基于风险选择主动监测:Movember基金会GAP3前列腺癌主动监测倡议结果
该研究基于Movember基金会GAP3联盟25个队列的个体参与者数据荟萃分析,纳入5530名男性,旨在识别并验证主动监测1年或4年疾病再分类的预测因素。1年再分类率为13%(范围0%-31%),4年为14%(范围3%-40%),重要预测因素包括年龄、PSA、前列腺体积、T分期和阳性活检针数。内部-外部验证的AUC在1年和4年分别为0.68和0.61,提示现有指南可考虑加入前列腺体积以改善个体化选择。
要点
- 主动监测1年和4年活检再分类率分别为13%和14%,研究间异质性大。
- 年龄、PSA、前列腺体积、T分期和阳性针数是再分类的重要预测因素。
- 预测模型区分度有限(AUC 0.68和0.61),需额外预测因素改善患者选择。
核心结论
- 主动监测开始后1年和4年活检再分类率分别为13%和14%,研究间异质性显著。
- 当前指南可考虑纳入前列腺体积以改善主动监测的个体化选择。
- 需额外预测因素以提升主动监测患者选择的准确性。
研究设计与样本
- 基于Movember基金会GAP3联盟25个已建立队列的个体参与者数据荟萃分析。
- 共纳入5530名男性,采用随机效应逻辑回归模型评估关联。
- 通过内部-外部验证程序使用AUC曲线评估预测因素组合的区分能力。
主要发现
- 1年评估的5570名男性中,815例活检Gleason分级组升高,合并再分类率13%(范围0%-31%)。
- 4年评估的1515名男性中,205例再分类,合并再分类率14%(范围3%-40%)。
- 重要预测因素包括年龄、前列腺特异性抗原、前列腺体积、T分期和阳性活检针数。
- 内部-外部验证的平均AUC在1年和4年再分类分别为0.68和0.61。
关键数值与推荐
- 再分类定义为活检时Gleason分级组任何升高。
- 1年再分类率13%(范围0%-31%),4年再分类率14%(范围3%-40%)。
- 预测模型AUC:1年0.68,4年0.61。
临床怎么用
- 选择主动监测患者时可结合年龄、PSA、前列腺体积、T分期和阳性针数进行风险评估。
- 考虑将前列腺体积纳入现有指南以优化个体化选择。
- 需注意现有预测模型区分度有限,应谨慎用于临床决策。
局限性
- 研究间异质性大(再分类率范围宽)。
- 预测模型区分能力有限(AUC 0.68和0.61)。
- 需额外预测因素改善患者选择。
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
PURPOSE: We sought to identify and validate known predictors of disease reclassification at 1 or 4 years to support risk-based selection of patients suitable for active surveillance. MATERIALS AND METHODS: An individual participant data meta-analysis using data from 25 established cohorts within the Movember Foundations GAP3 Consortium. In total 5,530 men were included. Disease reclassification was defined as any increase in Gleason grade group at biopsy at 1 and 4 years. Associations were estimated using random effect logistic regression models. The discriminative ability of combinations of predictors was assessed in an internal-external validation procedure using the AUC curve. RESULTS: Among the 5,570 men evaluated at 1 year, we found 815 reclassifications to higher Gleason grade group at biopsy (pooled reclassification rate 13%, range 0% to 31%). Important predictors were age, prostate specific antigen, prostate volume, T-stage and number of biopsy cores with prostate cancer. Among the 1,515 men evaluated at 4 years, we found 205 reclassifications (pooled reclassification rates 14%, range 3% to 40%), with similar predictors. The average areas under the receiver operating characteristic curve at internal-external validation were 0.68 and 0.61 for 1-year and 4-year reclassification, respectively. CONCLUSIONS: Disease reclassification occurs typically in 13% to 14% of biopsies at 1 and 4 years after the start of active surveillance with substantial between-study heterogeneity. Current guidelines might be extended by considering prostate volume to improve individualized selection for active surveillance. Additional predictors are needed to improve patient selection for active surveillance.
引用
- Bruinsma SM, Nieboer D, Roobol MJ, et al. Risk-Based Selection for Active Surveillance: Results of the Movember Foundation's Global Action Plan Prostate Cancer Active Surveillance (GAP3) Initiative. J Urol. 2021. PMID: 33617330. doi:10.1097/JU.0000000000001700
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。