Treatment-related adverse events of PD-1 and PD-L1 inhibitor-based combination therapies in clinical trials: a systematic review and meta-analysis

• Published in: The lancet oncology 2021-09, Vol.22 (9), p.1265-1274
• Main Authors: Zhou, Xiaoxiang, Yao, Zhuoran, Bai, Hua, Duan, Jianchun, Wang, Zhijie, Wang, Xin, Zhang, Xue, Xu, Jiachen, Fei, Kailun, Zhang, Zhen, Tan, Fengwei, Xue, Qi, Gao, Shugeng, Gao, Yibo, Wang, Jie, He, Jie
• Format: Article
• Language: English
• Subjects: Adverse events; Cancer therapies; Chemotherapy; Clinical trials; Combination therapy; Cytotoxicity; Dysphagia; Fatigue; Immunotherapy; Kidney cancer; Lung cancer; Lymphopenia; Medical innovations; Melanoma; Meta-analysis; Neutropenia; Patients; PD-1 protein; PD-L1 protein; Radiation therapy; Statistical analysis; Systematic review; Toxicity
• ISSN: 1470-2045; 1474-5488
• DOI: 10.1016/S1470-2045(21)00333-8

Numerous ongoing trials are testing anti-PD-1-based or anti-PD-L1-based cancer treatment combinations. Understanding the toxicity profiles of treatment-related adverse events is essential. The aim of this study was to comprehensively investigate the incidences and profiles of treatment-related adverse events across different combination therapies. We did a systematic review and meta-analysis comparing different chemotherapy, targeted therapy, immunotherapy, and radiotherapy combinations with PD-1 or PD-L1 inhibitors. We searched Pubmed, Embase, and Cochrane databases for articles published in English between Jan 1, 2000, and May 21, 2020, investigating globally approved PD-1 or PD-L1 inhibitor-based combination therapies. Only prospective trials reporting overall incidence or tabulated data of treatment-related adverse events were included. Trials investigating sequential therapies, comprising three or more classes of therapies, and enrolling less than ten patients were excluded. The primary outcomes were overall incidences and profiles for all-grade and grade 3 or higher treatment-related adverse events by random-effect models. Heterogeneity between studies was assessed with I2 statistics. The summary measures for main outcomes are incidences (95% CI). The 95% CI were calculated together with the incidence through a random-effects model with a logit transformation. The protocol is registered with PROSPERO (CRD42020189617). We identified 2540 records, of which 161 studies (17 197 patients) met the inclusion criteria. The overall incidence of treatment-related adverse events in the chemotherapy combination was 97·7% (95% CI 96·4–98·5; I2=75%) for all-grade adverse events and 68·3% (60·7–75·0; I2=93%) for grade 3 or higher adverse events; in the targeted therapy combination was 94·5% (90·7–96·8; I2=86%) for all-grade adverse events and 47·3% (37·3–57·5; I2=93%) for grade 3 or higher adverse events; in the immunotherapy combination was 86·8% (80·9–91·1; I2=94%) for all-grade adverse events and 35·9% (29·5–42·9; I2=92%) for grade 3 or higher adverse events; and in the radiotherapy combination was 89·4% (69·0–96·9; I2=74%) for all-grade adverse events and 12·4% (4·4–30·6; I2=73%) for grade 3 or higher adverse events. For these four combination therapies, the most common all-grade adverse events were anaemia (45.4% [95% CI 32·4–59·1]), fatigue (34·3% [27·5–41·9]), fatigue (26·4% [19·2–35·2]), and dysphagia (30·0% [18·7–44·5]), respectively, and the most common g