Residual tumor model in esophageal squamous cell carcinoma after neoadjuvant immunochemotherapy: Frequently involves the mucosa and/or submucosa

OBJECTIVES: The efficacy and safety of neoadjuvant immunochemotherapy (nICT) are widely explored in locally advanced esophageal squamous cell carcinoma (ESCC). Whether the “wait-and-see” strategy is applicable in ESCC after nICT is still lacking a theoretical basis. This study aimed to preliminarily explore the distribution of residual tumors and the regression pattern of ESCC after nICT. METHODS: Patients undergoing radical esophagectomy after nICT in Fujian Medical University Union Hospital between January 2020 and March 2022 were identified. The resection specimens were re-evaluated by one experienced pathologist. The pathological response was assessed by tumor regression grade (TRG) (modified Ryan scheme). The TRG grade was divided into grades 0 (pathological complete response), 1, 2, and 3. The pathological stage was evaluated in the Eighth Edition AJCC. In the non-pCR group, the residual model was divided into four types: Type I, regression towards the lumen; type II, regression towards the invasive front; type III, concentric regression; and type IV, scattered regression. RESULTS: A total of 95 consecutive patients were included for analysis. Seventy-six (80.0%) of 95 patients were in non-pCR (pathological complete response), and nine patients (9/76, 11.84%) had isolated residual tumors in lymph nodes. There was no significant difference in baseline characteristics between the pCR group and the non-pCR group (p > 0.05). The overall distribution of TRG for all esophageal wall layers was TRG 0 = 28 (28/95, 29.5%), TRG 1 = 17 (17/95, 17.9%), TRG 2 = 18 (18.9%, 18/95), and TRG 3 = 32 (32/95, 33.7%). In 67 patients with residual tumors in the esophageal wall (TRG ≧1), 63 (63/67, 94.0%) had residual tumor cells in the mucosa and/or submucosa, and four had isolated residual tumors in the muscle layer (4/67, 6.0%). Further analysis showed eight (8/67, 11.9%) patients with submucosal involvement but without mucosal involvement. The distribution of regression patterns was type I (n = 35, 52.2%), type II (n = 3, 4.5%), type III (n = 8, 11.9%), and type IV (n = 21, 31.3%). CONCLUSIONS: The mucosa and/or submucosa are frequently involved in residual malignancy, and the frequent regression models are regression toward the lumen and random regression. There is an opportunity to carefully test the residual tumors in a subgroup of the population with ESCC following nICT. However, some patients had residual tumors only in the muscle layer or lymph nodes. The clinical application of the wait-and-see strategy in ESCC after nICT should be explored using an appropriate evaluation protocol.