Morbidity and Mortality of Patients Who Underwent Minimally Invasive Esophagectomy After Neoadjuvant Chemoradiotherapy vs Neoadjuvant Chemotherapy for Locally Advanced Esophageal Squamous Cell Carcinoma: A Randomized Clinical Trial

IMPORTANCE: Safety and efficacy of neoadjuvant chemoradiotherapy (nCRT) vs neoadjuvant chemotherapy (nCT) for treatment of locally advanced esophageal squamous cell carcinoma (ESCC) remain uncertain given lack of high-level clinical evidence. OBJECTIVE: To compare safety and long-term survival of nCRT followed by minimally invasive esophagectomy (MIE) with that of nCT followed by MIE for patients with locally advanced ESCC. DESIGN, SETTING, AND PARTICIPANTS: A prospective, multicenter, open-label, randomized clinical trial that compared safety and efficacy of nCRT vs nCT followed by MIE for patients with locally advanced ESCC. From January 1, 2017, to December 31, 2018, 264 patients with ESCC of clinical stages from cT3 to T4aN0 to 1M0 were enrolled. Analysis was performed on an intention-to-treat basis from January 1, 2017, to August 30, 2020. INTERVENTIONS: Eligible patients were randomized to the nCRT group (n = 132) or the nCT group (n = 132) by a computer-generated random system. The chemotherapy, based on paclitaxel and cisplatin, was administered to both groups, while 40 Gy of concurrent radiotherapy was added for the nCRT group. At about 6 weeks after neoadjuvant therapy, MIE via thoracoscopy and laparoscopy was performed for the patients in both groups. MAIN OUTCOMES AND MEASURES: The primary outcome was 3-year overall survival. Secondary outcomes included postoperative complications, mortality, postoperative pathologic outcome, recurrence-free survival time, and quality of life. RESULTS: Among 264 patients (226 men [85.6%]; mean [SD] age, 61.4 [6.8] years), postoperative morbidity was 47.4% in the nCRT group (54 of 114) and 42.6% in the nCT group (46 of 108), with no significant difference between groups (difference, 4.8%; 95% CI, −8.2% to 17.5%; P = .48). Distribution of the severity of complications was similar between the 2 groups based on Clavien-Dindo classification. The 90-day perioperative mortality rate was 3.5% for the nCRT group (4 of 114) and 2.8% for the nCT group (3 of 108) (P = .94). The R0 resection rates were similar between groups (109 of 112 [97.3%] vs 100 of 104 [96.2%]; P = .92). However, patients in the nCRT group had a higher pathologic complete response (residual tumor, 0%) rate (40 of 112 [35.7%] vs 4 of 104 [3.8%]; P < .001) and a higher rate of negative lymph nodes (ypN0, 74 of 112 [66.1%] vs 48 of 104 [46.2%]; P = .03) than those in the nCT group. One-year overall survival using intention-to-treat analysis was 87.1% in the nCRT group (115 of 132) and 82.6% in the nCT group (109 of 132) (P = .30). Furthermore, deaths caused by tumor progression or recurrence were significantly less in the nCRT group than in the nCT group (9 of 132 [6.8%] vs 19 of 132 [14.4%]; P = .046); however, deaths from nontumor causes were similar (8 of 132 [6.1%] vs 4 of 132 [3.0%]; P = .24). CONCLUSIONS AND RELEVANCE: Initial results of the trial showed that nCRT followed by MIE has similar safety to and better histopathologic outcome than nCT followed by MIE for treatment of locally advanced ESCC. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03001596