Pathologic Complete Response in Esophageal Squamous Cell Cancer: Do Diverse Neoadjuvant Modalities Predict Prognostic Differences?

Abstract Background Pathological complete response (pCR) in patients with locally advanced esophageal cancer after neoadjuvant therapy substantially improves overall survival. Nevertheless, recurrence remains a significant concern. This study endeavors to develop a prediction model for a more accurate prognosis evaluation in esophageal squamous cell cancer (ESCC) patients attaining pCR following neoadjuvant treatments. Methods This multicenter prognostic study included 399 neoadjuvant therapy-induced pathological complete response (pCR) esophageal squamous cell carcinoma (ESCC) patients (2016–2022, three Chinese hospitals) as the training cohort, with external validation in 230 patients from two independent centers. Multivariate Cox regression with backward elimination (P 0.10) identified recurrence predictors, integrated into a dynamic nomogram. Discrimination was assessed via Harrell’s C-index and time-dependent ROC analysis; calibration used 45-degree plots(bootstrapped 1000×). Decision curve analysis (DCA) compared clinical utility (0–100% threshold probabilities) against AJCC 8th Edition staging. Results Multivariate Cox regression identified neoadjuvant therapy modality (chemotherapy nCT, chemoradiotherapy nCRT, or CT combined with immunotherapy nCI) as an independent predictor of recurrence, alongside BMI, alcohol history, clinical stage, and histologic grade. Patients achieving pCR exhibited significant survival variations depending on the therapeutic modality (nCRT vs. nCI: HR = 2.262, P = 0.014). The nomogram demonstrated robust discrimination (training cohort C-index: 0.835; external validation: 0.752) and time-dependent AUCs (1−/3−/5-year: 0.866/0.823/0.810 training; 0.701/0.738/0.814 validation). Decision curve analysis confirmed superior clinical net benefit over AJCC 8th Edition staging across threshold probabilities. Conclusion The proposed model exhibited excellent performance in predicting the prognosis of ESCC patients who achieved pCR after different neoadjuvant therapy modalities. Its integration of therapy-specific predictors and dynamic risk visualization may guide personalized surveillance strategies, particularly for high-risk subgroups. However, prospective validation in larger cohorts is needed to confirm generalizability, and future studies should explore molecular mechanisms underlying modality-driven prognostic disparities despite pCR status.