You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/49857460

Antibiotic dosing in critically ill patients with acute kidney injury

Article  in  Nature Reviews Nephrology · February 2011


DOI: 10.1038/nrneph.2011.12 · Source: PubMed

CITATIONS READS

79 3,557

2 authors:

Rachel F Eyler Bruce A Mueller


University of Connecticut University of Michigan
15 PUBLICATIONS   274 CITATIONS    182 PUBLICATIONS   4,226 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Pharmacogenetics View project

All content following this page was uploaded by Bruce A Mueller on 14 April 2016.

The user has requested enhancement of the downloaded file.


REVIEWS

Antibiotic dosing in critically ill patients


with acute kidney injury
Rachel F. Eyler and Bruce A. Mueller
Abstract | A common cause of acute kidney injury (AKI) is sepsis, which makes appropriate dosing of
antibiotics in these patients essential. Drug dosing in critically ill patients with AKI, however, can be
complicated. Critical illness and AKI can both substantially alter pharmacokinetic parameters as compared
with healthy individuals or patients with end-stage renal disease. Furthermore, drug pharmacokinetic
parameters are highly variable within the critically ill population. The volume of distribution of hydrophilic
agents can increase as a result of fluid overload and decreased binding of the drug to serum proteins, and
antibiotic loading doses must be adjusted upwards to account for these changes. Although renal elimination of
drugs is decreased in patients with AKI, residual renal function in conjunction with renal replacement therapies
(RRTs) result in enhanced drug clearance, and maintenance doses must reflect this situation. Antibiotic dosing
decisions should be individualized to take into account patient-related, RRT-related, and drug-related factors.
Efforts must also be made to optimize the attainment of antibiotic pharmacodynamic goals in this population.
Eyler, R. F. & Mueller, B. A. Nat. Rev. Nephrol. 7, 226–235 (2011); published online 22 February 2011; doi:10.1038/nrneph.2011.12

Introduction
Continuing Medical Education online
Sepsis is a common cause of acute kidney injury (AKI);1
This activity has been planned and implemented in accordance consequently, the proper dosing of antibiotics in these
with the Essential Areas and policies of the Accreditation Council patients is crucial. Drug dosing in critically ill patients
for Continuing Medical Education through the joint sponsorship of
Medscape, LLC and Nature Publishing Group. Medscape, LLC is
with AKI, however, can be complex as a patient’s renal
accredited by the ACCME to provide continuing medical education function is dynamic and difficult to quantify, their
for physicians. volume status also fluctuates, and drug doses need to be
Medscape, LLC designates this Journal-based CME for a maximum frequently reassessed. Pharmacokinetic studies to guide
of 1 AMA PRA Category 1 CreditsTM. Physicians should claim only the clinician through the complexities of drug dosing in
the credit commensurate with their participation in the activity.
patients with AKI have only been conducted for a limited
All other clinicians completing this activity will be issued a
number of antibiotics.2,3 Furthermore, the results of these
certificate of participation. To participate in this journal CME
activity: (1) review the learning objectives and author disclosures; trials might only be applicable to institutions with similar
(2) study the education content; (3) take the post-test and/or populations of patients and those using comparable renal
complete the evaluation at www.medscape.org/journal/nrneph; replacement therapy (RRT) techniques.
(4) view/print certificate.
Antibiotic dosing decisions should take into account
Released: 22 February 2011; Expires: 22 February 2012
the individual patient’s characteristics, the choice of RRT,
Learning objectives and drug-related factors. Critically ill patients with AKI
Upon completion of this activity, participants should be able to: exhibit altered pharmacokinetic parameters in response
1 Analyze alterations in pharmacokinetics among critically ill to antibiotic therapy, and interpatient variability is high.4
patients.
2 Distinguish the most important variable of renal
RRTs remove drugs from the circulation, and the extent
replacement therapy in altering antibiotic concentrations. of removal is dependent not only on drug-related prop-
3 Evaluate effective strategies for antibiotic prescribing among erties, but also on the RRT technique employed. For
critically ill patients with AKI. each individual patient, antibiotic drug regimens must
4 Specify how to prescribe certain antibiotics among critically
ill patients.
be adjusted to optimize pharmacodynamics and maxi-
Department of Clinical, mize treatment efficacy. In this Review, we describe
Social, and
Administrative the character­istics that should guide initial antibiotic
Sciences, College of dosing decisions in critically ill patients with AKI. We
Pharmacy, University of
Michigan, 428 Church
also discuss the need for dose adjustment to ensure that
Street, Ann Arbor, Competing interests adequate serum concentrations of antibiotics are consis-
MI 48109-1065, USA tently achieved despite changes in the patient’s clinical
R. F. Eyler declares associations with the following companies:
(R. F. Eyler,
B. A. Mueller).
Merck, Roche. B. A. Mueller declares associations with the status, particularly when difficult-to-treat pathogens are
following companies: Amgen, Cubist Pharmaceuticals, Gambro,
implicated. We focus on the role of evidence-based anti-
Correspondence to: Merck, Roche. See the article online for full details of the
B. A. Mueller relationships. The locum journal Chief Editor R. Ireland and the biotic dosing in attaining pharmacodynamic targets in
muellerb@umich.edu CME questions author C. P. Vega declare no competing interests. patients requiring RRT for AKI.

226  |  APRIL 2011  |  VOLUME 7  www.nature.com/nrneph


© 2011 Macmillan Publishers Limited. All rights reserved
FOCUS ON AKI IN CRITICAL CARE

Pharmacokinetic alterations Key points


Critically ill patients with AKI often exhibit different ■■ Altered drug pharmacokinetics in critically ill patients with acute kidney injury
pharmaco­kinetic profiles in response to drug treatment (AKI) and heterogeneous renal replacement therapy (RRT) techniques in
compared with healthy individuals or patients with end- intensive care units preclude standardized antibiotic dosing
stage renal disease (ESRD).2,4 The AKI popula­tion is hetero­ ■■ Most critically ill patients with AKI exhibit altered antibiotic pharmacokinetics
geneous, and, consequently, pharmaco­kinetic parameters that necessitate increased doses in spite of decreased renal clearance,
can be quite variable. Generally, the pharmaco­kinetic particularly when serious infections are implicated
changes can be categorized as alterations in absorption, ■■ Drug dosing decisions must take into account pharmacodynamic as well as
distribution, metabolism, and elimination. pharmacokinetic considerations
■■ Clinicians should compare their RRT protocols to those in published guidelines
Altered absorption and ensure that their recommendations are applicable to the individual
Oral drug absorption may be altered by gastro­intestinal patient’s clinical situation
dysmotility in critically ill patients. Dysmotility can ■■ Hybrid RRTs require the same antibiotic dosing alterations as do continuous
develop as a result of decreased perfusion of the gut, RRTs, but for hybrid therapies the dose timing must also be considered
particularly if the patient is taking vasoconstrictive
drugs.4,5 To ensure adequate drug exposure, intravenous
anti­biotics should be used whenever possible, although Table 1 | Volume of distribution data from pharmacokinetic studies in adults
medica­tions are sometimes switched from intravenous to Antibiotic Healthy volunteers (l/kg) Patients with AKI
oral administra­tion for economic reasons. Interestingly, receiving RRT (l/kg)
intestinal absorption of certain drugs is increased during Lipophilic drugs
renal failure; findings from animal models indicate that Ciprofloxacin 1.9876 1.60,77 1.6578
this effect might relate to an accumulation of uremic mol- Levofloxacin 0.96, 79
1.1380
1.02,81 1.5182
ecules, which cause deterioration of the integrity of the
Hydrophilic drugs
intestinal mucosa.6 For example, the bioavailability of a
37.5 mg/kg intraintestinal dose of propranolol increased Amikacin 0.1883 0.4484

from 54.7% in healthy control rats to 81.4% in rats with Daptomycin 0.10 85
0.2326
cisplatin-­induced acute renal failure.6 Although interesting, Meropenem 0.17,86 0.18,87 0.2788 0.26,89 0.35,28 0.3729
this concept has yet to be studied in humans. Piperacillin 0.15 90
0.14,91 0.1892
The enteral feeding status of a patient is tied to some
Vancomycin 0.39,93 0.59,94 0.6395 0.57,96 0.6597
important considerations that affect oral drug absorp-
Abbreviations: AKI, acute kidney injury; RRT, renal replacement therapy.
tion. A critically ill individual who does not receive
enteral feeding will develop intestinal atrophy in as
little as 3 days, 7 and their gut mass may decrease by increased capillary permeability, which cause displace-
50% within 7 days.5 These intestinal changes could alter ment of fluids from the vasculature into the interstitium.4
drug absorption, although this hypothesis has not been Consequently, the volume of distribution of hydrophilic
fully investigated. Enteral feeding improves blood flow drugs in critically ill patients with AKI can differ sub-
to the digestive tract, and should be used wherever fea- stantially from that reported in pharmacokinetic studies
sible to maintain the integrity and viability of the gut.5 of healthy individuals (Table 1).
Enteral feeding itself, however, can affect drug absorp- These fluid shifts into the interstitial space can be
tion. Commercial enteral feeds decrease the absorption substantial, owing to the large fluid volumes frequently
of many fluoroquinolone and tetracycline anti­biotics.8–10 given to patients with sepsis as part of resuscitation,
For example, in a study of 13 healthy volunteers, medication, and delivery of nutrition. Extravascular fluid
coadministra­tion of one such enteral feed decreased the gains can be even more marked in patients who have a
average bioavailability of a 750 mg oral dose of cipro­ decreased urine output secondary to AKI. In a review
floxacin by 72%.10 Critically ill patients also often receive of data on 81 critically ill adults receiving continuous
H2-receptor antagonists or proton pump inhibitors, RRTs, 38 patients (47%) had ≥10% fluid-related gains
which raise the pH of the stomach and can interfere with of body weight and 13 patients (16%) gained ≥20% of their
the absorption of weak bases that require a strongly acidic body weight at the time of starting RRT.12 These fluid-
environment for absorption.4 When 200 mg of itracon- related weight gains correlated closely with increased
azole (a weak base) was orally administered to 12 healthy mortality. Fluid overload can lead to hypoxia and the
patients with famotidine-induced hypochlorhydria, peak need for mechanical ventilation, as well as impaired
drug concentrations were decreased by 52.9% compared cardiac function.13 Furthermore, an increase in fluid
with values obtained in the absence of famotidine.11 volume can dilute serum creatinine concentration, result-
ing in inappropriately low measurements, and thus delay
Altered distribution the diagnosis of AKI and initiation of RRT.12,14
The volume of distribution is an estimate of the extent to Another possible contributor to the poor outcomes
which a drug will migrate into extravascular tissues, and associated with fluid overload could be under­treatment
is one of the most striking sources of pharmaco­kinetic with antibiotics. Just as the capillary leakage associ-
variability in critically ill patients with AKI. Sepsis can ated with sepsis can increase the volume of distribu-
lead to the development of endothelial damage and tion of creati­nine,15 it can also increase the volume of

NATURE REVIEWS | NEPHROLOGY VOLUME 7  |  APRIL 2011  |  227


© 2011 Macmillan Publishers Limited. All rights reserved
REVIEWS

distribution of hydrophilic antibiotics, such as amino­ dose to all patients.21 This consideration is especially true
glycosides, β-lactams, and glycopeptides.16 This increased for loading doses, which must be able to fill the volume
volume of distribution could lead to subtherapeutic of distribution.
plasma concentrations of these agents. The volume of Pharmacokinetic studies in obese individuals often
distribution of gentamicin is 0.48 l/kg in patients with focus on identifying a surrogate marker (such as total,
hyperdynamic sepsis, compared with 0.29 l/kg in a adjusted, or lean body weight) that will most accurately
control group of nonseptic postoperative patients.17 predict the volumes of distribution seen in the study
Factors associated with an increased volume of distribu- popula­tion. The surrogate marker of choice seems to
tion were identified as increased severity of illness, as cal- be total weight for vancomycin and daptomycin, and
culated by the Acute Physiology Score, and a high cardiac adjusted body weight for aminoglycosides.20 When pos-
index. Another study of gentamicin pharmacokinetics sible, drug monitoring should be used to guide subse-
in 14 critically ill patients with AKI receiving extended quent doses.20 For drugs that do not have weight-based
daily dialysis (mean APACHE [Acute Physiology and dosing recommendations, obese patients should usually
Chronic Health Evaluation] III score of 98) reported an receive doses toward the top end of the dosing range.
even larger volume of distribution of 0.55 l/kg.18 These Finally, critically ill patients with AKI often have
increased volumes of distribution must be ‘filled’ with increased antibiotic volumes of distribution owing to
drug if therapeutic serum concentrations are to be decreased binding of the drug to serum proteins. This
achieved. Consequently, initial gentamicin doses must reduction in protein binding could be due to decreased
be adjusted upwards to account for the increased volumes serum albumin synthesis or increased extracellular shifts
of distribution associated with sepsis, fluid overload, and of serum proteins.22 In addition, studies conducted in
AKI. The results of these studies17,18 indicate that, for patients with ESRD have indicated that decreased protein
some critically ill patients with AKI and fluid overload, binding could result from the accumulation of certain
gentamicin loading doses might need to be doubled to uremic molecules that can bind to these proteins and
attain therapeutic serum concentrations simply because displace the drug from its binding site.23 In patients with
of their altered volume of distribution. advanced stages of AKI, uremic molecules could also
As the patient’s clinical status improves and the degree contribute to decreased protein binding. The extent of
of tissue edema decreases, the volume of distribution of the decrease in protein binding in critically ill patients
gentamicin can decline dramatically within days, with AKI is difficult to quantify, however, as therapeutic
and drug dosages should be adjusted to account for drug monitoring in the hospital setting usually involves
this change. Triginer et al.19 measured the gentamicin measurement of total rather than free drug concentra-
volume of distribution in 40 critically ill patients with tions. Furthermore, the amount of protein-bound drug
suspected or confirmed sepsis owing to a Gram-negative may be dependent on its concentration in serum, and not
bacterial infection on day 2 of therapy (after aggressive remain constant throughout the dosing interval.24
fluid resuscitation), and then again on day 7 of therapy.
The volume of distribution decreased from 0.43 l/kg to Altered metabolism
0.29 l/kg (P <0.001) and the required gentamicin dosage Although they have rarely been studied directly, altera-
decreased from 5.14 mg/kg per 24 h to 3.98 mg/kg per 24 h tions in the metabolism and nonrenal clearance of
(P <0.001), despite the patients’ renal function remain- antibiotics have been observed in patients with AKI.
ing fairly stable.19 These ongoing changes should also be The metabolism of drugs that are largely extracted by the
expected in patients with AKI: not only will a patient’s liver, such as β‑blockers and midazolam, is highly influ-
capillary leakage status change, but after the initia­ enced by hepatic blood flow.4 In patients with advanced
tion of RRT, fluid overload will be gradually corrected, sepsis, the clearance of these drugs can be reduced as
and drug doses need to be reduced accordingly. hepatic blood flow decreases. In patients with hyper-
The increased incidence of obesity has presented dynamic sepsis, however, hepatic metabolism may be
another noteworthy source of pharmacokinetic vari­ preserved as blood is shunted to the liver and other vital
ability, because volumes of distribution can be vastly organs.4 Vasoconstrictive drugs, such as phenyl­ephrine,
different for a patient weighing 70 kg compared with norepinephrine, epinephrine, and dopamine, can also
a patient who is twice that weight. In obese patients decrease hepatic blood flow, although the extent of this
(BMI ≥30 kg/m2), lipophilic antibiotics (such as fluoro­ effect varies according to the specific agent used. 4,25
quinolones) tend to have a larger total volume of distri- Metabolic enzyme activity in the liver also seems to be
bution, although when the totals are corrected to take decreased in patients with AKI.26 Although some animal
body weight into account, the volume of distribution per models of AKI demonstrate variability in the expression
kg is lower than that in patients who are not obese.20 As of a number of cytochrome P450 isoenzymes, human
plasma volume correlates with body weight, obese data are currently lacking.26 Several drugs that are non-
patients usually also have higher total volumes of dis- renally eliminated have recognized alterations in their
tribution for hydrophilic drugs, although the increases hepatic clearance, including imipenem, meropenem, and
might be less pronounced than those seen with lipophilic vancomycin. In anuric patients with AKI, the total clear-
drugs.21 To account for the pharmacodynamic changes ance rate of imipenem is 90–95 ml/min—considerably
associated with obesity, drugs should always be adjusted lower than that reported in patients with normal renal
for body weight, rather than administering a standard function (130 ml/min), but higher than that reported in

228  |  APRIL 2011  |  VOLUME 7  www.nature.com/nrneph


© 2011 Macmillan Publishers Limited. All rights reserved
FOCUS ON AKI IN CRITICAL CARE

patients with ESRD (50 ml/min). These results indicate the percent­age of unbound drug can be used as a rough
that hepatic metabolism of imipenem is relatively pre- estimate of the RRT extraction coefficient.42 However,
served in patients with AKI compared with those with when calculating the extraction coefficient, clinicians
ESRD.26,27 Similarly, the rate of clearance of meropenem should use protein binding values reported in critically ill
is greater in patients with AKI than in those with ESRD indivi­duals rather than those from healthy persons.
(40–60 ml/min versus 30–35 ml/min). 26,28,29 The same RRT techniques vary widely,43 and each variation can
pattern of comparatively preserved nonrenal clearance also affect drug clearance. Diffusive therapies (such as
in patients with AKI is found with vancomycin; the clear- intermittent hemodialysis and continuous venovenous
ance rate of this drug in patients with normal renal func- hemodialysis) and convective therapies (such as continu-
tion is 40 ml/min, compared with 15 ml/min in patients ous venovenous hemofiltration) both efficiently remove
with AKI, and 5 ml/min in patients with ESRD. However, small solutes, but convective therapies are superior with
as AKI persists, clearance of vancomycin may decline regard to the removal of solutes with large molecular
to the rate seen in ESRD.26,30 Nonhepatic elimination weights.44,45 Furthermore, hemofilters themselves vary
pathways can also account for increased drug clearance. by material, surface area, pore size, water permeability,
Ciprofloxacin undergoes transintestinal excretion, which, and the efficiency of removal of low-molecular-weight and
in patients with renal failure, may represent a compensa- mid-molecular-weight solutes. High-flux hemofilters
tory clearance mechanism that prevents accumulation have increased permeability to mid-molecular-weight
of this drug.31,32 molecules and remove considerably more drug than
low-flux filters do.46–48 Almost all currently used con-
Altered renal elimination tinuous RRTs utilize high-flux hemofilters; however, the
Sepsis-induced AKI is not only associated with decreased hybrid RRTs used to treat AKI (which typically involve
glomerular filtration by the kidney, but also with impair- slow, low-efficiency daily dialysis or extended daily
ment of tubular secretion and reabsorption.33–35 These dialysis) sometimes utilize low-flux hemofilters.49,50
changes can influence drug dosing in ways that are not When low-flux filters are used, drug clearances may
always recognized. β‑lactams, for example, are thought to differ (sometimes dramatically) from those achieved
be excreted by the organic anionic transporter type 1 and, with high-flux membranes. An additional issue related
at least in theory, dosing methods that take into account to antibiotic clearance is drug adsorption to the filter.
the decrease in glomerular filtration but not decreased Polyacrylonitrile filters are particularly associated with
transport could produce a larger than anticipated drug this phenomenon, although the extent of adsorption is
exposure.35 The reabsorption of fluconazole, which can difficult to quantify.51,52
be substantial in patients with normal renal function, Another technical consideration that affects drug
is decreased in critically ill patients with AKI. In many clearance occurs with continuous RRTs that utilize a high
cases, these patients might require unadjusted doses of convective component, such as continuous venovenous
fluconazole, or even higher doses than are required for hemofiltration. The point at which fluid replacement
patients with normal renal function.36,37 occurs in these techniques alters drug clearance, because
A critically ill patient with AKI usually has some resid- dilution before filtration decreases the concentration of
ual renal function, which may change dynamically along drug available for removal by the filter.53,54 The differ-
with the patient’s clinical status. For some renally elimi- ences between predilution and postdilution modes of
nated drugs, patients with residual renal function will fluid replacement are predictable, and can be easily cal-
require higher antibiotic doses than their anuric counter­ culated from the rates of blood flow, fluid replacement,
parts.38 The patient’s urine output should, therefore, be and clearance.2
monitored closely, and drug doses adjusted upwards as In general, the most important RRT-related factor
renal function returns. If a patient is taking renally elimi- affecting drug removal is effluent volume, which is deter-
nated antibiotics that permit therapeutic drug monitor- mined by both flow rate and therapy duration.2 Effluent
ing, such as aminoglycosides or vanco­mycin, an increase rates vary between institutions, and contemporary efflu-
in the clearance of these drugs can also signal the return ent flow rates are often much higher than those used even
of renal function and the possible need for upward 10 years ago.55 Consequently, drug-dosing recommenda­
adjustment of other renally eliminated drugs for which tions for use with continuous RRT published in the past
monitoring might not be available. might not apply today. Clinicians prescribing RRTs at
high effluent rates should use higher antibiotic doses
Elimination by RRT than those recommended in guidelines developed for
In general, drugs that are primarily renally eliminated, low flow rates. For example, the 2007 edition of Drug
with a small molecular weight,39 small volume of dis- Prescribing in Renal Failure assumes a continuous RRT
tribution (<1 l/kg), and low degree of protein binding, effluent rate of 2 l/h.56 Thus, these dosing recommenda-
are likely to be removed by RRTs.40 Although decreased tions may not be applicable if the patient is being treated
protein binding (which is often exhibited by critically with substantially lower or higher effluent rates.
ill individuals) increases the volume of distribution
of a drug, elimination is also increased, because more Pharmacodynamic considerations
free drug is available to be removed by the kidney or Successful antibiotic treatment of a patient with AKI
RRTs. Although exceptions exist to this general rule,41 requires not only consideration of pharmacokinetic

NATURE REVIEWS | NEPHROLOGY VOLUME 7  |  APRIL 2011  |  229


© 2011 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 2 | Pharmacodynamic measures linked to antibacterial activity* Daptomycin is a concentration-dependent lipopeptide


Antibiotic Pharmacodynamic measure antibiotic primarily eliminated by the kidneys. Trough
plasma concentrations of daptomycin above 24.3 μg/ml
Time-dependent pharmacodynamic profile
have been associated with an increased risk of eleva-
β-Lactams: penicillins, cephalosporins, carbapenems %T>MIC tions in creatine phosphokinase levels, a serum marker
Glycopeptides: vancomycin AUC/MIC of myopathy.59 In critically ill patients with AKI who
Oxazolidinones: linezolid AUC/MIC were receiving continuous RRT at a mean effluent rate of
Concentration-dependent pharmacodynamic profile
33 ml/kg/h, the pharmacokinetics of a single daptomycin
dose of 8 mg/kg was studied during the subsequent 48 h.60
Aminoglycosides AUC/MIC, ratio of peak to MIC
This dose was selected based on the results of an in vitro
Colistin AUC/MIC investigation carried out by the same group.61 The team
Lipopeptides: daptomycin, telavancin AUC/MIC, ratio of peak to MIC‡ used the serum concentration–time data collected from
Fluoroquinolones AUC/MIC, ratio of peak to MIC each patient to generate a pharmacodynamic model, in
*Selected antibiotics commonly used in the intensive care unit. ‡Telavancin only. Abbreviations: AUC, area which they simulated the effects of daptomycin dosing
under the serum concentration–time curve; MIC, minimum inhibitory concentration; %T>MIC, the % of the regimens of 8 mg/kg every 48 h and 4 mg/kg every 24 h.60
dosing interval above MIC.
The 8 mg/kg dose every 48 h not only achieved higher
peaks at steady state (88.8 μg/ml versus 53.0 μg/ml),
profiles, but also careful attention to pharmaco­dynamics. but also produced lower troughs (7.2 μg/ml versus
The study of pharmacodynamics, as it applies to anti­ 12.3 μg/ml). These findings indicate that the 8 mg/kg
biotics, seeks to link measures of drug exposure (such every 48 h regimen not only optimizes pharmaco­dynamic
as peak and trough serum concentrations, and area goal attainment, but could, furthermore, reduce the risk
under the serum concentration–time curve [AUC]) of myopathy.
to bactericidal activity. For example, most anti­biotics Aminoglycoside antibiotics also have concentration-
fall into one of two pharmacodynamic categories: dependent activity, and peak:MIC ratios of 10–12 are
concentration-­dependent or time-dependent. In general, associated with the greatest bacterial killing effect.58
for concentration-­dependent antibiotics, a favorable high Prolonged courses of treatment and elevated trough
ratio of the peak serum concentration to the minimum levels of these agents are associated with nephrotoxic
inhibitory concentra­tion (MIC, the minimum concentra- and ototoxic effects.62 However, to fully optimize the
tion required to inhibit bacterial growth) is best associ- concentration-­dependent pharmacodynamic profile of
ated with clinical success, whereas for time-dependent amikacin in patients with sepsis who have an increased
antibiotics, the percentage of the dosing interval spent volume of distribution, initial doses that are higher than
above the MIC is the most important parameter to the conventional 15 mg/kg might be required. Amikacin
maximize (Table 2).57,58 A third parameter, AUC/MIC, is doses of 15 mg/kg and 20 mg/kg have produced sub­
closely related to the other two parameters, and is some- optimal mean peak concentrations of 33.5 mg/l and
times most predictive of clinical success for antibiotics of 33.8 mg/l, respectively, in critically ill patients with
either type (Table 2). However, an antibiotic’s pattern preserved renal function.63,64 Use of a loading dose of
of bactericidal activity (concentration-dependent or aminoglycosides, to ensure an adequate initial concentra­
time-dependent) is not the sole determinant of which tion peak in these often fluid-overloaded patients, fol-
pharmaco­dynamic parameter is most predictive of effi- lowed with close drug therapeutic monitoring of serum
cacy. The presence and duration of any postantibiotic concentra­tions, will help with pharmacodynamic goal
effect—defined as the period between antibiotic exposure attainment. Aggressive treatment to optimize pharmaco­
and the point at which the surviving microbes begin to dynamic parameters is particularly warranted in the
multiply—is also important.57 Decreased drug elimina­ presence of pathogens that have a high MIC to the anti-
tion as a result of AKI, increased volume of distribution biotic, or in patients with sepsis and large accumulations
from fluid overload and decreased protein binding, and of extravascular volume.
increased clearance owing to RRTs not only affect anti- The use of RRTs may aid pharmacodynamic goal
biotic pharmacokinetics, but also influence whether attainment for concentration-dependent antibiotics
optimal pharmacodynamic targets can be met. by preventing toxic effects associated with prolonged
high levels of antibiotic. For example, the drug clearance
Concentration-dependent antibiotics produced by continuous RRTs could help to prevent
Desirable pharmacodynamic characteristics for the persistence of high levels of aminoglycosides after
concentration-­d ependent antibiotics are high peak a loading dose. In patients managed with intermittent
concentra­tions that optimize bactericidal activity, fol- hybrid RRTs, some researchers have suggested that the
lowed by low troughs that minimize toxic effects. This antibiotic dose should be scheduled before dialysis,
knowledge can help guide dosing decisions. However, so that drug removal by the RRT can be used to avoid
owing to the increased volume of distribution in critically toxic effects. Roberts et al. 18 prospectively collected
ill patients with AKI, the high peaks and low troughs that pharmaco­kinetic data and then used population mod-
are desirable for concentration-dependent antibiotics can eling to evaluate several gentamicin dosing regimens
be difficult to attain with usual doses, as illustrated by the in patients receiving extended daily hemodiafiltration,
following examples. and concluded that 6 mg/kg of gentamicin administered

230  |  APRIL 2011  |  VOLUME 7  www.nature.com/nrneph


© 2011 Macmillan Publishers Limited. All rights reserved
FOCUS ON AKI IN CRITICAL CARE

30 min before RRT with dosing repeated every 48 h investigated in patients undergoing continuous RRT; a
was the optimal regimen for achieving the pharmaco­ 2 g loading dose and 3 g infusion over 24 h produced
dynamic goals (defined as a peak serum concentration mean steady state concentrations of 33.5 mg/l, effectively
of 10 mg/l, and trough below 1.5 mg/l). maintaining the level of antibiotic at ≥4 times the MIC
The use of RRT to meet pharmacodynamic goals of (4 mg/l) required to eliminate susceptible pathogens for
concentration-dependent antibiotics is not without 100% of the dosing interval.68 Continuous infusions of
limita­tions. A major problem is the variability of RRT β‑lactams have not been investigated with hybrid RRT.
treatments, which can be interrupted for a variety of Such continuous infusions, with an administration rate
reasons in the intensive care setting. When Roberts adjustment depending on whether the hybrid RRT is
et al. 18 investigated the effect of RRT on gentamicin running or not, might prove useful. When continuous
pharmacodynamic goal attainment, the assumption was infusions of β‑lactams are not available, an unadjusted
made that extended daily hemodiafiltration would last loading dose, followed by small, frequent maintenance
10 h—although in their study clotting and other factors doses is thought to maintain serum concentrations
limited the typical treatment duration to approximately above the MIC more effectively than large doses given at
6 h. If the timing of antibiotic doses is intended to be extended intervals. The development and implementa­
coordinated with that of RRT, dialysis must be per- tion of clinically useful assays for therapeutic drug
formed at the same time every day. The investigators monitoring could alleviate some of the difficulties with
of a pharmaco­kinetic study of daptomycin in patients dosing of β‑lactams. Roberts et al.69 performed twice-
undergoing hybrid RRT (blood and dialyzate flows of weekly therapeutic drug monitoring for β‑lactams in
160 ml/min; 8 h treatment) recommended a daily dose 236 critically ill patients who were prescribed this agent.
of 6 mg/kg, but emphasized that daptomycin must be Of the 36 (15.3%) patients in the study who were receiv-
given 8 h before the hemodialysis session.65 In institutions ing continuous venovenous hemodiafiltration (blood
that employ similar hybrid dialysis techniques, if dapto­ flow 200 ml/min, dialyzate flow rate 1 l/h, and ultra-
mycin is given less than 8 h before dialysis, increased filtration rate of 2 l/h), 25 required dose adjustments,
drug removal will lower the AUC and the drug might be among whom seven required dose increases after their
underdosed. If daptomycin is administered more than first monitoring session.69 Unfortunately, therapeutic
8 h before dialysis, drug clearance will be decreased and drug monitoring for β‑lactams is currently unavailable
the risk of toxic effects increased. to most clinicians.

Time-dependent antibiotics Individual drug-dosing strategies


For time-dependent antibiotics, such as β‑lactams, main- Individual patient-specific (body weight and volume
taining serum concentrations above the MIC optimizes status), RRT-specific (effluent rate, dialyzer flux,
therapeutic efficacy and prevents the develop­ment of and mode of fluid replacement), and drug-specific
microbial drug resistance.57 Unlike the disadvantages (pharmaco­kinetic and pharmacodynamic) character-
experienced with having to attain both high peaks and istics should guide initial dosing decisions, and doses
low troughs for concentration-dependent antibiotics, should be adjusted continually as a patient’s clinical
the reduced renal function in patients with AKI helps status changes. Drug dosing in patients with AKI can,
to prevent inadequate serum concentrations of time- therefore, be quite complex.
­d ependent antibiotics. RRTs that run at high efflu-
ent rates, however, can remove substantial amounts of Practical approaches to drug dosing
β‑lactams.32 Several modified methods of administra­ In nearly all situations, a loading dose that will achieve
tion, which include prolonged intermittent infusions, the target serum concentrations based on the expected
low-dose with short-interval regimens, and continuous volume of distribution should be given.70 No adjustments
infusions, have been developed to optimize the bacteri- need to be made for residual renal function or RRT for
cidal activity of β‑lactams.66 Continuous infusions are this initial dose. Therapeutic drug monitoring should
of particular interest for use in patients on RRTs, as they subsequently be used whenever possible, but is not
could theoretically be adjusted to administer the drug at available for many drugs. Among the antibiotics com-
a similar rate to that at which it is removed by dialysis. monly used in the intensive care unit setting (Table 3),
Examples of β‑lactam antibiotics for which continuous thera­peutic drug monitoring is available only for amino­
infusions have been tested in patients on continuous RRT glycosides and vancomycin in most hospitals, and in these
include meropenem and ceftazidime. Langgartner et al.67 two cases monitoring should, therefore, be used to guide
compared an intermittent meropenem regimen (1 g dosing. For hybrid RRTs, additional attention should be
infused every 12 h) with a continuous regimen (500 mg paid to the timing of drug administration relative to that
loading dose followed by an infusion of meropenem 2 g of the dialysis treatment. Monitoring should also take
over 24 h) in 12 critically ill patients receiving continu- into account that some drugs exhibit rebound, that is,
ous hemodiafiltration at an effluent rate of 25 ml/kg/h. an increase in drug concentrations in the blood that
The continuous infusion maintained steady state can occur when RRT ends and drug sequestered in the
serum concentrations of 18.6 mg/l, and produced an tissues redistributes back into the blood.70
AUC similar to that of the intermittent dosing regimen. Consulting the literature would be the most
Continuous infusions of ceftazidime have also been evidence-­b ased method of dose adjustment, as long

NATURE REVIEWS | NEPHROLOGY VOLUME 7  |  APRIL 2011  |  231


© 2011 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 3 | Dosing recommendations for selected intravenous antibiotics in patients on continuous RRT
Drug *Aronoff et al.56 ‡
Trotman et al.98 §
Heintz et al.99 ||
Sanford Guide100
Amikacin 7.5 mg/kg every 10 mg/kg LD, then 7.5 mg/kg 10 mg/kg LD then 7.5 mg/kg 7.5 mg/kg every 24 h
24–72 h every 24–48 h every 24–48 h
Ciprofloxacin 400 mg every 24 h 200–400 mg every 12 h 400 mg every 12–24 h 200–400 mg every 24 h
Daptomycin 8 mg/kg every 48 h 4 mg/kg or 6 mg/kg every 48 h 4 mg/kg or 6 mg/kg every 48 h No recommendation
Levofloxacin 500 mg every 48 h 500 mg LD, then 250 mg every 500–750 mg LD, then 750 mg LD, then
24 h 250–500 mg every 24 h 500 mg every 48 h
Meropenem 1–2 g every 12 h 1 g every 12 h 1 g LD then 0.5–1 g every 1 g every 12 h
8–12 h
Piperacillin– 4.5 g every 8 h 2.25–3.375 g every 6 h 2.25–3.375 g every 6 h 2.25 g every 6 h
tazobactam
Vancomycin 1 g every 24–96 h 15–20 mg/kg LD, then 1 g every 15–25 mg/kg LD, then 500 mg every 24–48 h
24 h 10–15 mg/kg every 24 h
*Recommendations based on dialyzate/ultrafiltrate/effluent rate of 2 l/h. ‡Recommendations for patients on continuous venovenous hemodialysis with a
dialyzate flow rate of 1 l/h. §Recommendations for patients on continuous venovenous hemodialysis with a dialyzate flow rate of 1–2 l/h. ||Flow rates not
specified. Abbreviations: LD, loading dose; RRT, renal replacement therapy.

as the clinician ensures that their patient’s clini- approach is not appropriate for all situations. Taking
cal situation is comparable to that of the population every one of the factors discussed in this Review into
of patients and the RRT modalities studied. Drug- account may, however, be time consuming and imprac-
dosing recommenda­t ions for continuous RRT are tical for the intensive care unit clinician. Further
available, although the conclusions vary according to compli­cating the issue is the variability of how RRT is
which source is consulted (Table 3). Before using these delivered, and the lack of sufficient pharmacokinetic
recom­mendations, it is crucial to verify that they are up and pharmacodynamic data provided in the drug man-
to date and based on dialyzate, ultrafiltrate, and efflu- ufacturer’s prescribing information.74 Hybrid therapies
ent rates that are similar to those being utilized. Our carry the added difficulty of ensuring that medications
institution, for example, often employs dialyzate flow are given with an optimal timing in relation to RRT.
rates of at least 2 l/h and, consequently, we tend to use Owing to the variation in intermittent, hybrid, and
higher antibiotic doses than are recommended in some continuous RRT techniques, published pharmaco­
of these published sources. kinetic trials may lack generalizability. Furthermore,
published trials are often missing information that is
Dosage adaptations vital to making appropriate clinical interpretations.
A review of the dosing techniques aimed at individual- Less than 90% of pharmacokinetic studies in patients
izing antibiotic therapy in patients on continuous RRT undergoing continuous RRT actually specified the
has been published.71 One commonly used drug-dosing continuous RRT dose used in their respective popula­
technique involves calculating the total creatinine clear- tions, and only 58% of studies in patients on continu-
ance rate by adding any estimated residual renal creati- ous venovenous hemodialysis specified whether fluid
nine clearance to the expected extracorporeal creatinine replacement was given before or after filtration. 75
clearance. The extracorporeal creatinine clearance rate Without such essential informa­t ion, the translation
can be assumed to be approximately equivalent to the of published continuous RRT studies to the clinical
dialyzate, ultrafiltrate, or effluent rate, and medication setting is very difficult.
dosing guidelines specified for the total creatinine clear- The review by Li et al.75 establishes an ideal data set
ance can be used to guide dose selection. In most patients that should published in all pharmacokinetic studies
receiving conventional continuous RRT, most drugs conducted in patients receiving RRTs. Efforts to con-
will fall within the 25–50 ml/min creatinine clearance vince drug manufacturers to conduct trials in critically
range. This method is useful, although it assumes that ill patients receiving RRTs, as well as to update drug-
drugs only undergo glomerular filtration, not tubular dosing recommendations in package inserts that were
secretion or reabsorption.70 For drugs that do undergo developed using outdated RRT techniques, will also
tubular secretion, this method could lead to increased improve the quality of the pharmacokinetic data avail-
drug exposures, and in patients with impaired reabsorp­ able. Population modeling could be a useful tool to
tion, underdosing can potentially occur. Several equa- improve the interpretability of the results of pharmaco­
tions have been developed to adjust drug doses, kinetic trials. Pharmacodynamic targets should also
although they do not take a drug’s pharmacodynamics be tied to the pharmacokinetic data in a way that is
into account.72,73 easily interpretable.

Barriers to effective dosing Conclusions


With a variety of factors influencing drug removal in a With the large variability in pharmacokinetic para­
particularly vulnerable population, a ‘one dose fits all’ meters reported in critically ill patients, increased

232  |  APRIL 2011  |  VOLUME 7  www.nature.com/nrneph


© 2011 Macmillan Publishers Limited. All rights reserved
FOCUS ON AKI IN CRITICAL CARE

antibiotic dosing is essential to ensure that adequate Review criteria


serum concentra­tions are achieved, particularly when
difficult-to-treat pathogens are implicated. Until thera- Studies for inclusion in this article were identified from a
search of the PubMed database using combinations of
peutic monitoring for a wider selection of antibiotics is
the following terms: “drug dosing”, “renal replacement
more readily available, much of antibiotic dosing will therapy”, “pharmacokinetics”, “critical illness”, and
continue to be based on estimates. Nonetheless, edu- “acute kidney injury”. The reference lists of published
cated, evidence-based dosing can make a large differ- articles and the personal records of the authors were
ence in attaining pharmaco­dynamic targets, preventing also searched for further relevant papers.
resistance, and optimizing patient outcomes.

1. Uchino, S. et al. Acute renal failure in critically ill pharmacokinetics in septic patients. Acta replacement therapy. Clin. Pharmacokinet. 46,
patients: a multinational, multicenter study. JAMA Anaesthesiol. Scand. 43, 726–730 (1999). 997–1038 (2007).
294, 813–818 (2005). 18. Roberts, J. A. et al. Using population 33. Schmidt, C., Höcherl, K., Schweda, F. &
2. Choi, G. et al. Principles of antibacterial dosing in pharmacokinetics to determine gentamicin Bucher, M. Proinflammatory cytokines cause
continuous renal replacement therapy. Crit. Care dosing during extended daily diafiltration in down-regulation of renal chloride entry pathways
Med. 37, 2268–2282 (2009). critically ill patients with acute kidney injury. during sepsis. Crit. Care Med. 35, 2110–2119
3. Li, A. M. et al. A systematic review of antibiotic Antimicrob. Agents Chemother. 54, 3635–3640 (2007).
dosing regimens for septic patients receiving (2010). 34. Sun, H., Frassetto, L. & Benet, L. Z. Effects of
continuous renal replacement therapy: do current 19. Triginer, C. et al. Gentamicin volume of renal failure on drug transport and metabolism.
studies supply sufficient data? J. Antimicrob. distribution in critically ill septic patients. Pharmacol. Ther. 109, 1–11 (2006).
Chemother. 64, 929–937 (2009). Intensive Care Med. 16, 303–306 (1990). 35. Miyazaki, H., Sekine, T. & Endou, H. The
4. Boucher, B. A., Wood, G. C. & Swanson, J. M. 20. Pai, M. P. & Bearden, D. T. Antimicrobial dosing multispecific organic anion transporter family:
Pharmacokinetic changes in critical illness. Crit. considerations in obese adult patients. properties and pharmacological significance.
Care Clin. 22, 255–271 (2006). Pharmacotherapy 27, 1081–1091 (2007). Trends Pharmacol. Sci. 25, 654–662 (2004).
5. Stechmiller, J. K., Treloar, D. & Allen, N. Gut 21. Falagas, M. E. & Karageorgopoulos, D. E. 36. Bergner, R. et al. Fluconazole dosing in
dysfunction in critically ill patients: a review of the Adjustment of dosing of antimicrobial agents for continuous veno-venous haemofiltration
literature. Am. J. Crit. Care 6, 204–209 (1997). bodyweight in adults. Lancet 375, 248–251 (CVVHF): need for a high daily dose of 800 mg.
6. Okabe, H. et al. The increased intestinal (2010). Nephrol. Dial. Transplant. 21, 1019–1023
absorption rate is responsible for the reduced 22. Fry, D. The importance of antibiotic (2006).
hepatic first-pass extraction of propranolol in rats pharmacokinetics in critical illness. Am. J. Surg. 37. Schetz, M. Drug dosing in continuous renal
with cisplatin-induced renal dysfunction. J. Pharm. 172, 20S–25S (1996). replacement therapy: general rules. Curr. Opin.
Pharmacol. 55, 479–486 (2003). 23. Vanholder, R., Van Landschoot, N., De Smet, R., Crit. Care 13, 645–651 (2007).
7. Hughes, C. A. & Dowling, R. H. Speed of onset of Schoots, A. & Ringoir, S. Drug protein binding in 38. Joy, M. S., Matzke, G. R., Frye, R. F. &
adaptive mucosal hypoplasia and hypofunction in chronic renal failure: evaluation of nine drugs. Palevsky, P. M. Determinants of vancomycin
the intestine of parenterally fed rats. Clin. Sci. Kidney Int. 33, 996–1004 (1988). clearance by continuous venovenous
(Lond.) 5, 317–327 (1980). 24. Crandon, J. L., Banevicius, M. A. & Nicolau, D. P. hemofiltration and continuous venovenous
8. Fagerman, K. E., McGuigan, D. & Pixley, B. Pharmacodynamics of tigecycline against hemodialysis. Am. J. Kidney Dis. 31, 1019–1027
Potential interaction between enteral feeding phenotypically diverse Staphylococcus aureus (1998).
solutions and oral tetracycline. Nutr. Clin. Pract. 1, isolates in a murine thigh model. Antimicrob. 39. Clark, W. R. & Ronco, C. CRRT efficiency and
257–258 (1986). Agents Chemother. 53, 1165–1169 (2009). efficacy in relation to solute size. Kidney Int.
9. Wright, D. H., Pietz, S. L., Konstantinides, F. N. & 25. Meier-Hellman, A. et al. Epinephrine impairs Suppl. 72, S3–S7 (1999).
Rotschafer, J. C. Decreased in vitro splanchnic perfusion in septic shock. Crit. Care 40. Churchwell, M. D. & Mueller, B. A. Drug dosing
fluoroquinolone concentrations after admixture Med. 25, 399–404 (1997). during continuous renal replacement therapy.
with an enteral feeding formulation. J. Parenter. 26. Vilay, A. M., Churchwell, M. D. & Mueller, B. A. Semin. Dial. 22, 185–188 (2009).
Enteral Nutr. 24, 42–48 (2000). Clinical review: drug metabolism and nonrenal 41. Bouman, C. S. et al. Discrepancies between
10. Mueller, B. A., Brierton, D. G., Abel, S. R. & clearance in acute kidney injury. Crit. Care 12, observed and predicted continuous venovenous
Bowman, L. Effect of enteral feeding with ensure 235 (2008). hemofiltration removal of antimicrobial agents in
on oral bioavailabilities of ofloxacin and 27. Mueller, B. A., Scarim, S. K. & Macias, W. L. critically ill patients and the effects on dosing.
ciprofloxacin. Antimicrob. Agents Chemother. 38, Comparison of imipenem pharmacokinetics in Intensive Care Med. 32, 2013–2019 (2006).
2101–2105 (1994). patients with acute or chronic renal failure 42. Golper, T. A. Drug removal during continuous
11. Lim, S. G., Sawyerr, A. M., Hudson, M., treated with continuous hemofiltration. Am. J. hemofiltration or hemodialysis. Contrib. Nephrol.
Sercombe, J. & Pounder, E. Short report: Kidney Dis. 21, 172–179 (1993). 93, 110–116 (1991).
the absorption of fluconazole and itraconazole 28. Giles, L. J. et al. Pharmacokinetics of 43. Uchino, S. et al. Continuous renal replacement
under conditions of low intragastric acidity. meropenem in intensive care unit patients therapy: a worldwide practice survey. The
Aliment. Pharmacol. Ther. 7, 317–321 (1993). receiving continuous veno-venous hemofiltration beginning and ending supportive therapy for the
12. Fülöp, T. et al. Volume-related weight gain and or hemodiafiltration. Crit. Care Med. 28, kidney (BEST kidney) investigators. Intensive
subsequent mortality in acute renal failure 632–637 (2000). Care Med. 33, 1563–1570 (2007).
patients treated with continuous renal 29. Ververs, T. F. et al. Pharmacokinetics and dosing 44. Jeffrey, R. F. et al. A comparison of molecular
replacement therapy. ASAIO J. 56, 333–337 regimen of meropenem in critically ill patients clearance rates during continuous hemofiltration
(2010). receiving continuous venovenous hemofiltration. and hemodialysis with a novel volumetric
13. Schrier, R. W. AKI: fluid overload and mortality. Crit. Care Med. 28, 3412–3416 (2000). continuous renal replacement system. Artif.
Nat. Rev. Nephrol. 5, 485 (2009). 30. Macias, W. L., Mueller, B. A. & Scarim, S. K. Organs 18, 425–428 (1994).
14. Mehta, R. L. et al. Nephrology consultation in Vancomycin pharmacokinetics in acute renal 45. Huang, Z., Letteri, J. J., Clark, W. R., Ronco, C. &
acute renal failure: does timing matter? Am. J. failure: preservation of nonrenal clearance. Clin. Gao, D. Operational characteristics of
Med. 113, 527–528 (2002). Pharmacol. Ther. 50, 688–694 (1991). continuous renal replacement modalities used
15. Edwards, K. D. Creatinine space as a measure of 31. Pea, F., Poz, D., Viale, P., Pavan, F. & Furlanut, M. for critically ill patients with acute kidney injury.
total body water in anuric subjects, estimated Which reliable pharmacodynamic breakpoint Int. J. Artif. Organs 31, 525–534 (2008).
after single injection and haemodialysis. Clin. Sci. should be advised for ciprofloxacin monotherapy 46. DeSoi, C. A., Sahm, D. F. & Umans, J. G.
18, 455–464 (1959). in the hospital setting? A TDM-based Vancomycin elimination during high-flux
16. Roberts, J. A. & Lipman, J. Pharmacokinetic retrospective perspective. J. Antimicrob. hemodialysis: kinetic model and comparison of
issues for antibiotics in the critically ill patient. Chemother. 58, 380–386 (2006). four membranes. Am. J. Kidney Dis. 20, 354–360
Crit. Care Med. 37, 840–850 (2009). 32. Pea, F., Viale, P., Pavan, F. & Furlanut, M. (1992).
17. Tang, G. J., Tang, J. J., Lin, B. S., Kong, C. W. & Pharmacokinetic considerations for 47. Agarwal, R. & Toto, R. D. Gentamicin clearance
Lee, T. Y. Factors affecting gentamicin antimicrobial therapy in patients receiving renal during hemodialysis: a comparison of high-

NATURE REVIEWS | NEPHROLOGY VOLUME 7  |  APRIL 2011  |  233


© 2011 Macmillan Publishers Limited. All rights reserved
REVIEWS
efficiency cuprammonium rayon and patients. J. Antimicrob. Chemother. 28, 753–764 Antimicrob. Agents Chemother. 45, 2949–2954
conventional cellulose ester hemodialyzers. Am. (1991). (2001).
J. Kidney Dis. 22, 296–299 (1993). 65. Kielstein, J. T. et al. Dosing of daptomycin in 82. Guenter, S. G., Iven, H., Boos, C., Bruch, H. P. &
48. Scott, M. K., Mueller, B. A. & Clark, W. R. intensive care unit patients with acute kidney Muhl, E. Pharmacokinetics of levofloxacin during
Vancomycin mass transfer characteristics of injury undergoing extended dialysis—a continuous venovenous hemodiafiltration and
high-flux cellulosic dialysers. Nephrol. Dial. pharmacokinetic study. Nephrol. Dial. Transplant. continuous venovenous hemofiltration in
Transplant. 12, 2647–2653 (1997). 25, 1537–1541 (2010). critically ill patients. Pharmacotherapy 22,
49. Ahern, J. W., Lai, C., Rebuck, J. A., 66. Perrott, J., Mabasa, V. H. & Ensom, M. H. 175–183 (2002).
Possidente, C. J. & Weidner, M. Experience with Comparing outcomes of meropenem 83. Lode, H., Grunert, K., Koeppe, K. P. &
vancomycin in patients receiving slow low administration strategies based on Langmaack, H. Pharmacokinetic and clinical
efficiency dialysis. Hosp. Pharm. 39, 138–143 pharmacokinetic and pharmacodynamic studies with amikacin, a new aminoglycoside
(2004). principles: a qualitative systematic review. Ann. antibiotic. J. Infect. Dis. 134, S316–S322
50. Fiaccadori, E. et al. Removal of linezolid by Pharmacother. 44, 557–564 (2010). (1976).
conventional intermittent hemodialysis, 67. Langgartner, J., Vasold, A., Glück, T., Reng, M. & 84. Kinowski, J. M. et al. Multiple-dose
sustained low-efficiency dialysis, or continuous Kees, F. Pharmacokinetics of meropenem during pharmacokinetics of amikacin and ceftazidime in
venovenous hemofiltration in patients with acute intermittent and continuous intravenous critically ill patients with septic multiple-organ
renal failure. Crit. Care Med. 32, 2437–2442 application in patients treated by continuous failure during intermittent hemofiltration.
(2004). renal replacement therapy. Intensive Care Med. Antimicrob. Agents Chemother. 37, 464–473
51. Choi, G. et al. The effect of adsorption, filter 34, 1091–1096 (2008). (1993).
material and point of dilution on antibiotic 68. Mariat, C. et al. Continuous infusion of 85. Benvenuto, M., Benziger, D. P., Yankelev, S. &
elimination by haemofiltration: an in vitro study ceftazidime in critically ill patients undergoing Vigliani, G. Pharmacokinetics and tolerability of
of levofloxacin. Int. J. Antimicrob. Agents 24, continuous venovenous haemodiafiltration: daptomycin at doses up to 12 milligrams per
468–472 (2004). pharmacokinetic evaluation and dose kilogram of body weight once daily in healthy
52. Tian, Q. et al. Effect of drug concentration on recommendation. Crit. Care 10, R26 (2006). volunteers. Antimicrob. Agents Chemother. 50,
adsorption of levofloxacin by polyacrylonitrile 69. Roberts, J. A. et al. Therapeutic drug monitoring 3245–3249 (2006).
haemofilters. Int. J. Antimicrob. Agents 28, of β-lactams in critically ill patients: proof of 86. Nilsson-Ehle, I., Hutchison, M., Haworth, S. J. &
147–150 (2006). concept. Int. J. Antimicrob. Agents 36, 332–339 Norrby, S. R. Pharmacokinetics of meropenem
53. Uchino, S., Cole, L., Morimatsu, H., (2010). compared to imipenem–cilastatin in young,
Goldsmith, D. & Bellomo, R. Clearance of 70. Schetz, M. Drug dosing in continuous renal healthy males. Eur. J. Clin. Microbiol. Infect. Dis.
vancomycin during high-volume haemofiltration: replacement therapy: general rules. Curr. Opin. 10, 85–88 (1991).
impact of pre-dilution. Intensive Care Med. 28, Crit. Care 13, 645–651 (2007). 87. Krueger, W. A. Evaluation by Monte Carlo
1664–1667 (2002). 71. Choi, G. et al. Principles of antibacterial dosing in simulation of the pharmacokinetics of two doses
54. Clark, W. R., Turk, J. E., Kraus, M. A. & Gao, D. continuous renal replacement therapy. Blood of meropenem administered intermittently or as
Dose determinants in continuous renal Purif. 30, 195–212 (2010). a continuous infusion in healthy volunteers.
replacement therapy. Artif. Organs 27, 815–820 72. Reetze-Bonorden, P., Böhler, J. & Keller, E. Drug Antimicrob. Agents Chemother. 49, 1881–1889
(2003). dosage in patients during continuous renal (2005).
55. Mueller, B. A., Pasko, D. A. & Sowinski, K. M. replacement therapy: pharmacokinetic and 88. Dreetz, M. et al. Serum bactericidal activities and
Higher renal replacement therapy dose delivery therapeutic considerations. Clin. Pharmacokinet. comparative pharmacokinetics of meropenem
influences on drug therapy. Artif. Organs 27, 24, 362–379 (1993). and imipenem–cilastatin. Antimicrob. Agents
808–814 (2003). 73. Kroh, U. F. Drug administration in critically ill Chemother. 40, 105–109 (1996).
56. Aronoff, G. R. et al. Drug prescribing in renal patients with acute renal failure. New Horiz. 3, 89. Krueger, W. A. et al. Pharmacokinetics of
failure: dosing guidelines for adults and children 748–759 (1995). meropenem in critically ill patients with acute
5th edn (American College of Physicians, 74. Mueller, B. A. & Smoyer, W. E. Challenges in renal failure treated by continuous
Philadelphia, 2007). developing evidence-based drug dosing hemodiafiltration. Antimicrob. Agents Chemother.
57. Ambrose, P. G. et al. Pharmacokinetics: guidelines for adults and children receiving renal 42, 2421–2424 (1998).
pharmacodynamics of antimicrobial therapy: replacement therapy. Clin. Pharmacol. Ther. 86, 90. Occhipinti, D. J. et al. Pharmacokinetics and
it’s not just for mice anymore. Clin. Infect. Dis. 479–482 (2009). pharmacodynamics of two multiple-dose
44, 79–86 (2007). 75. Li, A. M. et al. A systematic review of antibiotic piperacillin–tazobactam regimens. Antimicrob.
58. Owens, R. C. Jr & Shorr, A. F. Rational dosing of dosing regimens for septic patients receiving Agents Chemother. 41, 2511–2517 (1997).
antimicrobial agents: pharmacokinetic and continuous renal replacement therapy: 91. Capellier, G. et al. Removal of piperacillin in
pharmacodynamic strategies. Am. J. Health Syst. do current studies supply sufficient data? critically ill patients undergoing continuous
Pharm. 66, S23–S30 (2009). J. Antimicrob. Chemother. 64, 929–937 (2009). venovenous hemofiltration. Crit. Care Med. 26,
59. Bhavnani, S. M., Rubino, C. M., Ambrose, P. G. & 76. Wingender, W. et al. Pharmacokinetics of 88–91 (1998).
Drusano, G. L. Daptomycin exposure and the ciprofloxacin after oral and intravenous 92. van der Werf, T. S., Mulder, P. O., Zijlstra, J. G.,
probability of elevations in the creatine administration in healthy volunteers. Eur. J. Clin. Uges, D. R. & Stegeman, C. A. Pharmacokinetics
phosphokinase level: data from a randomized Microbiol. 3, 355–359 (1984). of piperacillin and tazobactam in critically ill
trial of patients with bacteremia and 77. Fish, D. N., Bainbridge, J. L. & Peloquin, C. A. patients with renal failure, treated with
endocarditis. Clin. Infect. Dis. 50, 1568–1574 Variable disposition of ciprofloxacin in critically ill continuous veno-venous hemofiltration (CVVH).
(2010). patients undergoing continuous arteriovenous Intensive Care Med. 23, 873–877 (1997).
60. Vilay, A. M. et al. Daptomycin pharmacokinetics hemodiafiltration. Pharmacotherapy 15, 93. Blouin, R. A., Bauer, L. A., Miller, D. D.,
in critically ill patients receiving continuous 236–245 (1995). Record, K. E. & Griffen, W. O. Jr. Vancomycin
venovenous hemodialysis. Crit. Care Med. 39, 78. Wallis, S. C., Mullany, D. V., Lipman, J., pharmacokinetics in normal and morbidly obese
19–25 (2011). Rickard, C. M. & Daley, P. J. Pharmacokinetics of subjects. Antimicrob. Agents Chemother. 21,
61. Churchwell, M. D., Pasko, D. A. & Mueller, B. A. ciprofloxacin in ICU patients on continuous veno- 575–580 (1982).
Daptomycin clearance during modeled venous hemodiafiltration. Intensive Care Med. 94. Healy, D. P., Polk, R. E., Garson, M. L., Rock, D. T.
continuous renal replacement therapy. Blood 27, 665–672 (2001). & Comstock, T. J. Comparison of steady-state
Purif. 24, 548–554 (2006). 79. Chien, S. C. et al. Pharmacokinetic profile of pharmacokinetics of two dosage regimens of
62. Turnidge, J. Pharmacodynamics and dosing of levofloxacin following once-daily 500-milligram vancomycin in normal volunteers. Antimicrob.
aminoglycosides. Infect. Dis. Clin. North Am. 17, oral or intravenous doses. Antimicrob. Agents Agents Chemother. 31, 393–397 (1987).
503–528 (2003). Chemother. 41, 2256–2260 (1997). 95. Boeckh, M. et al. Pharmacokinetics and serum
63. Beaucaire, G. et al. Clinical and bacteriological 80. Chow, A. T. et al. Safety and pharmacokinetics of bactericidal activity of vancomycin alone and in
efficacy, and practical aspects of amikacin given multiple 750-milligram doses of intravenous combination with ceftazidime in healthy
once daily for severe infections. J. Antimicrob. levofloxacin in healthy volunteers. Antimicrob. volunteers. Antimicrob. Agents Chemother. 32,
Chemother. 27 (Suppl. C), 91–103 (1991). Agents Chemother. 45, 2122–2125 (2001). 92–95 (1988).
64. Marik, P. E., Lipman, J., Kobilski, S. & 81. Malone, R. S., Fish, D. N., Abraham, E. & 96. Kielstein, J. T. et al. Pharmacokinetics and total
Scribante, J. A prospective randomized study Teitelbaum, I. Pharmacokinetics of levofloxacin elimination of meropenem and vancomycin in
comparing once- versus twice-daily amikacin and ciprofloxacin during continuous renal intensive care unit patients undergoing extended
dosing in critically ill adult and paediatric replacement therapy in critically ill patients. daily dialysis. Crit. Care Med. 34, 51–56 (2006).

234  |  APRIL 2011  |  VOLUME 7  www.nature.com/nrneph


© 2011 Macmillan Publishers Limited. All rights reserved
FOCUS ON AKI IN CRITICAL CARE
97. DelDot, M. E., Lipman, J. & Tett, S. E. 99. Heintz, B. H., Matzke, G. R. & Dager, W. E. Acknowledgments
Vancomycin pharmacokinetics in critically ill Antimicrobial dosing concepts and C. P. Vega, University of California, Irvine, CA, is the
patients receiving continuous venovenous recommendations for critically ill adult author of and is solely responsible for the content of
haemodiafiltration. Br. J. Clin. Pharmacol. 58, patients receiving continuous renal the learning objectives, questions and answers of the
259–268 (2004). replacement therapy or intermittent Medscape, LLC-accredited continuing medical
98. Trotman, R. L., Williamson, J. C., hemodialysis. Pharmacotherapy 29, 562–577 education activity associated with this article.
Shoemaker, D. M. & Salzer, W. L. Antibiotic (2009).
dosing in critically ill adult patients receiving 100. Gilbert, D. N. (Ed.) The Sanford Guide to Author contributions
continuous renal replacement therapy. Clin. Antimicrobial Therapy 40th edn (Sanford, R. F. Eyler and B. A. Mueller contributed equally to all
Infect. Dis. 41, 1159–1166 (2005). Sperryville, 2010). aspects of the manuscript.

NATURE REVIEWS | NEPHROLOGY VOLUME 7  |  APRIL 2011  |  235


View publication stats
© 2011 Macmillan Publishers Limited. All rights reserved

You might also like