目的:系统评价免疫检查点抑制剂伊匹木单抗联合化疗治疗晚期实体肿瘤的安全性。方法:检索有关数据库（截至2020年4月30日），收集伊匹木单抗联合化疗（试验组）与安慰剂联合相同方案化疗（对照组）治疗晚期实体肿瘤的随机对照试验（RCT），结局指标含治疗相关不良事件（AE），包括乏力、皮疹、恶心、腹泻、中性粒细胞减少、丙氨酸转氨酶（ALT）增高和天冬氨酸转氨酶（AST）增高等。采用Cochrane协作网偏倚风险评估工具进行方法学质量评价，采用RevMan 5.3统计软件及R软件进行meta分析。效应值为相对危险度（n RR）及其95%置信区间（n CI）。n 结果:纳入分析的RCT共5项（治疗非小细胞肺癌和小细胞肺癌各2项，转移性黑色素瘤1项），包括2 532例患者，其中试验组1 335例，对照组1 197例。质量评价结果显示，5项RCT均为高质量研究。meta分析结果显示，试验组治疗相关1~5级AE和3~5级AE发生率均明显高于对照组[87.27%（1 165/1 335）比82.04%（982/1 197），n RR=1.07，95n %CI：1.03~1.10，n P<0.001；50.26%（671/1 335）比37.68%（451/1 197），n RR=1.40，95n %CI：1.07~1.82，n P=0.010]；亚组分析结果显示，仅转移性黑色素瘤试验组与对照组之间治疗相关1~5级AE和3~5级AE发生率差异有统计学意义[98.79%（244/247）比94.02%（236/251），n RR=1.05，95n %CI：1.02~1.09，n P=0.005；56.28%（139/247）比27.89%（70/251），n RR=2.02，95n %CI：1.61~2.53，n P<0.001]。常见AE的分析结果显示，试验组皮疹、腹泻和肝损伤发生风险增加（均n P<0.05），2组治疗相关病死率差异无统计学意义[1.05%（14/1 335）比0.42%（5/1 197），n χ2=3.374，n P=0.066]。n 结论:伊匹木单抗联合化疗治疗晚期实体肿瘤AE发生风险高于单纯化疗，尤其是免疫相关AE，临床应用时应提高警惕。“,”Objective:To systematically evaluate the safety of immune checkpoint inhibitor ipilimumab combined with chemotherapy for advanced solid tumors.Methods:Randomized controlled trials (RCTs) of ipilimumab combined with chemotherapy (the trial group) versus placebo combined with chemotherapy (the control group) for advanced solid tumors were collected by searching related databases (up to April 30, 2020). The outcome indicators were treatment-related adverse events (AEs), including fatigue, rash, nausea, diarrhea, neutropenia, alanine aminotransferase (ALT) elevation, and aspartate aminotransferase (AST) elevation. The methodological quality of studies was evaluated using the Cochrane collaboration risk of bias tool. RevMan 5.3 software and R software was used in meta-analysis and the effect values were expressed as relative risk (n RR) and its 95% confidence interval (n CI).n Results:A total of 5 RCTs (2 for non-small cell lung cancer, 2 for small cell lung cancer, and 1 for metastatic melanoma) were entered in this study, involving 2 532 patients, in which 1 335 patients were in the trial group and 1 197 patients in the control group. The results of quality evaluation showed that 5 RCTs were all high-quality studies. The results of meta-analysis showed that the incidences of treatment-related grade 1-5 AEs and grade 3-5 AEs in the trial group were significantly higher than those in the control group [87.27% (1 165/1 335) n vs. 82.04% (982/1 197), n RR=1.07, 95n %CI: 1.03-1.10, n P<0.001; 50.26% (671/1 335) n vs. 37.68% (451/1 197), n RR=1.40, 95n %CI: 1.07-1.82, n P=0.010]. However, the results of subgroup analysis showed that the difference in incidences of treatment-related grade 1-5 AEs and grade 3-5 AEs were statistically significant only in patients with metastatic melanoma between the trial group and the control group [98.79% (244/247) n vs. 94.02% (236/251), n RR=1.05, 95n %CI: 1.02-1.09, n P=0.005; 56.28%(139/247) n vs. 27.89%(70/251), n RR=2.02, 95n %CI: 1.61-2.53, n P<0.001]. The analysis of common AEs showed that the risks of rash, diarrhea, and liver injury increased in the trial group (all n P<0.05) and the difference in treatment-related death between the 2 groups was not statistically significant [1.05%(14/1 335) n vs. 0.42%(5/1 197), n χ2=3.374, n P=0.066].n Conclusion:The risk of AEs in patients with advanced solid tumors treated with combination of ipilimumab and chemotherapy is higher than that with chemotherapy alone, especially the immune-related AEs, which deserves clinical vigilance.