Additional neoadjuvant immunotherapy does not increase the risk of anastomotic leakage after esophagectomy for esophageal squamous cell carcinoma: a multicenter retrospective cohort study.

Abstract

Neoadjuvant chemoimmunotherapy (nICT) is a novel and promising therapy model for locally advanced esophageal squamous cell carcinoma (LA-ESCC).The objective of this study aimed to assessed the impact of additional neoadjuvant immunotherapy on patients' short-term outcomes, particularly the incidence of anastomotic leakage (AL) and pathological response. Patients with LA-ESCC who received neoadjuvant chemotherapy(nCT)/ nICT combination with radical esophagectomy were enrolled from three medical centers in China. We used propensity score matching (PSM， ration：1:1, caliper=0.01) and inverse probability processing weighting (IPTW) to balance the baseline characteristics and compare the outcomes. Conditional logistic regression and weighted logistic regression analysis were used to further evaluate whether additional neoadjuvant immunotherapy would increased the risk of postoperative AL. A total of 331 patients getting partially advanced ESCC receiving nCT or nICT were enrolled from three medical centers in China. After PSM/IPTW, the baseline characteristics reached a equilibrium between the two groups. After matching, there were no significant difference in the AL incidence between the two groups(P=0.68, after PSM; P=0.97 after IPTW), and the incidence of AL in the two groups was 15.85% versus 18.29%, and 14.79% versus 15.01%, respectively. After PSM/IPTW, both groups were similar in pleural effusion and pneumonia. After IPTW, the nICT group had a higher incidence of bleeding(3.36% vs. 0.30%, P=0.01), chylothorax(5.79% 0.30%, P=0.001), and cardiac events(19.53% vs. 9.20%, P=0.04). recurrent laryngeal nerve palsy(7.85% vs. 0.54%, P=0.003). After PSM, both groups were similar in palsy of recurrent laryngeal nerve(1.22% vs. 3.66%, P=0.31) and cardiac events(19.51% vs. 14.63%, P=0.41). Weighted logistic regression analysis showed that additional neoadjuvant immunotherapy wasn't responsible for AL (OR=0.56, 95%CI: [0.17,1.71], after PSM; 0.74, 95%CI: [0.34,1.56], after IPTW). The nICT group had dramatically higher pCR in primary tumor than the nCT group (P=0.003, PSM; P=0.005, IPTW), 9.76% versus 28.05% and 7.72% versus 21.17%, respectively. Additional neoadjuvant immunotherapy could benefit pathological reactions without increasing the risk of AL and pulmonary complications.We require further randomized controlled research to validate whether additional neoadjuvant immunotherapy would make a difference in other complications, and determine whether pathologic benefits could translate into prognostic benefits, which would require longer follow-up.