Serum creatinine alone may miss aminoglycoside kidney injury in urosepsis, letter warns

A letter published in Annals of Intensive Care challenges the renal safety conclusion of a recent multicenter retrospective study that examined aminoglycoside use in critically ill patients with urosepsis. The letter, authored by Min Li and Min Xu and published on 18 November 2025, argues that the original study’s assessment strategy for kidney injury was too limited to support a conclusion that aminoglycosides were not associated with increased nephrotoxicity.

The core concern raised by the letter writers centers on the original study’s reliance on serum creatinine and the MAKE30 composite endpoint to define renal injury. Serum creatinine, they note, is a delayed and insensitive marker of acute kidney injury, particularly in critically ill patients. Creatinine levels can remain within the normal range until 24 to 48 hours after significant renal tubular damage has occurred, and the measurement is influenced by non-renal factors including volume status, age, and muscle mass. The letter asks whether these limitations could have led to systematic under-detection of early or mild aminoglycoside-induced nephrotoxicity in the study population.

A second criticism concerns urine output. The letter points out that the KDIGO criteria for diagnosing acute kidney injury explicitly require both serum creatinine and urine output. Oliguria, or reduced urine production, often precedes any rise in serum creatinine and serves as an early warning sign of renal dysfunction. According to the letter, the original study provided no data on urine output or the duration of oliguria, raising the question of whether the full spectrum of acute kidney injury was adequately captured.

The letter also highlights the absence of KDIGO staging in the original study. KDIGO stages 1 through 3 reflect the severity of kidney injury, and the letter writers argue that identifying even mild, stage 1 acute kidney injury in patients exposed to aminoglycosides is clinically important, given its known association with longer hospital stays and increased mortality. They ask whether the lack of staging data could conceal dose-response or severity patterns related to aminoglycoside exposure.

Beyond conventional clinical measures, the letter calls attention to tubular injury biomarkers. It notes that a growing body of literature supports the use of markers such as NGAL, or neutrophil gelatinase-associated lipocalin, KIM-1, or kidney injury molecule-1, and IL-18 to detect subclinical nephrotoxicity, particularly in the setting of aminoglycoside exposure. The letter cites one study in intensive care unit patients in which NGAL levels rose significantly within hours of aminoglycoside administration, even in the absence of changes in serum creatinine. The letter writers suggest that omitting such markers may have masked clinically relevant injury in the urosepsis study.

The clinical implications of the original study’s conclusion are a further point of concern for the letter authors. They warn that by concluding aminoglycosides are not associated with increased nephrotoxicity, the study may inadvertently promote overconfident clinical use of these antibiotics, especially in settings that lack routine therapeutic drug monitoring or early injury surveillance. Given the well-established nephrotoxic potential of aminoglycosides, the letter argues that any reassurance about renal safety should rest on comprehensive and sensitive renal assessment rather than conventional endpoints alone.

In its conclusion, the letter states that the use of crude renal endpoints, without urine output data, KDIGO staging, or biomarker integration, renders the original study’s renal safety conclusion incomplete at best and potentially misleading at worst. The authors strongly encourage future investigations to incorporate multi-dimensional renal assessments to better inform clinical practice regarding aminoglycoside use in urosepsis.

The exchange illustrates a broader methodological debate in critical care nephrology: how best to detect drug-induced kidney injury in patients whose creatinine measurements are confounded by dynamic changes in muscle mass and fluid balance. The letter, published as open access correspondence in Annals of Intensive Care, does not present new patient data of its own but instead scrutinizes the measurement strategy underlying the original retrospective analysis. Readers evaluating the safety of aminoglycosides in urosepsis are directed to both the original study and this correspondence, as the letter’s criticisms concern the adequacy of the renal outcome assessment rather than a new comparison of survival or renal outcomes.

Subject of Research: A critique of renal outcome assessment methods in a retrospective study of aminoglycosides in urosepsis

Article Title: Re: “impact of aminoglycosides on survival rate and renal outcomes in patients with urosepsis: a multicenter retrospective study”

Article References: Li, M., & Xu, M. (2025). Re: “impact of aminoglycosides on survival rate and renal outcomes in patients with urosepsis: a multicenter retrospective study”. Annals of Intensive Care, 15(1), Article 183. https://doi.org/10.1186/s13613-025-01504-5

Image Credits: AI Generated

DOI: 10.1186/s13613-025-01504-5

Keywords: aminoglycosides, urosepsis, acute kidney injury, serum creatinine, KDIGO criteria, urine output, NGAL, biomarkers, nephrotoxicity, critical care, Annals of Intensive Care, letter to the editor