Princess Margaret Cancer Centre, University Health Network, Ontario, Canada; Department of Radiation Oncology, University of Toronto, Ontario, Canada; Techna Institute, University Health Network, Ontario, Canada.
Princess Margaret Cancer Centre, University Health Network, Ontario, Canada.
Int J Radiat Oncol Biol Phys. 2019 Jan 1;103(1):84-91. doi: 10.1016/j.ijrobp.2018.08.030. Epub 2018 Aug 29.
The National Comprehensive Cancer Network (NCCN) has recently endorsed the stratification of intermediate-risk prostate cancer (IR-PCa) into favorable and unfavorable subgroups and recommend the addition of androgen deprivation therapy (ADT) to radiation therapy (RT) for unfavorable IR-PCa. Recently, more accurate prognostication was demonstrated by integrating a 22-feature genomic classifier (GC) to the NCCN stratification system. Here, we test the utility of the GC to better identify patients with IR-PCa who are sufficiently treated by RT alone.
We identified a novel cohort comprising 121 patients with IR-PCa treated with dose-escalated image guided RT (78 Gy in 39 fractions) without ADT. GC scores were derived from tumors sampled in diagnostic biopsies. Multivariable analyses, including both NCCN subclassification and GC scores, were performed for biochemical failure (prostate-specific antigen nadir + 2 ng/mL) and metastasis occurrence.
By NCCN subclassification, 33 (27.3%) and 87 (71.9%) of men were classified as having favorable and unfavorable IR-PCa, respectively (1 case unclassifiable). GC scores were high in 3 favorable IR-PCa and low in 60 unfavorable IR-PCa. Higher GC scores, but not NCCN risk subgroups, were associated with biochemical relapse (hazard ratio, 1.36; 95% confidence interval [CI], 1.09-1.71] per 10% increase; P = .007) and metastasis (hazard ratio, 2.05; 95% CI, 1.24-4.24; P = .004). GC predicted biochemical failure at 5 years (area under the curve, 0.78; 95% CI, 0.59-0.91), and the combinatorial NCCN + GC model significantly outperformed the NCCN alone model for predicting early-onset metastasis (area under the curve for 5-year metastasis of 0.89 vs 0.86 [GC alone] vs 0.54 [NCCN alone]).
We demonstrated the accuracy of the GC for predicting disease recurrence in IR-PCa treated with dose-escalated image guided RT alone. Our findings highlight the need to evaluate this GC in a prospective clinical trial investigating the role of ADT-RT in clinicogenomic-defined IR-PCa subgroups.
美国国家综合癌症网络（NCCN）最近将中危前列腺癌（IR-PCa）分为有利亚组和不利亚组，并建议对不利亚组的 IR-PCa 患者在放疗（RT）基础上加用雄激素剥夺治疗（ADT）。最近，通过整合一个包含 22 个特征的基因组分类器（GC）到 NCCN 分层系统，可更准确地预测预后。在此，我们检验了 GC 用于更好地识别仅接受 RT 充分治疗的 IR-PCa 患者的效用。
我们鉴定了一个新的队列，包含 121 例接受剂量递增图像引导 RT（78Gy/39 次）治疗且未接受 ADT 的 IR-PCa 患者。GC 评分来源于诊断性活检的肿瘤样本。进行了多变量分析，包括 NCCN 亚分类和 GC 评分，以评估生化失败（前列腺特异性抗原最低值+2ng/mL）和转移发生。
按照 NCCN 亚分类，33 例（27.3%）和 87 例（71.9%）患者被归类为有利和不利 IR-PCa，分别有 1 例无法分类（1 case unclassifiable）。GC 评分在 3 例有利 IR-PCa 中较高，在 60 例不利 IR-PCa 中较低。GC 评分较高，但不是 NCCN 风险亚组，与生化复发（每增加 10%的风险比为 1.36；95%置信区间[CI]为 1.09-1.71；P=0.007）和转移（风险比为 2.05；95%CI 为 1.24-4.24；P=0.004）相关。GC 预测了 5 年的生化失败（曲线下面积为 0.78；95%CI 为 0.59-0.91），且组合的 NCCN+GC 模型在预测早期转移方面明显优于单独的 NCCN 模型（5 年转移的曲线下面积为 0.89 对 0.86[GC 单独]对 0.54[NCCN 单独]）。
我们证实了 GC 对预测接受剂量递增图像引导 RT 治疗的 IR-PCa 疾病复发的准确性。我们的研究结果强调需要在一个前瞻性临床试验中评估该 GC，以研究 ADT-RT 在临床基因组定义的 IR-PCa 亚组中的作用。
