Improving prostate cancer screening: accounting for genetic determinants of PSA variation

By Dr. Chinta SidharthanJun 5 2023Reviewed by Lily Ramsey, LLM

In a recent study published in Nature Medicine, researchers conducted a genome-wide analysis of prostate-specific antigen (PSA) levels of men without prostate cancer to understand the non-cancer-related variation in PSA levels to improve decision-making during the diagnosis of prostate cancer.

Study: Genetically adjusted PSA levels for prostate cancer screening. Image Credit: luchschenF/Shutterstock.com

Background

The kallikrein 3 (KLK3) gene produces the enzyme PSA in the prostate gland, and PSA is responsible for the release of motile sperm. The disruption of the prostate gland epithelial tissue, such as in the case of a prostate gland tumor, increases the circulatory levels of PSA.

Other factors, such as an inflammation or infection of the prostate gland, age, and benign prostatic hyperplasia, can also increase the levels of PSA in circulation. Furthermore, while low PSA levels are generally associated with being overweight or obese, it cannot eliminate the possibility of prostate cancer.

While regular testing of PSA levels does help in reducing the overall mortality due to prostate cancer, the variation in PSA levels across individuals results in an overdiagnosis of prostate cancer in 20% to 60% of the cases.

Since the heritability of PSA is high and recent, genome-wide association studies have identified 40 loci that independently contribute to the variability in PSA levels; understanding the genetic variation in PSA levels is essential for accurately using PSA levels in prostate cancer prediction.

About the study

In the present study, the researchers used data from five previous studies that included men without prostate cancer diagnoses to conduct a genome-wide analysis of PSA levels. The results from the genome-wide analyses were then meta-analyzed across different ancestries and populations.

The genetic data used to analyze the PSA levels was obtained from deoxyribonucleic acid (DNA) extracted from tissue that was not prostatic, such as buccal swabs or blood samples.

Only individuals who were biologically male, identified as the male gender, and had no history of prostate cancer or surgical prostate resections were included in the analysis.

Furthermore, the analyses were limited to a PSA level range of greater than 0.01 ng per ml and less than or equal to 10 ng per ml, corresponding to functional prostate and low-risk prostate cancer, respectively, to reduce potential reverse causation.

The study populations included in the meta-analysis were the United Kingdom Biobank cohort, the Kaiser Permanente Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort, the participants in the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening clinical trial, Vanderbilt University Medical Center BioVU biobank, and the Malmö Diet and Cancer Study (MDCS) cohort.

The multi-ancestry meta-analyses included 95,768 participants from European, African, East Asian, and Latino or Hispanic ancestries.

The results of the meta-analyses, stratified according to ancestry, were then used to develop a genome-wide polygenic score (PGA) for PSA, which was then validated using the Prostate Cancer Prevention Trial (PCPT) and the Selenium and Vitamin E Cancer Prevention Trial (SELECT) cohorts.

Based on the results from the validation, the effectiveness of genetically adjusted PSA levels (PSA^G) in a clinical setting in detecting prostate cancer was evaluated using the GERA, PCPT, and SELECT cohorts.

Results

The results reported that the PGA for PSA developed based on the results of the meta-analyses and the 128 significant associations from the genome-wide analysis explained 9.61% of the variation in PSA levels.

Furthermore, using PSA^G would result in a 31% reduction in negative biopsies among men of European ancestry and reduce the need for biopsies in prostate cancer patients by 12%.

The genetically adjusted levels of PSA were more effective in predicting aggressive prostate cancers than the unadjusted PSA levels. They improved the probability of predicting aggressive cancer compared to a prostate cancer polygenic score alone.

The genome-wide association analysis also discovered 82 novel variants associated with PSA levels, with genes related to reproductive processes showing the strongest signal among these novel variants.

Conclusions

To summarize, the researchers conducted a genome-wide association study and meta-analysis to understand the genetic variation in PSA levels across a large, ancestrally diverse dataset of men without prostate cancer and develop a polygenic score for a more accurate prediction of prostate cancer.

Overall, the findings indicated that using genetically adjusted levels of PSA resulted in a more accurate prediction of aggressive prostate cancer compared to unadjusted PSA levels, and the polygenic scores for PSA can be used for more personalized screening for prostate cancers across different ancestries.