temporal trends in initiation of therapy with tumor
TRANSCRIPT
Accepted Manuscript
Temporal Trends in Initiation of Therapy With Tumor Necrosis Factor Antagonistsfor Patients With Inflammatory Bowel Disease: A Population-based Analysis
Laura E. Targownik, Aruni Tenakaroon, Stella Leung, Lisa M. Lix, Harminder Singh,Charles N. Bernstein
PII: S1542-3565(17)30195-7DOI: 10.1016/j.cgh.2017.01.035Reference: YJCGH 55117
To appear in: Clinical Gastroenterology and HepatologyAccepted Date: 31 January 2017
Please cite this article as: Targownik LE, Tenakaroon A, Leung S, Lix LM, Singh H, Bernstein CN,Temporal Trends in Initiation of Therapy With Tumor Necrosis Factor Antagonists for Patients WithInflammatory Bowel Disease: A Population-based Analysis, Clinical Gastroenterology and Hepatology(2017), doi: 10.1016/j.cgh.2017.01.035.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Title: Temporal Trends in Initiation of Therapy With Tumor Necrosis Factor Antagonists for Patients With Inflammatory Bowel Disease: A Population-based Analysis
Short Title: Trends in Anti-TNF Initiation
Authors: Laura E, Targownik1.3, Aruni Tenakaroon2, Stella Leung, Lisa M. Lix2,3, Harminder Singh1,3,
Charles N. Bernstein1,3
Affiliations:
1Department of Internal Medicine, University of Manitoba, Winnipeg, Manitoba, Canada
2Department of Community Health Sciences, University of Manitoba, Winnipeg, Manitoba, Canada.
3University of Manitoba IBD Clinical and Research Centre Grant Support: Supported by Crohn’s and Colitis Canada Grants in Aid of Research Abbreviations: TNF – tumor necrosis factor IBD – inflammatory bowel disease CD – Crohn’s Disease UC – Ulcerative Colitis IM - Immunomodulators
Send Correspondence to:
Laura Targownik, MD, MSHS
805G-715 McDermot Avenue
Winnipeg, Manitoba
R3E 3P4
email: [email protected]
Word Count: 3545 (manuscript), 623 (references)
Competing Disclosures:
Laura Targownik: Speaker’s Panel for Janssen Canada, Takeda Canada, Pfizer Canada. Grant
Support from Pfizer Canada and Abbvie Canada. Advisory Boards for Takeda Canada, Abbvie
Canada, Janssen Canada.
Charles Bernstein: Advisory Boards for Abbvie Canada, Janssen Canada, Shire Canada, Takeda
Canada, Pfizer, Cubist Pharmaceuticals. Educational grants from Abbvie Canada, Shire Canada,
Takeda Canada, Janssen Canada. Speaker’s panel for Abbvie Canada and Shire Canada
Harminder Singh: Advisory board Pendopharm; Grant support:Merck Canada
All other authors: No competing interests
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Author Responsibilities:
a) study concept and design: LT, LL, HS, CB b) acquisition of data: AT, LL c) analysis and interpretation of data: LT, LL, HS, CB d) drafting of the manuscript: LT, LL, HS, CB e) critical revision of the manuscript for important intellectual content, and statistical analysis:
LT, AT, LL, HS, CB
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Abstract: Background & Aims: Anti-tumor necrosis factor (anti-TNF) agents are effective treatments for Crohn’s disease (CD) and ulcerative colitis (UC). We aimed to determine their patterns of use and changes in these patterns over time, as well as use of immunomodulators and corticosteroids before anti-TNF therapy for persons with inflammatory bowel diseases (IBD). Methods: We used the University of Manitoba IBD Epidemiology Database to identify all anti-TNF users with CD and UC from 2001 through 2014. We assessed changes in the prevalence and incidence of anti-TNF use over different time periods (April 2001–March 2005, April 2005–March 2009, or April 2009 – March 2013). We also characterized patterns of corticosteroid use, corticosteroid dependence, and immunomodulator use before anti-TNF administration, and determined how these changed over time. The primary endpoint was change in time to first receipt of anti-TNF among the different time periods. Results: We identified 950 persons (761 with CD and 189 with UC) who received anti-TNF agents. The cumulative prevalence of persons with current or prior anti-TNF exposure in 2014 was 20.4% for CD and 6.0% for UC. In 2014, the cumulative incidence values of anti-TNF exposure within 5 years of diagnosis were 23.4% for patients with CD and 7.8% for patients with UC. Most users of anti-TNF agents had evidence of corticosteroid dependence (more than 2 g prednisone within any 12-month period) before initiation of anti-TNF therapy. Cumulative corticosteroid exposure before anti-TNF use decreased over time for patients with UC, but not significantly for patients with CD. There was no increase over time in the use of concomitant immunomodulators with anti-TNF therapy. Conclusion: Use of anti-TNF agents increased from 2001 through 2014, with a concomitant significant decrease in cumulative use of corticosteroids before anti-TNF therapy for patients with UC. However, there has been no reduction in cumulative use of corticosteroids before anti-TNF therapy for patients with CD, and no change in use of immunomodulators by patients with CD. These findings indicate a continuing need for optimization of anti-TNF therapy for patients with IBD. KEY WORDS: TNF inhibitor, IBD treatment, patient management, inflammation
INTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTION
Anti-tumor necrosis factor (anti-TNF) therapy has revolutionized the management of patients with
inflammatory bowel disease (IBD)1. In multiple randomized controlled trials, both infliximab (IFX) and
adalimumab (ADA) have demonstrated efficacy in reducing symptom burden2-4, inducing and
maintaining clinical remission, decreasing the risk of hospitalization and surgical resection5, 6, and
promoting mucosal healing in both Crohn’s disease (CD)7 and ulcerative colitis (UC)8. Moreover, the
use of anti-TNF drugs has also been advocated to reduce corticosteroid dependence2, 9
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Corticosteroid use remains highly prevalent in IBD management, despite the known risk of significant
side effects and complications both in the short and long-term10. We recently reported that there
has been no reduction in the cumulative dose of corticosteroids on average across the population of
IBD between 1995 to 2010, despite the increasing prevalence of immunomodulator use among
persons with IBD over that time10. However, we have not specifically assessed corticosteroid
requirements in persons who ultimately use anti-TNF drugs. The expedient introduction of
corticosteroid-sparing agents such as anti-TNF drugs in persons with corticosteroid dependence is an
important metric of quality of care in persons with IBD.
Therefore we sought to evaluate temporal changes in the patterns of anti-TNF use, combination
immunomodulatory therapy at the time of anti-TNF initiation and in the burden of corticosteroid use
prior to anti-TNF use.
METHODSMETHODSMETHODSMETHODS
Data Source: We used the University of Manitoba Inflammatory Bowel Disease Epidemiologic
Database (UMIBDED)11, which contains demographics, inpatient hospital stays, outpatient visits, and
drug use for all persons in Manitoba meeting a well-validated administrative case definition of IBD
from 1984 to 2014. Nearly 100% of residents of Manitoba are registered with Manitoba Health,
which is the single insurer for all medical services. A full description of the methodology used to
identify IBD patients has been previously published11. Data on all outpatient medications dispensed
to residents of Manitoba, including drug name, strength, doses dispensed, and anticipated duration
of therapy, has been captured since 1995. The UMIBDED does not contain endoscopic or histologic
data, and thus the extent of disease, endoscopic and/or histologic severity cannot be determined.
The UMIBDED does not contain information on symptom burden, health-related quality of life, or
personal behaviors (diet, tobacco use, exercise).
Anti-TNF Use: IFX and ADA are the only anti-TNF agents approved for the management of IBD in
Canada. IFX was first approved by Health Canada for CD in 2001, and for UC in 2007; ADA was
approved by Health Canada in 2006 for CD, and in 2011 for UC. Manitoba Health policy restricts
access to anti-TNF therapy to persons who had either not responded to or developed intolerance to
immunomodulators such as azathioprine or methotrexate; however, exceptions are regularly made
on request. There are no specific provincial formulary restrictions regarding the dose of anti-TNF
drug or on the frequency of dispensation. Though the provincial drug databases do not capture
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
medications dispensed by inpatient pharmacies, short-term inpatients that are using anti-TNFs
generally have them dispensed by an outpatient pharmacy and brought into hospital to be
administered. We were not able to track any dispensations of anti-TNF agents that were provided in
a clinical trial.
Study Population: We identified all persons with IBD who received at least one dose of anti-TNF
therapy. Analyses which evaluated time from diagnosis to initiation of anti-TNF therapy included only
incident cases of IBD diagnosed after anti-TNF therapy first became available in Manitoba (August
21st, 2001). We assumed a case of IBD to be incident if they had at least 3 years of continuous
registration with Manitoba Health where no IBD-related health care contacts occurred; the date of
the first IBD contact was assumed to be the date of IBD diagnosis. In analyses of corticosteroid
and/or immunomodulator use prior to anti-TNF initiation, we included only incident cases of IBD
diagnosed after drug use data first became available (April 1st, 1996).
We excluded all IBD patients with potential competing indications for anti-TNF therapy, and a history
of solid organ transplant (as they are often on chronic corticosteroids), defined as having at least 2
physician contacts coded for these diagnoses prior to initiation of anti-TNF therapy.
Point and Cumulative Prevalence: The point and cumulative prevalence of anti-TNF use among all
persons with IBD on all dates between 2001 (the year of anti-TNF introduction in Manitoba) and
2014 was determined for the entire IBD population. Point prevalence was defined as having an anti-
TNF dispensation in the previous 90 days, whereas cumulative prevalence was defined as having a
history of anti-TNF dispensation at any time following IBD diagnosis.
Time to First Use of Anti-TNF Drugs: We performed survival analyses to assess the time from IBD
diagnosis until initial anti-TNF drug dispensation: Subjects were censored at death, outmigration, or
end of follow-up (March 31st, 2014).
Corticosteroid Use Prior to Anti-TNF Dispensation: We assessed oral corticosteroid use prior to anti-
TNF therapy in a number of ways. Within the cohort of all incident IBD who received anti-TNF therapy,
we determined the cumulative dose of corticosteroids in the year prior to starting anti-TNFs, and in
the span between initial IBD diagnosis and first receipt of anti-TNFs. We also determined the
proportion of anti-TNF users with a history of corticosteroid dependence prior to beginning anti-TNF
therapy, defined as receipt of oral systemic corticosteroids of greater than 2000mg of prednisone
used within any 12 month period spread over >=2 separate dispensations. Corticosteroid doses
were converted to prednisone equivalents using a well-accepted conversion table12. We did not
include the use of oral budesonide
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Immunomodulator Use Prior to anti-TNF Dispensation: We assessed the use of immunomodulator
(IM) medications (6-mercaptopurine [6-MP], azathioprine [AZA] and methotrexate [MTX]) in the year
prior to anti-TNF use. We determined the proportion with an active dispensation of IM medication
(defined as a dispensation of an IM of sufficient duration to overlap initiation of anti-TNF) at the time
of anti-TNF initiation. Of the anti-TNF users with active IM use at the time of anti-TNF dispensation,
we estimated the proportion who intended to continue IM therapy after anti-TNF initiation, defined
as: a) dispensation of an IM in the 60 days following initial anti-TNF dispensation or b) IM medication
use for the first time within the 30 days prior to the start of anti-TNF therapy. We also identified all
instances of first IM use among incident cases of IBD, and performed survival analysis using Kaplan
Meier estimators to assess time from initial IM dispensation to anti-TNF initiation.
IBD-Related Hospitalization and Surgery prior to Anti-TNF Use: We determined the frequency of IBD-
related hospitalizations with at least one overnight stay and surgeries between the date of IBD
diagnosis and the initial dispensation of anti-TNF. A hospitalization was considered to be IBD-related
if either CD or UC appeared in the first 3 diagnostic positions on the hospital discharge abstract. IBD
related surgery was defined as any intestinal resection either appearing in the procedure list in the
hospital discharge abstract, or physician’s billing database.
Determination of Disease Severity: We stratified all persons in the cohort according to the severity of
their initial course of disease, I.e., Over the first 12 months of disease following diagnosis. We
defined a person of having a severe initial course of disease if there was either 2 or more IBD related
hospitalizations, or >1000mg of corticosteroids dispensed in the first 12 months following diagnosis
Persons with use of anti-TNF medications in the first 12 months of diagnosis were excluded, as were
subjects with 12 months or fewer of post-diagnosis follow-up. Survival analysis was performed to
determine the rates of anti-TNF exposure between those persons with and without a severe initial
course of IBD.
Statistical Analysis: All survival analyses were performed using non-parametric Kaplan Maier
estimators, with subjects censored at death, outmigration from the province, or the end of database
follow-up (i.e., March 31st, 2014). All analyses were performed separately for persons with CD and
UC.
The primary analyses focused on changes in time to first receipt of anti-TNF among people
diagnosed in different eras (April 2001- March 2005, [Era 1], April 2005 – March 2009 [Era 2],and
April 2009 – March 2013 [Era 3]), with analyses of UC limited to Eras 2 and 3, as anti-TNFs were
not approved in the treatment of UC in Manitoba prior to 2005. We also performed comparisons
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
within these survival analyses using other stratifications, including sex, age group at diagnosis (<18,
18-39, >=40), as well as comparisons between CD and UC. Differences in event-free survival time
were compared using the log-rank test, with Sidak’s correction for multiple comparisons when
appropriate13. We used the Kruskal-Wallis-Dunn test to compare non-normally distributed data,
specifically those outcomes related to cumulative corticosteroid dosing in the year prior to anti-TNF
initiation14.
It is noted that people diagnosed in the different eras will have differing lengths of person-time
available for follow-up, which means that persons diagnosed in later eras will have less measurable
time to accrue corticosteroid exposure. To avoid the biasing effect of this, we limited this analysis to
the same fixed percentage of all persons in each era who went on to receive anti-TNF therapy in
each of the three eras for CD, and the later 2 eras for UC. That percentage was defined as the
number of persons diagnosed in that era who received anti-TNFs, divided by the total number of
persons diagnosed in that era. This resulted in our comparing of the first 23.0% of persons of CD and
the first 9.3% of persons with UC who received anti-TNF each era.
This research was approved by the University of Manitoba Health Research Ethics Board and access
to data was granted by the Manitoba Health Information Privacy Committee
RESULTSRESULTSRESULTSRESULTS
Overall, there were 950 persons with IBD who received anti-TNF therapy, details of this cohort are
shown in Table 1. The point prevalence, gradually increased for both IFX and ADA over time, with IFX
use being more prevalent than ADA use (Figure 1). By the end of the study period, 20.4% and 6.0%
of persons with CD and UC, respectively, had been exposed to an anti-TNF at any time, with 14.2% of
CD and 4.1% of UC having received an anti-TNF drug in the previous 90 days. At the end of the study
period, there was no signficant difference in the cumulative prevalence of anti-TNF use between the
sexes (22.9% males vs. 18.6% females [CD], 6.5% males vs. 5.6% females [UC]). Anti-TNF use was
much more common in younger IBD cohorts with 38.5% of persons with CD diagnosed under age 18
having had anti-TNF exposure, compared to 22.2% aged 18-39, and 10.7% over age 40.
Among incident cases of IBD who were diagnosed during the era of anti-TNF availability. 7.3%,
14.5%, and 23.4% of persons with CD, and 2.2%, 4.4%, and 7.8% of UC were exposed to anti-TNF
within 1, 2, and 5 years of disease diagnosis (Figure 3a-d). There was no difference between the
sexes in terms of likelihood of having an initial dispensation of anti-TNF in the first 5 years following
diagnosis (23.6% males vs. 23.1 % females CD; 8.4% males vs. 7.2% females UC, p>0.2).
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Severity: A total of 2,791 persons (1222 CD, 1569 UC) were diagnosed with IBD in the era of anti-
TNF availability of whom 447 eventually received anti-TNF. Persons with indicators of severe disease
in the 1st year following diagnosis were more likely to receive anti-TNF therapy in the subsequent
years than those without (Supplemental Figures 1a-b). By the end of the 5th year following diagnosis.
25.7% of severe CD and 12.1% of severe UC had been prescribed an anti-TNF, compared with those
with mild disease (12.7% of CD, 2.5% of UC).
IM Use: Overall, 91.5% of CD and 87.1% of UC patients had a history of IM use prior to anti-TNF
initiation, with 56.5% of CD and 46.0% of UC being active users of IMs at the time of onset of anti-
TNF therapy. Of these active users, 65% and 59% of with CD and UC respectively had evidence of
continuing use of IM following anti-TNF initiation. The median time between first receiving an IM drug
and starting anti-TNF therapy was 481 days for CD (IQR: 202-1329) and 447 days for UC 181-1315,
p>0.2 for CD vs. UC). Among persons with CD who went on to receive an anti-TNF, persons
diagnosed under the age of 18 had a longer time from first IM to first anti-TNF, than those ages 18 to
40 (median time: 762 days (IQR 297-1758) vs. 408 days (IQR 163-1314), p<0.01; the difference
was not significant among those with UC (median time: 542.5 days (IQR: 190-2014) vs. 421.5 days
(IQR: 199-1112), p>0.2. Among persons with CD, the cumulative incidence of receiving an anti-TNF
within 1, 2, or 5 years of first being prescribed an IM was 20.3%, 30.4% and 40.3%; corresponding
percentages for UC were 13.1%, 19.2% 25.3%.
Corticosteroid Use and Dependence Prior to anti-TNF Use: The median dose of corticosteroids
dispensed in the year prior to anti-TNF initiation was higher among persons with UC than for CD
(2865mg vs. 830 mg of prednisone equivalents, p<0.0001), as well as in the time between
diagnosis and the first anti-TNF dispensation (6495mg vs. 3490mg of prednisone equivalents,
p<0.0001). In CD, persons diagnosed prior to age 18 were dispensed more corticosteroids prior to
starting anti-TNFs (median: 4405 mg, interquartile range (IQR):2000-10885mg) than did those aged
18-40 at diagnosis (3520mg , IQR 1000-6370 mg), p<0.05 vs. age at diagnosis<18) as well as
those diagnosed when 40 or older (median 2678, IQR 520-5405 mg), p<0.05 vs. age at diagnosis
<18). Men with CD also had a greater consumption of corticosteroids between diagnosis and anti-
TNF initiation than did women (men: 4150 mg, IQR 1595-8000 mg; women: 3000 mg, IQR 830-
6225 mg, p< 0.019); among persons with UC, no differences in cumulative corticosteroid use were
seen between the sexes or between age strata.
Corticosteroid dependence prior to anti-TNF initiation was seen in 72% of UC and 51% of CD. Among
those corticosteroid dependent, 21.5%, 29.6%, and 40.7% of CD and 8.5%, 13.5%, and 19.6% of UC
received an anti-TNF within 1, 2, and 5 years of their initial episode of corticosteroid dependence. In
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
both CD and UC , persons over aged 40 and older at diagnosis were more likely to receive anti-TNF
therapy later after becoming corticosteroid dependent than both those diagnosed at age 18-39 (p-
values: CD <0.0001; UC 0.0006) and those diagnosed when 18 or younger (p-values: CD 0.0097; UC
0.0028).
Time Trends in, Hospitalizations, IBD-Related Surgeries, Immunomodulator and Corticosteroid Use
Prior to Anti-TNF Therapy The proportion of persons with hospitalization in the year prior to anti-TNF
initiation did not change across eras for CD (21% Era 1, 30% in Era 2, 25% Era 3, P>0.2), and for UC
(18% Era 2 vs 29% Era 3, P=0.19). There was also no change in the proportion of persons with CD
with a history of IBD-related surgery across the eras (40% Era 1, 37% in Era 2, 35% Era 3, P>0.2)
Ongoing IM therapy among persons with active IM use immediately prior to anti-TNF initiation
decreased among more recently treated cohorts with CD (82% Era 1, 68% in Era 2, 58% Era 3,
p<0.001 for trend), and was also less prevalent among those using ADA an initial anti-TNF therapy
vs. IFX (68% vs. 57%, p=0.043). There was a significant decrease in the time to anti-TNF
dispensation following IM use between those diagnosed in Era 1 vs. Era 3 (P=0.0056), but all other
comparisons were not significantly different (Figure 4a). However, there was a significant difference
in the time to IM use and anti-TNF dispensation among those with UC comparing Era 2 with Era 3
(p=0.0097, Figure 4b), as well as in the time between onset of corticosteroid dependence and anti-
TNF initiation (p-value 0.0064 via log rank test, Figure 4c); these differences were not significantly
different in CD (Figure 5d)
Among persons with CD in each era who went on to receive anti-TNF, we did not detect a difference
in the cumulative dose of corticosteroids within the subcohort of persons in Era 1, Era 2, or Era 3
who received anti-TNF therapy (Figure 5a). In UC, there was a significant reduction in the cumulative
corticosteroid exposure between persons diagnosed in Era 2 and those diagnosed in Era 3 (Figure
5b)
DISCUSSIONDISCUSSIONDISCUSSIONDISCUSSION
This paper represents the largest published assessment of anti-TNF use among persons with IBD in
a population based setting. Herrinton et al assessed infliximab use in among persons with IBD
followed at Kaiser Permanente in 2004, and found a prevalence of around 4% of CD patients and
0.5% of persons with UC15. More recently, Park et al assessed IFX use among IBD patients followed
in the Stanford University Health care network, and found a slight increase in the prevalence of IFX
use from 15% in 2008 to 18% of all IBD patients in 201116. Similarly, Jeuring et al, recently
demonstrated in a Dutch population based cohor that 41.8% of all persons with CD diagnosed since
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
2006 were exposed to anti-TNF agents within 5 years of diagnosis, compared with 3.1% for those
diagnosed between 1991 and 1998. Identifying the patterns and predictors of anti-TNF use is
important, given the high direct costs of biologic therapy for persons with IBD. Several economic
analyses have suggested the costs of biologic therapy in the modern era exceed all other drug costs,
as well as those for hospitalization and surgery. Bernstein et al have previously reported that anti-
TNF use was responsible for over half of all drug costs among IBD patients in 200617; a high
proportional cost of biologics has also been seen in European analyses18. Understanding that the
use of biologic therapy is increasing in time and also is being used earlier in the course of disease
will assist payers in estimating the impacts of coverage of these medications on future budgets.
We found that biologic use was much more common among persons diagnosed before age 18, when
compared to those diagnosed between ages 18-39 and those diagnosed past age 40. It is
recognized that persons diagnosed with IBD as children or adolescents may have a more aggressive
course of disease, which likely is a major driver of the need for biologic therapy19, 20. Still, given the
high rates of corticosteroid dependence prior to anti-TNF therapy , it is conceivable that increased
attention to reducing corticosteroid use and/or recognizing the need for recurrent corticosteroid use
as an indication for biologic therapy may lead to even greater utilization of anti-TNF therapies in
these cohorts.
Surprisingly, while we were able to detect a reduction in overall corticosteroid use and in the time
between either becoming corticosteroid dependent or starting immunomodulators prior to anti-TNF
use in UC, we did not see a statistically significant time trend effect on these outcomes among those
with CD. Furthermore, symptoms of ongoing or recurrent inflammation in UC are generally easier to
recognize than they are with CD21; therefore, given the secular trend towards greater utilization of
anti-TNF therapy in general, clinicians may be more likely to step-up to anti-TNF therapy. However,
only 40.7% of persons with CD used an anti-TNF within 5 years of having evidence of corticosteroid
dependence, suggesting opportunities for accessing anti-TNFs earlier in the course of disease and
avoiding significant corticosteroid exposure.
Interestingly, we did not detect any increase in concomitant immunomodulator use among persons
starting biologic therapy in the more recent eras of treatment. The benefit of combination therapy
was definitely demonstrated with the SONIC study in 20107. However, we were not able to detect
increased uptake in combination therapy among anti-TNF initiators. This may be related to fears
about an increased risk of infectious or malignant complications with combination therapy23, though
the benefits of combination therapy almost certainly outweigh the theoretical increased risks24.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Alternatively, this may represent a knowledge translation gap between evidence and practice,
suggesting the need for further education and incentivization for patients and clinicians
The major strength of this work is that is represents a large population based cohort of persons with
IBD with longitudinal data extending from the initial release of anti-TNF medication to the recent
past. As such, we are able to provide a panoramic view on the natural history of IBD in the anti-TNF
era, and evaluate changes in prescribing practice over that period. This study does have some
noteworthy limitations. As this dataset contains only outpatient drug dispensations, it is possible that
measured corticosteroid exposure may be underestimated, as corticosteroids provided to inpatients
would not have been counted. Conversely, persons dispensed corticosteroids may not necessarily
consume every dose, which could lead to dose overestimation. This dataset also does not contain
data on dietary habits, tobacco exposure, and other behavioral risk factors which may also influence
anti-TNF dispensation.
In summation, anti-TNFs use has been increasing over time, and they are used earlier in the course
of disease. The burden of cumulative corticosteroid use prior to starting anti-TNF therapy appears to
be decreasing over time, suggesting that clinicians are less likely to delay the use of anti-TNFs in
clinical scenarios where they are likely beneficial. However, a majority of persons with IBD do not
receive anti-TNFs within 5 years of first becoming corticosteroid dependent, suggesting the need for
improving quality of care in persons with IBD. Further, barriers preventing the use of combination
therapy in persons initiating anti-TNF drugs should be sought out in order to promote optimal
outcomes and best-practices.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Table 1: Baseline Characteristics of Anti-TNF Users
CD (n=721) UC (n=229) p-value Mean Age at 1st Use (mean±SD) 37.86 ± 14.6 39.29 ± 16.15 0.39 % female 379 (52.57) 107 (46.72) 0.13 Era of Diagnosis* (n=432, CD=293 UC=139) Aug 2001-Apr 2005 88 (20.37) 34 (24.46) Apr 2005-Mar 2009 113 (38.57) 47 (33.81) 0.051 Apr 2009-Mar 2013 92 (31.40) 58 (41.73) Era of anti-TNF Use Aug 2001-Apr 2005 136 (18.86) <6 Apr 2005-Mar 2009 206 (28.57) 44 (19.21) <0.0001 Apr 2009-Mar 2014 379 (52.57) 182 (79.48) Type of Anti-TNF Infliximab 539 (74.76) 222 (96.94) <0.0001 Adalimumab 182 (25.24) 7 (3.06) % with IBD-related hospitalization Within 30 days after diagnosis (n=868, CD=660 UC=208)
166 (25.15) 32 (15.38) 0.0032
Within Year Previous to Anti-TNF Initiation 242 (33.56) 68 (29.69) 0.29 % Use of Other Medications (Year before anti-TNF Initiation)
5-ASA 274 (38.00) 158 (69.00) <0.0001 Systemic Corticosteroids 407 (56.45) 197 (86.03) <0.0001 Oral budesonide 132 (18.31) 7 (3.06) <0.0001 Azathioprine/6-MP 528 (73.23) 154 (67.25) 0.092 Methotrexate 79 (10.96) 18 (7.86) 0.21 % With Use of Other IBD Medications (Any Time before anti-TNF)
5-ASA 561 (77.81) 220 (96.07) <0.0001 Systemic Corticosteroids 619 (85.85) 221 (96.51) <0.0001 Oral budesonide 245 (33.98) 16 (6.99) <0.0001 Azathioprine/6-MP 635 (88.07) 194 (84.72) 0.2103 Methotrexate 126 (17.48) 22 (9.61) 0.0034 % with Previous IBD surgery Within 1 year Prior to Anti-TNF Initiation 41 (5.69) <6 0.0036
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
Figure 1: Point Prevalence of Anti-TNF Use by Calendar Date
Figure 2: Cumulative Prevalence of Anti-TNF Use by Calendar Date
Figure 3: Time Between Initial Diagnosis of IBD and Initial Receipt of Anti-TNF
Figure 4: Time Until Initial Receipt of Anti-TNF From First Immunomodular Dispensation or First
Instance of Corticosteroid Dependence, by Era of Diagnosis
Figure 5: Comparison of Individual Cumulative Corticosteroid Dosing Between Eras, Matched on Time
from Diagnosis Until Anti-TNF Use
Supplemental Fiigure 1: Time From Disease Onset to Time to 1st anti-TNF, by Severity of Disease in
1st Year Following Diagnosis
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
REFERENCESREFERENCESREFERENCESREFERENCES
1. Terdiman JP, Gruss CB, Heidelbaugh JJ, et al. American Gastroenterological Association Institute guideline on the use of thiopurines, methotrexate, and anti-TNF-alpha biologic drugs for the induction and maintenance of remission in inflammatory Crohn's disease. Gastroenterology 2013;145:1459-63.
2. Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance infliximab for Crohn's disease: the ACCENT I randomised trial. Lancet 2002;359:1541-9.
3. Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn's disease: the CHARM trial. Gastroenterology 2007;132:52-65.
4. Sandborn WJ, Hanauer SB, Rutgeerts P, et al. Adalimumab for maintenance treatment of Crohn's disease: results of the CLASSIC II trial. Gut 2007;56:1232-9.
5. Leombruno JP, Nguyen GC, Grootendorst P, et al. Hospitalization and surgical rates in patients with Crohn's disease treated with infliximab: a matched analysis. Pharmacoepidemiol Drug Saf 2011;20:838-48.
6. Khanna R, Bressler B, Levesque BG, et al. Early combined immunosuppression for the management of Crohn's disease (REACT): a cluster randomised controlled trial. Lancet 2015.
7. Colombel JF, Sandborn WJ, Reinisch W, et al. Infliximab, azathioprine, or combination therapy for Crohn's disease. N Engl J Med 2010;362:1383-95.
8. Panaccione R, Ghosh S, Middleton S, et al. Combination therapy with infliximab and azathioprine is superior to monotherapy with either agent in ulcerative colitis. Gastroenterology 2014;146:392-400 e3.
9. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for induction and maintenance therapy for ulcerative colitis. N Engl J Med 2005;353:2462-76.
10. Targownik LE, Nugent Z, Singh H, et al. Prevalence of and outcomes associated with corticosteroid prescription in inflammatory bowel disease. Inflamm Bowel Dis 2014;20:622-30.
11. Bernstein CN, Blanchard JF, Rawsthorne P, et al. Epidemiology of Crohn's disease and ulcerative colitis in a central Canadian province: a population-based study. Am J Epidemiol 1999;149:916-24.
12. Mager DE, Lin SX, Blum RA, et al. Dose equivalency evaluation of major corticosteroids: pharmacokinetics and cell trafficking and cortisol dynamics. J Clin Pharmacol 2003;43:1216-27.
13. Šidàk Z. Rectangular confidence region for themeans of multivariate normal distributions. Journal of the American Statistical Association 1967;62:626-633
14. Dunn OJ. Multiple contrasts using rank sums. Technometrics 1964;5:241-52
15. Herrinton LJ, Liu L, Fireman B, et al. Time trends in therapies and outcomes for adult
inflammatory bowel disease, Northern California, 1998-2005. Gastroenterology 2009;137:502-11.
16. Park KT, Sin A, Wu M, et al. Utilization trends of anti-TNF agents and health outcomes in adults and children with inflammatory bowel diseases: a single-center experience. Inflamm Bowel Dis 2014;20:1242-9.
17. Bernstein CN, Longobardi T, Finlayson G, et al. Direct medical cost of managing IBD patients: a Canadian population-based study. Inflamm Bowel Dis 2012;18:1498-508.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
18. van der Valk ME, Mangen MJ, Leenders M, et al. Healthcare costs of inflammatory bowel disease have shifted from hospitalisation and surgery towards anti-TNFalpha therapy: results from the COIN study. Gut 2014;63:72-9.
19. Romberg-Camps MJ, Dagnelie PC, Kester AD, et al. Influence of phenotype at diagnosis and of other potential prognostic factors on the course of inflammatory bowel disease. Am J Gastroenterol 2009;104:371-83.
20. Beaugerie L, Seksik P, Nion-Larmurier I, et al. Predictors of Crohn's disease. Gastroenterology 2006;130:650-6.
21. Targownik LE, Sexton KA, Bernstein MT, et al. The Relationship Among Perceived Stress, Symptoms, and Inflammation in Persons With Inflammatory Bowel Disease. Am J Gastroenterol 2015;110:1001-12; quiz 1013.
22. Baert F, Noman M, Vermeire S, et al. Influence of immunogenicity on the long-term efficacy of infliximab in Crohn's disease. N Engl J Med 2003;348:601-8.
23. Deepak P, Stobaugh DJ, Ehrenpreis ED. Infectious complications of TNF-alpha inhibitor monotherapy versus combination therapy with immunomodulators in inflammatory bowel disease: analysis of the Food and Drug Administration Adverse Event Reporting System. J Gastrointestin Liver Dis 2013;22:269-76.
24. Siegel CA, Finlayson SR, Sands BE, et al. Adverse events do not outweigh benefits of combination therapy for Crohn's disease in a decision analytic model. Clin Gastroenterol Hepatol 2012;10:46-51.
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT
MANUSCRIP
T
ACCEPTED
ACCEPTED MANUSCRIPT