Predictive model of postoperative pneumonia after neoadjuvant immunochemotherapy for esophageal cancer

BackgroundPostoperative pneumonia (PP) is the most common pulmonary complication of esophagectomy. It is of great importance to identify any high-risk factors and prevent pulmonary complications to improve the prognosis of patients with esophageal cancer undergoing esophagectomy. Thus, we established a predictive model of PP in patients with neoadjuvant immunochemotherapy for resectable esophageal squamous cell carcinoma (ESCC), and provide suggestions for the best strategy for the perioperative period of the patients.MethodWe retrospectively analyzed 78 patients who underwent esophagectomy for squamous cell carcinoma after neoadjuvant immunochemotherapy between September 2019 and August 2021.We used the “glmnet” language package in R to perform least absolute shrinkage and selection operator (LASSO) regression to screen the best predictors of PP, and nomograms predicting PP were constructed utilizing screened factors. The performance of nomograms was internally validated by calibration curves, concordance index (C-index), and the Brier score for overall performance.ResultsTwenty-six patients (33.3%) had postoperative pneumonia. After LASSO regression, the factors that were independently associated with PP were diffusing capacity of the lungs for carbon monoxide (DLCO) (P=0.0002), white blood cell (WBC) difference before vs. after neoadjuvant immunochemotherapy (P=0.0133). We constructed a prediction model, plotted the nomogram, and verified its accuracy. Its Brier score was 0.147, its calibration slope was 0.98, and its C-index was 0.85 (95% CI: 0.75–0.95). Internal validation demonstrated a good discrimination power that the actual probability corresponds closely with the predicted probability.ConclusionsOur prediction model can predict the possibility of PP in patients with neoadjuvant immunochemotherapy for resectable esophageal squamous cell carcinoma and may facilitate physicians’ efforts to reduce the incidence of postoperative pneumonia.