Chao Yencheng, Zhou Jiebai, Hsu Shujung, Ding Ning, Li Jiamin, Zhang Yong, Xu Xiaobo, Tang Xinjun, Wei Tianchang, Zhu Zhengfei, Chu Qian, Neal Joel W, Wu Julie Tsu-Yu, Song Yuanlin, Hu Jie
Department of Pulmonary Medicine, Zhongshan Hospital, Fudan University, Shanghai, China.
Shanghai Key Laboratory of Lung Inflammation and Injury, Shanghai, China.
Transl Lung Cancer Res. 2022 Feb;11(2):295-306. doi: 10.21037/tlcr-22-72.
Immune checkpoint inhibitors (ICIs) have led to dramatic improvements in survival a subset of patients with non-small cell lung cancer (NSCLC); however, they have been shown to cause life-threatening toxicity such as immune checkpoint inhibitor-related pneumonitis (CIP). Our previous studies have shown that chronic obstructive pulmonary disease (COPD) and circulating cytokines are associated with clinical outcomes in NSCLC patients receiving ICIs. However, the relationship between these factors and the development of CIP is unclear. In this study, we retrospectively assessed NSCLC patients receiving ICIs to identify CIP risk factors.
This retrospective cohort study reviewed medical records of NSCLC patients receiving ICIs targeting programmed cell death 1 (PD-1) or its ligand PD-L1 between March 2017 and December 2020 at Zhongshan Hospital Fudan University. CIP was diagnosed by the treating investigator. Clinical characteristics and baseline plasma cytokines were collected. Logistic regression was used to compare clinical characteristics and circulating cytokine levels between patients with and without CIP to identify CIP risk factors.
Of 164 NSCLC patients who received ICIs, CIP developed in 20 cases (12.2%). The presence of COPD [odds ratio (OR), 7.194; 95% confidence interval (CI): 1.130 to 45.798; P=0.037] and PD-L1 expression of ≥50% (OR, 7.184; 95% CI: 1.154 to 44.721; P=0.035) were independently associated with a higher incidence of CIP, whereas a higher baseline level of interleukin-8 (IL-8) was associated with a lower incidence of CIP (OR, 0.758; 95% CI: 0.587 to 0.978; P=0.033). The independent risk factors from final multivariate analysis were incorporated into a nomogram to predict the incidence of CIP. The nomogram model receiver operating characteristic (ROC) curve had a good predictive accuracy of 0.883 (95% CI: 0.806 to 0.959).
Increased risk of CIP independently associated with history of COPD, tumor PD-L1 expression ≥50%, and low baseline IL-8 level. The nomogram may hold promise for CIP risk assessment in the administration of ICIs.
免疫检查点抑制剂（ICIs）已使部分非小细胞肺癌（NSCLC）患者的生存率得到显著提高；然而，它们已被证明会导致危及生命的毒性，如免疫检查点抑制剂相关肺炎（CIP）。我们之前的研究表明，慢性阻塞性肺疾病（COPD）和循环细胞因子与接受ICIs治疗的NSCLC患者的临床结局相关。然而，这些因素与CIP发生之间的关系尚不清楚。在本研究中，我们回顾性评估了接受ICIs治疗的NSCLC患者，以确定CIP的危险因素。
这项回顾性队列研究回顾了2017年3月至2020年12月在复旦大学附属中山医院接受靶向程序性细胞死亡蛋白1（PD-1）或其配体PD-L1的ICIs治疗的NSCLC患者的病历。CIP由主治研究者诊断。收集临床特征和基线血浆细胞因子。采用逻辑回归比较发生CIP和未发生CIP患者的临床特征及循环细胞因子水平，以确定CIP的危险因素。
在164例接受ICIs治疗的NSCLC患者中，20例（12.2%）发生了CIP。COPD的存在[比值比（OR），7.194；95%置信区间（CI）：1.130至45.798；P=0.037]和PD-L1表达≥50%（OR，7.184；95%CI：1.154至44.721；P=0.035）与CIP的较高发生率独立相关，而白细胞介素-8（IL-8）的较高基线水平与CIP的较低发生率相关（OR，0.758；95%CI：0.587至0.978；P=0.033）。最终多因素分析中的独立危险因素被纳入列线图以预测CIP的发生率。列线图模型的受试者工作特征（ROC）曲线具有良好的预测准确性，为0.883（95%CI：0.806至0.959）。
CIP风险增加与COPD病史、肿瘤PD-L1表达≥50%以及低基线IL-8水平独立相关。该列线图在ICIs给药中进行CIP风险评估方面可能具有前景。
