Abstract
Background and Objectives: While many factors are known to influence adverse outcomes after head and neck cancer ablation with free flap reconstruction, two low-cost inflammatory-nutritional indices-the Controlling Nutritional Status (CONUT) score and the Systemic Immune-Inflammation Index (SII)-are not widely used. This pilot study evaluated their accuracy for predicting surgical complications, including flap-site leakage, and explored data-driven cutoff values. Materials and Methods: We retrospectively analyzed 115 consecutive patients undergoing free-flap reconstruction. Data from medical records were obtained. Preoperative CONUT and SII were computed, receiver-operating characteristic (ROC) curves derived optimal thresholds, and associations with outcomes were tested using univariable analyses and multivariable logistic regression adjusted for age, defect area, previous radiotherapy, and flap size. Classification metrics were calculated. Results: SII (continuous) remained independently associated with leakage (adjusted OR ≈ 22 per log-unit increase, p ≈ 0.001), and CONUT (continuous) with surgical complications (adjusted OR 1.39 per point, p = 0.009). ROC analyses showed AUCs of 0.740 (95% CI 0.622-0.858; p < 0.001) for log(SII) and 0.685 (0.553-0.816; p = 0.006) for CONUT for overall surgical complications, and 0.780 (0.686-0.874; p < 0.001) for log(SII) and 0.688 (0.569-0.806; p = 0.002) for CONUT for leakage, without significant difference in both indices. Cutoff values for both dependent variables were 2.875 (log(SII) (≈SII 750-760) and ≥4 (CONUT). Conclusions: Preoperative SII and CONUT demonstrated moderate predictive performance, with notably high negative predictive values. Both indices showed no significant differences in discriminating risk, and, although the result is largely exploratory, given their high negative predictive values, they may be useful for excluding anastomotic leakage and overall postoperative complications after head and neck free-flap reconstruction.