Genetic profiling in PCPT: prostate cancer risk, PSA levels, and chemoprevention

描述(申请人提供)：全基因组关联研究发现，有十几个SNPs与前列腺癌(PCA)风险有关。这些已报道的与前列腺癌风险相关的SNP，如果不是由PSA检测偏差驱动的，或者与PSA不完全相关，可以用来改进PSA和其他现有的预测阳性前列腺癌(即前列腺癌)的临床变量。然而，大多数PCA病例对照研究无法解决这两个重要问题，因为一些病例是在PSA水平升高的基础上诊断出来的。只有前列腺癌预防试验(PCPT)这样的研究才能用来剖析这两个问题。在PCPT试验中，无论PSA水平如何，男性在试验结束时都会接受活组织检查。这项研究的总体假设是，当多种基因变异结合在一起时，可以用来预测患前列腺癌风险增加的男性。具体地说，我们假设1)部分遗传变异与PCa风险相关，而不仅仅是由于PSA检测偏差，2)这些遗传变异结合在一起，与PCa风险密切相关，3)遗传变异可以补充PSA和其他现有的临床变量，以提高对PCa和侵袭性PCa的预测性能，以及4)非那雄胺的化学预防作用在不同遗传风险的男性中不同，对于遗传风险较高的男性，非那雄胺对PCa诊断的减少幅度更大。为了验证这些假设，我们将使用PCPT研究的数据和样本，这是一项非那雄胺预防前列腺癌的III期随机、双盲、安慰剂对照试验。我们有三个具体目标。目的1是测试在没有PSA检测偏差的情况下，已报道的前列腺癌风险相关变异是否与前列腺癌风险相关。目的2是使用遗传变异以及PSA和其他现有的临床变量来评估联合预测前列腺癌诊断、总体前列腺癌和侵袭性前列腺癌的性能。目的3是评估非那雄胺对前列腺癌遗传风险较高或较低男性前列腺癌诊断的不同化学预防效果。这项研究的结果可能会让数百万男性受益。患前列腺癌风险最高的男性可以在很小的时候就被确定为进行密集筛查和化学预防，如非那雄胺。基因变异也可以与PSA和其他现有的临床变量结合使用，以显著提高它们对阳性前列腺活检的预测准确性。公共卫生相关性：数十种前列腺癌风险相关变种结合在一起，可以用来识别前列腺癌风险最高的男性。这样的男性可以在很小的时候就被识别出来，进行密集的筛查和化学预防。前列腺癌风险相关变量也可以与PSA和其他现有的临床变量结合使用，以显著提高它们对阳性前列腺活检的预测准确性。1

DESCRIPTION (provided by applicant): More than a dozen SNPs have been found to be associated with prostate cancer (PCa) risk by genome-wide association studies (GWAS). These reported PCa risk associated SNPs, if not driven by PSA detection bias or not completely correlated with PSA, could be used to improve PSA and other existing clinical variables in predicting positive prostate biopsy (i.e. PCa). These two important questions, however, cannot be addressed by most PCa case-control studies because some cases were diagnosed on the basis of elevated PSA levels. Only studies such as the Prostate Cancer Prevention Trial (PCPT) where men at the end of the trial were biopsied regardless of PSA levels can be used to dissect these two questions. The overall hypothesis of the study is that multiple genetic variants, when combined, can be used to predict men at increased risk for PCa. Specifically, we hypothesize that 1) a subset of genetic variants are associated with PCa risk and are not solely due to PSA detection bias, 2) these genetic variants, when combined, are strongly associated with PCa risk, 3) genetic variants can supplement PSA and other existing clinical variables to improve the predictive performance for PCa and aggressive PCa, and 4) the chemoprevention effect of finasteride is different among men with different genetic risks and the reduction in PCa diagnosis by finasteride is larger for men with higher genetic risk. To test these hypotheses, we will use data and samples from the PCPT study, a phase III randomized, double-blind, placebo-controlled trial of finasteride in the prevention of prostate cancer. We have three specific aims. Aim 1 is to test whether reported prostate cancer risk associated variants from GWAS are associated with PCa risk, free of PSA detection bias. Aim 2 is to estimate the joint predictive performance for prostate cancer diagnosis, overall PCa and aggressive PCa, using genetic variants as well as PSA and other existing clinical variables. Aim 3 is to assess the differential chemoprevention effect of finasteride on prostate cancer diagnosis among men with higher or lower genetic risk for PCa. Results from this study may potentially benefit millions men. Men at highest risk for PCa could be identified at an early age for intensive screening and chemoprevention such as finasteride. Genetic variants could also be used in combination with PSA and other existing clinical variables to considerably improve their predictive accuracy for positive prostate biopsy. PUBLIC HEALTH RELEVANCE: Dozens of prostate cancer risk associated variants, when combined, could be used to identify men at highest risk for prostate cancer. Such men could be identified at an early age for intensive screening and chemoprevention. Prostate cancer risk associated variants could also be used in combination with PSA and other existing clinical variables to considerably improve their predictive accuracy for positive prostate biopsy. 1