LAUSR

Copyright © 2020 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved. 932 www.ccmjournal.org June 2020 • Volume 48 • Number 6 In this issue of Critical Care Medicine, Flannery et al (1) present a systematic review and meta-analysis evaluating the risk of acute kidney injury (AKI) in critically ill adults receiving continuous infusion (CI) versus intermittent infusion (II) of vancomycin. Despite being one of the most commonly prescribed antibiotics in the inpatient setting for many years, there are a lack of well-designed randomized clinical data to guide clinicians on how to maximize the efficacy of treatment while minimizing toxicity. In recent years, a subtle but significant increase in minimal inhibitory concentration (MIC) of methicillin-resistant Staphylococcus aureus (MRSA) and the emergence of heteroresistant strains (2) have prompted experts to decrease the breakpoint of vancomycin susceptibility from 4 to 2 μg/mL and to recommend targeting serum vancomycin trough levels of 15–20 μg/mL for the treatment of MRSA pneumonia (3). However, attempts to optimize vancomycin exposure and hence antibacterial effectiveness by using higher loading and maintenance doses are associated with an increased prevalence of nephrotoxicity (4). Defining the prevalence and risk factors of AKI associated with CI is important given the availability of alternative anti– Gram-positive agents that are believed to be less nephrotoxic (5). CI of vancomycin has been proposed as a logistically and pharmacodynamically more convenient alternative to intermittent administration (2). However, the risk of developing AKI during CI remains poorly examined, especially in the critically ill (2). Pharmacokinetic/pharmacodynamic target attainment in this population is made particularly challenging, thanks to large variations in distribution volume during resuscitation, enlargement of the extracellular space, and significant fluctuations in renal clearance during the course of admission (2, 3). In an ideal world, it would be feasible to have continuous therapeutic drug monitoring in order to constantly adjust the dose of vancomycin; unfortunately, this is simply impossible to achieve (2). With their review, Flannery et al (1) try to shed light on this important topic. They put forward a plausible scientific rationale for their findings by avoiding high peak concentrations of vancomycin through delivery via CI rather than II dosing, the risk of AKI could be minimized (6). They conclude that when administered via a CI, vancomycin is associated with a 53% reduction in the odds of AKI and a 2.6-fold higher odds of pharmacokinetic target attainment when compared with II without influencing overall mortality (1). The review has several flaws. The most important is that only two of the included studies were randomized controlled trials, and of the remaining nine studies three were prospective observational and six were retrospective cohort studies. In addition, definitions of AKI in the included studies were not uniform, and this may be a source of significant bias (1). Important risk factors for AKI, such as hypertension, diabetes, receipt of IV contrast, and other nephrotoxic drugs, were not reported in many of the studies. In the reviewed studies, there was a large difference in dosing targets between CI (most commonly 20–25 mg/L) and intermittent dosing (trough of 15–20 mg/L). This is not without consequence as AKI has been more frequently found in patients with vancomycin levels between 25 and 30 μg/mL than in those with levels not exceeding 25 μg/mL (9 [24%] vs 3 [8%] patients; odds ratio, 9.75; CI, 2.41–39.52; p < 0.0001) (2, 7). The prevalence of AKI sharply increases when vancomycin concentrations exceed 30 μg/mL, as compared with values below this level (12 [32%] vs 26 [68%] patients; odds ratio, 30.69; CI, 10.49–89.83; p < 0.0001) (2, 7). The use of trough levels to guide treatment was the recommendation of guidelines at the time the included studies were performed; however, the parameter that best predicts vancomycin activity is the 24-hour area under the concentration curve over the MIC. It is recommended that this ratio be kept between 350 and 400 to achieve a good clinical and microbiological response against MRSAs with MIC values less than or equal to 1 μg/mL (8). Loading doses were more commonly noted in the CI group, and this is a notable difference when compared with II in terms of pharmacokinetic target attainment but not in toxicity (9). In the last 20 years, loading and maintenance doses have changed considerably – the new loading dose for vancomycin in CI is 35 mg/kg over 4 hours followed by an increased maintenance dose (10). In this context, studies that compare continuous with intermittent administration of vancomycin did not show a convincing difference in occurrence of nephrotoxicity. Ingram et al (11) retrospectively studied 167 patients *See also p. 912. Key-words: acute kidney injury; continuous infusion; daptomycin; linezolid; new lipoglycopeptides; vancomycin All authors designed the article and participated in drafting and reviewing. All authors read and approved the final version of the article. The authors have disclosed that they do not have any potential conflicts of interest. Patrick M. Honore, MD, PhD, FCCM Sebastien Redant, MD David De Bels, MD Department of ICU Centre Hospitalier Universitaire Brugmann Brussels, Belgium Continuous Versus Intermittent Infusion of Vancomycin: Toward the End of the Controversy or Even Closer to the Swan Song?*