Neoadjuvant PD-1 blockade with toripalimab with or without celecoxib for patients with mismatch repair-deficient or microsatellite instability-high, locally advanced, colorectal cancer (PICC): long-term outcomes of a single-centre, parallel-group, non-comparative, randomised phase 2 trial

作者：Xutao Shen, Yue Cai, Weiwei Li, Lishuo Shi, Jianwei Zhang, Xiaoyu Xie, Zehua Wu, Zhuoxin Zheng, Wuteng Cao, Yanhong Deng, Huabin Hu · 发表于：EClinicalMedicine · 年份：2025 · DOI：10.1016/j.eclinm.2025.103499 · 被引用次数：6 · 研究领域：Genetic factors in colorectal cancer、Cancer Immunotherapy and Biomarkers、Colorectal Cancer Treatments and Studies

Background: Neoadjuvant PD-1 blockade has demonstrated high rates of pathological complete response in patients with mismatch repair-deficient or microsatellite instability-high, locally advanced colorectal cancer. However, prospective data on long-term survival remain limited. Here, we report the 5-year outcomes from the phase 2 PICC study (NCT03926338) evaluating neoadjuvant toripalimab with or without celecoxib in this population. Methods: The PICC study was a single-centre, open-label, parallel-group, non-comparative, randomised, phase 2 study. A total of 34 patients with stage II-III mismatch repair-deficient or microsatellite instability-high colorectal cancer were enrolled between May 1, 2019, and April 1, 2021, and randomly assigned (1:1) to receive neoadjuvant toripalimab plus celecoxib or toripalimab alone every 14 days for six cycles before surgery. The primary endpoint was pathological complete response, which was previously met. Secondary endpoints included long-term oncologic outcomes, with 5-year overall, cancer-specific, disease-free, and event-free survival assessed in the modified intention-to-treat population. Quality of life was also evaluated as a secondary endpoint in this analysis. Data were analysed with a cutoff date of June 22, 2025. Findings: Of the 34 patients enrolled, all were assessable and randomised to toripalimab plus celecoxib (n = 17) or toripalimab monotherapy (n = 17). At a median follow-up of 61.9 months (IQR, 55.8-63.6), no disease recu...