Prostate Cancer

Utility of Single Nucleotide Polymorphisms in Prostate Biopsy Decisions

Literature reviews News and Views From the Literature Prostate Cancer Utility of Single Nucleotide Polymorphisms in Prostate Biopsy Decisions Reviewed by Stacy Loeb, MD,1 R. Scott Braithwaite, MD, MSc,2 Richard B. Hayes, DDS, PhD, MPH3 Departments of Urology, 2General Internal Medicine, and Environmental Medicine, New York University School of Medicine, New York, NY [Rev Urol. 2012;14(3/4):115-117 doi: 10.3909/riu0551] genetic data improve decision making compared with current risk assessment tools or are useful to guide prostate cancer screening protocols. Because personal genomic testing has become commercially available to consumers, additional data are urgently needed on the incremental prognostic value of genetic information beyond established predictors of prostate cancer risk, such as prostate-specific antigen (PSA). This article reviews recent studies on the performance characteristics of genetic variables in combination with PSA to predict prostate cancer risk. 1 3 © 2013 MedReviews®, LLC Combining 33 Genetic Variants With PSA for Prediction of Prostate Cancer: Longitudinal Study Johansson M, Holmström B, Hinchliffe SR, et al. I n recent years, numerous single nucleotide polymorphisms (SNPs) have been identified that are associated with prostate cancer risk.1-6 Although each individual allele is associated with a relatively small increase in prostate cancer risk, there appears to be a cumulative relationship,7 as previously reviewed.8 Risk alleles may also be combined with information on family history to identify men with a significantly greater risk of prostate cancer.9 However, the demonstration of SNP-prostate cancer associations in genome-wide association studies (GWAS) does not prove that they are clinically useful.10,11 Additional study is needed regarding whether Int J Cancer. 2012;130:129-137. This study evaluated the performance of genetic variants compared to PSA in a nested case-control study within the Northern Sweden Health and Disease Cohort.12 In this longitudinal population-based cohort, men provided blood samples at initial recruitment that were frozen and later used for genotyping. The authors identified 520 incident prostate cancer cases with stored blood samples from a median of 7 years prior to diagnosis. As a comparison group, 988 prostate cancer-free control subjects were randomly selected and matched Vol. 14 No. 3/4 • 2012 • Reviews in Cardiovascular Medicine • 115 40041700002_RIU0551.indd 115 12/02/13 2:29 PM Prostate Cancer continued on age and date of venipuncture. The median age was 59 years at the blood draw. Men subsequently diagnosed with prostate cancer had a significantly higher median PSA level at the initial blood draw compared with controls (3.6 vs 1.1 ng/mL). Next, the authors created a genetic risk score based on carrier status for 33 previously published SNPs from the literature. On receiver operating characteristic analysis, a model including demographics, baseline PSA, and the ratio of free to total PSA (%fPSA) had an area under the curve (AUC) of 0.862 for predicting prostate cancer; the addition of the genetic risk score to this model led to a modest statistically significant improvement in the AUC to 0.872 (P 5 .002). However, the genetic risk score did not lead to a meaningful improvement in the prediction of high-risk prostate cancer (defined as stage T3/T4, Gleason $ 8 or World Health Organization grade 3, lymph node or bone metastases or PSA . 20 ng/mL) beyond PSA and %fPSA. Specifically, the AUC for high-risk disease was 0.893 for PSA alone, 0.860 for %fPSA alone, compared with 0.634 for the genetic risk score. The combined AUC was 0.901 for total and %fPSA, and 0.906 with the addition of the genetic risk score. When evaluating these data, it is important to consider that the population in this study was relatively homogeneous (Northern Swedish men), and therefore the generalizability of these results to other ethnic groups is uncertain. It is also noteworthy that not all of the SNPs included in the model were validated in the current population, which may have affected the performance characteristics for the combined risk score. Despite these inherent limitations, this study provides novel insights into the performance of genetic variants compared with PSA and free PSA measurements. Although there was a small statistically significant improvement in the prediction of overall prostate cancer, genetic variants were not incrementally useful for identifying high-risk prostate cancer. Polygenic Risk Score Improves Prostate Cancer Risk Prediction: Results From the Stockholm-1 Cohort Study Aly M, Wiklund F, Xu J, et al. Eur Urol. 2011;60:21-28. Another recent report examined the utility of SNPs in prostate cancer risk assessment using a different methodology.13 From a population of 5241 men from Stockholm, Sweden, undergoing prostate biopsy from 2005 to 2007, the authors identified 2542 with PSA levels , 10 ng/mL and a blood sample available for genotyping. Considering 35 SNPs, they calculated a genetic risk score by adding together the number of risk alleles for each participant weighted by the logarithm of each SNP’s odds ratio for prostate cancer risk. Overall, they demonstrated that this genetic risk score was significantly associated with having a positive prostate biopsy (P , .0001). With regard to performance characteristics, a model including log PSA, log %fPSA, age at biopsy, and family history had an AUC of 0.642 for prostate cancer detection overall, which improved marginally to 0.674 with the addition of the genetic risk score (P 5 .014). Additionally, the genetic risk score led to net reclassification improvement (P , .0001) suggesting that it may be useful to reduce unnecessary biopsies. For example, using a cutoff of 25% predicted prostate cancer risk, the genetic model would avoid 480 (22.7%) biopsies (compared with 10.0% using the nongenetic model), while missing 3% of aggressive cancers in the nonbiopsied group. As in the previous study,12 the genetic risk score did not improve the discrimination of aggressive prostate cancer (P 5 .83). This study was limited to men designated for biopsy based on PSA (, 10 ng/mL) and other clinical characteristics. Limitations of this study also include possible selection bias, in that , 50% of men who had a biopsy in Stockholm during that time period were ultimately included in the analysis. As in the study by Johansson and colleagues,12 evaluation of the results of genetic risk assessment in ethnically diverse populations will also be necessary. Discussion Despite rapid advances in SNP identification through GWAS, these studies are among the first to address whether and how to integrate this genetic information into clinical practice.14,15 Indeed, the studies by both Johansson and colleagues12 and Aly and associates13 suggest that genetic variants may provide incremental risk information beyond PSA alone and may help reduce unnecessary biopsies. However, given the current problem with overdiagnosis of clinically insignificant prostate cancer, predictors of high-risk disease are greatly needed and SNP profiles were not shown to be useful in this regard. Nevertheless, our knowledge of prostate cancer genetics continues to expand rapidly, and investigation into the underlying biologic pathways is actively underway. Indeed, a rare new mutation in the HOXB13 gene has just been discovered that confers a 20-fold increased risk for prostate cancer in a small proportion 116 • Vol. 14 No. 3/4 • 2012 • Reviews in Urology 40041700002_RIU0551.indd 116 12/02/13 2:29 PM Prostate Cancer of subjects.16 It is possible that other new variants will be identified in the future that could improve on the performance characteristics of genetic models or specifically enhance the prediction of life-threatening disease. There may also be specific high-risk subgroups in the population for whom genetic testing may be more cost-effective, such as men with a positive family history.10 Finally, additional investigation is needed to demonstrate that performing these tests results in net benefit and that the appropriate resources are in place before genomic testing should be considered for use in daily clinical practice. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References 1. 2. 3. Amundadottir LT, Sulem P, Gudmundsson J, et al. A common variant associated with prostate cancer in European and African populations. Nat Genet. 2006;38:652-658. Eeles RA, Kote-Jarai Z, Giles GG, et al. Multiple newly identified loci associated with prostate cancer susceptibility. Nat Genet. 2008;40:316-321. Gudmundsson J, Sulem P, Manolescu A, et al. Genome-wide association study identifies a second prostate cancer susceptibility variant at 8q24. Nat Genet. 2007;39:631-637. 14. 15. 16. Thomas G, Jacobs KB, Yeager M, et al. Multiple loci identified in a genome-wide association study of prostate cancer. Nat Genet. 2008;40:310-315. Eeles RA, Kote-Jarai Z, Al Olama AA, et al. Identification of seven new prostate cancer susceptibility loci through a genome-wide association study. Nat Genet. 2009;41:1116-1121. Schumacher FR, Berndt SI, Siddiq A, et al. Genome-wide association study identifies new prostate cancer susceptibility loci. Hum Molec Genetics 2011;20:3867-3875. Zheng SL, Sun J, Wiklund F, et al. Cumulative association of five genetic variants with prostate cancer. N Engl J Med. 2008;358:910-919. Loeb S, Partin AW. Single nucleotide polymorphisms and prostate cancer susceptibility. Rev Urol. 2008;10:304-305. Xu J, Sun J, Kader AK, et al. Estimation of absolute risk for prostate cancer using genetic markers and family history. Prostate. 2009;69:1565-1572. Kraft P, Wacholder S, Cornelis MC, et al. Beyond odds ratios—communicating disease risk based on genetic profiles. Nat Rev Genet. 2009;10:264-269. McGuire AL, Burke W. An unwelcome side effect of direct-to-consumer personal genome testing: raiding the medical commons. JAMA. 2008;300:2669-2671. Johansson M, Holmström B, Hinchliffe SR, et al. Combining 33 genetic variants with prostate-specific antigen for prediction of prostate cancer: longitudinal study. Int J Cancer. 2012;130:129-137. Aly M, Wiklund F, Xu J, et al. Polygenic risk score improves prostate cancer risk prediction: results from the Stockholm-1 cohort study. Eur Urol. 2011;60:21-28. Nam RK, Zhang WW, Trachtenberg J, et al. Utility of incorporating genetic variants for the early detection of prostate cancer. Clin Cancer Res. 2009;15:1787-1793. Lindström S, Schumacher F, Cox DG, et al. Common genetic variants in prostate cancer risk prediction—results from the NCI Breast and Prostate Cancer Cohort Consortium (BPC3). Cancer Epidemiol Biomarkers Prev. 2012;21:437-444. Ewing CM, Ray AM, Lange EM, et al. Germline mutations in HOXB13 and prostatecancer risk. N Engl J Med. 2012;366:141-149. Vol. 14 No. 3/4 • 2012 • Reviews in Urology • 117 40041700002_RIU0551.indd 117 12/02/13 2:29 PM