Presentation Authors: Craig Labbate*, Chicago, IL, Zhuqing Shi, Shanghai, China, People's Republic of, Mary Helen Black, Shuwei Li, Holly Laduca, Aliso Viejo, CA, Chi-Hsiung Wang, Hongjie Yu, Brian Helfand, Evanston, IL, William Isaacs, Baltimore, MD, Jianfeng Xu, Evanston, IL
Introduction: The genetic risk score (GRS) using multiple single nucleotide polymorphisms has been well described to confer increased risk for prostate cancer (PCa). Currently, guidelines only acknowledge family history and race as modifiers to guide early prostate cancer screening. Using a cohort of screen-detected cancer cases and non-PCa control subjects, we sought to determine whether GRS predicts earlier age of disease onset.
Methods: Study subjects were 3,075 Caucasian men, including 1,694 PCa cases who underwent radical surgery at Johns Hopkins Hospital and 1,381 controls who did not have a PCa diagnosis from Johns Hopkins Hospital, NorthShore University HealthSystem, and Ambry Genetics. Seventy-two established PCa risk-associated SNPs were sequenced using ProstateNext, a targeted next-generation sequencing assay by Ambry Genetics. A population-standardized GRS was calculated using SNP-specific odds ratios and population-specific risk allele frequencies. Because GRS is a population standardized, it can be interpreted as relative risk to the general population. A GRS of 1, < 1 or >1 indicates average, lower or higher risk compared to the general population. Family history (FH) was defined as having â‰¥1 first- or second-degree relative with PCa. GRS and FH associations with PCa were estimated using logistic regression, and associations with time to diagnosis were examined with Kaplan Meier (KM) survival analysis. Proportional hazards assumption was assessed with Shoenfeld residuals.
Results: In the study cohort, 22% and 6% men had GRS at 1.5-3 and >3, respectively. Compared to men with GRS < 1.5, men with GRS 1.5-3.0 and >3.0 had significantly increased risk for PCa; OR=2.22 (P=1x10-16) and 4.17 (P=8x10-13), respectively. Similar and significant association was found after adjusting for age and FH. Kaplan-Meier survival analysis revealed that the age at median PCa diagnosis-free survival was 65, 61, and 59 years for men with GRS < 1.5, 1.5-3.0 and >3.0, respectively, P < 2x10-16 (Figure 1a). In comparison, the age at median PCa diagnosis-free survival was 63 and 60 years old for men with a negative and positive FH, respectively, P < 2x10-16 (Figure 1b). A Cox regression analysis found that both GRS (P=1x10-16) and FH (P=8x10-17) were significantly associated with age at PCa diagnosis.
Conclusions: Germline genetic risk as measured by the GRS is associated with earlier onset of PCa. Independent of family history, the GRS provides valuable information to inform the decision for early screening of PCa.
Source of Funding: Partially funded by Ambry Genetics