Linear dose-response relationship between vancomycin trough concentration and acute kidney injury in sepsis: effect modification by hyperlactatemia.

BACKGROUND: Vancomycin trough concentration (VTC) is commonly monitored to guide therapy in severe infections, yet its association with acute kidney injury (AKI) in septic patients remains inadequately characterized, particularly regarding the shape of the dose-response relationship and potential effect modifiers. This study aimed to investigate the relationship between initial VTC and AKI risk in sepsis, and to identify whether this association is modified by key clinical subgroups. METHODS: We conducted a multicenter retrospective cohort study using the multicenter eICU Collaborative Research Database. Adults with sepsis who received intravenous vancomycin and had an initial trough concentration (VTC) measured were included; patients with acute kidney injury (AKI) diagnosed before the initial VTC measurement were excluded. The primary outcome was AKI during ICU stay, defined by KDIGO creatinine criteria. Multivariable logistic regression, generalized additive models, and subgroup analyses were used to evaluate the association between initial VTC and AKI, testing for interactions with prespecified covariates. Aminoglycoside use was rare (n = 3, 0.3%) and was therefore not included in the analysis. RESULTS: Among 946 eligible patients, the overall AKI incidence was 9.7%. After multivariable adjustment, each 1 mg/L increase in initial VTC was associated with 6% higher odds of AKI (adjusted OR 1.06, 95% CI 1.02-1.11, P = 0.002). Patients with VTC >20 mg/L had a 3.5-fold higher risk compared with those with VTC ≤10 mg/L. The dose-response relationship was approximately linear (edf = 1.00, P = 0.002), with limited evidence for a distinct threshold. A significant interaction was observed for lactate level (P for interaction = 0.025): in patients with lactate >2.0 mmol/L, each 1 mg/L increase in VTC was associated with a 13% higher odds of AKI (aOR 1.13, 95% CI 1.05-1.21, P = 0.001), whereas no significant association was found in those with lactate ≤2.0 mmol/L (aOR 1.03, 95% CI 0.98-1.08, P = 0.258). CONCLUSION: In septic patients, higher initial VTC is independently and linearly associated with increased AKI risk without a discrete safety threshold. This association is significantly stronger in patients with hyperlactatemia, generating the hypothesis that lactate level identifies a subgroup with heightened susceptibility to vancomycin-associated nephrotoxicity, which requires prospective validation.