Table 2_Efficacy of therapeutic interventions for idiopathic recurrent pregnancy loss: a systematic review and network meta-analysis.docx

BackgroundApproximately 50% of cases of recurrent pregnancy loss (RPL) remain unexplained, and there is a lack of consensus concerning the effective treatments for idiopathic RPL. We used network meta-analyses to evaluate the efficacy of several prophylactic therapeutic interventions used in women with idiopathic RPL.
Materials and methodsWe conducted a systematic literature search using several databases from their inceptions to 20 July 2023. References from key articles were also manually searched. Randomized controlled trials assessing the efficacy and safety of any prophylactic intervention that were conducted in adult women with RPL were included. Studies with known causes of RPL were excluded. Two reviewers independently extracted data and assessed the risk of bias. Primary outcomes were live births and miscarriage rates. Secondary outcomes included serious adverse/adverse events and trial discontinuation. The network meta-analyses used a Bayesian hierarchical model with direct and indirect comparisons. Rank probabilities (assessed by surface under the cumulative ranking curve [SUCRA]) and certainty of evidence (assessed by Grading Recommendations Assessment, Development, and Evaluation [GRADE]) were also assessed.
ResultsThirty-eight studies (6,379 participants) were evaluated. No statistically significant differences in live birth rates among the interventions were found. The three best-ranked interventions for this outcome were prednisone plus progesterone plus aspirin (83%), leukocyte immune therapy (74%), and prednisolone (65%). Women who were treated with progesterone plus human chorionic gonadotrophin (instead of a placebo) presented an increase in miscarriage odds (odds ratio [OR] 3.83, 95% credible intervals [CrIs] 1.04–14.38). The three best-ranked interventions for miscarriage rate were prednisone plus progesterone plus aspirin (SUCRA = 81%), hydroxychloroquine (SUCRA = 79%), and intralipid (SUCRA = 65%). Overall, under placebo, 59% (95% confidence interval [CI] 51–67; I2 = 92%) of participants underwent successful live births, and 35% (95% CI 30–42, I2 = 86%) underwent miscarriages. We found no evidence of statistically significant differences between interventions (the top three interventions were low-molecular-weight heparin, granulocyte colony-stimulating factor, and leukocyte immune therapy) in those who discontinued trial participation.
ConclusionOur results suggest that none of the analyzed interventions led to improvements in the live birth rate or a reduction in the miscarriage rate in women with idiopathic RPL.
Systematic review registrationhttps://www.crd.york.ac.uk/prospero, identifier CRD42023455668.

背景 约50%的复发性流产（Recurrent Pregnancy Loss, RPL）病例仍病因不明，且针对特发性复发性流产（idiopathic RPL）的有效治疗方案尚未达成共识。本研究采用网络meta分析，评估了用于特发性复发性流产女性的多种预防性治疗干预措施的疗效。
材料与方法 本研究对多个数据库自建库至2023年7月20日的文献进行系统检索，并手动检索关键文献的参考文献。纳入针对复发性流产成年女性的、评估任意预防性干预措施疗效与安全性的随机对照试验，排除已知病因明确的复发性流产相关研究。由2名评价员独立提取数据并对偏倚风险进行评估。主要结局指标为活产率与流产率，次要结局指标包括严重不良/不良事件及试验脱落率。本研究的网络meta分析采用贝叶斯分层模型进行直接与间接比较，同时评估了排序概率（通过累积排序曲线下面积[SUCRA]计算）与证据质量（通过推荐分级的评估、制定与评价[GRADE]体系评估）。
结果 本研究共纳入38项研究，涉及6379名受试者。未发现各干预措施间的活产率存在统计学显著差异。该结局指标排名前三的干预措施依次为泼尼松+孕酮+阿司匹林（83%）、白细胞免疫治疗（74%）及泼尼松龙（65%）。接受孕酮+人绒毛膜促性腺激素（而非安慰剂）治疗的受试者，其流产风险升高（比值比[Odds Ratio, OR] = 3.83，95%可信区间[Credible Intervals, CrIs]：1.04~14.38）。针对流产率排名前三的干预措施依次为泼尼松+孕酮+阿司匹林（SUCRA=81%）、羟氯喹（SUCRA=79%）及脂肪乳（SUCRA=65%）。整体而言，安慰剂组受试者的活产成功率为59%（95%置信区间[Confidence Interval, CI]：51~67；I²=92%），流产率为35%（95% CI：30~42；I²=86%）。在试验脱落受试者中，未发现各干预措施间存在统计学显著差异，排名前三的干预措施分别为低分子肝素、粒细胞集落刺激因子及白细胞免疫治疗。
结论 本研究结果显示，所评估的所有干预措施均未提升特发性复发性流产女性的活产率，亦未降低其流产率。
系统评价注册 本研究已在https://www.crd.york.ac.uk/prospero平台完成注册，注册标识符为CRD42023455668。