AISI, SIRI, and MLR in Predicting Surgical Outcomes After Radical Cystectomy: Revisiting Inflammatory Risk Markers.

Background and Objectives: This study aimed to evaluate the predictive value of systemic inflammatory response markers-namely, the Systemic Inflammatory Response Index (SIRI), Aggregate Index of Systemic Inflammation (AISI), and Monocyte-to-Lymphocyte Ratio (MLR)-in determining the occurrence of major complications following radical cystectomy. Materials and Methods: A retrospective analysis was conducted on 200 patients who underwent open radical cystectomy with ileal conduit diversion. Demographic, clinical, and laboratory variables, including albumin, creatinine, eGFR, smoking, and ASA score, were collected. SIRI, AISI, and MLR were calculated from preoperative blood counts. Major complications and their subtypes (infectious, wound, cardiopulmonary, thrombotic, and anastomotic) were adjudicated independently. Statistical analyses included multivariable logistic regression, ROC curves, calibration (Hosmer-Lemeshow, intercept, slope, and plots), bootstrap resampling (B = 2000), linearity checks (restricted cubic splines and Box-Tidwell), incremental value metrics (ΔAUC, IDI, and NRI), and decision-curve analysis (DCA). Results: Major complications occurred in 57 patients (28.5%). SIRI values were significantly higher in patients with major complications (median 2.12 vs. 1.63, p = 0.006), whereas AISI and MLR did not differ. SIRI remained an independent predictor in multivariable analysis (OR 1.37, 95% CI 1.01-1.86, p = 0.045). An AUC of 0.624 (95% CI 0.538-0.709) with a negative predictive value of 83.3% was observed for SIRI. The baseline clinical model yielded an AUC of 0.648, and an AUC of 0.672 was obtained when SIRI was added (ΔAUC = +0.024, 95% CI -0.022-0.071, p = 0.16). Calibration was excellent (intercept = 0.07, slope = 1.08), and superior net benefit was demonstrated for the SIRI-augmented model within threshold probabilities of 0.15-0.45 in DCA. A statistically significant improvement in IDI (0.024, p = 0.024) was identified, while NRI was positive but not significant. Subtype analyses indicated that the strongest associations of SIRI were with infectious and wound complications. Conclusions: SIRI was found to be an independent predictor of major complications after open radical cystectomy. Although gains in discrimination were modest, incremental analyses demonstrated improved calibration and net clinical benefit when SIRI was incorporated into a clinical model. External validation is required before translation into clinical practice.