PO.CL02.01 · 临床研究
辅助CDK4/6抑制剂在早期乳腺癌中的疗效比较：一项使用NetMetaEasy平台的基于风险比的网状Meta分析
Comparative efficacy of adjuvant CDK4/6 inhibitors in early breast cancer: A hazard ratio-based network meta-analysis using the NetMetaEasy platform
该海报暂无可下载的资料
AACR 官方页面
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景：辅助CDK4/6抑制剂在激素受体阳性、HER2阴性早期乳腺癌的III期试验中显示出不一致的结果，且缺乏直接比较来阐明其相对疗效。网状Meta分析（NMA）能够整合直接和间接证据，同时比较多种治疗方案。在本研究中，我们应用基于风险比的NMA，以浸润性无病生存期（iDFS）为主要终点，评估abemaciclib、ribociclib和palbociclib的疗效比较。内分泌治疗（ET）作为各试验中的标准治疗参照。
方法：分析了来自四项随机III期试验的数据（病例数n=17,741）。治疗节点包括abemaciclib+内分泌治疗（ET）、ribociclib+ET、palbociclib+ET和单用ET。分析使用NetMetaEasy平台（https://www.metaanalysisonline.com/netmetaeasy）进行，该平台是一个基于R/Shiny的实现，支持二分类、连续和汇总效应输入。采用逆方差法并使用DerSimonian-Laird估计量估计研究间方差（τ²），进行了频率学派随机效应NMA。生成了标准的NMA输出，包括森林图、网络几何结构、治疗排序（P-scores）、不一致性评估和漏斗图。
结果：ribociclib+ET相较于ET显示出最强的效应（HR 0.72；95% CI 0.62-0.83）。abemaciclib+ET也改善了iDFS（HR 0.73；95% CI 0.66-0.82）。palbociclib+ET与ET相比无显著差异（HR 0.89；95% CI 0.79-1.00）。间接比较结果为：abemaciclib对比palbociclib的HR为0.82（95% CI 0.70-0.97），palbociclib对比ribociclib的HR为1.25（95% CI 1.03-1.50），abemaciclib对比ribociclib的HR为1.03（95% CI 0.85-1.23）。P-scores将ribociclib+ET排在最高（0.86），其次为abemaciclib+ET（0.80）和palbociclib+ET（0.33）。漏斗图检查未提示小样本研究效应。
结论：在这项基于风险比的NMA中，abemaciclib和ribociclib相较于内分泌治疗显著改善了iDFS，而palbociclib则未显示出改善。我们的研究结果说明了在缺乏头对头试验的情况下，NMA用于疗效比较评估的实用性。
查看英文原文 English abstract
BACKGROUND: Adjuvant CDK4/6 inhibitors have shown variable results across phase III trials in hormone receptor-positive, HER2-negative early breast cancer, and no direct comparisons exist to clarify their relative efficacy. Network meta-analysis (NMA) enables the integration of direct and indirect evidence to compare multiple treatments simultaneously. In this study we applied a hazard ratio-based NMA to evaluate the comparative effectiveness of abemaciclib, ribociclib, and palbociclib using invasive disease-free survival (iDFS) as the primary endpoint. Endocrine therapy (ET) served as the standard-of-care reference across trials.
METHODS: Data from four randomized phase III trials (case n=17,741) were analyzed. Treatment nodes included abemaciclib+endocrine therapy (ET), ribociclib+ET, palbociclib+ET, and ET alone. Analyses were performed using the NetMetaEasy platform (https://www.metaanalysisonline.com/netmetaeasy), an R/Shiny-based implementation supporting binary, continuous, and summary-effect inputs. A frequentist random-effects NMA was conducted using the inverse-variance method with the DerSimonian-Laird estimator for between-study variance (τ²). Standard NMA outputs were generated, including forest plots, network geometry, treatment ranking (P-scores), inconsistency assessments, and funnel plots.
RESULTS: Ribociclib+ET showed the strongest effect versus ET (HR 0.72; 95% CI 0.62-0.83). Abemaciclib+ET also improved iDFS (HR 0.73; 95% CI 0.66-0.82). Palbociclib+ET did not differ significantly from ET (HR 0.89; 95% CI 0.79-1.00). Indirect comparisons yielded HR 0.82 (95% CI 0.70-0.97) for abemaciclib versus palbociclib, HR 1.25 (95% CI 1.03-1.50) for palbociclib versus ribociclib, and HR 1.03 (95% CI 0.85-1.23) for abemaciclib versus ribociclib. P-scores ranked ribociclib+ET highest (0.86), followed by abemaciclib+ET (0.80) and palbociclib+ET (0.33). Funnel-plot inspection did not indicate small-study effects.
CONCLUSIONS: In this hazard ratio-based NMA, abemaciclib and ribociclib significantly improved iDFS compared with endocrine therapy, whereas palbociclib did not. Our findings illustrate the utility of NMA for comparative evaluation in the absence of head-to-head trials.
利益披露 Disclosure
J. Fekete, None..
B. Gyorffy, None.