不同生物标志物和终点在改善疑似前列腺癌转介中的效果：TARGET 研究（分层综合检测以早期诊断有临床意义的前列腺肿瘤）。
The efficacy of different biomarkers and endpoints to refine referrals for suspected prostate cancer: the TARGET study (Tiered integrAted tests for eaRly diaGnosis of clinically significant ProstatE Tumours).
机构信息
Division of Population Health, Health Services Research & Primary Care, University of Manchester, Manchester, UK.
Division of Urology, Department of Surgery, University of Cambridge, Cambridge, UK.
出版信息
BMC Med. 2024 Oct 8;22(1):440. doi: 10.1186/s12916-024-03667-7.
BACKGROUND
The majority of men referred with a raised PSA for suspected prostate cancer will receive unnecessary tertiary investigations including MRI and biopsy. Here, we compared different types of biomarkers to refine tertiary referrals and when different definitions of clinically significant cancer were used.
METHODS
Data and samples from 798 men referred for a raised PSA (≥ 3 ng/mL) and investigated through an MRI-guided biopsy pathway were accessed for this study. Bloods were acquired pre-biopsy for liquid biomarkers and germline DNA. Variables explored included PSA + Age (base model), free/total PSA (FTPSA), Prostate Health Index (phi), PSA density (PSAd), polygenic risk score (PRS) and MRI (≥ LIKERT 3). Different diagnostic endpoints for significant cancer (≥ grade group 2 [GG2], ≥ GG3, ≥ Cambridge Prognostic Group 2 [CPG2], ≥ CPG3) were tested. The added value of each biomarker to the base model was evaluated using logistic regression models, AUC and decision curve analysis (DCA) plots.
RESULTS
The median age and PSA was 65 years and 7.13 ng/mL respectively. Depending on definition of clinical significance, ≥ grade group 2 (GG2) was detected in 57.0% (455/798), ≥ GG3 in 27.5% (220/798), ≥ CPG2 in 61.6% (492/798) and ≥ CPG3 in 42.6% (340/798). In the pre-MRI context, the PSA + Age (base model) AUC for prediction of ≥ GG2, ≥ GG3, ≥ CPG2 and ≥ CPG3 was 0.66, 0.68, 0.70 and 0.75 respectively. Adding phi and PSAd to base model improved performance across all diagnostic endpoints but was notably better when the composite CPG prognostic score was used: AUC 0.82, 0.82, 0.83, 0.82 and AUC 0.74, 0.73, 0.79, 0.79 respectively. In contrast, neither FTPSA or PRS scores improved performance especially in detection of ≥ GG3 and ≥ CPG3 disease. Combining biomarkers did not alter results. Models using phi and PSAd post-MRI also improved performances but again benefit varied with diagnostic endpoint. In DCA analysis, models which incorporated PSAd and phi in particular were effective at reducing use of MRI and/or biopsies especially for ≥ CPG3 disease.
CONCLUSION
Incorporating phi or PSAd can refine and tier who is referred for tertiary imaging and/or biopsy after a raised PSA test. Incremental value however varied depending on the definition of clinical significance and was particularly useful when composite prognostic endpoints are used.
背景
大多数因疑似前列腺癌而 PSA 升高而被转介的男性将接受不必要的三级检查，包括 MRI 和活检。在这里，我们比较了不同类型的生物标志物，以在使用不同的临床显著癌症定义时细化三级转诊。
方法
本研究共纳入了 798 名因 PSA（≥3ng/ml）升高而接受 MRI 引导活检的男性。在研究前获取了血液样本以进行液体生物标志物和种系 DNA 检测。探索的变量包括 PSA+年龄（基础模型）、游离/总 PSA（FTPSA）、前列腺健康指数（phi）、PSA 密度（PSAd）、多基因风险评分（PRS）和 MRI（≥LIKERT 3）。测试了不同的临床显著癌症（≥分级组 2[GG2]、≥分级组 3[GG3]、≥剑桥预后组 2[CPG2]、≥剑桥预后组 3[CPG3]）的诊断终点。使用逻辑回归模型、AUC 和决策曲线分析（DCA）图评估每个生物标志物对基础模型的附加价值。
结果
中位年龄和 PSA 分别为 65 岁和 7.13ng/ml。根据临床意义的定义，≥分级组 2（GG2）在 57.0%（455/798）中检测到，≥分级组 3（GG3）在 27.5%（220/798）中检测到，≥剑桥预后组 2（CPG2）在 61.6%（492/798）中检测到，≥剑桥预后组 3（CPG3）在 42.6%（340/798）中检测到。在 MRI 之前，PSA+年龄（基础模型）预测≥GG2、≥GG3、≥CPG2 和≥CPG3 的 AUC 分别为 0.66、0.68、0.70 和 0.75。在基础模型中添加 phi 和 PSAd 可以提高所有诊断终点的性能，但当使用复合 CPG 预后评分时，效果更为显著：AUC 0.82、0.82、0.83、0.82 和 AUC 0.74、0.73、0.79、0.79。相比之下，FTPSA 或 PRS 评分均不能提高性能，尤其是在检测≥GG3 和≥CPG3 疾病时。联合使用生物标志物并没有改变结果。MRI 后使用 phi 和 PSAd 的模型也提高了性能，但再次受益取决于诊断终点。在 DCA 分析中，尤其将 PSAd 和 phi 纳入模型可以有效地减少 MRI 和/或活检的使用，特别是对于≥CPG3 疾病。
结论
在 PSA 升高的情况下，纳入 phi 或 PSAd 可以细化和分层转介接受三级影像学和/或活检的患者。然而，增量价值取决于临床显著意义的定义，在使用复合预后终点时尤其有用。
Zotero 插件