背景与目的:PD-1/PD-L1 抑制剂已被广泛应用于临床，是肺癌的重要治疗手段 之一。免疫治疗导致的免疫相关不良反应有其独特的机制和临床表现。随机对照试 验的结果与真实世界证据相冲突的情况并非罕见，真实世界研究的潜在效益和相关 性越来越引起人们的重视。我们收集了接受 PD-1/PD-L1 单药或联合方案治疗肺癌 患者的 TRAEs 及其管理。希望有助于临床医师更好的理解真实世界里更为复杂情 况下肺癌免疫治疗的风险与收益比。 材料与方法:回顾性分析自 2018-1 月-2021-6 月我院接受 PD-1 或 PD-L1 抑制剂 单药或联合方案治疗的 264 例肺癌患者，对所有级别 TRAEs 进行了收集，并根据 治疗方案将患者分为 ICIs 单药治疗组，ICIs 联合抗血管治疗组和 ICIs 联合化疗组； 同时还根据肺癌的病理分型将患者分为小细胞肺癌组（SCLC）和非小细胞肺癌组 （NSCLC）。计算了不同级别 TRAEs 的发生率，并在 ICIs 联合化疗组对 PD-1 VS PD-L1 组患者 TRAEs 的发生率进行了比较。 结果:所有级别的 TRAEs 的总体发生率为 94.3%，3 级及以上 TRAEs 发生率为 28.9%。有 25（9.5%）例患者出现 TRAEs 导致的治疗中断，3（1.1%）例患者因 TRAEs 死亡。ICIs 单药治疗组至少一种所有级别 TRAEs 的发生率是 78.3%。ICIs 联合抗血 管治疗组至少一种所有级别 TRAEs 的发生率是 90%，ICIs 联合化疗组至少一种所 有级别 TRAEs 的发生率是 96.6%，为三组患者中最高；绝大部分（24/25，96%） TRAEs 导致的治疗中断和全部 TRAEs 导致的患者死亡都在本组。所有级别的 TRAEs 的发生率在 PD-1 抑制剂联合化疗的患者组中为 96.1%，在 PD-L1 抑制剂联 合化疗的患者组中为 100%，p 值为 1。3 级及以上 TRAEs 发生率在 PD-1 抑制剂联 合化疗的患者组中为 33.3%，在 PD-L1 抑制剂联合化疗的患者组中为 40%，p 值为 0.62。出现 TRAEs 造成治疗中断的患者及 TRAEs 相关死亡均在 PD-1 抑制剂联合化 疗组。NSCLC 和 SCLC 患者所有级别及 3 级及以上级别的 TRAEs 发生率没有统计 学差异。3 级及以上 TRAEs 的患者的多变量分析提示 TPS 状态（TPS≥50%）、既 往接受系统治疗和治疗方案（ICIs 联合化疗）是 3 级及以上 TRAEs 的独立风险因 素。对于患者出现的不同系统 TRAEs，按照临床判断可能分别来源于化疗药物、抗 血管药物、ICIs 药物或不同药物共同造成，并给予相应处理。接受皮质类固醇激素 的患者共有 57 例，其中 12 例接受了预防性抗生素治疗。中位激素使用时间为 12 （4-45）天。 1
北京协和医学院中国医学科学院博士研究生学位论文英文缩略词 结论:ICIs 在肺癌的治疗领域已经受到广泛应用，治疗方式多样。免疫联合治疗 的 TRAEs 发生率可达 90%以上，但总体安全可耐受。ICIs 联合化疗组发生的所有 级别 TRAEs 和 3 级及以上 TRAEs 较 ICIs 联合抗血管治疗组及 ICIs 单药治疗组为 高。PD-1 抑制剂联合化疗组与 PD-L1 抑制剂联合化疗组相比，所有级别 TRAEs 和 3 级及以上 TRAEs 均无显著统计学差异。然而在接受免疫联合化疗的患者中，所有 TRAEs 导致的治疗中断及患者死亡均在 PD-1 抑制剂联合化疗组，提示 PD-L1 抑制 剂联合化疗的安全性较好。在避免血液学检查错漏和提高不良反应的记录和上报等 方面仍需要进一步改进。

Background and Purpose：PD-1/PD-L1inhibitors have been widely used in the clinic and are one of the important treatments for lung cancer.Immunotherapy leads to immune-related adversereactions with unique mechanisms and clinical manifestations.It is not uncommon for the results of randomized controlled trials to conflict with real-world evidence,and the potential benefits and relevance of real-world studies are receiving increasing attention.We collected TRAEs and the irmanagement in patients receiving PD-1/PD-L1monotherapy or combination regimens for lung cancer.We hope to help clinicians better understand the risk-benefit ratio of immunotherapy for lung cancer.

Materials and methods：Retrospective analysis of 264 lung cancer patients treated with PD-1/PD-L1 inhibitor monotherapy or combination regimens from 2018-January to 2021-June at our hospital was performed, and all grades of TRAEs were collected, and patients were divided into ICIs monotherapy group, ICIs combined with targeted therapy group and ICIs combined with chemotherapy group according to treatment regimens. The incidence of TRAEs at different grades was calculated and compared in the ICIs combined chemotherapy group for patients in the anti-PD-1 VS anti-PD-L1 group.

Results：The overall incidence of all grades of TRAEs was 94.3%, and grade 3 and higher TRAEs occurred in 28.9%.
The incidence of at least one all-grade TRAEs in the ICIs monotherapy group was 78.3%, the incidence of at least one all-grade TRAEs in the ICIs combined with targeted therapy group was 90%, and the incidence of at least one all-grade TRAEs in the ICIs combined with chemotherapy group was 96.6%, the highest among the three groups; the vast majority (24/25, 96%) of treatment interruptions due to TRAEs and all patient deaths due to TRAEs were in this group. The incidence of all grades of TRAEs was 96.1% in the group of patients treated with PD-1 inhibitors in combination with chemotherapy and 100% in the group of patients treated with PD-L1 inhibitors in combination with chemotherapy, with a p value of 1. Grade 3 and higher TRAEs occurred in 33.3% in the group of patients treated with PD-1 inhibitors in combination with chemotherapy and 40% in the group of patients treated with PD-L1 inhibitors in combination with chemotherapy, with a p value of 0.62. patients who developed TRAEs resulting in treatment discontinuation and TRAEs-related deaths were in the PD-1 inhibitor combination chemotherapy group. there was no statistical difference in the incidence of 3TRAEs across all grades and grades 3 and above in patients with NSCLC and SCLC. multivariate analysis of patients with grade 3 and above TRAEs suggested that TPS status (TPS ≥ 50%), prior receipt of systemic therapy and treatment regimen (ICIs combined with chemotherapy) were independent risk factors for grade 3 and above TRAEs. For patients presenting with different systemic TRAEs, they may originate from chemotherapy drugs, targeted drugs, ICIs drugs or different drugs together, respectively, according to clinical judgment, and were treated accordingly. A total of 57 patients received corticosteroids, 12 of whom received prophylactic antibiotics. The median duration of hormone use was 12 (4-45) days.

Conclusions:ICIs have been widely used in the treatment of lung cancer with a variety of treatment modalities. The incidence of TRAEs with immunotherapy can be over 90%but is generally safe and tolerable. All-grade TRAEs and grade 3 and above TRAEs occurred more in the ICIs combination chemotherapy group than in the ICIs combination targeted and ICIs monotherapy groups, and there was no statistically significant difference in all-grade TRAEs and grade 3 and above TRAEs in the PD-1 inhibitor combination chemotherapy group compared with the PD-L1 inhibitor combination chemotherapy group. However, in patients receiving ICIs combination chemotherapy, all TRAEs resulting in treatment discontinuation and patient death were in the anti-PD-1 group,which seems to suggest a better safety profile for anti-PD-L1 combination chemotherapy.Further improvements are still needed, for example, avoiding hematologic errors and improving the documentation and reporting of adverse reactions.