Feed

Objective: To investigate the incidence of immune-related adverse events (irAEs) and corresponding perioperative outcomes among patients with colorectal cancer who received neoadjuvant immunotherapy and underwent curative radical resection. Methods: This was a retrospective, multicenter, real-world cohort study. A total of 452 patients with pathologically confirmed colorectal adenocarcinoma who completed neoadjuvant immunotherapy and curative radical surgery were enrolled from five tertiary hospitals between January 1, 2020 and December 31, 2024. The participating institutions included Beijing Friendship Hospital, Capital Medical University; Union Hospital, Tongji Medical College, Huazhong University of Science and Technology; The Sixth Affiliated Hospital, Sun Yat-sen University; The First Affiliated Hospital of Nanchang University; and The Second Affiliated Hospital, Zhejiang University. The median age of the overall cohort was 58.5 years. In terms of clinical TNM staging, 86 patients (19.0%) were at stage Ⅱ and 366 (81.0%) were at stage Ⅲ. Based on MMR/MSI status, 116 patients (25.7%) were identified as deficient mismatch repair/high microsatellite instability (dMMR/MSI-H), and the remaining 336 (74.3%) as proficient mismatch repair/microsatellite stable (pMMR/MSS). Regarding treatment modalities, 92 patients (20.4%) received single-agent immune checkpoint inhibitor (ICI) therapy, 22 (4.9%) dual ICI combination therapy, 30 (6.6%) immunotherapy combined with chemotherapy, and 308 (68.1%) immunotherapy combined with chemoradiotherapy. All irAEs were graded in accordance with the Common Terminology Criteria for Adverse Events Version 5.0 (CTCAE v5.0), and grade ≥3 irAEs were defined as severe irAEs. The primary outcomes were the overall incidence and organ-specific distribution of irAEs. Secondary outcomes covered irAE risks stratified by treatment regimens and MMR/MSI status, overall pathological complete response (pCR) rate, as well as perioperative clinical outcomes. Temporal trend analysis was conducted after logical validation of the initial onset time of irAEs. Results: The overall incidence of all-grade irAEs reached 24.6% (111/452), including 73 cases (16.2%) of grade 1, 24 (5.3%) of grade 2, 7 (1.5%) of grade 3, 5(1.1%) of grade 4 and 2 (0.4%) of grade 5. Severe irAEs (grade ≥3) occurred in 14 patients, accounting for 3.1% of the total cohort. A total of 157 irAE episodes were documented, among which single-organ involvement was observed in 65 patients (58.6%) and multi-organ involvement in 46 patients (41.4%). The most frequently affected organ systems were skin and mucous membranes (40 episodes, 25.5%), endocrine and thyroid system (36 episodes, 22.9%), hematopoietic system (34 episodes, 21.7%), hepatobiliary system (21 episodes, 13.4%), and gastrointestinal tract (9 episodes, 5.7%). Other involved systems comprised musculoskeletal system (6 episodes, 3.8%), urinary and renal system (6 episodes, 3.8%), cardiovascular system (4 episodes, 2.5%), and general systemic reactions (1 episode, 0.6%). One fatal grade 5 event consisting of fulminant myocarditis complicated with malignant arrhythmia and liver failure was recorded. Stratified by treatment strategy, the incidence of all-grade irAEs was 23.9% (22/92) for single-agent ICI, 27.3% (6/22) for dual ICI combination, 23.3% (7/30) for immunochemotherapy and 24.7% (76/308) for immunochemoradiotherapy; the corresponding incidence of severe irAEs was 3.3%, 4.5%, 3.3%, and 2.9%, respectively. Univariate and multivariate regression analyses confirmed that MMR/MSI status was the only independent risk factor for irAE occurrence (OR=2.626,95%CI:1.566-4.402,P<0.001). Validated complete onset time data were available for 85 patients, who were included in the temporal analysis, with 110 irAE episodes including 14 severe events identified. Approximately 58.8% (50/85) of irAEs emerged within 84 days after the first ICI administration, and 52.9% (45/85) developed one month after treatment cessation, including two grade 5 fatal events: severe myelosuppression/overlap syndrome following immunochemoradiotherapy, and fulminant myocarditis accompanied by malignant arrhythmia and liver failure after immunochemotherapy. The median time to first irAE onset was 61 (35-100) days. Among 14 severe irAEs, 9 events occurred after immunochemoradiotherapy, 3 after single-agent ICI therapy, 1 after dual ICI combination, and 1 after immuneochemotherapy. Apart from the 2 fatal cases, the other 11 severe irAEs were effectively alleviated via drug discontinuation, glucocorticoid intervention and/or supportive treatment, whereas one case of immune-related enteritis progressed into a chronic condition. No surgical delay of 14 days or longer attributed to irAEs was noted in the entire cohort. The median length of hospital stay was 14 (12-18) days in patients with irAEs versus 13 (11-16) days in those without irAEs, with no significant intergroup difference (Z=1.80, P=0.072). The postoperative complication rates were 18.9% (21/111) and 18.5% (63/341) in the irAEs and non-irAEs groups, respectively, showing no statistical discrepancy (χ²=0.01, P>0.999). The overall pCR rate of all enrolled patients was 49.8% (225/452). Conclusions: IrAEs, which are mostly mild-to-moderate and manifest as multi-organ involvement, are not rare during neoadjuvant immunotherapy for colorectal cancer. Skin, endocrine and hematopoietic systems are the predominantly affected sites, and the majority of irAEs are clinically manageable. Although severe irAEs remain uncommon, they are featured with delayed onset and life-threatening potential; in particular, cardiovascular toxicities such as fulminant myocarditis warrant close clinical vigilance. The onset of irAEs is not confined to the conventional neoadjuvant treatment period, hence sustained safety monitoring is required throughout the perioperative phase and even after treatment completion.