CSF1R抑制剂治疗复发或不可切除腱鞘巨细胞瘤的疗效与安全性：一项Ⅲ期随机对照试验的系统评价和Meta分析。
完整摘要 · 中文译文
目的：腱鞘巨细胞瘤（TGCT）是一种罕见的、局部侵袭性肿瘤，由集落刺激因子1（CSF1）过表达驱动。我们对CSF1受体抑制剂（CSF1Ri）与安慰剂在复发性或不可切除TGCT（定义为因手术会导致功能限制加重或严重并发症的症状性疾病）中的疗效和安全性进行了荟萃分析。 方法：检索了PubMed/MEDLINE、Scopus、Google Scholar、CENTRAL和ClinicalTrials.gov注册库，时间从建库至2026年3月25日，纳入比较任何CSF1Ri与安慰剂在经组织学确诊的成人TGCT患者中的III期随机对照试验（RCT）。主要结局为客观缓解率（ORR）；次要结局为身体功能（PROMIS-PF）、关节活动度（ROM）、疼痛、健康相关生活质量（EQ-VAS）以及3级或以上不良事件。评估了偏倚风险（RoB 2）和证据确定性（GRADE）；使用随机效应模型（RevMan 5.4.1）合并效应估计，并进行了固定效应、Hartung-Knapp-Sidik-Jonkman、风险差和逐项剔除敏感性分析。 结果：纳入3项随机对照试验（ENLIVEN、MOTION、MANEUVER；n = 337），均存在低偏倚风险。CSF1Ri显著改善了ORR（RR 25.55 [95% CI 6.43-101.52]；风险差0.43 [0.36-0.50]）、PROMIS-PF（MD 3.81 [2.28-5.34]）、相对ROM（MD 11.43 [6.53-16.34]个百分点）和EQ-VAS（MD 7.63 [4.03-11.23]）（所有I2 = 0%，p < 0.001），并减轻了疼痛（SMD -1.03 [-1.65至-0.42]；p = 0.0009；I2 = 76%）。3级或以上不良事件在CSF1Ri组更常见。 (RR 4.08 [2.31-7.21]; p < 0.001; I2 = 0%)。功能结局的确定性为高，ORR及3级或以上不良事件的确定性为中度，疼痛的确定性为低。 结论：在复发性或不可切除的腱鞘巨细胞瘤（TGCT）中，与安慰剂相比，CSF1Ri可带来肿瘤缓解、功能及生活质量方面具有临床意义的改善，但代价是严重不良事件发生频率更高，且不同试验中严重不良事件的性质存在差异。所纳入的试验未对各药物进行直接比较，因此不同CSF1Ri药物间的相对疗效与安全性仍不确定。
完整摘要 · 英文原文
AIM: Tenosynovial giant cell tumor (TGCT) is a rare, locally aggressive neoplasm driven by overexpression of colony-stimulating factor 1 (CSF1). We performed a meta-analysis of the efficacy and safety of CSF1 receptor inhibitors (CSF1Ri) versus placebo in recurrent or unresectable TGCT, defined as symptomatic disease for which surgery would cause worsening functional limitation or severe morbidity. METHODS: PubMed/MEDLINE, Scopus, Google Scholar, CENTRAL, and the ClinicalTrials.gov registry were searched from inception to March 25, 2026 for phase III randomized controlled trials (RCTs) comparing any CSF1Ri with placebo in adults with histologically confirmed TGCT. The primary outcome was objective response rate (ORR); secondary outcomes were physical function (PROMIS-PF), range of motion (ROM), pain, health-related quality of life (EQ-VAS), and Grade 3 or higher adverse events. Risk of bias (RoB 2) and certainty (GRADE) were assessed; estimates were pooled using random-effects models (RevMan 5.4.1), with fixed-effect, Hartung-Knapp-Sidik-Jonkman, risk-difference, and leave-one-out sensitivity analyses. RESULTS: Three RCTs were included (ENLIVEN, MOTION, MANEUVER; n = 337), all at low risk of bias. CSF1Ri significantly improved ORR (RR 25.55 [95% CI 6.43-101.52]; risk difference 0.43 [0.36-0.50]), PROMIS-PF (MD 3.81 [2.28-5.34]), relative ROM (MD 11.43 [6.53-16.34] percentage points), and EQ-VAS (MD 7.63 [4.03-11.23]) (all I2 = 0%, p < 0.001), and reduced pain (SMD -1.03 [-1.65 to -0.42]; p = 0.0009; I2 = 76%). Grade 3 or higher adverse events were more frequent with CSF1Ri (RR 4.08 [2.31-7.21]; p < 0.001; I2 = 0%). Certainty was high for functional outcomes, moderate for ORR and Grade 3 or higher adverse events, and low for pain. CONCLUSIONS: In recurrent or unresectable TGCT, CSF1Ri produce clinically meaningful improvements in tumor response, function, and quality of life compared with placebo, at the cost of more frequent severe adverse events, the nature of which differed between trials. The included trials did not compare agents directly, so the relative efficacy and safety of individual CSF1Ri remain uncertain.