Association and incremental predictive value of preoperative AISI and CALLY for postoperative pulmonary complications after McKeown esophagectomy following neoadjuvant chemoimmunotherapy

Background: Postoperative pulmonary complications (PPCs) remain a major source of morbidity after McKeown esophagectomy for esophageal squamous cell carcinoma (ESCC), particularly in patients receiving neoadjuvant chemoimmunotherapy (nICT). Readily available preoperative biomarkers may improve risk stratification. This study evaluated the predictive value of the aggregate index of systemic inflammation (AISI), the C-reactive protein-albumin-lymphocyte (CALLY) index, and their combined use for PPCs after McKeown esophagectomy following nICT. Methods: We retrospectively analyzed 412 consecutive ESCC patients who underwent McKeown esophagectomy after nICT between January 2019 and December 2025. The primary endpoint was PPCs within 30 days after surgery. Univariable and multivariable logistic regression analyses were used to examine associations between preoperative biomarkers and PPCs. A combined AISI-CALLY model was constructed using binary logistic regression, and its predictive performance was assessed by receiver operating characteristic (ROC) analysis, DeLong testing, calibration analysis, and decision curve analysis. Propensity score matching (PSM) was performed as a sensitivity analysis. Results: = 0.001). The biomarker-only AISI-CALLY model achieved an AUC of 0.689, compared with 0.650 for AISI alone and 0.665 for CALLY alone. The final integrated model incorporating clinical variables, AISI, and CALLY showed the best discrimination (AUC 0.712, 95% CI 0.658-0.762) and provided greater net benefit on decision curve analysis. In the matched cohort, both biomarkers remained independently associated with PPCs, although discrimination was attenuated. Conclusion: Preoperative AISI and CALLY were independently and complementarily associated with PPCs after McKeown esophagectomy following nICT. Their combined use provided only modest incremental predictive value and may serve as an accessible adjunct, rather than a stand-alone tool, for preoperative PPC risk stratification.