Validation of Digital Morphometry for Cancer Risk in Benign Prostate Biopsies

描述(申请人提供)：在PCPT和Reduced试验中，非那雄胺和度他雄胺显著降低了前列腺癌(PCa)的检测率，分别降低了23%和25%，从而确立了51-还原酶抑制剂(5ARI)是第一类被证明为前列腺癌化学预防药物。这些药物的实际疗效因人而异，即使在考虑遵守处方剂量后也是如此。更好地了解对5ARI化学预防敏感的基础将：1)为利用这一既定机制开发耐受性更好、更有效的药物铺平道路；2)使临床医生能够针对有反应的男性，从而降低终生剂量的成本和发病率；以及3)验证特定的组织生物标记物作为第二阶段试验的替代终点。我们基于直接DNA染色的初步数据表明，表征5ARI反应的核形态计量学特征也可能定义一种场效应，预测未经治疗的活检阴性男性的PCA。2009年完成的RECESS为解决这些问题提供了一个独特的机会，因为收集了中间的、研究中的活组织检查样本(第2年和第4年是强制性的)。先前的研究表明，5ARI(短期内高剂量)会导致良性前列腺的变化，类似于不完全萎缩。我们的总体目标是应用尖端成像方法，结合模式识别和多光谱分析的机器学习，开发和验证良性组织中的中间终点生物标记物，以表征对5ARI化学预防的反应以及阴性活检男性的前列腺癌风险。我们将从随机抽样的RECESS参与者中获得第二年的幻灯片和组织块，结果各不相同。目的1：确定度他雄胺(与安慰剂相比)对良性组织的核和结构特征的影响。目的2：确定服用度他雄胺期间发生前列腺癌的受试者与未服用度他雄胺的受试者之间的多变量治疗反应评分是否不同，以及目标3：确定良性活检中的核表型与未接受治疗的高风险男性随后发生前列腺癌的风险之间的关联程度。总而言之，我们将使用3种技术来评估细胞形态：a)基于核大小、形状和质地的形态计分；b)通过量子点成像表达p300和核仁蛋白(两种对染色质模式和核形态有主要影响的蛋白质)；以及c)通过可培训软件绘制结构特征(例如，上皮面积和高度)。这将是第一项将5ARI的细胞形态效应与随后的癌症发生联系起来的工作，5ARI长期给予化学预防剂量水平。这一结果将对化学预防研究和临床实践产生影响，包括大量美国男性在前列腺活检阴性后仍处于风险之中。

DESCRIPTION (provided by applicant): In the PCPT and REDUCE trials, finasteride and dutasteride significantly reduced the detection of prostate cancer (PCa) by 23 and 25% respectively, and thus established that 51-reductase inhibitors (5ARI) are the first class of drugs proven as chemopreventive agents for PCa. The actual efficacy of these drugs varies among individuals, even after considering compliance with the prescribed dosage. Better understanding of the basis for sensitivity to 5ARI chemoprevention will: 1) pave the way for development of better-tolerated and more effective agents that exploit this established mechanism, 2) allow clinicians to target responsive men, thus reducing the cost and morbidity of life-long dosage, and 3) validate specific tissue biomarkers as surrogate endpoints for Phase II trials. Our preliminary data based on direct DNA staining suggest that nuclear morphometric features that characterize 5ARI response may also define a field effect that predicts PCa in untreated men with negative biopsies. REDUCE, which was completed in 2009, provides a unique opportunity to address these questions because intermediate, on-study biopsy samples (mandatory at Years 2 and 4) were collected. Previous research indicated that 5ARIs (at higher doses for short periods) produce changes in benign prostate that resemble incomplete atrophy. Our overall goal is to apply cutting- edge imaging approaches, incorporating machine-learning for pattern recognition and multispectral analysis, to the development and validation of intermediate endpoint biomarkers in benign tissue that characterize the response to 5ARI chemoprevention as well as the risk of PCa among men with negative biopsies. We will obtain Year 2 slides and tissue blocks from a random sample of REDUCE participants with various outcomes. Aim 1: To determine the effects of dutasteride (vs. placebo) on both nuclear and architectural features in benign tissue. Aim 2: To determine whether a multivariable treatment-response score differs between subjects who develop PCa while on dutasteride and those who do not, and, Aim 3: To determine the magnitude of association between nuclear phenotype in benign biopsies, and subsequent risk of PCa in untreated men at elevated risk. Altogether, we will use 3 techniques to assess cytomorphology: a) a morphometric score based on nuclear size, shape and texture, b) expression -via quantum dot imaging - of p300 and nucleolin (two proteins with major effects on chromatin pattern and nuclear morphology) and c) mapping of architectural features (e.g., epithelial area and height) via trainable software. This will be the first work to relate the cytomorphological effects of a 5ARI, given at a chemopreventive dose level for a lengthy period, to subsequent cancer occurrence. The results will have implications for chemoprevention research and clinical practice, including the large number of U.S. men who remain at risk following a negative prostate biopsy.