Acute kidney injury after propofol or sevoflurane anaesthesia for colorectal cancer surgery: a secondary analysis

Acute kidney injury is a frequent postoperative complication associated with substantial morbidity and mortality 1. Although haemodynamic instability and surgical complexity are well-established risk factors, the influence of anaesthetic maintenance technique on acute kidney injury incidence and severity remains poorly understood 1-3. Experimental and clinical data suggest a plausible biological rationale for differential renal effects between drugs, including altered renal perfusion and sympathetic activation associated with volatile inhalational anaesthesia 4. To evaluate whether the choice of anaesthetic drug influences postoperative acute kidney injury, we performed a secondary analysis of a multicentre randomised controlled trial 5. Patients undergoing colorectal cancer surgery with available peri-operative creatinine measurements were included. The primary outcome was postoperative acute kidney injury within 7 days, as defined by KDIGO creatinine criteria 6. Logistic regression was used to assess the association between anaesthetic technique and acute kidney injury, adjusting for predefined intra-operative variables reflecting haemodynamic and procedural stress (hypotension, blood loss, fluid balance and duration of anaesthesia). Sensitivity analyses using penalised regression models (ridge, lasso and elastic net) yielded consistent results. Of 3256 patients assessed in the parent trial, 3229 had available peri-operative creatinine measurements and were included in the acute kidney injury incidence analysis. There were then 3213 patients with complete data for all prespecified covariates included in the adjusted analyses (Table 1). There were 2769 (86%) patients classified as ASA physical status 1–2 and 249 (8%) of the overall cohort developed postoperative acute kidney injury. There was no significant difference in acute kidney injury incidence between patients receiving propofol- and sevoflurane-based anaesthesia (121, 8% vs. 128, 8%, p = 0.700). The timing of acute kidney injury diagnosis within 7 postoperative days as assessed using Kaplan–Meier methods did not differ between groups. Adjusted analyses showed no association between anaesthetic drug and acute kidney injury for any KDIGO stage (Fig. 1). For acute kidney injury stages 2 and 3, event numbers were low, resulting in wide confidence intervals compatible with both increased and decreased risk. In this cohort, the precision of the estimates allows exclusion of absolute between-group differences in postoperative acute kidney injury of ≥ 2.8 percentage points in absolute risk, although smaller differences cannot be ruled out. This secondary analysis of a large, international randomised trial of anaesthetic drugs found no evidence of a clinically meaningful difference in postoperative acute kidney injury between propofol- and sevoflurane-based anaesthesia in patients undergoing laparoscopic colorectal cancer surgery. The precision of the estimates allows exclusion of moderate differences in overall acute kidney injury risk, particularly for stage 1 acute kidney injury. Because allocation to anaesthetic drug was randomised in the parent trial, confounding by indication was minimised thus strengthening causal inference regarding agent-specific renal effects. From a clinical perspective, these findings suggest that the choice of anaesthetic maintenance technique alone should not be expected to influence postoperative acute kidney injury risk in this setting, supporting equipoise between propofol and sevoflurane for renal outcomes in low-risk colorectal surgery and allowing anaesthetic choice to be guided by other clinical priorities. These results should be interpreted in the context of previous studies reporting inconsistent findings when comparing volatile and intravenous anaesthesia. Experimental and physiological data provide a plausible biological rationale for differential renal effects between agents, including altered renal perfusion and sympathetic activation under volatile anaesthesia 4, 7, 8. However, clinical studies in gastrointestinal surgery have yielded mixed results 2, 3, and our findings support the notion that such mechanistic differences do not necessarily translate into clinically detectable acute injury in well-optimised, low-risk patients. The absence of an observed difference may be influenced by the characteristics of the study population. Most patients (86%) were ASA physical status 1–2 and underwent laparoscopic colorectal surgery, a setting characterised by relatively limited surgical trauma and blood loss. In such contexts, a possible difference in renal blood regulation between anaesthetic drugs may be insufficient to produce measurable creatinine-based acute kidney injury. Our study also has limitations. Acute kidney was defined using plasma creatinine alone, which may underestimate mild events, and the analysis was not designed to assess high-risk subgroups or severe acute kidney injury. Accordingly, potential differences between anaesthetic drugs may still be clinically relevant in patients with impaired renal reserve, greater haemodynamic instability or higher surgical risk. These populations warrant targeted investigation in future studies. This secondary analysis was reviewed and approved separately by the Swedish Ethical Review Authority and registered at ClinicalTrials.gov (NCT05585866). We thank the CAN study investigators for granting access to the trial database, which made this analysis possible. We also acknowledge the Department of Anaesthesiology and Intensive Care, Växjö Central Hospital, Region Kronoberg, for supporting MT. Our analysis was supported by the Swedish Research Council. ME has received consulting fees from Nimbelle AB, honoraria for lectures from Mälardalen University, is on the board of the ENCORE trial and is the initiator and the principal investigator for the CAN trial. SF declares a non-financial interest with Thioredoxin Systems AB and Sangair AB. MT and RF have no conflicting interests to declare. No other competing interests declared.