk equivalent (i.e. type 2resistance.[182] In the VA-HIT trial in patients withdiabetes carries a risk for new CHD that is equal toCHD and low HDL-C levels, gemfibrozil reducedthe risk of recurrent CHD in patients with estab-recurrent events to a greater extent in those withlished CHD), and levels of prothrombotic and proin-diabetes or insulin resistance than in those with-

out.[184] flammatory factors tend to be higher in patients with



diabetes than in nondiabetics, contributing to CHD known.[143] It seems that the beneficial effect cannotrisk.[1] be explained by changes in glycaemic control (as

HbA1c was essentially unaltered during the study),The placebo-controlled FIELD trial is the largestor by changes in concomitant drug use or BP (only aclinical outcomes trial to be completed to date insmall reduction in systolic and diastolic BP ofpatients with type 2 diabetes (table VIII). It ex-≈5mm Hg was seen with fenofibrate, compared withamined the use of fenofibrate in patients with type 2≈3mm Hg with placebo).[143] The pleiotropic effectsdiabetes who did not have a clear indication forof fenofibrate (e.g. its beneficial effects on inflam-lipid-lowering therapy. This study was predomi-mation and endothelial dysfunction in patients withnantly a primary prevention study, as >75% of pa-type 2 diabetes) [section 2.3] may be one possibletients had no history of CVD. Fenofibrate did notexplanation for the microvascular benefits seen inreduce the incidence of CHD mortality or total CHDthe FIELD trial. Further studies are needed to con-events, although a reduction was seen in the inci-firm these initial findings.dence of non-fatal MI and total CVD events (section

4.3.2). As discussed previously, the fact that more Overall, it appears that statins remain the lipid-placebo than fenofibrate recipients received con- lowering treatment of choice for reducing CHDcomitant statin therapy may have masked part of the events in patients with type 2 diabetes.[186] However,beneficial effect of fenofibrate. Further subgroup as previously mentioned, even with statin therapy,analyses revealed greater reductions in total CVD patients with type 2 diabetes are likely to haveevents and CHD events in patients with no previous residual CHD risk. Thus, the concomitant use ofCVD than in patients with CVD, suggesting that fenofibrate in patients with type 2 diabetes mayfenofibrate may be of particular benefit in patients provide additional benefits. Indeed, it has been sug-with early type 2 diabetes without CVD. gested that the main role for fibrates in patients with

diabetes may be in combination with statins.[5,143] AInterestingly, fenofibrate had beneficial effectstrial comparing the effect of fenofibrate plus simvas-on the progression of albuminuria and the need fortatin with simvastatin alone on CVD event rates inretinal laser surgery in the FIELD study (sectionpatients with type 2 diabetes (the ACCORD trial) is4.3.3); this is the first major study to show a benefi-currently underway.[187] This randomised, double-cial effect of lipid-lowering therapy on microvascu-blind, multicentre trial enrolled ≈10 000 patientslar outcomes. The mechanism by which fenofibratewith type 2 diabetes and will examine the effects ofimproved these microvascular endpoints is not yetintensive glycaemic control, lipid-lowering therapyand intensive BP control on CVD events.[23] In termsof lipid-lowering effects, ACCORD will examinewhether in the context of good glycaemic control,improving TG and HDL-C levels through the addi-tion of fenofibrate to simvastatin will further reduceCVD event rates. Results are expected in late 2009.

In other studies, fenofibrate monotherapy wasassociated with improvements from baseline in lipidlevels in patients with diabetic dyslipidaemia (sec-tion 4.3). Moreover, combination therapy with fe-nofibrate plus a statin generally improved athero-genic dyslipidaemia to a greater extent thanmonotherapy in such patients (section 4.3).

In terms of current treatment guideline recom-mendations, NCEP guidelines recommend the use

Table VIII. Fenofibrate in type 2 diabetes mellitus: summary ofFIELD trial findings[143]

Main findings

No significant reduction in the risk of total coronary events(primary outcome)

Significant reduction in the risk of total CVD events, primarilythrough the prevention of non-fatal MI and coronaryrevascularisation

Significant reduction in microvascular-associated complications

Generally well tolerated, both as monotherapy and in combinationwith a statin

Potential implications

Potential role in early type 2 diabetes, in patients without CVD

May reduce both total CVD events and the progression ofmicroangiopathy

May provide additional benefit when used in combination with astatinCVD = cardiovascular disease; MI = myocardial infarction.


146 Keating & Croom

of fibrates in patients with CHD who have low LDL- or pancreatitis in the FIELD trial (section 5). Pan-creatitis has been noted previously in fibrate recipi-C levels and atherogenic dyslipidaemia; for use inents,[24] although an increase in the risk of thrombo-combination with statins in patients with elevatedembolism has not been previously reported withLDL-C levels and atherogenic dyslipidaemia; and infenofibrate.[143]patients with very high TG levels (to reduce the risk

of acute pancreatitis).[1] Recent guidelines from the The difference in musculoskeletal tolerability be-American Heart Association/American College of tween fenofibrate and gemfibrozil when given inCardiology also support the use of fibrates to reduce combination with a statin may be explained by theirlevels of non-HDL-C after LDL-C-lowering ther- differing propensity for pharmacokinetic interac-apy.[188] In terms of treating the atherogenic dys- tion.[190] Fenofibrate did not generally have a clini-lipidaemia component of the metabolic syndrome, cally significant effect on the pharmacokinetics ofNCEP guidelines recommend that the non-HDL-C statins (section 3.2). Even though fenofibrate is as-level become a secondary target of therapy in pa- sociated with mild to moderate inhibition oftients with high TG levels (table II); high-dose sta- CYP2C9, the pharmacokinetics of statinstins or moderate-dose statins plus a fibrate are op- metabolised by this isozyme (e.g. fluvastatin andtions in patients with elevated non-HDL-C levels rosuvastatin) were not altered to a clinically signifi-after the LDL-C target has been reached.[1] Both the cant extent.[105] However, gemfibrozil appears toNCEP[1] and American Diabetes Association[15]

inhibit statin glucuronidation,[190,191] leading to inhi-guidelines recommend fibrates as an option in pa- bition of statin elimination and increased statin con-tients with diabetic dyslipidaemia who have low centrations.[105] Gemfibrozil also inhibitsLDL-C levels and atherogenic dyslipidaemia; com- CYP2C8[192] (cerivastatin is a CYP2C8 substrate, asbination therapy with a fibrate and a statin is an are repaglinide and rosiglitazone).[105] These mecha-option in patients with diabetic dyslipidaemia in nisms may explain the myotoxicity seen with com-whom LDL lowering is required. European guide- bination therapy with gemfibrozil plus a statin.[162]

lines recommend target lipid levels of <2.5 mmol/L Thus, it has been suggested that gemfibrozil befor LDL-C and <4.5 mmol/L for TC in patients with avoided when considering combination therapy withdiabetes.[189] These guidelines recommend that treat- a fibrate plus a statin.[193] Other recommendationsment choice should be guided by HDL-C and TG include only using combination therapy with alevels, as well as by LDL-C levels. fibrate plus a statin if renal function is normal; using

usual, not high, statin dosages; avoiding concomi-Fenofibrate is generally well tolerated in thetant drugs that could increase either fibrate or statintreatment of dyslipidaemia (section 5). Combinationconcentrations; checking baseline CPK and tran-therapy (e.g. with a statin plus a fibrate) is currentlysaminase levels and regularly monitoring CPK, tran-underutilised in patients with dyslipidaemia.[165]

saminase and creatinine levels; and monitoring eld-One reason for this may be concern over the safetyerly patients particularly carefully.[193]

of such treatment, particularly in regard to muscle-related adverse events.[105,165] However, the risk of In terms of the tolerability of other lipid-loweringadverse events such as rhabdomyolysis appears low agents, ezetimibe appears to be generally well toler-in patients receiving fenofibrate plus a statin, espe- ated, with the most common adverse events in place-cially compared with patients receiving gemfibrozil bo-controlled trials including back pain, arthralgia,plus a statin (section 5). Moreover, the FIELD trial diarrhoea, sinusitis and abdominal pain (occurringsuggests that combination therapy with fenofibrate in 3.0–4.1% of ezetimibe recipients).[194] However,and a statin is generally well tolerated in patients the use of nicotinic acid is somewhat limited by thewith type 2 diabetes (section 5). Although inci- occurrence of adverse events such as flushing,dences were low, significantly more fenofibrate than hyperglycaemia, hyperuricaemia, gastrointestinal

symptoms and hepatotoxicity.[1,16] Bile acid seques-placebo recipients developed pulmonary embolism



trants are most commonly associated with gastroin- Disclosuretestinal adverse events such as constipation, abdom-

During the peer review process, the manufacturer of theinal pain, bloating and nausea; these agents seem toagent under review was also offered an opportunity to com-lack systemic toxicity.[1]ment on this article; changes based on any comments received

In conclusion, fenofibrate improves lipid levels were made on the basis of scientific and editorial merit.

(in particular TG and HDL-C levels) in patients withprimary dyslipidaemia. Its lipid-lowering profile

Referencesmeans that fenofibrate is particularly well suited for 1. National Cholesterol Education Program Expert Panel. Thirdreport of the National Cholesterol Education Program (NCEP)use in atherogenic dyslipidaemia, which is common-Expert Panel on detection, evaluation, and treatment of highly seen in patients with the metabolic syndrome and blood cholesterol in adults (Adult Treatment Panel III): finalreport. Circulation 2002 Dec 17; 106 (25): 3143-421type 2 diabetes. Indeed, fenofibrate improves the

2. Fruchart J-C, Duriez P. Mode of action of fibrates in thecomponents of atherogenic dyslipidaemia in pa- regulation of triglycerides and HDL-cholesterol metaboism.

Drugs Today 2006; 42 (1): 39-64tients with these conditions, including a shift from3. Jellinger PS. The American Association of Clinical Endocri-

small, dense LDL particles to larger, more buoyant nologists medical guidelines for clinical practice for the diag-nosis and treatment of dyslipidemia and prevention of ather-LDL particles. Greater improvements in lipid levelsogenesis: 2002 amended version. Endocr Pract 2000 Mar/Apr;

are seen when fenofibrate is administered in combi- 6 (2): 1-524. Grundy SM, Cleeman JI, Merz CNB, et al. Implications ofnation with a statin or ezetimibe, compared with

recent clinical trials for the National Cholesterol Educationmonotherapy with these agents. In the DAIS study, Program Adult Treatment Panel III guidelines. Circulation

2004 Jul 13; 110: 227-39fenofibrate significantly slowed the angiographic5. Bloomfield HE. The role of fibrates in a statin world. Arch

progression of focal coronary atherosclerosis in pa- Intern Med 2006 Apr 10; 166: 715-66. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolictients with type 2 diabetes. In terms of clinical

syndrome among US adults: findings from the third Nationaloutcomes, although no significant reduction in the Health and Nutrition Examination Survey. JAMA 2002 Jan 16;

287 (3): 356-9risk of coronary events was seen with fenofibrate in7. Ford ES, Giles WH, Mokdad AH, et al. Increasing prevalence of

the FIELD trial in patients with type 2 diabetes, the metabolic syndrome among US adults. Diabetes Care 2004Oct; 27 (10): 2444-249treatment was associated with a significantly re-

8. Grundy SM, Brewer HB, Cleeman JI, et al. Definition of meta-duced risk of total CVD events, primarily through bolic syndrome: report of the National Heart, Lung, and Blood

Institute/American Heart Association Conference on Scientificthe prevention of non-fatal MI and coronary revas-Issues Related to Definition. Circulation 2004 Jan 27; 109:

cularisation. Subgroup analyses revealed significant 433-89. Grundy SM. Drug therapy of the metabolic syndrome: minimiz-reductions in total CVD events and CHD events in

ing the emerging crisis in polypharmacy. Nat Rev Drug Discovpatients with no previous CVD, suggesting a poten- 2006 Apr; 5 (4): 295-309

10. Wild S, Roglic G, Green A, et al. Global prevalence of diabetes:tial role for primary prevention with fenofibrate inestimates for the year 2000 and projections for 2030. Diabetes

patients with early type 2 diabetes. Improvements Care 2004 May; 27 (5): 1047-5311. Beckman JA, Creager MA, Libby P. Diabetes and atherosclero-were also seen in microvascular outcomes with fe-

sis: epidemiology, pathophysiology, and management. JAMAnofibrate in the FIELD trial. Fenofibrate is generally 2002 May 15; 287 (19): 2570-81

12. Nesto RW. Beyond low-density lipoprotein: addressing thewell tolerated, both as monotherapy and when ad-atherogenic lipid triad in type 2 diabetes mellitus and the

ministered in combination with a statin. Combina- metabolic syndrome. Am J Cardiovasc Drugs 2005; 5 (6): 379-87tion therapy with fenofibrate plus a statin appears to

13. National Institute of Diabetes and Digestive and Kidney Diseas-be associated with a low risk of rhabdomyolysis; no es. National Diabetes Statistics [online]. Available from URL:

http://diabetes.niddk.nih.gov/dm/pubs/statistics [Accessedcases of rhabdomyolysis were reported in patients2006 May 12]

receiving such therapy in the FIELD trial. Thus, 14. Donnelly R, Emslie-Smith AM, Gardner ID, et al. Vascularcomplications of diabetes. Br Med J 2000 Apr 15; 320: 1062-6fenofibrate is a valuable lipid-lowering agent, par-

15. American Diabetes Association. Standards of medical care indiabetes. Diabetes Care 2005 Jan; 28 Suppl. 1: S4-36ticularly in patients with atherogenic dyslipidaemia.


148 Keating & Croom

16. Khera A, McGuire DK. Management of diabetic dyslipidemia: normalizes the atherogenic, dense LDL profile in combinedneed for reappraisal of the goals. Am J Cardiovasc Drugs hyperlipidemia. Arterioscler Thromb Vasc Biol 1996 Jun; 162005; 5 (2): 83-91 (6): 763-72

17. Munoz A, Guichard JP, Reginault P. Micronised fenofibrate. 35. Shepherd J, Caslake MJ, Lorimer AR, et al. Fenofibrate reducesAtherosclerosis 1994 Oct; 110 Suppl.: S45-8 low density lipoprotein catabolism in hypertriglyceridemic

subjects. Arteriosclerosis 1985 Mar/Apr; 5 (2): 162-818. Sauron R, Wilkins M, Jessent V, et al. Absence of a food effectwith a 145 mg nanoparticle fenofibrate tablet formulation. Int J 36. Caslake MJ, Packard CJ, Gaw A, et al. Fenofibrate and LDLClin Pharmacol Ther 2006 Feb; 44 (2): 64-70 metabolic heterogeneity in hypercholesterolemia. Arterioscler

Thromb 1993 May; 13 (5): 702-1119. Fournier Pharmaceuticals Ltd. Lipantil Micro 67: summary ofproduct characteristics [online]. Available from URL: http:// 37. Ikewaki K, Tohyama J, Nakata Y, et al. Fenofibrate effectivelyemc.medicines.org.uk [Accessed 2006 Apr 4] reduces remnants, and small dense LDL, and increases HDL

particle number in hypertriglyceridemic men: a nuclear mag-20. Fournier Pharmaceuticals Ltd. Lipantil Micro 200: summarynetic resonance study. J Atheroscler Thromb 2004; 11 (5):of product characteristics [online]. Available from URL: http://278-85emc.medicines.org.uk [Accessed 2006 Mar 31]

38. Winkler K, Weltzien P, Friedrich I, et al. Qualitative effect of21. Fournier Pharmaceuticals Ltd. Lipantil Micro 267: summaryfenofibrate and quantitative effect of atorvastatin on LDLof product characerteristics [online]. Available from URL:profile in combined hyperlipidaemia with dense LDL. Exphttp://emc.medicines.org.uk [Accessed 2006 Apr 6]Clin Endocrinol Diabetes 2004; 112: 241-722. Fournier Pharmaceuticals Ltd. Supralip 160mg: summary of

39. Berthou L, Duverger N, Emmanuel F, et al. Opposite regulationproduct characteristics [online]. Available from URL: http://of human versus mouse apolipoprotein A-I by fibrates inemc.medicines.org.uk [Accessed 2006 Apr 6]human apolipoprotein A-I transgenic mice. J Clin Invest 1996;23. Data on file, Fournier Pharmaceuticals Ltd, 200697 (11): 2408-1624. Abbott Laboratories. Tricor 48mg and 145mg (fenofibrate

40. Vu-Dac N, Schoonjans K, Kosykh V, et al. Fibrates increasetablets): prescribing information [online]. Available fromhuman apolipoprotein A-II expression through activation ofURL: http://tricortablets.com/ [Accessed 2006 Apr 3]the peroxisome proliferator-activated receptor. J Clin Invest25. Schoonjans K, Martin G, Staels B, et al. Peroxisome prolifer-1995; 96: 741-50ator-activated receptors, orphans with ligands and functions.

41. Ji J, Barrett PHR, Johnson AG, et al. Lipid transfer proteins andCurr Opin Lipidol 1997 Jun; 8 (3): 159-66apolipoprotein B-100 metabolism in the metabolic syndrome26. Arakawa R, Tamehiro N, Nishimaki-Mogami T, et al. Fe-treated with fenofibrate [abstract no. We-P11:111]. Athero-nofibric acid, an active form of fenofibrate, increases apoli-sclerosis 2006 Jun; 7 (3 Suppl.): 370poprotein A-I-mediated high-density lipoprotein biogenesis by

42. Kiyanagi T, Miyazaki T, Kume A, et al. Decrease in CETPenhancing transcription of ATP-binding cassette transporteractivity by fenofibrate may increase LDL particle size mea-A1 gene in a liver X receptor-dependent manner. Arteriosclersured by HPLC method in patients with coronary artery dis-Thromb Vasc Biol 2005 Jun; 25 (6): 1193-7ease [abstract no. Th-P16:301]. Atherosclerosis 2006 Jun; 7 (327. Duez H, Lefebvre B, Poulain P, et al. Regulation of humanSuppl.): 559apoA-I by gemfibrozil and fenofibrate through selective per-

43. McPherson R, Agnani G, Lau P, et al. Role of Lp A-I and Lp A-oxisome proliferator-activated receptor α modulation. Arteri-I/A-II in cholesteryl ester transfer protein-mediated neutraloscler Thromb Vasc Biol 2005 Mar; 25 (3): 585-91lipid transfer: studies in normal subjects and in hypertriglycer-28. Chapman MJ. Fibrates: therapeutic review. Br J Diabetes Vascidemic patients before and after fenofibrate therapy. Arterios-Dis 2006 Jan-Feb; 6 (1): 11-18, 20cler Thromb Vasc Biol 1996 Nov; 16 (11): 1340-629. Schoonjans K, Peinado-Onsurba J, Lefebvre A-M, et al. PPARα

44. Bertolini S, Elicio N, Daga A, et al. Effect of a single daily doseand PPARγ activators direct a distinct tissue-specific transcrip-treatment of fenofibrate on plasma lipoproteins in hyperli-tional response via a PPRE in the lipoprotein lipase gene.poproteinaemia IIb. Eur J Clin Pharmacol 1988; 34: 25-8EMBO J 1996; 15 (19): 5336-48

45. Elisaf M, Tsimichodimos V, Bairaktari E, et al. Effect of mi-30. Haubenwallner S, Essenburg AD, Barnett BC, et al. Hy-cronized fenofibrate and losartan combination on uric acidpolipidemic activity of select fibrates correlates to changes inmetabolism in hypertensive patients with hyperuricemia. Jhepatic apolipoprotein C-III expression: a potential physiolog-Cardiovasc Pharmacol 1999 Jul; 34 (1): 60-3ic basis for their mode of action. J Lipid Res 1995 Dec; 36

(12): 2541-51 46. Genest Jr J, Nguyen N-H, Theroux P, et al. Effect of micronizedfenofibrate on plasma lipoprotein levels and hemostatic pa-31. Staels B, Vu-Dac N, Kosykh VA, et al. Fibrates downregulaterameters of hypertriglyceridemic patients with low levels ofapolipoprotein C-III expression independent of induction ofhigh-density lipoprotein cholesterol in the fed and fasted state.peroxisomal acyl coenzyme A oxidase: a potential mechanismJ Cardiovasc Pharmacol 2000 Jan; 35 (1): 164-72for the hypolipidemic action of fibrates. J Clin Invest 1995

Aug; 95: 705-12 47. Haak T, Haak E, Kusterer K, et al. Fenofibrate improvesmicrocirculation in patients with hyperlipidemia. Eur J Med32. Schoonjans K, Staels B, Auwerx J. The peroxisome proliferatorRes 1998 Feb 21; 3 (1-2): 50-4activated receptors (PPARs) and their effects on lipid metabo-

lism and adipocyte differentiation. Biochim Biophys Acta 48. Insua A, Massari F, Rodriguez Moncalvo JJ, et al. Fenofibrate1996; 1302: 93-109 or gemfibrozil for treatment of types IIa and IIb primary

hyperlipoproteinemia: a randomized, double-blind, crossover33. Schoonjans K, Staels B, Auwerx J. Role of the peroxisomestudy. Endocr Pract 2002 Mar/Apr; 8 (2): 96-101proliferator-activated receptor (PPAR) in mediating the effects

of fibrates and fatty acids on gene expression. J Lipid Res 49. Kiortsis DN, Millionis H, Bairaktari E, et al. Efficacy of combi-1996; 37: 907-25 nation of atorvastatin and micronised fenofibrate in the treat-

34. Guerin M, Bruckert E, Dolphin PJ, et al. Fenofibrate reduces ment of severe mixed hyperlipidemia. Eur J Clin Pharmacolplasma cholesteryl ester transfer from HDL to VLDL and 2000 Dec; 56 (9-10): 631-5



50. Ko HS, Kim CJ, Ryu WS. Effect of fenofibrate on lipoprotein(a) 65. Athyros VG, Papageorgiou AA, Athyrou VV, et al. Atorvastatinin hypertriglyceridemic patients: impact of change in triglycer- and micronized fenofibrate alone and in combination in type 2ide level and liver function. J Cardiovasc Pharmacol 2005 Oct; diabetes with combined hyperlipidemia. Diabetes Care 200246 (4): 405-11 Jul; 25 (7): 1198-202

66. Cavallero E, Dachet C, Assadolahi F, et al. Micronized fe-51. Koh KK, Ahn JY, Han SH, et al. Effects of fenofibrate onnofibrate normalizes the enhanced lipidemic response to a fatlipoproteins, vasomotor function, and serological markers ofload in patients with type 2 diabetes and optimal glucoseinflammation, plaque stabilization, and hemostasis. Athero-control. Atherosclerosis 2003 Jan; 166 (1): 151-61sclerosis 2004 Jun; 174 (2): 379-83

67. Feher MD, Caslake M, Foxton J, et al. Atherogenic lipoprotein52. Koh KK, Quon MJ, Han SH, et al. Additive beneficial effects ofphenotype in type 2 diabetes: reversal with micronised fe-fenofibrate combined with atorvastatin in the treatment ofnofibrate. Diabetes Metab Res Rev 1999 Nov-Dec 31; 15 (6):combined hyperlipidemia. J Am Coll Cardiol 2005 May 17; 45395-9(10): 1649-53

68. Playford DA, Watts GF, Best JD, et al. Effect of fenofibrate on53. Koh KK, Han SH, Quon MJ, et al. Beneficial effects of fe-brachial artery flow-mediated dilatation in type 2 diabetesnofibrate to improve endothelial dysfunction and raise adi-mellitus. Am J Cardiol 2002 Dec 1; 90: 1254-7ponectin levels in patients with primary hypertriglyceridemia.

69. Forcheron F, Cachefo A, Thevenon S, et al. Mechanisms of theDiabetes Care 2005 Jun; 28 (6): 1419-24triglyceride- and cholesterol-lowering effect of fenofibrate in54. Koh KK, Quon MJ, Han SH, et al. Additive beneficial effects ofhyperlipidemic type 2 diabetic patients. Diabetes 2002 Dec;fenofibrate combined with candesartan in the treatment of51: 3486-91hypertriglyceridemic hypertensive patients. Diabetes Care

70. Chinetti G, Gbaguidi FG, Griglio S, et al. CLA-1/SR-BI is2006 Feb; 29 (2): 195-201expressed in atherosclerotic lesion macrophages and regulated55. Krempf M, Rohmer V, Farnier M, et al. Efficacy and safety ofby activators of peroxisome proliferator-activated receptors.micronised fenofibrate in a randomised double-blind studyCirculation 2000 May 23; 101 (20): 2411-7comparing four doses from 200 mg to 400 mg daily with

71. Zambon A, Gervois P, Pauletto P, et al. Modulation of hepaticplacebo in patients with hypercholesterolemia. Diabetes Metabinflammatory risk markers of cardiovascular diseases by2000 May; 26 (3): 184-91PPAR-α activators: clinical and experimental evidence. Arter-

56. Lemieux I, Salomon H, Despres J-P. Contribution of apo CIIIioscler Throm Vasc Biol 2006 May; 26 (5): 977-86

reduction to the greater effect of 12-week micronized fe-72. Milionis HJ, Papakostas J, Kakafika A, et al. Comparativenofibrate than atorvastatin therapy on triglyceride levels and

effects of atorvastatin, simvastatin, and fenofibrate on serumLDL size in dyslipidemic patients. Ann Med 2003; 35 (6): 442-homocysteine levels in patients with primary hyperlipidaemia.8J Clin Pharmacol 2003 Aug; 43 (8): 825-30

57. Malik J, Melenovsky V, Wichterle D, et al. Both fenofibrate and 73. Wang T-D, Chen W-J, Lin J-W, et al. Efficacy of fenofibrateatorvastatin improve vascular reactivity in combined hyper- and simvastatin on endothelial function and inflammatorylipidaemia (fenofibrate versus atorvastatin trial - FAT). Cardi- markers in patients with combined hyperlipidemia: relationsovasc Res 2001 Nov; 52 (2): 290-8 with baseline lipid profiles. Atherosclerosis 2003 Oct; 170 (2):

58. Paragh G, Seres I, Harangi M, et al. The effect of micronised 315-23fenofibrate on paraoxonase activity in patients with coronary 74. Yesilbursa D, Serdar A, Saltan Y, et al. The effect of fenofibrateheart disease. Diabetes Metab 2003 Dec; 29 (6): 613-8 on serum paraoxonase activity and inflammatory markers in

59. Raslova K, Dubovska D, Mongiellova V, et al. Relationship patients with combined hyperlipidemia. Kardiol Pol 2005 Jun;between plasma fenofibric acid levels and the effect of micron- 62 (6): 526-30ized fenofibrate on cholesterol, low-density-lipoprotein cho- 75. Bissonnette R, Treacy E, Rozen R, et al. Fenofibrate raiseslesterol and apolipoprotein B in patients with primary plasma homocysteine levels in the fasted and fed states. Ath-hypercholesterolemia. Eur J Clin Pharmacol 1997 Apr; 52: erosclerosis 2001 Apr; 155 (2): 455-62101-6 76. Capell WH, DeSouza CA, Poirier P, et al. Short-term triglycer-

60. Sasaki J, Yamamoto K, Ageta M. Effects of fenofibrate on high- ide lowering with fenofibrate improves vasodilator function indensity lipoprotein particle size in patients with hyper- subjects with hypertriglyceridemia. Arterioscler Thromb Vasclipidemia: a randomized, double-blind, placebo-controlled, Biol 2003 Feb; 23: 307-13multicenter, crossover study. Clin Ther 2002 Oct; 24 (10): 77. Coban E, Ozdogan M, Yazicioglu G, et al. The effect of fe-1614-26 nofibrate on the levels of high sensitivity C-reactive protein in

61. Tsimihodimos V, Kostoula A, Kakafika A, et al. Effect of dyslipidaemic hypertensive patients. Int J Clin Pract 2005 Apr;fenofibrate on serum inflammatory markers in patients with 59 (4): 415-8high triglyceride values. J Cardiovasc Pharmacol Ther 2004 78. Dierkes J, Westphal S, Luley C. Serum homocysteine increasesMar; 9 (1): 27-33 after therapy with fenofibrate or bezafibrate [letter]. Lancet

62. Tsimihodimos V, Tambaki A, Tzovaras V, et al. Comparison of 1999 Jul 17; 354 (9174): 219-20the effects of atorvastatin and fenofibrate on apolipoprotein B- 79. de Lorgeril M, Salen P, Paillard F, et al. Lipid-lowering drugscontaining lipoprotein subfractions in patients with combined and homocysteine [letter]. Lancet 1999 Jan 16; 353 (9148):dyslipidemia. Hellenic J Cardiol 2004; 45 (4): 225-30 209-10

63. Watts GF, Barrett PHR, Ji J, et al. Differential regulation of 80. Ellen RLB, McPherson R. Long-term efficacy and safety oflipoprotein kinetics by atorvastatin and fenofibrate in subjects fenofibrate and a statin in the treatment of combined hyper-with the metabolic syndrome. Diabetes 2003 Mar; 52: 803-11 lipidemia. Am J Cardiol 1998 Feb 26; 81 (4A): 60-5B

64. Watts GF, Ji J, Chan DC, et al. Relationships between changes 81. Kowalski J, Okopien B, Madej A, et al. Effects of atorvastatin,in plasma lipid transfer proteins and apolipoprotein B-100 simvastatin, and fenofibrate therapy on monocyte chemoat-kinetics during fenofibrate treatment in the metabolic syn- tractant protein-1 secretion in patients with hyperlipidemia.drome. Clin Sci (Lond) 2006 May 16; 111 (3): 193-9 Eur J Clin Pharmacol 2003 Jul; 59 (3): 189-93


150 Keating & Croom

82. Kowalski J, Okopien B, Madej A, et al. Effects of fenofibrate and hemostasis in patients with impaired glucose tolerance. Jand simvastatin on plasma sICAM-1 and MCP-1 concentra- Clin Endocrinol Metab 2006 May; 91 (5): 1770-8tions in patients with hyperlipoproteinemia. Int J Clin 100. Genest J, Frohlich J, Steiner G. Effect of fenofibrate-mediatedPharmacol Ther 2003 Jun; 41 (6): 241-7 increase in plasma homocysteine on the progression of corona-

83. Mackness MI, Phuntuwate W, Suthisisang C, et al. Effect of ry artery disease in type 2 diabetes mellitus. Am J Cardiol 2004fenofibrate treatment on paraoxonase 1 [abstract no. A68]. Apr 1; 93 (7): 848-53Diabet Med 2006 Mar 1; 23 Suppl. 2: 18-9 101. Rosenson RS, Huskin AL, Wolff DA. Adiponectin and the anti-

84. Noguchi Y, Tatsuno I, Suyama K, et al. Effect of fenofibrate on inflammatory effects of fenofibrate in patients with hypertrig-uric acid metabolism in Japanese hyperlipidemic patients. J lyceridemia and the metabolic syndrome [abstract no. Tu-Atheroscler Thromb 2004; 11 (6): 335-40 W27:5]. Atherosclerosis 2006 Jun; 7 (3 Suppl.): 175

85. Okopien B, Cwalina L, Lebek M, et al. Effects of fibrates on 102. Diabetes Atherosclerosis Intervention Study Investigators. Ef-plasma prothrombotic activity in patients with type IIb dys- fect of fenofibrate on progression of coronary-artery disease inlipidemia. Int J Clin Pharmacol Ther 2001 Dec; 39 (12): 551-7 type 2 diabetes: the Diabetes Atherosclerosis Intervention

Study, a randomised study. Lancet 2001 Mar 24; 357 (9260):86. Okopien B, Krysiak R, Haberka M, et al. Effect of monthly905-10atorvastatin and fenofibrate treatment on monocyte chemoat-

tractant protein-1 release in patients with primary mixed dys- 103. Bergman AJ, Murphy G, Burke J, et al. Simvastatin does notlipidemia. J Cardiovasc Pharmacol 2005 Apr; 45 (4): 314-20 have a clinically significant pharmacokinetic interaction with

fenofibrate in humans. J Clin Pharmacol 2004 Sep; 44 (9):87. Okopien B, Kowalski J, Krysiak R, et al. Monocyte suppressing1054-62action of fenofibrate. Pharmacol Rep 2005; 57 (3): 367-72

104. Martin PD, Dane AL, Schneck DW, et al. An open-label,88. Pfutzner A, Ambrosch A, Forst T, et al. The influence ofrandomized, three-way crossover trial of the effects of coad-micronized fenofibrate on cardiovascular risk factors [in Ger-ministration of rosuvastatin and fenofibrate on the pharma-man]. Z Allg Med 1994; 70: 510-5cokinetic properties of rosuvastatin and fenofibric acid in89. Ramjattan BR, Callaghan DJG, Theiss U. Efficacy and tolera-healthy male volunteers. Clin Ther 2003 Feb; 25 (2): 459-71bility of a ‘suprabioavailable’ formulation of fenofibrate in

105. Davidson MH. Statin/fibrate combination in patients with meta-patients with dyslipidemia: a pooled analysis of two open-labelbolic syndrome or diabetes: evaluating the risks of pharma-trials. Clin Ther 2002 Jul; 24 (7): 1105-16cokinetic drug interactions. Expert Opin Drug Saf 2006; 5 (1):90. Saklamaz A, Comlekci A, Temiz A, et al. The beneficial effects145-56of lipid-lowering drugs beyond lipid-lowering effects: a com-

106. Whitfield LR, Abel R, Hartman D, et al. Effects of co-adminis-parative study with pravastatin, atorvastatin, and fenofibrate intration of gemfibrozil and fenofibrate on the pharmacokineticpatients with type IIa and type IIb hyperlipidemia. Metabolismprofile of atorvastatin and major metabolites [abstract no. 547-2005 May; 54 (5): 677-81P]. Diabetes 2005 Jun; 54 Suppl. 1: A13591. Staels B, Koenig W, Habib A, et al. Activation of human aortic

107. Gustavson LE, Schweitzer SM, Koehne-Voss S, et al. Thesmooth-muscle cells is inhibited by PPARα but not by PPARγeffects of multiple doses of fenofibrate on the pharmacokinet-activators. Nature 1998 Jun 25; 393 (6687): 790-3ics of pravastatin and its 3α-hydroxy isomeric metabolite. J92. Undas A, Celinska-Lowenhoff M, Domagala TB, et al. EarlyClin Pharmacol 2005 Aug; 45 (8): 947-53antithrombotic and anti-inflammatory effects of simvastatin

108. Pan W-J, Gustavson LE, Achari R, et al. Lack of a clinicallyversus fenofibrate in patients with hypercholesterolemia.significant pharmacokinetic interaction between fenofibrateThromb Haemost 2005 Jul 1; 94 (1): 193-9and pravastatin in healthy volunteers. J Clin Pharmacol 2000;93. Westphal S, Dierkes J, Luley C. Effects of fenofibrate and40: 316-23gemfibrozil on plasma homocysteine [letter]. Lancet 2001 Jul

109. Gustavson LE, Schweitzer SM, Burt DA, et al. Evaluation of the7; 358 (9275): 39-40potential for pharmacokinetic interaction between fenofibrate94. Yang T-L, Chen M-F, Xia X, et al. Effect of fenofibrate on theand ezetimibe: a phase I, open-label, multiple-dose, threelevel of asymmetric dimethylarginine in individuals withperiod crossover study in healthy subjects. Clin Ther 2006hypertriglyceridemia. Eur J Clin Pharmacol 2006; 62: 179-84Mar; 28 (3): 373-8795. Ye P, Li J-J, Su G, et al. Effects of fenofibrate on inflammatory

110. Kosoglou T, Statkevich P, Fruchart J-C, et al. Pharmacodynam-cytokines and blood pressure in patients with hypertriglycer-ic and pharmacokinetic interaction between fenofibrate andidemia [letter]. Clin Chim Acta 2005 Jun; 356 (1): 229-32ezetimibe. Curr Med Res Opin 2004 Aug; 20 (8): 1197-20796. Belfort RS, Berria R, DeFronzo R, et al. Fenofibrate improves

111. Jones MR, Baker BA, Mathew P. Effect of colesevelam HCl onthe atherogenic lipid profile and markers of vascular inflam-single-dose fenofibrate pharmacokinetics. Clin Pharmacokinetmation independent of changes in insulin sensitivity in hyper-2004; 43 (13): 943-50triglyceridemic subjects with the metabolic syndrome [abstract

112. Kajosaari LI, Backman JT, Neuvonen M, et al. Lack of effect ofno. 620]. Diabetologia 2004 Aug; 47 Suppl. 1: A225-6bezafibrate and fenofibrate on the pharmacokinetics and phar-97. Idzior-Walus B, Sieradzki J, Rostworowski W, et al. Effects ofmacodynamics of repaglinide. Br J Clin Pharmacol 2004 Oct;comicronised fenofibrate on lipid and insulin sensitivity in58 (4): 390-6patients with polymetabolic syndrome X. Eur J Clin Invest

113. Boissonnat P, Salen P, Guidollet J, et al. The long-term effects2000 Oct; 30 (10): 871-8of the lipid-lowering agent fenofibrate in hyperlipidemic heart98. Rosenson RS, Huskin AL, Wolff DA. Fenofibrate reduces post-transplant recipients. Transplantation 1994 Jul; 58 (2): 245-7prandial cytokine production in metabolic syndrome patients

114. Julius U, Schwartz T. Micronized fenofibrate in lipometabolism[abstract no. 3789]. Circulation 2005 Oct 25; 112 (17 Suppl.):disorders [in German]. Therapiewoche 1994; 44 (25): 1442-5818. Plus poster presented at the 78th Scientific Sessions of the

American Heart Association; 2005 Nov 13-16; Dallas (TX) 115. Kornitzer M, Dramaix M, Vandenbroek MD. Efficacy and99. Okopien B, Krysiak R, Herman ZS. Effects of short-term fe- tolerance of 200 mg micronised fenofibrate administered over

nofibrate treatment on circulating markers of inflammation a 6-month period in hyperlipidaemic patients: an open Belgian



multicenter study. Atherosclerosis 1994 Oct; 110 Suppl.: S49- fluvastatin + fibrate (bezafibrate, fenofibrate, or gemfibrozil).54 Am J Cardiol 2003 Oct 1; 92 (7): 794-7

116. Marz W, Schmitz H, Bach G, et al. Micronised fenofibrate: 132. Stefanutti C, Bucci A, Di Giacomo S, et al. Efficacy, safety andefficacy and tolerability in the treatment of isolated and com- tolerability of combined low-dose simvastatin-fenofibratebined dyslipidemia [in German]. Munch Med Wochenschr treatment in primary mixed hyperlipidaemia. Clin Drug Invest1994 Nov 11; 136 (45): 690-4 2004; 24 (8): 465-77

117. Kirchgassler KU, Schmitz H, Bach G. Effectiveness and tolera- 133. Liamis G, Kakafika A, Bairaktari E, et al. Combined treatmentbility of 12-week treatment with micronised fenofibrate with fibrates and small doses of atorvastatin in patients with200mg in a drug-monitoring programme involving 9884 pa- mixed hyperlipidemia. Curr Med Res Opin 2002; 18 (3): 125-8tients with dyslipidaemia. Clin Drug Invest 1998 Mar; 15 (3): 134. McKenney JM, Farnier M, Lo K-W, et al. Safety and efficacy of197-204 long-term co-administration of fenofibrate and ezetimibe in

118. Zhu J, Ye P. A survey on the efficacy and tolerability of patients with mixed hyperlipidemia. J Am Coll Cardiol 2006micronized fenofibrate in patients with dyslipidemia. Chin Apr 18; 47 (8): 1584-7Med J 2003 Jun; 116 (6): 840-3 135. Wysocki J, Belowski D, Kalina M, et al. Effects of micronized

119. Ducobu J, VanHaelst L, Salomon H. Comparison of micronized fenofibrate on insulin resistance in patients with metabolicfenofibrate and pravastatin in patients with primary hyper- syndrome. Int J Clin Pharmacol Ther 2004 Apr; 42 (4): 212-7lipidemia. J Cardiovasc Pharmacol 2003 Jan; 41 (1): 60-7 136. Nieuwdorp M, Stroes ESG, Kastelein JJP, on behalf of the

120. Farnier M, Bonnefous F, Debbas N, et al. Comparative efficacy Fenofibrate/Metformin Study Group. Normalisation of meta-and safety of micronized fenofibrate and simvastatin in pa- bolic syndrome using fenofibrate, metformin or their combina-tients with primary type IIa or IIb hyperlipidemia. Arch Intern tion. Diabetes Obes Metab. In pressMed 1994 Feb 28; 154: 441-9 137. Vega GL, Ma PTS, Cater NB, et al. Effects of adding fenofibrate

121. Steinmetz A, Schwartz T, Hehnke U, et al. Multicenter compari- (200 mg/day) to simvastatin (10 mg/day) in patients withson of micronized fenofibrate and simvastatin in patients with combined hyperlipidemia and metabolic syndrome. Am Jprimary type IIA or IIB hyperlipoproteinemia. J Cardiovasc Cardiol 2003 Apr 15; 91: 956-60Pharmacol 1996; 27 (4): 563-70 138. Rosenson RS, Huskin AL, Wolfe DA. Fenofibrate markedly

122. Farnier M, Dejager S, and the French Fluvastatin Study Group. reduces postprandial dyslipidemia in metabolic syndrome pa-Effect of combined fluvastatin-fenofibrate therapy compared tients [abstract no. 979-P]. Diabetes 2005 Jun; 54 Suppl. 1:with fenofibrate monotherapy in severe primary hypercholes- A239terolemia. Am J Cardiol 2000 Jan 1; 85 (1): 53-7 139. Filippatos TD, Kiortsis DN, Liberopoulos EN, et al. Effect of

123. Grundy SM, Vega GL, Yuan Z, et al. Effectiveness and tolera- orlistat, micronised fenofibrate and their combination on meta-bility of simvastatin plus fenofibrate for combined hyper- bolic parameters in overweight and obese patients with thelipidemia (the SAFARI Trial). Am J Cardiol 2005 Feb 15; 95 metabolic syndrome: the FenOrli study. Curr Med Res Opin(4): 462-8 2005 Dec; 21 (12): 1997-2006

124. Farnier M, Freeman MW, Macdonell G, et al. Efficacy and 140. McIvor ME. In combination with statins, fenofibrate, niacin andsafety of the coadministration of ezetimibe with fenofibrate in rosiglitazone effectively treat the atherogenic dyslipidemia ofpatients with mixed hyperlipidaemia. Eur Heart J 2005 May; the metabolic syndrome [abstract no. 1084]. Circulation 200526 (9): 897-905 Oct 25; 112 (17 Suppl.): 209. Plus poster presented at the 78th

125. McKenney J, Jones M, Abby S. Safety and efficacy of coleseve- Scientific Sessions of the American Heart Association; 2005lam hydrochloride in combination with fenofibrate for the Nov 13-16; Dallas (TX)treatment of mixed hyperlipidemia. Curr Med Res Opin 2005 141. Filippatos TD, Gazi IF, Liberopoulos EN, et al. The effect ofSep; 21 (9): 1403-12 orlistat and fenofibrate, alone or in combination, on small

126. Despres J-P, Lemieux I, Salomon H, et al. Effects of micronized dense LDL and lipoprotein-associated phospholipase A2 infenofibrate versus atorvastatin in the treatment of dys- obese patients with metabolic syndrome. Atherosclerosis.lipidaemic patients with low plasma HDL-cholesterol levels: a Epub 2006 Aug 1212-week randomized trial. J Intern Med 2002 Jun; 251 (6): 142. Tan CE, Chew LS, Tai ES, et al. Benefits of micronised fe-490-9 nofibrate in type 2 diabetes mellitus subjects with good

127. Roth E, Farnier M, Gil-Extremera B, et al. Coadministration of glycemic control. Atherosclerosis 2001 Feb 1; 154 (2): 469-74ezetimibe/simvastatin and fenofibrate is an efficacious and 143. The FIELD study investigators. Effects of long-term fenofibratewell-tolerated treatment for patients with mixed hyper- therapy on cardiovascular events in 9795 people with type 2lipidemia [abstract no. 2246-PO]. Diabetes 2006 Jun; 55 Sup- diabetes mellitus (the FIELD study): randomised controlledpl. 1: 519 trial. Lancet 2005 Nov 26; 366 (9500): 1849-61

128. Packard KA, Backes JM, Lenz TL, et al. Comparison of 144. Frost RJA, Otto C, Geiss HC, et al. Effects of atorvastatin versusgemfibrozil and fenofibrate in patients with dyslipidemic coro- fenofibrate on lipoprotein profiles, low-density lipoproteinnary heart disease. Pharmacotherapy 2002 Dec; 22 (12): 1527- subfraction distribution, and hemorheologic parameters in type32 2 diabetes mellitus with mixed hyperlipoproteinemia. Am J

129. Levin A, Duncan L, Djurdjev O, et al. A randomized placebo- Cardiol 2001 Jan 1; 87 (1): 44-8controlled double-blind trial of lipid-lowering strategies in 145. Sarano N, Kozlov S, Tvorogova M, et al. Combined therapypatients with renal insufficiency: diet modification with or with fluvastatin and fenofibrate in patients with hyper-without fenofibrate. Clin Nephrol 2000 Feb; 53 (2): 140-6 lipidaemia and non-insulin-dependent diabetes mellitus [ab-

130. Keating GM, Ormrod D. Micronised fenofibrate: an updated stract no. 111]. Heart 2000 Jun; 83 Suppl. II: 28review of its clinical efficacy in the management of dys- 146. Durrington PN, Tuomilehto J, Hamann A, et al. Rosuvastatinlipidaemia. Drugs 2002; 62 (13): 1909-44 and fenofibrate alone and in combination in type 2 diabetes

131. Farnier M, Salko T, Isaacsohn JL, et al. Effects of baseline level patients with combined hyperlipidaemia. Diabetes Res Clinof triglycerides on changes in lipid levels from combined Pract 2004 May; 64 (2): 137-51


152 Keating & Croom

147. Muhlestein JB, May HT, Jensen JR, et al. The reduction of Atherosclerosis (MESA): gender, ethnicity, and coronary ar-inflammatory biomarkers by statin, fibrate, and combination tery calcium. Circulation 2006 Feb 7; 113: 647-56therapy among diabetic patients with mixed dyslipidemia: the 164. Beaton SJ, Nag SS, Gunter MJ, et al. Adequacy of glycemic,DIACOR (Diabetes and Combined Lipid Therapy Regimen) lipid, and blood pressure management for patients with diabe-study. J Am Coll Cardiol 2006 Jul 18; 48 (2): 396-401 tes in a managed care setting. Diabetes Care 2004 Mar; 27 (3):

148. Derosa G, Cicero AFG, Bertone G, et al. Comparison of fluvas- 694-8tatin + fenofibrate combination therapy and fluvastatin 165. Davidson MH. Reducing residual risk for patients on statinmonotherapy in the treatment of combined hyperlipidemia, therapy: potential role of combination therapy. Am J Cardioltype 2 diabetes mellitus, and coronary heart disease: a 12- 2005 Nov 7; 96 (9A): 3-13Kmonth, randomized, double-blind, controlled trial. Clin Ther 166. Steiner G. A new perspective in the treatment of dyslipidemia:2004 Oct; 26 (10): 1599-607 can fenofibrate offer unique benefits in the treatment of type 2

149. Pearson RR, Thomas H, Jensen JR, et al. Triple-therapy with a diabetes mellitus? Treat Endocrinol 2005; 4 (5): 311-7statin, fibrate and ezetimibe safely allows more diabetic pa- 167. Cholesterol Treatment Trialists’ (CTT) Collaborators. Efficacytients with mixed dyslipidemias to reach national cholesterol and safety of cholesterol-lowering treatment: prospectiveeducation program guidelines: an analysis of the DIACOR meta-analysis of data from 90 056 participants in 14 randomis-study [abstract no. 808-6 plus oral presentation]. 55th Annual ed trials of statins. Lancet 2005 Oct 8; 366 (9493): 1267-78Scientific Session of the American College of Cardiology;

168. Ballantyne CM, Herd A, Ferlic LL, et al. Influence of low HDL2006 Mar 11-14; Atlanta (GA)on progression of coronary artery disease and response to

150. Keech A, on behalf of the FIELD Study Investigators. Effects of fluvastatin therapy. Circulation 1999 Feb 16; 99: 736-43long-term fenofibrate therapy on cardiovascular events among

169. Woodman RJ, Chew GT, Watts GF. Mechanisms, significance9795 people with type 2 diabetes mellitus: the FIELD study, aand treatment of vascular dysfunction in type 2 diabetes mel-randomised controlled trial [abstract no. We-S15:2]. Athero-litus: focus on lipid-regulating therapy. Drugs 2005; 65 (1):sclerosis 2006 Jun; 7 (3 Suppl.): 34231-74151. Ansquer J-C, Foucher C, Rattier S, et al. Fenofibrate reduces

170. Robins SJ, Collins D, Wittes JT, et al. Relation of gemfibrozilprogression to microalbuminuria over 3 years in a placebo-treatment and lipid levels with major coronary events: VA-controlled study in type 2 diabetes: results from the DiabetesHIT: a randomized controlled trial. JAMA 2001 Mar 28; 285Atherosclerosis Intervention Study (DAIS). Am J Kidney Dis(12): 1585-912005 Mar; 45 (3): 485-93

171. Robins SJ. Cardiovascular disease with diabetes or the metabol-152. Hottelart C, El Esper N, Achard J-M, et al. Fenofibrate increasesic syndrome: should statins or fibrates be first line lipid ther-blood creatinine, but does not change the glomerular filtrationapy? Curr Opin Lipidol 2003 Dec; 14 (6): 575-83rate in patients with mild renal insufficiency [in French].

172. Pyorala K, Olsson AG, Pedersen TR, et al. Cholesterol loweringNephrologie 1999; 20 (1): 41-4with simvastatin improves prognosis of diabetic patients with153. Hottelart C, El Esper N, Rose F, et al. Fenofibrate increasescoronary heart disease: a subgroup analysis of the Scandinavi-creatininemia by increasing metabolic production of creati-an Simvastatin Survival Study. Diabetes Care 1997 Apr; 20nine. Nephron 2002; 92 (3): 536-41(4): 614-20154. Graham DJ, Staffa JA, Shatin D, et al. Incidence of hospitalized

173. Haffner SM, Alexander CM, Cook TJ, et al. Reduced coronaryrhabdomyolysis in patients treated with lipid-lowering drugs. Jevents in simvastatin-treated patients with coronary heart dis-Am Med Asso 2004 Dec 1; 292 (21): 2585-90ease and diabetes or impaired fasting glucose levels: subgroup155. Barker BJ, Goodenough RR, Falko JM. Fenofibrate monother-analyses in the Scandinavian Simvastatin Survival Study. Archapy induced rhabdomyolysis. Diabetes Care 2003 Aug; 26 (8):Intern Med 1999 Dec 13/27; 159: 2661-72482-3

174. Goldberg RB, Mellies MJ, Sacks FM, et al. Cardiovascular156. Jacob SS, Jacob S, Williams C, et al. Simvastatin, fenofibrate,events and their reduction with pravastatin in diabetic andand rhabdomyolysis. Diabetes Care 2005 May; 28 (5): 1258glucose-intolerant myocardial infarction survivors with aver-157. Kursat S, Alici T, Colak HB. A case of rhabdomyolysis inducedage cholesterol levels: subgroup analyses in the Cholesterolacute renal failure secondary to statin-fibrate-derivative com-and Recurrent Events (CARE) trial. Circulation 1998 Dec 8;bination and occult hypothyroidism. Clin Nephrol 2005 Nov;98: 2513-964 (5): 391-3

175. Heart Protection Study Collaborative Group. MRC/BHF Heart158. Ireland JHE, Eggert CH, Arendt CJ, et al. Rhabdomyolysis withProtection Study of cholesterol-lowering with simvastatin incardiac involvement and acute renal failure in a patient taking5963 people with diabetes: a randomised placebo-controlledrosuvastatin and fenofibrate. Ann Intern Med 2005 Jun 7; 142trial. Lancet 2003 Jun 14; 361 (9374): 2005-16(11): 949-50

176. Keech A, Colquhoun D, Best J, et al. Secondary prevention of159. Ghosh B, Sengupta S, Bhattacharjee B, et al. Fenofibrate-cardiovascular events with long-term pravastatin in patientsinduced myopathy. Neurol India 2004 Jun; 52 (2): 268-9with diabetes or impaired fasting glucose: results from the160. Ledl M, Hohenecker J, Francesconi C, et al. Acute myopathy inLIPID trial. Diabetes Care 2003 Oct; 26 (10): 2713-21a type 2 diabetic patient on combination therapy with

metformin, fenofibrate and rosiglitazone. Diabetologia 2005 177. Sever PS, Dahlof B, Poulter NR, et al. Prevention of coronaryOct; 48 (10): 1996-8 and stroke events with atorvastatin in hypertensive patients

who have average or lower-than-average cholesterol concen-161. Alsheikh-Ali AA, Kuvin JT, Karas RH. Risk of adverse eventstrations, in the Anglo-Scandinavian Cardiac Outcomes Trial –with fibrates. Am J Cardiol 2004 Oct 1; 94 (7): 935-8Lipid Lowering Arm (ASCOT-LLA): a multicentre randomis-162. Jones PH, Davidson MH. Reporting rate of rhabdomyolysised controlled trial. Lancet 2003 Apr 5; 361 (9364): 1149-58with fenofibrate + statin versus gemfibrozil + any statin. Am J

Cardiol 2005 Jan 1; 95 (1): 120-2 178. Colhoun HM, Betteridge DJ, Durrington PN, et al. Primary163. Goff Jr DC, Bertoni AG, Kramer H, et al. Dyslipidemia preva- prevention of cardiovascular disease with atorvastatin in type 2

lence, treatment, and control in the Multi-Ethnic Study of diabetes in the Collaborative Atorvastatin Diabetes Study



(CARDS): multicentre randomised placebo-controlled trial. http://www.clinicaltrials.gov/ct/gui/show/NCT00000620?ordLancet 2004 Aug 21; 364 (9435): 685-96 er=2 [Accessed 2006 Feb 4]

179. Knopp RH, D’Emden M, Smilde JG, et al. Efficacy and safety 188. Smith Jr SC, Allen J, Blair SN, et al. AHA/ACC guidelines forof atorvastatin in the prevention of cardiovascular end points in secondary prevention for patients with coronary and othersubjects with type 2 diabetes: the Atorvastatin Study for Pre- atherosclerotic vascular disease: 2006 update. Circulationvention of Coronary Heart Disease Endpoints in Non-insulin- 2006 May 16; 113: 2363-72dependent diabetes mellitus (ASPEN). Diabetes Care 2006 Jul;

189. De Backer G, Ambrosioni E, Borch-Johnsen K, et al. European29 (7): 1478-85guidelines on cardiovascular disease prevention in clinical180. Wanner C, Krane V, Marz W, et al. Atorvastatin in patients withpractice: third joint task force of European and other societiestype 2 diabetes mellitus undergoing haemodialysis. N Engl Jon cardiovascular disease prevention in clinical practice (con-Med 2005 Jul 21; 353 (3): 238-48stituted by representatives of eight societies and by invited181. Tenkanen L, Manttari M, Kovanen PT, et al. Gemfibrozil in theexperts). Atherosclerosis 2004 Apr; 173 (2): 381-91treatment of dyslipidemia: an 18-year mortality follow-up of

the Helsinki Heart Study. Arch Intern Med 2006 Apr 10; 166: 190. Prueksaritanont T, Tang C, Qiu Y, et al. Effects of fibrates on743-8 metabolism of statins in human hepatocytes. Drug Metab

Dispos 2002; 30 (11): 1280-7182. Tenenbaum A, Fisman EZ, Boyko V, et al. Attenuation ofprogression of insulin resistance in patients with coronary 191. Prueksaritanont T, Zhao JJ, Ma B, et al. Mechanistic studies onartery disease by bezafibrate. Arch Int Med 2006 Apr 10; 166:

metabolic interactions between gemfibrozil and statins. J737-41

Pharmacol Exp Ther 2002; 301 (3): 1042-51183. Tenenbaum A, Motro M, Fisman EZ, et al. Bezafibrate for the

192. Wang J-S, Neuvonen M, Wen X, et al. Gemfibrozil inhibitssecondary prevention of myocardial infarction in patients withCYP2C8-mediated cerivastatin metabolism in human livermetabolic syndrome. Arch Intern Med 2005 May 23; 165:microsomes. Drug Metab Dispos 2002; 30 (12): 1352-61154-60

184. Rubins HB, Robins SJ, Collins D, et al. Diabetes, plasma 193. Farnier M. Combination therapy with an HMG-CoA reductaseinsulin, and cardiovascular disease: subgroup analysis from the inhibitor and a fibric acid derivative: a critical review ofDepartment of Veterans Affairs High-Density Lipoprotein In- potential benefits and drawbacks. Am J Cardiovasc Drugstervention Trial (VA-HIT). Arch Intern Med 2002 Dec 9/23; 2003 Jan; 3 (3): 169-78162: 2597-604

194. Merck/Schering-Plough Pharmaceuticals. Zetia (ezetimibe)185. UK Prospective Diabetes Study (UKPDS) Group. Effect oftablets: prescribing information [online]. Available from URL:intensive blood-glucose control with metformin on complica-http://www.fda.gov [Accessed 2006 May 10]tions in overweight patients with type 2 diabetes (UKPDS 34).

Lancet 1998 Sep 12; 352: 854-65

186. Colhoun H. After FIELD: should fibrates be used to preventCorrespondence: Gillian M. Keating, Wolters Kluwercardiovascular disease in diabetes? Lancet 2005 Nov 26; 366Health | Adis, 41 Centorian Drive, Private Bag 65901,(9500): 1829-31Mairangi Bay, Auckland 1311, New Zealand.187. ClinicalTrials.gov. Strategies to reduce cardiovascular diseaseE-mail: [email protected] in individuals with diabetes [online]. Available from URL:


2007, volume 67, issue 1

Documents