articulo antibiotico 2011

Antibiotic dosing during sustained low-efficiency dialysis: Specialconsiderations in adult critically ill patients*

Kimberly N. Bogard, PharmD, BCPS; Nicole T. Peterson, PharmD, BCPS; Troy J. Plumb, MD;Michael W. Erwin, PharmD, BCPS; Patrick D. Fuller, PharmD, BCPS; Keith M. Olsen, PharmD, FCCP, FCCM

Critically ill patients demon-strate alterations in drugpharmacokinetics due to mul-tiorgan dysfunction syn-drome, acute changes in volume status(both intra- and extravascular), and alter-ations in metabolism, absorption, andelimination. Chief among these altera-

tions is the development of acute kidneyinjury often leading to an increased rateof morbidity and mortality. The incidenceof acute kidney injury in critically ill pa-tients has been estimated at 1% to 25% ofadmissions with a resulting mortalityranging from 28% to 90% (1, 2). In anattempt to reduce refractory fluid over-load and metabolic acidosis and the asso-ciated morbidity and mortality of acutekidney injury, both intermittent and con-tinuous renal replacement therapies(CRRTs) have been utilized.

Continuous renal replacement thera-pies are not new to the intensive care unit(ICU) as they have been used for 25 yrs(3, 4). Continuous therapies are preferredamong critically ill patients who are he-modynamically unstable. In addition,CRRT is preferred over intermittent he-modialysis (IHD) techniques due to im-proved efficacy in severely catabolic pa-tients (5). Despite these advantages,

morbidity and mortality remain high,and better dialysis techniques have beensought. Marshall et al (5) first describedsustained low-efficiency dialysis (SLED),a hybrid dialysis modality that utilizesIHD equipment in conjunction with re-duced blood and dialysate flow rates.SLED offers several advantages over con-ventional CRRT, including higher soluteclearance and decreased supply costs.When first introduced, SLED was utilizedfor periods of 612 hrs. However, at ourinstitution and others, SLED is being uti-lized as a continuous therapy (6). Despitethe adoption of SLED by some nephrolo-gists as the dialysis modality of choice inhemodynamically unstable patients, littleis known regarding the effect of SLED ondrug clearance.

Few studies have examined the phar-macokinetics and resulting pharmacody-namics of drug therapy during CRRT, andeven fewer have evaluated the impact of

*See also p. 602.From the Department of Pharmaceutical and Nu-

trition Care (KNB, MWE, PDF, KMO), The NebraskaMedical Center; Departments of Pharmacy Practice(KNB, KMO, PDF, NTP) and Internal Medicine, Divisionof Nephrology (TJP), University of Nebraska MedicalCenter, Omaha, NE; and Department of Pharmacy(NTP), Shands at the University of Florida, Gainesville, FL.

The authors have not disclosed any potential con-flicts of interest.

For information regarding this article, E-mail:[email protected]

Copyright 2011 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/CCM.0b013e318206c3b2

Objective: To address issues of antibiotic dosing during sus-tained low-efficiency dialysis by using available pharmacokineticdata, intermittent and continuous renal replacement therapy di-alysis guidelines, and our experience with sustained low-effi-ciency dialysis.

Data Resources: Published clinical trials, case reports, andreviews of antibiotic dosing in humans during sustained low-efficiency dialysis.

Data Extraction: A search of electronic databases (MEDLINE,PubMed, and Ovid) was conducted by using key words of ex-tended daily dialysis, sustained low-efficiency dialysis, antibiot-ics, antimicrobial agents, and pharmacokinetics. MEDLINE iden-tified 32 sustained low-efficiency dialysis articles, and PubMedidentified 33 articles. All papers describing antibiotic clearanceprospectively in patients were considered for this article.

Data Synthesis: We identified nine original research articlesand case reports that determined the impact of sustained low-efficiency dialysis on antibiotic clearance in patients. The bloodand dialysate flow rates, duration of dialysis, type of filter, and thepharmacokinetic parameters were extracted from each article. Ifmultiple articles on the same drug were published, they werecompared for consistency with the aforementioned dialysis pa-rameters and then compared with forms of continuous renal

replacement therapy. Antibiotic clearance by sustained low-effi-ciency dialysis was determined to be similar or higher thancontinuous renal replacement therapy therapies. The estimatedcreatinine clearance during sustained low-efficiency dialysis wasapproximately 60 mL/min to 100 mL/min depending on the bloodand dialysate flow rates and the type of filter used.

Conclusions: The potential for significant drug removal duringan 8-hr-or-longer sustained low-efficiency dialysis session isevident by the limited number of studies available. Becausesignificant amounts of drug may be removed by sustained low-efficiency dialysis combined with altered pharmacokinetic vari-ables in critically ill patients, the risk for suboptimal drug con-centrations and pharmacodynamics must be considered.Appropriate dose and calculation of dosing intervals is essentialto provide adequate antibiotic therapy in these patients. It isrecommended that institutions who utilize sustained low-effi-ciency dialysis establish dosing guidelines for all pharmacists andphysicians to follow to provide consistent delivery of antibiotics atadequate concentrations. (Crit Care Med 2011; 39:560570)

KEY WORDS: pharmacokinetics; dialysis; antibiotics; anti-infec-tives; sustained low-efficiency dialysis; continuous renal replace-ment therapy; critical care

560 Crit Care Med 2011 Vol. 39, No. 3

SLED on drug removal (79). Most earlydrug pharmacokinetic studies with CRRTutilized low hemofiltration rates and di-alysate flows, which provide little infor-mation in critically ill patients who re-ceive much higher hemofiltration andflow rates. Although SLED uses lowerflow rates than IHD, the duration ofSLED often results in higher clearancerates, and thus has a potentially greaterimpact on drug clearance. The goals ofantibiotic dosing are to maximize thepharmacodynamic effect while limitingdrug-associated toxicity. This is especiallytrue among critically ill patients who al-ready have significant alterations in phar-macokinetic and pharmacodynamic rela-tionships associated with underlyingdisease processes, before the introductionof dialysis to their treatment regimen(10, 11). The primary objectives of thisreview paper are to provide a greater un-derstanding of SLED and the differencesfrom CRRT, to critically review the avail-able literature on antibiotic pharmacoki-netics during SLED, and to share ourexperiences, interpretations, and recom-mendations for antibiotic dosing duringSLED.

METHODS

To identify relevant articles, a search ofelectronic databases was conducted (PubMed,MEDLINE, and Ovid) from January 1950 toMay 2010. Search results were limited to Eng-lish-language articles only. The following keywords were used: extended daily dialysis,SLED, slow low-efficiency dialysis, antibiotics,antibacterial agents, and pharmacokinetics.Articles were assessed for relevance by a singleinvestigator. Excluded articles failed to ad-dress drug dosing, kinetics, or solute removal.Ovid did not identify any additional papers.Included articles were reviewed and informa-tion regarding dialysis specifications, drugdosing and elimination, pharmacokinetics,and patient characteristics were extracted.

RESULTS

The MEDLINE search yielded the fol-lowing results: Extended daily dialysisprovided 26 articles, SLED yielded 32 ar-ticles, and slow low-efficiency dialysissupplied two articles. The PubMed searchprovided slightly different numbers of re-sults: Extended daily dialysis yielded 51articles, SLED provided 33 articles, andslow low-efficiency dialysis supplied 16articles. When the dialysis search termswere combined with the terms antibioticsor antibacterial agents, nine relevant re-sults were obtained. Those results aresummarized in a later section.

SLED

IHD is currently the standard of carefor the treatment of both acute andchronic renal failure (12). IHD utilizeshigh blood and dialysate flow rates toremove large quantities of toxins ordrugs over a short period of time, usu-ally between 3 and 4 hrs. Blood flowrates are set at 300 500 mL/min, anddialysate flow rates are set at 500 800mL/min. Due to large fluid shifts in ashort duration of time, session lengthscan be extended and blood rates re-duced if standard IHD causes hemody-namic instability.

In comparison, CRRTs operate in acontinuous mode and are only inter-rupted for clotted circuits, patient proce-dures, and routine dialysis circuitchanges whereas IHD treatments typi-cally last 34 hrs. These CRRT therapiesinvolve either diffusive or convectiveclearance and use similar mechanisms asin IHD. However, by using a slower rateof fluid or solute removal, they are con-sidered to be better tolerated. DifferentCRRT modalities utilize hemofiltration,hemodialysis, or a combination of thetwo (hemodiafiltration) for solute re-

moval. Diffusion is the movement of sol-utes from a higher concentration to alower concentration across the semiper-meable dialysis membrane and is the pri-mary mechanism to remove metabolicwaste. Convective clearance occurs assolutes are dragged across the semiper-meable dialysis membrane during ultra-filtration. The method of hemofiltrationutilizes large volumes of dilutional fluidand high ultrafiltration rates to removesolute.

SLED is a hybrid dialysis modality inwhich conventional dialysis equipment isused with the reduced blood and dialysateflow design of CRRT. It was first utilizedin the United States at the University ofArkansas for Medical Sciences in 1998and was designed to provide adequate sol-ute control and better hemodynamic sta-bility with low cost and ease of use fornursing personnel (5). Although some in-stitutions utilize SLED as a continuousmodality, it was originally designed torun overnight (812 hrs), thus providingbetter access to the patient for proce-dures during the day. Blood flow and di-alysate flow rates are generally set be-tween 100 and 300 mL/min. These flowrates are much lower than those utilizedfor IHD. The blood flow rate is similar tothose used for continuous venovenous he-modialysis (CVVHD), whereas the dialysateflow falls in between. These dialysis meth-ods are compared in Table 1. Like IHD,drug removal is achieved primarily throughdiffusion, thus favoring the removal ofwaste products and solutes with low molec-ular weights (13). However, if hemofiltra-tion is added solutes may be transported viaconvection as well, which would allow mol-ecules of greater molecular weights to beremoved (13). Therefore, an important con-sideration for drug dosing during SLED isthe type of dialysis and whether hemofiltra-tion is employed.

Comparison of DialysisModalities

In comparison with traditional IHD,SLED provides similar or better soluteremoval while minimizing the hemody-namic instability that often results fromshort-term, large fluid shifts (14). Al-though it utilizes the same equipment asconventional dialysis, SLED does havesome additional requirements. First, al-though SLED can be run on the conven-tional dialysis machine without addi-tional software, some recalibration oftemperature settings is required. There

Table 1. A comparison of different methods of dialysis

ParameterIntermittentHemodialysis

Continuous VenovenousHemodialysis

SustainedLow-Efficiency

Dialysis

Duration 34 hrs Days 8 hrs to daysDialysate flow (mL/min) 500800 1632 100300Blood flow (mL/min) 300500 100200 160200Urea clearance (mL/min) 150250 2025 7080Hemofiltration Not used in the

United StatesYes Yes

Hemofiltration flow rates Not applicable 2040 mL/kg/hr 100 mL/min

References 6, 15, 18, 67, and 68.

561Crit Care Med 2011 Vol. 39, No. 3

are now software programs available thatprovide a continuous mode that elimi-nates the necessity of recalibration (5).Second, as the blood and dialysate flowrates are slower with SLED than withIHD, the extracorporeal circuitry is at anincreased risk of clotting during the dial-ysis session. Therefore, anticoagulationwith systemic heparinization or citrateregional anticoagulation is often requiredto minimize this risk (5). Either methodis effective at reducing the risk of clottingin the dialysis circuit, but both carrysome inherent risk for adverse effects.

In comparison to other CRRT modal-ities, SLED provides some distinct advan-tages. It does not require specializedequipment, does not require the samelevel of manpower to operate, and doesnot require specialized dialysate solutions(15, 16). In addition, due to its higherdialysate flows relative to CVVHD, it mayprovide a higher drug and solute clear-ance rate (17). Despite these differencesbetween techniques, no one dialysis mo-dality has been proven superior to theothers in terms of patient mortality (5).

Antibiotic Clearance and DosingDuring SLED

When evaluating published literaturefor drug clearance in SLED, one must becognizant of the specific settings or proto-col used for the procedure. The principlesof antimicrobial drug dosing in CRRT haverecently been discussed in detail in thisjournal (10). However, some basic princi-ples will be reviewed here. Like CVVHD andIHD, the most important factor determin-ing drug removal is the degree of proteinbinding. Operating characteristics varyconsiderably and can greatly affect the drugremoval. As shown in Table 2, review of theavailable studies on SLED drug clearanceindicate blood flows vary from 160 to 200mL/min, and dialysate flows vary from 100to 300 mL/min. Unlike CVVHD where lowdialysate flows limit solute removal, thehigher dialysate flow rates used in SLEDmarkedly increase urea clearance rates.Whereas typical CVVHD urea clearancerates are 2035 mL/min, urea and creati-nine clearances have recently been mea-sured at 76 and 75 mL/min, respectively,

using continuous SLED with a blood flowof 200 mL/min and dialysis flow of 100mL/min (18, 19).

A variety of dialyzers have been usedfor SLED (Table 2). With the exception ofone vancomycin investigation, studieshave utilized high-flux polysulfone mem-branes with surface areas between 0.8and 1.6 m2 (Table 2). However, the influ-ence of filter surface area on drug clear-ance is diminished at the low blood anddialysate flow rates used in SLED. In ad-dition, membrane flux can have a signif-icant impact on middle molecular weightdrugs of 500 to 5000 Da (e.g., antibiotics)while high molecular weight drugs(5000 Da) are not effectively removedby low-flux or high-flux dialysis mem-branes (20). High-flux membranes read-ily remove non-proteinbound drug bydiffusion, and drug clearance correlateswith urea clearance (19, 21). Hence,when evaluating drug clearance studiesin patients undergoing SLED, it is impor-tant to not only assess the blood anddialysate flow rates, but also the dialyzermembrane surface area and flux.

Table 2. Review of published pharmacokinetic studies and case reports with antimicrobial agents during sustained low-efficiency dialysis

Drug Study Design PatientsSLED

Duration (hr)Qd

(mL/min)Qb

(mL/min)

Anidulafungin (66) Singledose pharmacokinetic 1 8 180 180

Daptomycin (25) Single-dose pharmacokinetic 1 12 100 200

Daptomycin (26) Single-dose pharmacokinetic 10 8 160 160

Gentamicin (38) Prospective, single-dose pharmacokineticHome dialysis patients

8 8 300 200

Gentamicin (32) Single- and multidose pharmacokinetic,Monte Carlo simulation

14 6 (designed for 10 hrs) 300 300

Critically ill patientsErtapenem (51) Prospective, single-dose pharmacokinetic 6 8 160 160

Meropenem (47) Prospective, single-dose pharmacokinetic 10 8 160 160

Vancomycin (47) Prospective, single-dose pharmacokinetic 10 8 160 160

Vancomycin (48) Prospective, single-dose pharmacokinetic 11 24 100 200

Linezolid (58) Prospective, single-dose pharmacokinetic 5 8-9 100 200

Moxifloxacin (61) Prospective pharmacokinetic 15 8 160 160

Levofloxacin (61) Prospective pharmacokinetic 15 8 160 160

SLED, sustained low-efficiency dialysis; Qd, dialysate flow; Qb, blood flow; IHD, intermittent dialysis; Cl, clearance; Vd, volume of distribution; PK,pharmacokinetics.


Antibiotics that require dose adjust-ment for SLED share some commonproperties. In general, antibiotics thathave increased clearance by SLED arethose that have low protein binding,small molecular size, high water solubil-ity, or high dependence on renal clear-ance. These agents often require adjust-ing the timing of administration orsupplementing the usual dose after IHD.It would follow that SLED patients mayalso require adjusted dosing. The variousantibiotic agents in each class often sharesimilar pharmacokinetic properties, andtherefore, dosing recommendations forSLED patients can be based on availablestudy data. However, there are subtle dif-ferences between certain agents in anti-biotic classes that may lead to differencesin clearance, making the continued studyof pharmacokinetics in these patients es-sential to providing the best outcomes inthe critically ill populations. Many anti-biotics require loading doses to assureadequate initial concentrations; however,none of the published articles with SLEDutilized them. We recommend the use ofsound clinical judgment for each patient

and applying loading doses in a similarmanner as recommend for CRRT (11).

Providing adequate concentrations ofantibiotics to the site of infection is cru-cial to the successful treatment and pos-itive outcomes of all patients, especiallythose residing in an ICU. Antibiotic phar-macodynamic parameters are easily al-tered in the ICU population not only bythe underlying disease, but also by acutephysiologic changes resulting from se-vere infections. SLED may significantlyincrease antibiotic clearance and alterpharmacodynamic parameters resultingin diminished bacterial eradication andclinical response rates. In Table 2, allpublished studies and case reports of an-tibiotic pharmacokinetics during SLEDare summarized. The following discus-sion not only addresses these studies ingreater detail, but also addresses otheragents in the same antibiotic class andprovides dosing recommendations basedon the current understanding of SLED.For the purpose of consistency in thisreview, the term extended daily dialysisis replaced by SLED. Blood and dialysateflow rates and the use of low- and high-

flux membranes differ from one report tothe next and therefore, the dosing recom-mendations in this review must be tai-lored to the dialysis parameters used inindividual institutions.

Daptomycin. Daptomycin, a lipopep-tide antibiotic, has a relatively long half-life of 89 hrs, is primarily eliminated inthe urine (78%; 50% as unchanged), hasa small volume of distribution (0.1 L/kg),and is highly protein bound at 92%,which suggests the drug remains withinplasma and interstitial fluid (22). At leastone report indicates daptomycin is morebioavailable than the high protein bind-ing would imply (22). Predicting dapto-mycin clearance during dialysis is diffi-cult with a low volume of distributionsuggesting significant clearance, but thehigh protein binding suggests otherwise.However, the plasma protein binding isweak, and critically ill patients often havedecreased plasma proteins, which indi-cate a significant clearance by dialysiswould be expected.

Limited data are available regardingthe impact of hemodialysis on daptomy-cin clearance. Dvorchik et al (23) de-

Table 2. Review of published pharmacokinetic studies and case reports with antimicrobial agents during sustained low-efficiency dialysis

Filter Type Outcomes Study Dosing Recommendations

Fresenius F60S polysulfone; high-fluxsurface areas 1.3 m2

PK data comparable to healthy adults No dosage adjustment necessary


PK data indicates SLED effectively eliminateddaptomycin to a larger extent than IHD

Dosing interval every 48 hrs as recommended in IHD mayresults in under-dosing


Cl 63 9 mL/min with a mean drug clearedby one dialysis treatment of 23.3%

6 mg/kg daily to prevent under dosing


PK data indicates SLED cleared 70% ofgentamicin dose, with significantly shorterdrug half-life during SLED sessions

2-2.5 mg/kg ABW after each SLED session to provide peakapproximately 7.5 g/mL and trough approximately 0.7g/mL


PK data indicated rapid Cl by SLED (13.6 hrshalf-life)

Gentamicin 6 mg/kg lean body weight every 48 hrsfollowing a 30-min infusion and given 1 hr before SLED


Total Cl on dialysis was comparable to controlswith normal renal function not requiring renalreplacement therapy with no difference in Vd1

Intensive care unit patients undergoing SLED should notreceive reduced doses to ensure optimal drugconcentrations


Significant removal by high-flux SLED Current IHD dosage adjustments for meropenem may resultin serious underdosing; exact doses must be tailored toillness severity, bacterial minimum inhibitoryconcentration, and drug levels


Significant removal by high-flux SLED Vancomycin concentrations reach subtherapeutic atapproximately 30 hrs

Fresenius F4 and F5; polysulfone; low-flux surface area 0.8 and 1.0 m2

Similar removal as seen with other continuousrenal replacement therapy modalities

Recommend obtaining a random vancomycin level 24 hrsafter infusion

Fresenius F7HPS polysulfone; low-flux surface area 1.6 m2

Significant drug removal by SLED Doses should be administered after SLED sessions whenpossible


Increased Cl by SLED Moxifloxacin requires a standard dose after SLED session


Increased Cl by SLED Levofloxacin may require a dose adjustment after SLEDsession


scribed the disposition and clearancefrom a group of subjects with varied renalfunction by creatinine clearance (Clcr80 mL/min to 30 mL/min) and agroup with end-stage renal disease re-ceiving either IHD or continuous ambu-latory peritoneal dialysis. IHD had a mod-est impact on daptomycin with 15% ofdrug removed over a 4-hr session,prompting the investigators to recom-mend dosing after dialysis without addi-tional supplemental doses (23). Limiteddata are available describing the impactof CRRT on daptomycin clearance. An invitro study evaluated the impact of con-tinuous venovenous hemofiltration(CVVH) on daptomycin clearance in asimulated dialysis model (24). CVVHclearance of daptomycin from wholeblood exceeded the physiologic clearancedescribed in patients with normal renalfunction. A review of antibiotic dosing inpatients receiving CRRT contradictedthese findings by recommending dapto-mycin dosing of 46 mg/kg every 48 hrs inpatients receiving CVVH, CVVHD, or con-tinuous venovenous hemodiafiltration(CVVHDF) (11). Based on the extensiveclearance, the authors recommended dap-tomycin blood concentrations be moni-tored in patients receiving CRRT (24).

Two studies have determined the ef-fect of SLED on daptomycin clearance(Table 2) (25, 26). A 67-yr-old male (110kg total body weight; albumin 2.1 g/dL)with infective endocarditis was treatedwith daptomycin 6 mg/kg infused over 30mins. He received SLED over a 12-hrperiod, with a dialysate flow rate of 100mL/min and a blood flow rate of 200mL/min by using a high-flux polysulfonedialyzer (F60S, surface area 1.3 m2). Asingle blood sample was obtained beforedosing, and serial blood samples were ob-tained after a daptomycin dose with anadditional sample of the total dialysateobtained to calculate the absoluteamount of drug removed. The authorscompared this case study with a previouspaper in which patients received dapto-mycin followed by IHD (27). The dapto-mycin clearance over a 12-hr SLED ses-sion was substantially higher comparedwith IHD with a calculated half-life of 9.4hrs vs. 29.32 hrs, respectively (27). Thedata from the single case report suggestthat SLED eliminates daptomycin effec-tively and to a larger extent than IHD. Arecent study evaluated ten critically illpatients (mean Acute Physiology Assess-ment and Chronic Health Evaluation IIscore 35) with vancomycin-resistant en-

terococci endocarditis and bacteremiatreated with daptomycin 6 mg/kg and re-ceiving SLED with a high-flux polysul-fone dialyzer (F60S; surface area 1.3 m2)(26). A prospective single-dose pharma-cokinetic study was completed with bloodand dialysis flow rates of 160 mL/minwith a dialysis time of 480 mins. Thedialyzer clearance of daptomycin was63 9 mL/min with a mean amount ofdrug cleared by one dialysis treatment of23.3%. Thus, the authors of both studiesconcluded that dosing daptomycin every48 hrs as recommended with IHD wouldresult in significant under dosing andsuggested a more frequent dosing inter-val. Following the 8-hr SLED session,Kielstein et al (26) recommend a dailydaptomycin dose of 6 mg/kg to preventunder dosing. We suggest a schedule thatallows daptomycin be dosed daily at stan-dard doses on the days a patient receivesSLED, which should account for the ob-served increased clearance. We also sug-gest the daptomycin dose be adjusted tothe patients renal function on nondialy-sis days. Increased monitoring for ad-verse events and creatinine phosphoki-nase levels may be warranted anddiscontinuation of daptomycin should beconsidered if elevations occur.

Aminoglycosides. All aminoglycosides(amikacin, gentamicin, and tobramycin)are small hydrophilic molecules, and innon-ICU patients have short half-lives(mean 2.7 hrs), low protein binding(30%), and a low volume of distribution(mean 0.25 L/kg) and are highly depen-dent upon renal clearance (half-life nor-mal 23 hrs vs. renal failure 5070 hrs)(28, 29). It is these same properties thatpredict aminoglycoside removal duringIHD, CVVH, and SLED.

IHD provides an effective mechanismfor removal of aminoglycosides in acuterenal injury. Some authors have recom-mended a loading dose of 23 mg/kg forgentamicin and tobramycin and a 10-mg/kg loading dose for amikacin due tothe alterations in volume of distributionin the critically ill (10, 11, 30). However,based on the current pharmacokineticdata, a greater understanding of amino-glycoside pharmacodynamics, andchanges in bacterial susceptibility, it isunlikely that these loading doses will pro-vide adequate treatment to maximize ef-ficacy. To maximize aminoglycoside bac-terial killing, a ratio of the peak serumconcentration or the total area under-the-serum-concentration-time curve

from 0 to 24 hrs (AUC024) to the mini-mum inhibitory concentration (MIC) ofthe pathogen should equal or exceed 10(peak serum concentration/MIC) and 70to 120 (AUC0-24/MIC) (31). The volume ofdistribution is significantly alteredamong critically ill patients with meanvalues ranging from 0.36 to 0.668 L/kg(3237). The clinical impact of an in-creased volume of distribution is that theinitial dose must be increased or inade-quate pharmacodynamic ratios will beobserved. To achieve these aforemen-tioned pharmacodynamic ratios, someclinicians have suggested initial loadingdoses of 57 mg/kg for gentamicin ortobramycin (32, 33). Extrapolation ofthese recommendations from gentamicin/tobramycin to amikacin would result in a15- to 20-mg/kg initial dose (32, 33, 36). Allaminoglycoside maintenance dose recom-mendations are dependent upon serumconcentration values and pre- or postdialy-sis timing of the serum level.

CVVH, CVVHD, and CVVHDF (CRRTs)also effectively clear aminoglycosides, andtheir continuous nature may provide accel-erated drug removal (11). Loading dosesare recommended for all aminoglycosides ifa patient is receiving CRRTs. An initial one-time dose of 23 mg/kg for gentamicin andtobramycin or 10 mg/kg for amikacinshould help to provide adequate serumconcentrations (10, 11). As with IHD, theseloading doses are unlikely to achieve ade-quate pharmacodynamic ratios to maxi-mize efficacy against targeted pathogensand may need to be increased depending onthe infection type.

Maintenance dosing in IHD is deter-mined by drug clearance over the dosageinterval (10, 30). Usual maintenance doserecommendations for gentamicin and to-bramycin are 12.5 mg/kg every 4872hrs depending on indication, with thehigher doses recommended for systemicGram-negative infections (10, 11). Ami-kacin dosing usually ranges from 5 to 7.5mg/kg every 4872 hrs for IHD patients,again depending on indication (10, 11).Maintenance dosing in CRRT tends to beat shorter intervals relative to IHD, al-though it remains dependent on serumconcentrations. In general, gentamicinand tobramycin may be dosed at 12.5mg/kg every 2448 hrs, whereas amika-cin may be dosed at 7.5 mg/kg every2448 hrs during CRRT (10, 11). Thesedosing recommendations assume a CRRTdialysate rate of 12 L/hr and may requireadjustment to be more aggressive shouldthe ultrafiltration rate be higher.


SLED dosing is not well determinedfor aminoglycosides. Gentamicin wasevaluated in a single-dose study involvingeight non-critically ill patients receiving8-hr home hemodialysis (Table 2) (38).Patients received a gentamicin dose of 0.6mg/kg actual body weight, infused over30 mins after an 8-hr session of SLED.Blood and dialysate flow rates were set at200 mL/min and 300 mL/min, respec-tively, and a high-flux polysulfone dialysisfilter was used (F50S; surface area 0.8m2). Based on the data analysis, a phar-macokinetic model was generated, whichindicated approximately 70% of the gen-tamicin dose was removed during SLED.The elimination half-life of gentamicinwas significantly different during SLED(3.7 0.8 hrs) vs. between SLED ses-sions (20.4 4.7 hrs). The authors alsonoted a lack of gentamicin concentrationrebound after SLED stopped. This is a keydifference between SLED and IHD genta-micin pharmacokinetics. In IHD, the re-moval rate of gentamicin is faster thanthe rate at which gentamicin is releasedfrom the tissues. However, due to theslower blood and dialysate flow rates inSLED, the gentamicin release rate is sim-ilar to the overall SLED removal rate.Therefore, because the tissue concentra-tions and serum concentrations duringSLED are estimated to be at approximateequilibrium, there is little re-equilibra-tion observed after the SLED session(31). The authors felt this difference maycause inappropriate evaluation of genta-micin levels after SLED, leading to poten-tial dosing problems. Based on the find-ings from the pharmacokinetic model,the authors recommend a gentamicindose of 22.5 mg/kg actual body weightafter each 8-hr SLED session to providean estimated peak level of 7.5 g/mL andan estimated trough level of 0.7 g/mL.

The effect of SLED on gentamicinpharmacokinetics was also evaluated in aprospective pharmacokinetic trial in crit-ically ill patients with acute kidney injury(32). Data were collected from 28 genta-micin doses in 14 critically ill patientsreceiving a targeted SLED session of 10hrs with a high-flux polysulfone dialyzer(F60S; surface area 0.6 m2) with bloodand dialysate flow rates of 300 mL/min,with a predilution rate of 50 mL/min.Following serial plasma collection forgentamicin concentration determination,a population pharmacokinetic model wasused to describe gentamicin pharmacoki-netics, and a Monte Carlo simulation wasperformed with doses between 3 mg/kg of

body weight and 7 mg/kg (lean bodyweight; 70-kg patient normalized to 55kg). The Monte Carlo simulation wasused to determine the percentage ofdoses that would achieve the targetedpharmacodynamic parameter. The pa-tients had a mean age of 66 yrs, werepredominately male, weighed a mean92.5 kg, and had a mean Acute PhysiologyAssessment and Chronic Health Evalua-tion III score of 98. A two-compartmentpharmacokinetic model best describedgentamicin with a half-life of 13.8 hrsduring SLED and 153.4 hrs withoutSLED. The Monte Carlo analysis indi-cated a dose of 6 mg/kg (lean bodyweight) every 48 hrs as a 30-min infusion1 hr before SLED resulted in 100% oftargeted (70 120) AUC0 24/MIC ratios.An interval of 48 hrs was required toallow trough concentrations to approach1.5 mg/liter. The authors concluded thatgentamicin should be administered atdoses of 6 mg/kg lean body weight every48 hrs with appropriate therapeutic mon-itoring to guide and optimize subsequentdoses (32). Due to the similar pharmaco-kinetics for tobramycin, the same dosingrecommendations could be utilized. Al-though no studies have been conductedin SLED with amikacin, extrapolating thegentamicin data to amikacin, a dose of1520 mg/kg lean body weight every 48hrs could be considered.

Following the initial loading dose forpatients on continuous SLED, it may beappropriate to consider a scheduled ami-noglycoside regimen due to more exten-sive SLED clearance. Importantly, serumconcentration monitoring should guidedosing adjustments. In addition, obesepatients require an adjusted body weightbased dose as actual body weight maylead to overdosing (11, 32).

Vancomycin. Vancomycin is a largemolecule, but generally is well distrib-uted into most body tissues (3941). Thevolume of distribution is between 0.62L/kg and 0.8 L/kg, and is 55% proteinbound in patients without renal dysfunc-tion or low albumin concentrations. Inpatients with severe renal dysfunction,the amount of protein binding is reducedapproaching values of 18% to 20%. Van-comycin is primarily renally eliminated,with 80% removed through glomerularfiltration, which results in a mean half-life of 6 hrs in subjects with normal renalfunction compared with 168 hrs inanephric patients. Due to significant re-nal elimination, vancomycin requirescareful pharmacokinetic monitoring in

patients with renal dysfunction (40). Dos-ing recommendations in patients withnormal renal function typically rangefrom 15 to 20 mg/kg every 812 hrs de-pending on indication and body weight.Recent guideline recommendations sug-gest more vigilant monitoring of vanco-mycin concentrations may be required toprevent treatment failures (40, 41). Sev-eral factors affect the clearance of vanco-mycin including the hemodialysis filtertype and residual renal function (42).Vancomycin is poorly removed by low-flux membranes, whereas high-fluxmembranes significantly remove thedrug by 30% to 50% (11, 43). Most dosingregimens follow the principle of admin-istering a loading dose of 1520 mg/kgpostdialysis session on day 1 of therapy,followed by 500- to 1000-mg doses aftersubsequent dialysis sessions dependingon predialysis vancomycin concentra-tions (1, 43). For example, Pai and Pai(44) recommend holding the dose of van-comycin for a predialysis concentration20 g/mL, administering 500 mg forconcentrations between 520 g/mL, andadministering 1000 mg if the vancomycinconcentration falls below 5 g/mL.

CRRTs effectively remove vancomycin(30). Clearance by CRRT methods rangefrom 11.5 to 19.3 L/day (39). Similar toHD, a vancomycin-loading dose of 1520mg/kg is recommended. Vancomycinmaintenance dosing regimens may rangefrom 500 mg intravenously (IV) every 12hrs for CVVHDF to 1015 mg/kg IV every2448 hrs for CVVH and CVVHD (42, 45,46). Specifically, CVVHDF removes van-comycin to a greater degree than otherCRRT methods (45).

Two pharmacokinetic studies havebeen published evaluating vancomycinclearance during extended daily dialysis(Table 2) (47, 48). Kielstein et al (47)evaluated ten critically ill patients withacute renal failure receiving vancomycintherapy. Vancomycin (1 g) was adminis-tered 12 hrs before SLED was initiated toobtain pharmacokinetic parameters bothon and off SLED (40). SLED sessionslasted approximately 8 hrs and were per-formed with high-flux polysulfone dialyz-ers (F60S; surface area 1.3 m2) with bloodand dialysate flow rates of 160 mL/min.After administration of IV vancomycin,blood samples were drawn at several in-tervals throughout SLED. Dialysate sam-ples were also collected to evaluate fortotal drug removal. Half-life and volumeof distribution of vancomycin were 11.2hrs and 0.57 L/kg, respectively, during


SLED, in comparison to a half-life of 37.3hrs while off SLED. Drug removal by one8-hr SLED session was estimated to bebetween 8% and 26%. Concentrationsreached values of 10 g/mL at 30 hrsafter dose. The investigators of this studyrecommended administering a 20- to 25-mg/kg initial dose and obtaining druglevels no sooner than 12 hrs after dose toguide further therapy (47).

A second study evaluated the pharma-cokinetics of vancomycin during a 24-hrsession of SLED (48). SLED therapy wasperformed by using a low-flux filter (F4and F5; 0.8 and 1.0 m2) with blood anddialysate flow rates of 200 mL/min and100 mL/min, respectively. Patients re-ceived vancomycin 15 mg/kg, based onactual body weight, as a single dose eitherbefore SLED or during SLED. Blood lev-els were drawn at 6, 12, and 24 hrs afterdrug infusion. If levels at 24 hrs were 20g/mL, then vancomycin was redosed at15 mg/kg. The volume of distribution ofvancomycin in these patients rangedfrom 0.58 to 1.24 L/kg, with a mean of0.84 L/kg. The half-life of vancomycinranged from 18.8 to 96 hrs, with a meanof 43.1 hrs. The authors attributed thehalf-life variability to the volume of dis-tribution differences in these patients. Ofnote, the investigators did not find thatresidual renal function had any effect onthe half-life of vancomycin. The investi-gators recommended a vancomycin level24 hrs after infusion due to the variabilityin half-life and inability to predict whichpatients require more frequent dosing.

SLED sessions of 1224 hrs by usinghigh-flux membranes would potentiallyrequire redosing more frequently thanevery 24 hrs. We recommend that a ran-dom vancomycin level be obtained be-tween 12 hrs and 18 hrs after infusionduring longer SLED sessions utilizinghigh-flux membranes to ensure concen-trations do not fall below 1020 g/mL atany point during dialysis therapy.

Carbapenems. Carbapenems are onlyadministered IV and are widely distrib-uted into most bodily fluids, includingthe cerebral spinal fluid (49). Clearanceduring SLED has only been assessed forertapenem and meropenem. The individ-ual pharmacokinetics parameters of er-tapenem and meropenem are substan-tially different. Meropenem has limitedprotein binding of approximately 2%whereas ertapenem is highly proteinbound ranging from 90% to 95%. Thehalf-lives for meropenem and ertapenemare 1 hr and 4 hrs, respectively. Both

drugs exhibit extensive renal elimination.Meropenem is metabolized into a singleactive metabolite, with 70% of the drugexcreted unchanged into the urine, andertapenem is 80% renally excreted with38% excreted unchanged. Doripenemand imipenem share a similar pharmaco-kinetic profile with meropenem with re-spect to half-life, renal elimination andprotein binding. All carbapenems exceptertapenem require dosage adjustment forrenal impairment (Clcr 50 mL/min) andare extensively removed by IHD (11, 50).Despite the high protein binding of ertap-enem, 30% of the drug is cleared in pa-tients with end-stage renal disease receiv-ing hemodialysis necessitating asupplemental dose following dialysis (51).Following IHD, an imipenem dose of250 500 mg every 12 hrs and mero-penem and doripenem dose of 500 mgevery 24 hrs are recommended. AllCRRTs increase carbapenem clearancebeyond IHD with normal doses given forimipenem every 8 hrs and every 812 hrsfor doripenem and meropenem (11).

Meropenem clearance during SLEDhas been evaluated in a single pharmaco-kinetic study of ten patients (Table 2)(47). One gram of meropenem was in-fused over 30 mins, 6 hrs before the ini-tiation of SLED. A high-flux polysulfonefilter (F60S; surface area 1.3 m2) wasused with blood and dialysate flow ratesof 160 mL/min. Blood samples weredrawn at the start of infusion, 0.5, 1, 2, 4,and 6 hrs after the administration ofmeropenem; before SLED started; duringSLED at 2, 4, and 6 hrs; at the completionof the 8-hr SLED session, and 0.5, 1, 3,and 8 hrs after SLED completion. Thehalf-life of meropenem during SLED wassimilar to the half-life of the drug on IHDas compared with a longer half-life dur-ing CVVH. Although the results werevariable, up to 51% of meropenem wasremoved by SLED over the 8 hrs. Thus, amore aggressive dosing regimen of 500-1000 mg IV every 8 hrs is recommendedin patients receiving high-flux SLED.

Ertapenem was also evaluated in aSLED pharmacokinetic study (Table 2)(51). Six patients undergoing SLED for 8hrs with a high-flux polysulfone dialyzer(surface area 1.3 m2), with blood and di-alysate flow rates of 160 mL/min, re-ceived 1000 mg of ertapenem IV as asingle dose. Blood samples were obtainedat various time points before and afterertapenem administration. Due to highprotein binding, the investigators evalu-ated both total and free ertapenem con-

centrations. Ertapenem-free concentra-tions were then evaluated in relation toMIC90 data for several common pathogens.The investigators concluded that 1000 mgof ertapenem daily results in sufficient con-centrations above the MIC90 for most com-mon pathogens for the entire dosing inter-val. Based on the half-life and MIC dataprovided by Burkhardt et al (51), we con-clude that with high-flux SLED equal to or8 hrs duration, once daily dosing of er-tapenem should be sufficient to maintainadequate concentrations.

The clearances of doripenem and imi-penem have not been evaluated in patientsduring SLED. However, based on theirpharmacokinetic similarities to mero-penem, we suggest utilizing a more aggres-sive dosing strategy than currently recom-mended for IHD during high-flux SLED.Imipenem doses of 500 mg IV every 6 hrs inpatients 70 kg and doripenem doses of500 mg IV every 8 hrs may be required toensure adequate serum drug concentra-tions. Further data are needed to confirmthese recommendations.

Linezolid. Linezolid is a water-solublecompound with a volume of distributionapproximating total body water (4050 L)and is readily distributed to most tissues(5254). The half-life is approximately4.55.5 hrs with the primary route ofelimination being hepatic metabolism.Approximately 30% to 35% of the parentcompound is excreted unchanged intothe urine. Because of the water solubilityand low protein binding (31%) of lin-ezolid, significant clearance by dialysiswould be anticipated (5557).

The IHD-related pharmacokinetics oflinezolid and the two major metaboliteswere evaluated in patients with end-stagerenal disease and in patients admitted tothe ICU (56, 57). The pharmacokinetics of600 mg of linezolid IV were determined infive male critically ill patients with an AcutePhysiology Assessment and Chronic HealthEvaluation II score range of 23 to 29 withsepsis and renal failure (56). Approximatelyone third of the drug was removed; al-though there was substantial variability inthe starting serum concentrations from pa-tient to patient (56).

Linezolid clearance was evaluated in asingle study performed in 15 critically illpatients with oliguric renal failure receiv-ing SLED, CVVH, or IHD (Table 2) (57).Conventional IHD lasted 34 hrs with ablood flow rate of 300 mL/min and dialy-sate flow rate of 500 mL/min comparedwith SLED sessions of 89 hrs with bloodflow rates of 200 mL/min and dialysate


flow rates of 100 mL/min. A low-fluxpolysulfone filter (F7HPS; surface area1.6 m2) was used for both IHD and SLED.CVVH was performed in two patients,both in the predilution mode with a re-infusion rate of 35 mL/min and bloodflow of 150 mL/min at 10.5 hrs and 12 hrseach (58). Mean linezolid elimination was193.7 mg (32.3% of the dose adminis-tered) with IHD, 205 mg (33.9%) withSLED, and 105 mg (17.5%) followingCVVH. Because this was a single-dosestudy, the authors declined to make con-clusive recommendations, although thedata suggest linezolid clearance by 8- to9-hr SLED sessions approximates that ofIHD necessitating the re-administrationof drug. With IHD, no dosage adjustmentis recommended; however, IHD is usuallyonly administered three times weekly. IfSLED is utilized on a daily basis, an ad-ditional one third of the linezolid dosewould be removed for each 12 hrs ofdialysis and the serum concentration maydrop below breakpoint values for entero-cocci, streptococci, and Staphylococcusaureus. We recommend that for a singleSLED session of 812 hrs that no addi-tional linezolid dose adjustment is neces-sary. However, if repeated sessions orprolonged (12 hrs) periods of SLED areutilized, the clinician should consider anadditional administration of 30% of thelinezolid dose per each 12 hrs of SLEDtime.

Fluoroquinolones. The three mostcommonly used fluoroquinolones in thecritically ill include ciprofloxacin, levo-floxacin, and moxifloxacin (39, 59, 60).Although some similarities exist amongthe pharmacokinetic parameters of theseagents, there are enough differences thatdialysis clearance varies widely. In gen-eral, fluoroquinolones are lipophilicagents with low molecular weights andlarge volumes of distribution that areminimally affected in the critically ill (11,28). They also all exhibit excellent oralbioavailability. However, whereas levo-floxacin is primarily cleared renally, cip-rofloxacin is cleared by both renal andhepatic mechanisms, and moxifloxacin ishepatically eliminated (61). In addition,ciprofloxacin and levofloxacin exhibitlower protein binding than moxifloxacin,although all range from 20% to 50% (11).Therefore, based on the pharmacokineticproperties of these agents, one would ex-pect minimal dialysis clearance of moxi-floxacin, with somewhat greater clear-ance of ciprofloxacin and substantialclearance of levofloxacin.

IHD dosing for these agents is reflec-tive of their pharmacokinetics. Moxifloxa-cin requires no dose adjustment and maycontinue to be administered at 400 mgevery 24 hrs. Ciprofloxacin should bedosed at 400 mg IV every 24 hrs in thecritically ill to ensure attainment of tar-get AUC/MIC ratios (11). Levofloxacin,with its greater extent of renal clearanceand significantly prolonged half-life in re-nal disease (normal half-life 68 hrs, re-nal disease half-life 20 hrs), is dosed at250500 mg every 48 hrs, depending ondisease state and severity (11). All dosesshould be given after dialysis sessions.

CRRTs require the use of adjusteddoses for fluoroquinolones, except formoxifloxacin, which remains dosed at400 mg every 24 hrs. However, due to thedifferences in clearance mechanisms be-tween CRRTs dosing needs vary. CVVHdoses tend to be lower than those re-quired for CVVHDF, with CVVHD dosingin the middle, although due to the con-tinuous nature of these modalities, thedosing approaches that of normal renalfunction patients (11). Ciprofloxacin dos-ing for CVVH is 200400 mg IV every1224 hrs, with CVVHDF dosing at 400mg every 12 hrs. Levofloxacin dosing mayrange from 250 mg every 24 hrs for CVVHto 250750 mg every 24 hrs for CVVHDF(11). A loading dose of 500750 mg mayalso be given for levofloxacin to assist inearlier attainment of pharmacodynamictargets (11).

Fluoroquinolone dosing in SLED re-mains under investigation. To date, onesingle-dose study of critically ill anuricpatients receiving moxifloxacin or levo-floxacin and SLED has been completed(Table 2) (61). SLED sessions were com-pleted using a high-flux polysulfone filter(F60S; surface area 1.3 m2) over 8 hrswith mean blood and dialysate flow ratesof 161 4 mL/min. In one study arm, tenpatients received 400 mg of moxifloxacininfused over 60 mins, 8 hrs before SLEDwas initiated. Six of those ten patientshad liver disease with a Child-Pugh scoreof class C. The clearance of moxifloxacinwas estimated at 15.7 L/hr with a half-lifeof 12.3 hrs off SLED. Despite the lack ofrenal clearance for moxifloxacin, SLEDitself added 23.1 L/hr of clearance, re-sulting in a half-life of 6 hrs duringSLED. Even with the change in half-life,the pharmacokinetic parameters of moxi-floxacin in the critically ill anuric pa-tients appeared similar to those ofhealthy subjects and the presence of liverimpairment did not seem to impact the

drug clearance. Therefore, the authorsrecommended moxifloxacin at 400 mgevery 24 hrs for patients receiving 8-hrSLED sessions.

In the second study arm, five patientsreceived levofloxacin (250 or 500 mg) in-fused over 60 mins, 12 hrs before theSLED session was started (61). It shouldbe noted that two of these five patientshad received levofloxacin over 9 days be-fore enrollment and were at steady-state.Drug clearance was calculated both onand off SLED. The estimated clearance oflevofloxacin was 3.07 L/hr with a half-lifeof 34.5 hrs off SLED and a clearance of2.933.12 L/hr with a half-life of 10.3 hrsfor the 8-hr SLED session. The fraction ofdrug estimated to be removed by SLEDwas 17%27%, which is similar to thepercentage removed by IHD; however,the calculated half-life on SLED is similarto that in CVVH. The authors recom-mended adjusting the dose in SLED andadministering the drug after SLED butdid not provide specific dosing details(61). Based on the pharmacokinetic prop-erties observed in this study and informa-tion available on CVVH dosing, we wouldrecommend dosing levofloxacin at 250mg every 24 hrs. If SLED is utilized as acontinuous modality, the dose may needto be increased to account for additionalclearance. Although ciprofloxacin sharessome similar pharmacokinetic propertiesto other fluoroquinolones, there are in-sufficient data in SLED at this time todetermine effects on those parameters.Based on the study discussed above andon CVVH data, dosing ciprofloxacin sim-ilar to CVVHD (400 mg IV every 1224hrs) could be considered at this time un-til more information is known.

Other Anti-Infectives

Echinocandins. All three echinocan-dins, anidulafungin, caspofungin, andmicafungin, are highly protein bound(range 84%99%) and have small vol-umes of distributions (volume of distri-bution at steady state 0.15 0.6 L/kg) andrelatively long half-lives (926 hrs) (6165). Although limited data are available,no dosage adjustment during or afterIHD is required with any of the echino-candins, principally due to their highprotein binding (61 65). No prospectivetrials have been published that evaluatethe effect of CRRT dosing in patients re-ceiving anidulafungin, caspofungin, ormicafungin. Anidulafungin pharmacoki-netics have been evaluated in a single


patient undergoing SLED (Table 2) (66).A 63-yr-old male with Candida albicanscholecystitis associated with septic shockand acute renal failure was given 200 mgof IV anidulafungin over 30 mins followedby serial blood samples for analysis (66).SLED was performed over an 8-hr periodwith a dialysate flow of 180 mL/min and ablood flow rate of 180 mL/min with ahigh-flux polysulfone dialyzer (F60S; sur-face area 1.3 m2). SLED had no impact onthe plasma concentrations, which werecomparable to healthy adults, and nomeasurable drug amount was found inthe dialysate. The authors concludedSLED does not impact the pharmacoki-netics of anidulafungin; therefore no doseadjustment is necessary, and cliniciansshould follow product label dosage rec-ommendations for reduced renal func-tion. We recommend following this state-ment with all echinocandins untiladditional information is available.

Ideal Study Design

Prospective pharmacokinetic studiesin ICU patients can present a challengedue to the uncertainty of the disease

course, the multiple procedures a patientmay require that would interrupt dialysis,and difficulties obtaining consent forstudies in critically ill patients. The ma-jority of antibiotic studies in patients re-quiring SLED are single-dose pharmaco-kinetic trials that utilize various bloodand dialysate flows (25, 26, 32, 38, 47, 48,51, 58, 63, 66). Although these studiesprovide a foundation for dosing recom-mendations in clinical practice, there arelimitations to the applicability of thedata. Most patients in the ICU requiringSLED will be receiving more than onedose of an antibiotic agent and will re-ceive repeated SLED sessions. Therefore,a study that examines the pharmacoki-netics of multiple antibiotic doses bothon-SLED and off-SLED would allow moreaccurate assessment of the impact ofSLED on clearance. To facilitate this typeof study, multiple blood and dialysatesamples would be needed. If such a studywere to be conducted, samples of bloodand dialysate while the patient is onSLED, both before and after antibioticadministration, would provide adequatedata to assess drug removal during SLED.

It would be important to account for thepatients intrinsic clearance by includinga urine output parameter in the studyinclusion criteria. It would also be impor-tant to account for a minimum durationof SLED to ensure adequate time for drugremoval. Operational characteristics ofSLED (membrane type, surface area,method of CRRT, pre- and postfilter re-placement fluid, sieving, and saturationcoefficients) particularly blood and dialy-sate flow should also be standardized tobest assess clearance, because those pa-rameters will affect solute removal rates.Blood samples would also be required inthe 24 hrs following the cessation ofSLED to evaluate the re-equilibration ofserum concentrations once the patienthas returned to a drug-dosing regimenfor reduced renal function (e.g., Clcr 10mL/min). In addition, measured clear-ances of endogenous markers such asurea and/or -2 microglobulin should bereported. This type of study would pro-vide an improved approximation of theantibiotic dosing observed in clinicalpractice and may provide insight intodrug accumulation effects and re-equili-bration phenomena to allow for moreprecise dosing recommendations.

Institutional PracticeConsiderations

Although the study of antibiotic phar-macokinetics for patients receiving SLEDis increasing, the available literature re-mains limited. Thus, there are numerousantibiotics for which no data are availablenecessitating dosing empirically or basedon extrapolated data. At The NebraskaMedical Center, it is our practice to ad-just those antibiotics for which renal dos-ing is normally necessary when the pa-tient is on SLED, despite the lack ofpublished pharmacokinetic data (Table3). The standard blood flow and dialysateflow rates utilized for SLED at our insti-tution are 200 mL/min and 100 mL/min,respectively. Dosage regimens are deter-mined by using an estimated creatinineclearance of approximately 60 mL/min,which is based on data from SLED ureaclearance studies (5, 6, 13). This may be aconservative estimate for antibiotic clear-ance, but we feel that it allows our criti-cally ill patients to receive more aggres-sive dosing regimens and minimizes therisk of treatment failure from inadequateserum concentrations while also mitigat-ing the risk of achieving supra-therapeu-tic serum concentrations. Therefore, we

Table 3. Antimicrobial dosing recommendations during sustained low-efficiency dialysis

Drug Nebraska Medical Center Dosing Recommendationsa,b

Anidulafungin (66) No dosing adjustments necessaryDaptomycin (25, 26) Dose every 24 hrs on SLED

Dose every 48 hrs off SLEDGentamicin (32, 38) Initial loading dose 6 mg/kg lean body weight every 48 hrs as a 30-

min infusion. Give 1 hr before SLED. Doses should be adjustedaccording to serum concentrations observed during and after theSLED sessionc

Non-ICU patients, dose every 24 hrs on SLED at approximately 2-2.5mg/kg/dose (would recommend using ideal or adjusted body weight)

Dose per levels off SLEDTobramycinsame recommendationsAmikacin 15-20 mg/kg lean body weight every 48 hrs. Doses should be

adjusted according to serum concentrations observed during andafter the SLED session.

Ertapenem (51) No dose adjustment neededMeropenem, Meropenem 500-1000 mg IV every 8 hrs while on SLED

impenem (47) Imipenem 500 mg IV every 6 hrs while on SLED (if 70 kg)Vancomycin (47, 48) Dose every 12-18 hrs while on SLED

Dose per levels off SLEDLinezolid (58) No dose adjustment needed

May consider supplemental dose or every 8 hrs dosing if continuousSLED for some organisms

Moxifloxacin, Moxifloxacinno adjustment neededlevofloxacin (61) Levofloxacinadjustment may be necessary, no specific

recommendation

SLED, sustained low-efficiency dialysis; IV, intravenous; ICU, intensive care unit.aThese recommendations are used in our institution based on evaluation and interpretation of the

literature and known properties of SLED. This information is not intended to replace clinical judgmentor experience in individual patients; bdialysis parameters include blood and dialysate flow rates eachset at 200 mL/min and the use of a high-flux polysulfone filter; cdialysis parameters include blood anddialysate flow rates each set at 300 mL/min.


utilize an on SLED-/off SLED-dosing reg-imen in which we dose the antibioticswith an estimated Clcr of 60 mL/minwhile on SLED and a Clcr of 10 mL/minin anuric patients while the patient is offSLED. However, off SLED dosing must beindividualized if the patient is continuedon IHD or to the patients residual renalfunction if not receiving dialysis.

CONCLUSIONS

SLED poses a significant challengewith antibiotic dosing due to the ex-tended duration and lower blood and di-alysate flow rates. The potential for sig-nificant drug clearance and subsequentalterations in pharmacodynamics couldpotentially alter the treatment of bacte-rial and fungal infections. It is importantfor the clinician to have a clear under-standing of how SLED differs from otherforms of dialysis and its impact on drugclearance. Despite the potential for sig-nificant drug clearance, the dosing rec-ommendations made within this articleare based off very limited data, mostlysingle-dose pharmacokinetic studies inone patient, our extrapolation of the lit-erature, and known pharmacokinetic andpharmacodynamic parameters. Often,our recommendations are based off ourown experience; however, this shouldsupplement sound clinical decision mak-ing. More clinical investigations areneeded to support our extrapolations andrecommendations for dosing antibioticagents during SLED.

ACKNOWLEDGMENTS

We thank the SLED research team fordata collection and manuscript review:Christopher J. Bultsma, PharmD, BCPS,and Steven W. Nissen, PharmD, BCPS,and Anna Damgaard-Selden for adminis-trative support.

REFERENCES

1. Bellomo R, Roncom C, Kellum JA, et al:Acute renal failure-definition, outcome mea-sures, animal models, fluid therapy and in-formation technology needs: The Second In-ternational Consensus Conference of theAcute Dialysis Quality Initiative (ADQI)Group. Crit Care 2004; 8:R204R212

2. Lameire N, Van Biesen W, Vanholder R:Acute renal failure. Lancet 2005; 365:417430

3. Joy MS, Matzke GR, Armstrong DK, et al: Aprimer on continuous renal replacementtherapy for critically ill patients. Ann Phar-macother 1998; 32:362375

4. Kramer P, Wigger W, Reiger J, et al: [Arte-riovenous haemofiltration: A new and simplemethod for treatment of over-hydrated pa-tients resistant to diuretics]. Klin Wochen-schr 1977; 55:11211122

5. Marshall MR, Golper TA, Shaver MJ, et al:Sustained low-efficiency dialysis for criticallyill patients requiring renal replacement ther-apy. Kidney Int 2001; 60:777785

6. Salahudeen AK, Kumar V, Madan N, et al:Sustained low efficiency dialysis in the con-tinuous mode (C-SLED): dialysis efficacy,clinical outcomes, and survival predictors incritically ill cancer patients. Clin J Am SocNephrol 2009; 4:13381346

7. Veltri MA, Neu AM, Fivush BA, et al: Drugdosing during intermittent hemodialysis andcontinuous renal replacement therapy. Spe-cial considerations in pediatric patients. Pe-diatr Drugs 2004; 6:4565

8. Schetz M: Drug dosing in continuous renalreplacement therapy: General rules. CurrOpin Crit Care 2007; 13:645651

9. Mueller BA, Pasko DA, Sowinski KM: Higherrenal replacement therapy dose delivery in-fluences on drug therapy. Artif Organs 2003;27:808814

10. Choi G, Gomersall CD, Tian Q, et al: Princi-ples of antibacterial dosing in continuousrenal replacement therapy. Crit Care Med2009; 37:22682282

11. Heintz BH, Matzke GR, Dager WE: Antimi-crobial dosing concepts and recommenda-tions for critically ill adult patients receivingcontinuous renal replacement therapy or in-termittent hemodialysis. Pharmacotherapy2009; 29:562577

12. Mehta R: Supportive therapies: Intermittenthemodialysis, continuous renal replacementtherapies, and peritoneal dialysis. In: Atlas ofDisease of the Kidney. Schrier R (Ed). Phil-adelphia, PA, Current Medicine, 1998, pp116

13. Huang Z, Letteri JJ, Clark WR, et al: Opera-tional characteristics of continuous renal re-placement modalities used for critically illpatients with acute kidney injury. Int J ArtifOrgans 2008; 31:525534

14. Davenport A: Renal replacement therapy inacute kidney injury: Which method to use inthe intensive care unit? Saudi J Kidney DisTranspl 2008; 19:529536

15. Marshall MR, Golper TA, Shaver MJ, et al:Urea kinetics during sustained low-efficiencydialysis in critically ill patients requiring re-nal replacement therapy. Am J Kidney Dis2002; 39:556570

16. Kumar VA, Craig M, Depner TA, et al: Ex-tended daily dialysis: A new approach to renalreplacement for acute renal failure in theintensive care unit. Am J Kidney Dis 2000;36:294300

17. Marshall MR, Golper TA: Sustained low effi-ciency or extended daily dialysis. In: UpTo-Date, Post TW (Ed). Waltham, MA, UpTo-Date, 2010

18. Marshall MR, Ma T, Galler D, et al: Sustainedlow-efficiency daily dialfiltration (SLEDD-f)

for critically ill patients requiring renal re-placement therapy: Towards an adequatetherapy. Nephrol Dial Transplant 2004; 19:877884

19. Scott MK, Mueller BA, Clark WR: Vancomy-cin mass transfer characteristics of high-fluxcellulosic dialysers. Nephrol Dial Transplant1997; 12:24672653

20. Schaedeli F, Uehlinger DE: Urea kinetics anddialysis treatment time predict vancomycinelimination during high-flux hemodialysis.Clin Pharmacol Ther 1998; 63:2638

21. Mushatt DM, Mihm LB, Dreisbach AW, et al:Antibiotic dosing in slow extended dialysis.Clin Infect Dis 2009; 49:433437

22. Rybak MJ: The efficacy and safety of dapto-mycin: First in a new class of antibiotics forGram-positive bacteria. Clin Microbiol Infect2006; 12(Suppl 1):S24S32

23. Dvorchik B, Sica D, Gehr T: Pharmacokinet-ics and safety of single-dose daptomycin insubjects with graded renal insufficiency andend-stage renal disease. 42nd InterscienceConference on Antimicrobial Agents andChemotherapy. San Diego, CA, 2002, Ab-stract A-1387

24. Wagner CC, Steiner I, Zeitlinger M: Dapto-mycin elimination by CVVH in vitro: Evalu-ation of factors influencing sieving andmembrane adsorption. Int J Clin PharmacolTher 2009; 47:178186

25. Burkhardt O, Joukhadar C, Traunmuller F,et al: Elimination of daptomycin in a patientwith acute renal failure undergoing extendeddaily dialysis. J Antimicrob Chemother 2007;60:224225

26. Kielstein JT, Eugbers C, Bode-Boeger SM, etal: Dosing of daptomycin in intensive careunit patients with acute kidney injury under-going extended dialysisa pharmacokineticstudy. Nephrol Dial Transplant 2010; 25:15371541

27. Dvorchik B, Arbeit RD, Chung J, et al: Pop-ulation pharmacokinetics of daptomycin. An-timicrob Agents Chemother 2004; 48:27992807

28. Roberts JA, Lipman J: Pharmacokinetic is-sues for antibiotics in the critically ill pa-tient. Crit Care Med 2009; 37:840851; quiz859

29. Drusano GL, Ambrose PG, Bhavnani SM, etal: Back to the future: Using aminoglycosidesagain and how to dose them optimally. ClinInfect Dis 2007; 45:753760

30. Trotman RL, Williamson JC, Shoemaker DM,et al: Antibiotic dosing in critically ill adultpatients receiving continuous renal replace-ment therapy. Clin Infect Dis 2005; 41:11591166

31. Begg EJ, Barclay ML, Duffull SB: A suggestedapproach to once-daily aminoglycoside dos-ing. Br J Clin Pharmacol 1995; 39:605609

32. Roberts JA, Field J, Visser A, et al: Usingpopulation pharmacokinetics to determinegentamicin dosing during extended daily dia-filtration in critically ill patients with acutekidney injury. Antimicrob Agents Chemother2010; 54:36353640


33. Rea RS, Capitano B, Bies R, et al: Suboptimalaminoglycoside dosing in critically ill pa-tients. Ther Drug Monit 2008; 30:674681

34. Dasta JF, Armstrong DK: Variability in ami-noglycoside pharmacokinetics in critically illsurgical patients. Crit Care Med 1988; 16:327330

35. Triginer C, Izquierdo I, Fernandez R, et al:Gentamicin volume of distribution in criti-cally ill septic patients. Intensive Care Med1990; 16:303306

36. Gustavo L, Castaneda-Hernandez G: Amika-cin Bayesian forecasting in critically ill pa-tients with sepsis and cirrhosis. Ther DrugMonit 1997; 19:271276

37. Marik PE: Aminoglycoside volume of distri-bution and illness severity in critically illseptic patients. Anesth Intensive Care 1993;21:172173

38. Manley HJ, Bailie GR, McClaran ML, et al:Gentamicin pharmacokinetics during slowdaily home hemodialysis. Kidney Int 2003;63:10721078

39. Clinical Pharmacology [database online]. Tampa,FL: Gold Standard, Inc.; updated January, 2011.Available at: http://www.clinicalpharmacology.com. Accessed September 5, 2010

40. Rybak M, Lomaestro B, Rotschafer JC, et al:Therapeutic monitoring of vancomycin inadult patients: A consensus review of theAmerican Society of Health-System Pharma-cists, the Infectious Diseases Society ofAmerica, and the Society of Infectious Dis-eases Pharmacists. Am J Health Syst Pharm2009; 66:8298

41. Rybak MJ: The pharmacokinetic and phar-macodynamic properties of vancomycin. ClinInfect Dis 2006; 42:S35S39

42. Pallotta KE, Manley HJ: Vancomycin use inpatients requiring hemodialysis: A literaturereview. Semin Dial 2008; 21:6370

43. Ariano RE, Fine A, Sitar DS, et al: Adequacyof a vancomycin dosing regimen in patientsreceiving high-flux hemodialysis. Am J Kid-ney Dis 2005; 46:681687

44. Pai AB, Pai MP: Vancomycin dosing in highflux hemodialysis: A limited-sampling algo-rithm. Am J Health Syst Pharm 2004; 61:18121816

45. Reference deleted

46. DelDot ME, Lipman J, Tett SE: Vancomycinpharmacokinetics in critically ill patients re-ceiving continuous venovenous hemodiafil-tration. Br J Clin Pharmacol 2004; 58:259268

47. Kielstein JT, Czock D, Schopke T, et al: Phar-macokinetics and total elimination of mero-penem and vancomycin in intensive careunit patients undergoing extended daily di-alysis. Crit Care Med 2006; 34:5156

48. Ahern JW, Lai C, Rebuck JA, et al: Experiencewith vancomycin in patients receiving slowlow-efficiency dialysis. Hosp Pharm 2004; 39:138143

49. Zhanel GG, Wiebe R, Dilay L, et al: Compar-ative review of the carbapenems. Drugs 2007;67:10271052

50. Mistry GC, Majumdar AK, Swan S, et al:Pharmacokinetics of ertapenem in patientswith varying degrees of renal insufficiencyand in patients on hemodialysis. J Clin Phar-macol 2006; 46:11281138

51. Burkhardt O, Hafer C, Langhoff A, et al:Pharmacokinetics of ertapenem in criticallyill patients with acute renal failure undergo-ing extended daily dialysis. Nephrol DialTransplant 2009; 24:267271

52. Adembri C, Fallani S, Cassetta MI, et al: Li-nezolid pharmacokinetics/pharmacodynamicprofile in critically ill septic patients: Inter-mittent versus continuous infusion. Int J An-timicrob Agents 2008; 31:122129

53. Moellering RC: Linezolid: The first oxazolidi-none antimicrobial. Ann Intern Med 2003;138:135142

54. Bain KT, Wittbrodt ET: Linezolid for thetreatment of resistant gram-positive cocci.Ann Pharmacother 2001; 35:566575

55. Brier ME, Stalker DJ, Aronoff GR, et al: Phar-macokinetics of linezolid in subjects withrenal dysfunction. Antimicrob Agents Che-mother 2003; 47:27752780

56. Fiaccadori E, Maggiore U, Rotelli C, et al:Does haemodialysis significantly affect se-rum linezolid concentrations in critically illpatients with renal failure? A pilot investiga-tion. Nephrol Dial Transplant 2006; 21:14021406

57. Mauro LS, Peloquin CA, Schmude K, et al:Clearance of linezolid via continuous venous

hemodiafiltration. Am J Kidney Dis 2006;47:e83e86

58. Fiaccadori E, Maggiore U, Rotelli C, et al:Removal of linezolid by conventional inter-mittent hemodialysis, sustained low-effi-ciency dialysis, or continuous venovenoushemofiltration in patients with acute renalfailure. Crit Care Med 2004; 32:24372442

59. Wright DH, Brown GH, Peterson ML, et al:Application of fluoroquinolone pharmacody-namics. J Antimicrob Chemother 2000; 46:669683

60. Olsen KM, Gentry-Nielsen M, Yue M: Effectof ethanol on fluoroquinolone efficacy in arat model of pneumococcal pneumonia. An-timicrob Agents Chemother 2006; 50:210219

61. Czock D, Husig-Linde C, Langhoff A, et al:Pharmacokinetics of moxifloxacin and levo-floxacin in intensive care unit patients whohave acute renal failure and undergo ex-tended daily dialysis. Clin J Am Soc Nephrol2006; 1:12631268

62. Morris MI, Villmann M: Echinocandins in themanagement of invasive fungal infections,part 1. Am J Health Syst Pharm 2006; 63:16931703

63. Merck: Caspofungin [package insert]. White-house Station, NJ, Merck, February 2005

64. Astellas Pharma US: Micafungin [package in-sert]. Deerfield, IL, Astellas Pharma US, April2005

65. Pfizer: Eraxis [package insert]. New York,NY, Pfizer, February 2006

66. Burkhardt O, Kaever V, Burhenne H, et al:Extended daily dialysis does not affect thepharmacokinetics of anidulafungin. Int J An-timicrob Chemother 2009; 34:282283

67. Awdishu L: Drug issues in renal replacementtherapies. In: Pharmacotherapy Self-Assess-ment Program. Sixth Edition. Kansas City,MO, American College of Clinical Pharmacy,2007

68. Khan E, Huggan P, Celi L, et al: Sustainedlow-efficiency dialysis with filtration(SLEDD-f) in the management of acute so-dium valproate intoxication. Hemodial Int2008; 12:211214


http://www.clinicalpharmacology.comhttp://www.clinicalpharmacology.com

articulo antibiotico 2011

Documents