新辅助化疗放疗（nCRT）后食管癌手术的最佳时间尚不清楚，在临床实践中一直是4-6周。观察性研究表明，在nCRT后长时间延迟手术（最长三个月）之后，尤其在组织学反应方面，疗效更好。NeoRes II试验是第一个比较标准和延迟时间对nCRT后食管癌手术的随机对照试验。对于可切除的局部晚期食管癌患者，随机分配为nCRT后标准手术延迟4-6周或延迟10-12周。以腺癌患者中原发肿瘤组织学完全反应为主要终点。次要终点包括组织学肿瘤反应、切缘复发、总生存期和无进展生存期，对所有患者进行分层分析。在瑞典、挪威和德国的十个参与中心，从2015年2月到2019年3月，共随机分配了249名患者：125名为标准组，124名为延迟手术组。腺癌患者中，标准组（21%）与延迟组（26%）之间的组织学完全反应没有显著差异（P=0.429）。两组之间的肿瘤回归、切缘复发、切除淋巴结数量和总及转移淋巴结数量没有差异。标准手术时间组的患者的第一四分位总生存期为26.5个月，而延迟手术组为14.2个月（P=0.003），延迟手术后的总死亡风险在随访期内增加了35%（风险比为1.35，95%可信区间为0.94-1.95，P=0.107）。延迟手术并不会改善组织学完全反应或其他病理终点，但明显趋向于较差的生存期，提示在nCRT后通常不要延迟手术超过六周。

The optimal time to surgery after neoadjuvant chemoradiotherapy (nCRT) for oesophageal cancer is unknown and has traditionally been 4-6 weeks in clinical practice. Observational studies have suggested better outcomes, especially in terms of histological response, after prolonged delay of up to three months after nCRT. The NeoRes II trial is the first randomised trial to compare standard to prolonged time to surgery after nCRT for oesophageal cancer.Patients with resectable, locally advanced oesophageal cancer were randomly assigned to standard delay of surgery of 4-6 weeks or prolonged delay of 10-12 weeks after nCRT. The primary endpoint was complete histological response of the primary tumour in patients with adenocarcinoma. Secondary endpoints included histological tumour response, resection margins, overall and progression-free survival in all patients and stratified by histologic type.Between February 2015 and March 2019, 249 patients from ten participating centres in Sweden, Norway and Germany were randomised: 125 to standard and 124 to prolonged time to surgery. There was no significant difference in complete histological response between adenocarcinoma patients allocated to standard (21%) compared to prolonged (26%) time to surgery (P=0.429). Tumour regression, resection margins, number of resected lymph nodes, total and metastatic, did not differ between the allocated interventions. First quartile overall survival in patients allocated to standard time to surgery was 26.5 months compared to 14.2 months after prolonged time to surgery (P=0.003) and the overall risk of death during follow-up was 35% higher after prolonged delay (hazard ratio 1.35, 95% CI: 0.94-1.95, P=0.107).Prolonged time to surgery did not improve histological complete response or other pathological endpoints, while there was a strong trend towards worse survival, suggesting caution in routinely delaying surgery for more than six weeks after nCRT.Copyright © 2023 The Author(s). Published by Elsevier Ltd.. All rights reserved.