文献速递妇科

绝经泌尿生殖综合征现代治疗方式的比较疗效与安全性:系统评价与网状Meta分析

Comparative Efficacy and Safety of Modern Treatment Modalities for Genitourinary Syndrome of Menopause: A Systematic Review and Network Meta-Analysis.
PubMed 文献 · BJOG · 2026-11-08

该系统评价与网状Meta分析纳入24个试验家族共4958名随机参与者,比较绝经泌尿生殖综合征的多种治疗。与安慰剂/假干预/对照相比,仅局部阴道雌激素的置信区间排除无效(SMD 0.34,95%CI 0.02-0.66)。作者认为局部阴道雌激素仍是症状性GSM证据最充分的基准治疗,数据支持个体化阶梯治疗而非确定性治疗排序。

要点

  • 与安慰剂/假干预/对照相比,仅局部阴道雌激素的置信区间排除无效(SMD 0.34,95%CI 0.02-0.66)。
  • 普拉睾酮、透明质酸、奥培米芬、保湿剂/润滑剂及CO2激光的估计值均不精确。
  • 作者认为CO2激光和射频不应被定位为等效或更优的一线治疗。

核心结论

  • 局部阴道雌激素仍是症状性GSM证据最充分的基准治疗。
  • 透明质酸、普拉睾酮和奥培米芬有临床相关证据,但跨类别估计不精确且排序不稳定。
  • 作者认为基于现有证据,CO2激光和射频不应被定位为等效或更优的一线治疗。
  • 数据支持个体化阶梯治疗而非确定性治疗排序。

研究设计与样本

  • 系统评价与类别级网状Meta分析,纳入随机对照试验。
  • 检索MEDLINE/PubMed、Scopus、Cochrane CENTRAL、Web of Science、Google Scholar、ClinicalTrials.gov和WHO ICTRP,从建库至2026年3月,并进行前后向引文检索。
  • 纳入标准为绝经后女性伴有症状性GSM或等同诊断的随机对照试验。
  • 26篇疗效出版物代表24个独特试验家族,经分母裁定后共4958名随机参与者。
  • 主要症状网状Meta分析纳入15个试验家族、7个节点和3415名参与者。

主要发现

  • 与安慰剂/假干预/对照相比,局部阴道雌激素是唯一置信区间排除无效的类别(SMD 0.34,95%CI 0.02-0.66)。
  • 普拉睾酮估计值为0.37(95%CI -0.10至0.84),透明质酸为0.31(95%CI -0.10至0.72),奥培米芬为0.24(95%CI -0.38至0.87),保湿剂/润滑剂为0.15(95%CI -0.37至0.67),CO2激光为0.03(95%CI -0.56至0.61),均不精确。
  • 异质性仍然显著(τ2=0.097;I2=65.6%)。
  • 节点拆分未显示统计学可检测的局部不一致性(所有p≥0.35),全局设计-by-治疗交互分解也无统计学意义(Q=4.90,df=3,p=0.179)。
  • 局部雌激素和普拉睾酮的P评分均为0.704,排序具有敏感性依赖性。

关键数值与推荐

  • 主要结局为统一的患者报告症状负担,采用预设层级:全局症状评分、最困扰症状、性交痛和阴道干涩。
  • 采用频率学派随机效应一致性网状Meta分析,生成Hedges g标准化均数差(SMD)、95%置信区间(CI)、预测区间和P评分。
  • 异质性通过REML估计;节点拆分、设计-by-治疗交互分解和敏感性分析评估不一致性和影响。
  • 所有21个网络比较的置信度通过CINeMA使用经验证的百分比贡献矩阵评估。

临床怎么用

  • 作者认为局部阴道雌激素仍是症状性GSM证据最充分的基准治疗。
  • 作者认为透明质酸、普拉睾酮和奥培米芬有临床相关证据,但跨类别估计不精确且排序不稳定。
  • 作者认为基于现有证据,CO2激光和射频不应被定位为等效或更优的一线治疗。
  • 作者认为数据支持个体化阶梯治疗而非确定性治疗排序。

局限性

  • 异质性仍然显著(τ2=0.097;I2=65.6%)。
  • 局部阴道雌激素和普拉睾酮 versus 安慰剂/假干预/对照的CINeMA置信度为低,其余19个网络比较为极低。
  • 跨类别估计不精确且排序不稳定。
  • 射频未纳入主要症状网状Meta分析,因为无法获得具有可提取离散度的层级兼容连续症状效应。
  • 总体RoB 2:2个试验家族为低,11个为一些担忧,11个为高。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Genitourinary syndrome of menopause (GSM) is a chronic consequence of hypoestrogenism affecting vaginal, sexual and lower urinary tract health. Treatments include local vaginal oestrogen, intravaginal prasterone, oral ospemifene, hyaluronic acid preparations, moisturisers/lubricants and energy-based interventions, but comparative evidence is heterogeneous. OBJECTIVES: To compare the relative efficacy and safety of contemporary GSM therapies using a systematic review and class-level network meta-analysis (NMA) of randomised controlled trials. SEARCH STRATEGY: MEDLINE/PubMed, Scopus, Cochrane CENTRAL, Web of Science, Google Scholar, ClinicalTrials.gov and WHO ICTRP were searched from inception to March 2026, with backward and forward citation searching. SELECTION CRITERIA: Randomised controlled trials enrolling postmenopausal women with symptomatic GSM or equivalent diagnoses. DATA COLLECTION AND ANALYSIS: Two reviewers independently screened, extracted and linked reports from the same trial family and assessed RoB 2. The primary outcome was harmonised patient-reported symptom burden, using a prespecified hierarchy of global symptom score, most bothersome symptom, dyspareunia and vaginal dryness. A frequentist random-effects consistency NMA generated; Hedges g standardised mean differences (SMDs), 95% confidence intervals (CIs), prediction intervals and P-scores. Heterogeneity was estimated by REML; node splitting, a design-by-treatment interaction decomposition and sensitivity analyses assessed incoherence and influence. Confidence in all 21 network comparisons was evaluated with CINeMA using a validated percentage contribution matrix. RESULTS: Twenty-six efficacy publications represented 24 unique trial families and 4958 randomised participants after denominator adjudication. The primary symptom NMA included 15 trial families, seven nodes and 3415 participants. Versus placebo/sham/control, local vaginal oestrogen was the only class with a CI excluding the null (SMD 0.34, 95% CI 0.02-0.66). Estimates were imprecise for prasterone (0.37, -0.10 to 0.84), hyaluronic acid (0.31, -0.10 to 0.72), ospemifene (0.24, -0.38 to 0.87), moisturiser/lubricant (0.15, -0.37 to 0.67) and CO2 laser (0.03, -0.56 to 0.61). Heterogeneity remained material (τ2 = 0.097; I2 = 65.6%). Node splitting showed no statistically detectable local incoherence (all p ≥ 0.35) and the global design-by-treatment interaction decomposition was also non-significant (Q = 4.90, df = 3, p = 0.179). P-scores were similar for local oestrogen (0.704) and prasterone (0.704) and rankings were sensitivity-dependent. CINeMA confidence was low for local oestrogen versus placebo/sham/control and prasterone versus placebo/sham/control and very low for the remaining 19 network comparisons. Radiofrequency was not included in the primary symptom NMA because a hierarchy-compatible continuous symptom effect with extractable dispersion was unavailable. Overall RoB 2 was low for 2 trial families, some concerns for 11 and high for 11. CONCLUSIONS: Local vaginal oestrogen remains the best-supported benchmark for symptomatic GSM. Hyaluronic acid, prasterone and ospemifene have clinically relevant evidence, but cross-class estimates are imprecise and rankings are unstable. CO2 laser and radiofrequency should not be positioned as equivalent or superior first-line treatments on the available evidence. The data support individualised stepped care rather than a definitive treatment hierarchy. TRIAL REGISTRATION: PROSPERO CRD420261332305.

来源:BJOG,查看原文。版权归原作者所有。

引用

  1. Randombage PJS, Akurugodagama MS. Comparative Efficacy and Safety of Modern Treatment Modalities for Genitourinary Syndrome of Menopause: A Systematic Review and Network Meta-Analysis. BJOG. 2026. PMID: 43000041. doi:10.1111/1471-0528.70370

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。