基于限制性平均生存时间的一线治疗转移性透明细胞肾细胞癌的网络荟萃分析。
完整摘要 · 中文译文
比较基于风险比（HR）与基于限制性平均生存时间差值（dRMST）的网络荟萃分析，用于转移性透明细胞肾细胞癌的一线免疫检查点抑制剂方案。我们还引入了一个基于RMST的框架，将生存获益与毒性、皮质类固醇负担和药品采购成本结合起来进行背景分析。纳入了五项比较ICI联合方案与舒尼替尼治疗mccRCC的III期试验。已发表的试验水平HR用于基于HR的NMA，而重建的个体患者数据用于估计OS在40至64个月及PFS在24至48个月的dRMST。采用频率学派共同效应模型。对排序一致性进行描述性评估。将≥3级治疗相关不良事件（TRAEs）、大剂量皮质类固醇使用以及增量药品采购成本相对于获得的RMST月数进行评估。所有方案相对于舒尼替尼均显著改善了OS和PFS的dRMST，分别除外avelumab联合阿昔替尼和纳武利尤单抗联合伊匹木单抗。对于OS，纳武利尤单抗联合伊匹木单抗按HR排序第一，而纳武利尤单抗联合卡博替尼在每个dRMST时间点上均排序第一。仑伐替尼联合帕博利珠单抗在两种方法下均对PFS排序第一。在各时间点上，HR与dRMST的一致性在PFS中更高，并且两种终点的一致性均随时间增加。dRMST提供了更清晰的时间分离，并区分了获益的早期累积与后期累积。纳武利尤单抗联合伊匹木单抗的≥3级TRAE风险最低，且每获得一个RMST月的药品采购成本最低。基于RMST的NMA得出了tr 治疗排序与基于风险比（HR）的分析大体一致，同时对不同方案之间的区分更为清晰。这些发现凸显了限制性平均生存时间（RMST）在治疗决策中整合疗效与毒性、皮质类固醇负担及费用的临床价值。
完整摘要 · 英文原文
To compare hazard ratio (HR)- and difference in restricted mean survival time (dRMST)-based network meta-analyses of first-line immune checkpoint inhibitor-based regimens for metastatic clear-cell renal cell carcinoma. We also introduce a RMST-based framework contextualizing survival benefit alongside toxicity, corticosteroid burden, and drug-acquisition cost. Five phase III trials comparing ICI-based combinations with sunitinib in mccRCC were included. Published trial-level HRs informed HR-based NMAs, while reconstructed individual patient data were used to estimate dRMST for OS at 40 to 64 months and PFS at 24 to 48 months. Frequentist common-effect models were used. Ranking concordance was assessed descriptively. Grade ≥3 treatment-related adverse events (TRAEs), high-dose corticosteroid use, and incremental drug-acquisition cost were evaluated relative to RMST-months gained. All regimens significantly improved dRMST vs. sunitinib for OS and PFS, except avelumab plus axitinib and nivolumab plus ipilimumab, respectively. For OS, nivolumab plus ipilimumab ranked first by HR, whereas nivolumab plus cabozantinib ranked first at every dRMST horizon. Lenvatinib plus pembrolizumab ranked first for PFS under both approaches. HR-dRMST concordance was greater in PFS across horizons, and concordance increased over time for both endpoints. dRMST provided clearer temporal separation and distinguished early vs. later accumulation of benefit. Nivolumab plus ipilimumab had the lowest grade ≥3 TRAE risk and the lowest drug-acquisition cost per RMST-month gained. RMST-based NMA yielded treatment rankings largely concordant to HR-based analysis while providing clearer distinction between regimens. These findings highlight the clinical value of RMST for integrating efficacy with toxicity, corticosteroid burden, and cost in treatment decision-making.