Abstract: Background　Multiple myeloma（MM）is a malignant plasma cell neoplasm，which is refractory，and prone to relapse.Moreover，there is no cure for it.Chimeric antigen receptor T-cell（CAR-T）is an innovative immunotherapy with great potentials in the treatment of MM，but its efficacy and safety are not well evaluated currently.Objective　To evaluate the efficacy and safety of CAR-T in the treatment of MM systematically，providing evidence-based medical basis for safe and efficient application of the therapy.Methods　Databases of PubMed，EMBase，and the Cochrane Library were systematically searched for published clinical trials articles in English about CAR-T for MM from inception to November 2019.The overall response rate，incidence of severe adverse events（≥ grade 3）and incidence of cytokine release syndrome were used as major therapeutic outcome indicators.R software was used to perform a Meta-analysis of major therapeutic outcomes of patients with CAR-T.Meta-regression analysis was conducted to further analyze major therapeutic outcomes in patients with CAR-T according to pretreatment regimen，costimulatory molecule and the antigen epitope type of CAR-T.Results　A total of 8 articles were included，including 191 patients.Meta-analysis showed that the overall response rate of CAR-T treatment of MM was 0.88 〔95% CI（0.83，0.93）〕，the incidence of serious adverse events was 0.90 〔95%CI（0.83，0.98）〕，and the incidence of cytokine release syndrome was 0.92〔95%CI（0.85，1.00）〕.Further Meta regression analysis suggested：（1）The incidence of severe adverse reactions and cytokine release syndrome differed significantly by pretreatment regimen（P=0.013，P<0.001）.To be specific，Flu+Cy ，Cy pretreatment and no pretreatment subgroup had higher incidence of severe adverse reactions HDM+ASCT pretreatment subgroups（P<0.05）.（2）Costimulatory molecule 4-1BB subgroup had higher overall response rate but lower incidence of severe adverse reactions than costimulatory molecule CD28 subgroup（P<0.05）.（3）The overall response rate，and incidence of severe adverse reactions and cytokine release syndrome in MM patients with different antigen epitope types were statistically significant（P=0.004，0.049，0.023）.BCMA subgroup had lower overall response rate than LCAR-B38M and BCMA +CD19 subgroups（P<0.05），and had higher incidence of severe adverse reactions and cytokine release syndrome than CD19 subgroups（P<0.05）.The incidence of severe adverse reactions and cytokine release syndrome in BCMA+CD19 subgroup was higher than that of CD19 subgroup（P<0.05）.Conclusion　CAR-T may have a good effect on MM，although there are certain controllable safety risks.The pretreatment regimen，costimulatory molecule，and CAR-T antigen type may be factors associated with its efficacy and safety.Due to small sample size and different qualities of the included studies，the above conclusions need to be further verified by large-sample，high-quality randomized controlled clinical trials.

Key words: Multiple myeloma, Chimeric antigen receptor T cell, Overall response rate, Adverse events, Cytokine release syndrome, Meta-analysis, System evaluation

摘要： 背景　多发性骨髓瘤（MM）是一种浆细胞恶性肿瘤，目前尚无法治愈，面临复发与难治的困境。嵌合抗原受体T淋巴细胞（CAR-T）是一种在MM治疗中有巨大潜力的新型免疫疗法，但目前关于其疗效与安全性的评价较少。目的　系统评价CAR-T治疗MM的疗效及安全性，为其安全高效应用提供循证医学证据。方法　计算机检索PubMed、EMBase、The Cochrane Library数据库，筛选自建库至2019年11月已发表的CAR-T治疗MM的临床试验英文文献，以总反应率、严重不良事件（AEs）发生率（≥3级）、细胞因子释放综合征（CRS）发生率为结局指标，使用R软件进行单组率的Meta分析，并按照预处理方案、共刺激分子、CAR-T抗原类型分组进行亚组分析及Meta回归分析。结果　共纳入8项临床研究，包括191例患者，Meta分析结果显示：CAR-T治疗MM的总反应率为0.88〔95%CI（0.83，0.93）〕，严重AEs发生率为0.90〔95%CI（0.83，0.98）〕，CRS发生率为0.92〔95%CI（0.85，1.00）〕。亚组分析及Meta回归分析结果提示：不同预处理方案MM患者严重AEs发生率、CRS发生率比较，差异均有统计学意义（P=0.013，P<0.001）；其中，Flu+Cy组、Cy组、无预处理组严重AEs发生率、CRS发生率高于HDM+ASCT组（P<0.05）。共刺激分子4-1BB组总反应率高于CD28组，严重AEs发生率低于CD28组（P<0.05）。不同抗原类型MM患者总反应率、严重AEs发生率、CRS发生率比较，差异均有统计学意义（P=0.004、0.049、0.023）；其中，BCMA组总反应率低于LCAR-B38M组、BCMA+CD19组，严重AEs发生率、CRS发生率高于CD19组（P<0.05）；LCAR-B38M组、BCMA+CD19组严重AEs发生率、CRS发生率高于CD19组（P<0.05）。结论　CAR-T治疗MM具有良好的疗效，虽然有一定的可控安全风险。预处理方案、共刺激分子以及CAR-T抗原类型均是影响疗效及安全性的因素。因纳入研究样本量较少、质量不一，上述结论尚待大样本、高质量随机对照临床试验进一步验证。

关键词: 多发性骨髓瘤, 嵌合抗原受体T淋巴细胞, 总反应率, 不良事件, 细胞因子释放综合征, Meta分析, 系统评价