Imipenem/cilastatin fails to show a consistent acute kidney injury advantage in the ICU
In critically ill patients who need a carbapenem, imipenem/cilastatin did not deliver a reliable reduction in acute kidney injury across three large intensive care datasets. The question matters because cilastatin has long carried a biologic reputation that extends beyond its role as a renal dehydropeptidase inhibitor paired with imipenem. Preclinical work has suggested kidney protection, and earlier clinical observations had raised the possibility that the combination might blunt nephrotoxicity during severe illness. In this analysis, that signal did not hold up consistently once the data were tested across multiple ICU populations.
The study used retrospective cohorts from MIMIC-III, MIMIC-IV, and the Amsterdam University Medical Center database, bringing the total sample to 3,932 patients. Investigators compared patients treated with imipenem/cilastatin against those receiving non-cilastatin carbapenems and applied propensity score-based overlap weighting to reduce measured baseline imbalances. The design is important because ICU prescribing is rarely random: clinicians often choose one antibiotic over another based on severity, infection source, organ dysfunction, and local practice patterns, all of which can distort a naive comparison. By using multiple databases and adjustment methods, the analysis pushed the question closer to the bedside than a single-center retrospective report ever could.
A sharp split appeared in one database and vanished in the others.
One cohort suggested protection, but the broader pattern did not
In MIMIC-III, AKI development or progression was less common in the imipenem/cilastatin group, with an odds ratio of 0.52 and a 95% confidence interval of 0.31 to 0.86. That result points toward a possible protective association in that specific ICU population. Yet MIMIC-IV showed no meaningful difference between groups, with an odds ratio of 1.22 and a 95% confidence interval of 0.81 to 1.86. The Amsterdam cohort was similarly neutral, with an odds ratio of 1.06 and a 95% confidence interval of 0.76 to 1.47.
The pooled estimate across all datasets also failed to support a differential kidney signal. Random-effects meta-analysis produced an overall odds ratio of 0.90 with a 95% confidence interval of 0.55 to 1.47. In practical terms, the confidence interval crosses no-effect, and the data do not support a dependable kidney-protective advantage from pairing cilastatin with imipenem in this setting. For ICU clinicians, the finding narrows the room for assuming that one carbapenem choice is intrinsically safer for the kidneys than another.
Why the biologic rationale still did not translate cleanly
Cilastatin remains an intriguing compound because animal experiments have linked it to reduced kidney injury, and the mechanism has a plausible pharmacologic foundation. In theory, protecting the renal tubule from toxic stress should matter most in the ICU, where sepsis, hemodynamic instability, vasopressors, contrast exposure, and multi-drug therapy can all converge on the kidney at once. But a plausible mechanism does not guarantee a measurable clinical benefit when the exposure is modest, the competing causes of AKI are numerous, and baseline risk varies sharply from one ICU to another.
The study design also raises an unavoidable interpretive issue: the approved cilastatin dose given with imipenem is lower than the dose used in some animal models. If a nephroprotective effect depends on higher exposure, the clinical formulation may simply not provide enough drug to make a reproducible difference. Another possibility is that any signal is overwhelmed by the broader physiology of critical illness, where renal injury often reflects systemic inflammation, perfusion failure, and cumulative treatment burden more than a single antibiotic choice.
What the ICU data mean for prescribing
This analysis does not suggest that imipenem/cilastatin is harmful to the kidney. It says something more restrained and more useful: within these ICU datasets, it did not consistently alter the odds of AKI compared with other carbapenems. That matters because drug selection in intensive care often turns on efficacy, spectrum, local resistance patterns, and dosing logistics rather than on unproven renal advantages. The findings argue against choosing imipenem/cilastatin primarily to gain kidney protection in critically ill patients who already require carbapenem therapy.
The result also fits the broader reality of ICU nephrology. AKI in this setting is a syndromic outcome, not a single-cause event, and even a drug with a renal-protective mechanism can be too small a piece of the injury pathway to change outcomes on its own. When patients are already exposed to shock, infection, organ support, and fluctuating renal perfusion, the kidney’s vulnerability is driven by a layered physiologic assault. A modest pharmacologic shield may help in experimental systems, but the clinical signal can disappear once those forces accumulate.
Interpretation is strongest where the data are weakest
The three databases did not behave identically, and that variation is part of the story. MIMIC-III produced a favorable association, while MIMIC-IV and AUMC did not. Differences in patient mix, ICU practice, antibiotic selection, AKI definitions, and temporal changes in critical care could all influence the observed effect. A retrospective design can control for measured confounding, but it cannot eliminate hidden differences between treatment groups, especially in a population as unstable as the critically ill.
The authors also worked within the limits of anonymized database research, and the study was not classified as human subjects research by the institutional review board. One author disclosed patent involvement related to cilastatin as a renal protector, a disclosure that deserves careful reader attention even though the analysis itself rests on the data from the three ICU cohorts. None of that changes the central result: the overall evidence here does not support a reliable AKI advantage for imipenem/cilastatin over non-cilastatin carbapenems in critically ill patients.
Future work will need to resolve whether any subgroup truly benefits, whether a different dosing strategy would matter, and whether a prospective design can isolate renal effects from the background noise of critical illness. Until then, the clearest reading is practical and specific: in ICU patients requiring carbapenem therapy, cilastatin should not be assumed to confer broad protection against acute kidney injury.