摘要Although there are challenges in treating traumatic central nervous system diseases, mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) have recently proven to be a promising non-cellular therapy. We comprehensively evaluated the efficacy of mesenchymal stem cell-derived extracellular vesicles in traumatic central nervous system diseases in this meta-analysis based on preclinical studies. Our meta-analysis was registered at PROSPERO (CRD42022327904, May 24, 2022). To fully retrieve the most relevant articles, the following databases were thoroughly searched: PubMed, Web of Science, The Cochrane Library, and Ovid-Embase (up to April 1, 2022). The included studies were preclinical studies of mesenchymal stem cell-derived extracellular vesicles for traumatic central nervous system diseases. The Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE)'s risk of bias tool was used to examine the risk of publication bias in animal studies. After screening 2347 studies, 60 studies were included in this study. A meta-analysis was conducted for spinal cord injury (n = 52) and traumatic brain injury (n = 8). The results indicated that mesenchymal stem cell-derived extracellular vesicles treatment prominently promoted motor function recovery in spinal cord injury animals, including rat Basso, Beattie and Bresnahan locomotor rating scale scores (standardized mean difference [SMD]: 2.36, 95％ confidence interval [CI]: 1.96–2.76, P < 0.01, I2 = 71％) and mouse Basso Mouse Scale scores (SMD = 2.31, 95％ CI: 1.57–3.04, P = 0.01, I2 = 60％) compared with controls. Further, mesenchymal stem cell-derived extracellular vesicles treatment significantly promoted neurological recovery in traumatic brain injury animals, including the modified Neurological Severity Score (SMD = –4.48, 95％ CI: –6.12 to –2.84, P < 0.01, I2 = 79％) and Foot Fault Test (SMD = –3.26, 95％ CI: –4.09 to –2.42, P = 0.28, I2 = 21％) compared with controls. Subgroup analyses showed that characteristics may be related to the therapeutic effect of mesenchymal stem cell-derived extracellular vesicles. For Basso, Beattie and Bresnahan locomotor rating scale scores, the efficacy of allogeneic mesenchymal stem cell-derived extracellular vesicles was higher than that of xenogeneic mesenchymal stem cell-derived extracellular vesicles (allogeneic: SMD = 2.54, 95％ CI: 2.05–3.02, P = 0.0116, I2 = 65.5％; xenogeneic: SMD: 1.78, 95％CI: 1.1–2.45, P = 0.0116, I2 = 74.6％). Mesenchymal stem cell-derived extracellular vesicles separated by ultrafiltration centrifugation combined with density gradient ultracentrifugation (SMD = 3.58, 95％ CI: 2.62–4.53, P < 0.0001, I2 = 31％) may be more effective than other EV isolation methods. For mouse Basso Mouse Scale scores, placenta-derived mesenchymal stem cell-derived extracellular vesicles worked better than bone mesenchymal stem cell-derived extracellular vesicles (placenta: SMD = 5.25, 95％ CI: 2.45–8.06, P = 0.0421, I2 = 0％; bone marrow: SMD = 1.82, 95％ CI: 1.23–2.41, P = 0.0421, I2 = 0％). For modified Neurological Severity Score, bone marrow-derived MSC-EVs worked better than adipose-derived MSC-EVs (bone marrow: SMD = –4.86, 95％ CI: –6.66 to –3.06, P = 0.0306, I2 = 81％; adipose: SMD = –2.37, 95％ CI: –3.73 to –1.01, P = 0.0306, I2 = 0％). Intravenous administration (SMD = –5.47, 95％ CI: –6.98 to –3.97, P = 0.0002, I2 = 53.3％) and dose of administration equal to 100 μg (SMD = –5.47, 95％ CI: –6.98 to –3.97, P < 0.0001, I2 = 53.3％) showed better results than other administration routes and doses. The heterogeneity of studies was small, and sensitivity analysis also indicated stable results. Last, the methodological quality of all trials was mostly satisfactory. In conclusion, in the treatment of traumatic central nervous system diseases, mesenchymal stem cell-derived extracellular vesicles may play a crucial role in promoting motor function recovery.