目的·探讨肺癌患者发生免疫检查点抑制剂相关心肌炎（immune checkpoint inhibitor-associated myocarditis，ICI-M）后，永久停药与继续免疫治疗的安全性差异。方法·采用单中心回顾性队列研究，纳入上海交通大学医学院附属胸科医院2016年6月至2024年8月确诊ICI-M的肺癌患者357例。根据ICI-M后的免疫治疗策略分为永久停药组（n=131）与继续治疗组（n=226）。继续治疗组包括从未停药患者以及暂时停药后重启患者。描述性统计ICI-M复发率及新发其他免疫相关不良事件（immune-related adverse events，irAEs）发生率，比较首次ICI-M与复发ICI-M的严重程度，并探索ICI-M复发或新发其他类型irAEs的危险因素。利用多因素Cox回归分析继续免疫治疗对主要心血管不良事件（major adverse cardiovascular events，MACE）的影响。采用倾向性评分匹配进行敏感性分析，以增强研究结果的稳健性。结果·ICI-M复发率为8.8%（20/226），复发ICI-M时的临床症状严重程度及心肌损伤标志物峰值并未高于首次ICI-M。13.3%（30/226）患者新发其他类型irAEs。ICI-M复发或新发其他类型irAEs的患者首次ICI-M发生时间更早（47.0 d vs 71.0 d，P=0.006），肌酸激酶同工酶MB（creatine kinase-MB，CK-MB）峰值更高（4.8 ng/mL vs 2.8 ng/mL，P=0.019）。多因素Cox回归显示，继续免疫治疗并未显著增加ICI-M患者的远期MACE发生风险（HR=0.75，95%CI 0.28~1.98，P=0.556），且该结果在各亚组中均保持一致（P_相互作用>0.05）。采用倾向性评分匹配进行敏感性分析，结论依然保持稳健（匹配后HR=0.93，95%CI 0.32~2.70，P=0.894）。结论·肺癌患者发生ICI-M后继续免疫治疗总体安全性良好，ICI-M复发率较低，且继续免疫治疗并不显著增加MACE发生风险。研究结果为临床决策提供初步数据支持，未来仍需大规模前瞻性研究对结论加以确认与验证。

Objective ·To compare the safety outcomes between immunotherapy continuation and permanent immunotherapy discontinuation in patients with lung cancer who developed immune checkpoint inhibitor-associated myocarditis (ICI-M). Methods ·This single-center retrospective cohort study enrolled 357 lung cancer patients who were diagnosed with ICI-M between June 2016 and August 2024 in Shanghai Chest Hospital, Shanghai Jiao Tong University School of Medicine. Based on post-ICI-M immunotherapy strategies, patients were categorized into a permanent discontinuation group (n=131) and a continuation group (n=226). The continuation group comprised patients who never interrupted immunotherapy and those who temporarily discontinued but later resumed ICI treatment. The ICI-M recurrence rate and the incidence of other new-onset immune-related adverse events (irAEs) were reported. The severity of initial and recurrent ICI-M was compared, and risk factors for ICI-M recurrence or new-onset irAEs were explored. Multivariable Cox regression analysis was used to evaluate the impact of immunotherapy continuation on major adverse cardiovascular events (MACE). Propensity score matching was performed as a sensitivity analysis to enhance the robustness of the study findings. Results ·The ICI-M recurrence rate was 8.8% (20/226). The clinical severity and peak levels of myocardial injury biomarkers were not higher in recurrent ICI-M than in initial ICI-M. A total of 13.3% (30/226) of patients developed other new-onset irAEs. Patients who experienced ICI-M recurrence or new-onset irAEs had a significantly earlier onset of initial ICI-M (47.0 d vs 71.0 d, P=0.006) and a higher peak creatine kinase-MB (CK-MB) level (4.8 ng/mL vs 2.8 ng/mL, P=0.019). Multivariable Cox regression analysis indicated that immunotherapy continuation did not significantly increase the risk of long-term MACE (HR=0.75, 95%CI 0.28‒1.98, P=0.556), and this result remained consistent across all subgroups (P_interaction>0.05). Sensitivity analysis using propensity score matching confirmed the robustness of this result (after matching: HR=0.93, 95%CI 0.32‒2.70, P=0.894). Conclusion ·Continued immunotherapy after ICI-M in lung cancer patients shows an overall favorable safety profile, with a low ICI-M recurrence rate and no significant increase in MACE incidence. This study provides preliminary evidence to support clinical decision-making; however, large-scale prospective studies are still needed for further confirmation and validation.