Deciphering the genetic basis of prostate-specific antigen (PSA) levels may improve their utility to screen for prostate cancer (PCa). We thus conducted a transcriptome-wide association study (TWAS) of PSA levels using genome-wide summary statistics from 95,768 PCa-free men, the MetaXcan framework, and gene prediction models trained in Genotype-Tissue Expression (GTEx) project data. Tissue-specific analyses identified 41 statistically significant (p < 0.05/12,192 = 4.10e-6) associations in whole blood and 39 statistically significant (p < 0.05/13,844 = 3.61e-6) associations in prostate tissue, with 18 genes associated in both tissues. Cross-tissue analyses that combined associations across 45 tissues identified 155 genes that were statistically significantly (p < 0.05/22,249 = 2.25e-6) associated with PSA levels. Based on conditional analyses that assessed whether TWAS associations were attributable to a lead GWAS variant, we found 20 novel genes (11 single-tissue, 9 cross-tissue) that were associated with PSA levels in the TWAS. Of these novel genes, five showed evidence of colocalization (colocalization probability > 0.5): EXOSC9, CCNA2, HIST1H2BN, RP11-182L21.6, and RP11-327J17.2. Six of the 20 novel genes are not known to impact PCa risk. These findings yield new hypotheses for genetic factors underlying PSA levels that should be further explored toward improving our understanding of PSA biology.

/pdf/transcriptome-wide-association-analysis-identifies-novel-3rbh8g0v.pdf

Transcriptome-Wide Association Analysis Identifies Novel Candidate Susceptibility Genes for Prostate-Specific Antigen Levels in Men Without Prostate Cancer

Polygenic risk scores (PRS) are rapidly emerging as aggregated measures of disease-risk associated with many genetic variants. Understanding the interplay of PRS with environmental factors is critical for interpreting and applying PRS in a wide variety of settings. We develop an efficient method for simultaneously modeling gene-environment correlations and interactions using PRS in case-control studies. We use a logistic-normal regression modeling framework to specify the disease risk and PRS distribution in the underlying population and propose joint inference across the two models using the retrospective likelihood of the case-control data. Extensive simulation studies demonstrate the flexibility of the method in trading-off bias and efficiency for the estimation of various model parameters compared to the standard logistic regression or a case-only analysis for gene-environment interactions, or a control-only analysis for gene-environment correlations. Finally, using simulated case-control datasets within the UK Biobank study, we demonstrate the power of the proposed method for its ability to recover results from the full prospective cohort for the detection of an interaction between long-term oral contraceptive use and PRS on the risk of breast cancer. This method is computationally efficient and implemented in a user-friendly R package.

/pdf/joint-modeling-of-gene-environment-correlations-and-2frvds4c.pdf

Joint Modeling of Gene-Environment Correlations and Interactions using Polygenic Risk Scores in Case-Control Studies

Prostate-specific antigen (PSA) screening for prostate cancer remains controversial because it increases overdiagnosis and overtreatment of clinically insignificant tumors. We evaluated the potential of genetic determinants of PSA levels to improve screening utility by accounting for variation in PSA not due to cancer. A multi-ancestry genome-wide meta-analysis of 95,768 men without prostate cancer discovered 128 PSA-associated index variants (P<5 x 10-8), including 82 novel signals. The resulting 128-variant polygenic score for PSA (PGSPSA) explained 7.3-8.8% of PSA variation in external validation cohorts. Adjusting PSA values using PGSPSA enabled more accurate diagnostic decisions, such as avoiding 17-20% of negative prostate biopsies. Genetically adjusted PSA was more predictive of aggressive prostate cancer (odds ratio (OR)=3.04, P=3.3 x 10-7; AUC=0.716) than unadjusted PSA (OR=2.80, P=3.2 x 10-6; AUC=0.684) and improved detection of aggressive disease when combined with a prostate cancer PGS (AUC: 0.732 vs. 0.645, P=3.3 x 10-4). We further showed that PSA-related selection bias distorts genetic associations with prostate cancer and hampers PGS performance. Our findings provide a roadmap towards personalizing cancer biomarkers and screening.

https://www.medrxiv.org/content/medrxiv/early/2022/10/07/2022.04.18.22273850.full.pdf

N. D. Freadman

Papers

Incorporating Genetic Determinants of Prostate-Specific Antigen Levels Improves Prostate Cancer Screening