Disparity in genomic application
Although PRS predictive power is high, PRS studies suffer from a significant deficit in the inclusion of African Americans, which represent only around 3% of the participants for GWASs published through 2015 32 33. The insufficient inclusion of African Americans leads to a disproportion in identifying risk-associated SNPs and compromises the predictive potential of PRS over this minority group19. As a result, there might be an imbalance in how PRS can improve care for patients of European descent versus patients of African descent32. In a multi-ethnic study of 80,481 participants, prostate cancer PRS performance was superior in those with genetically defined European ancestry than in those with African ancestry, which comprised 89.3% and 7.8% of the study population respectively28. This disparity is inevitable considering the bias introduced in European-dominated GWAS28 34. Known health disparity might also contribute to the diminished predictive power of PRS in minority groups. For instance, limited healthcare access leads to missing diagnosis information and failure of early detection of PCa, paving way for systemic differences in age of diagnosis across different ethnic groups and leading to inequitable risk stratification28 34.
A recent trans-ancestry GWAS of 127,006 controls and 107,247 prostate cancer cases discovered 86 new risk variants, totaling the known risk variants to 26919. In this study, PRS is shown to have a larger contribution to overall PCa risk for AAM because variants with odds ration >1.10, which have a greater effect on PRS, are more common in African American participants. Notwithstanding, AAM also have mean PRS that are approximately 2.18 times higher than that of EAM19. These findings are consistent with the conclusion that known risk variants substantially accounted for the estimated 75% higher prostate cancer incidence in African Americans when compared to non-Hispanic white35.
There has been considerable effort to attenuate the disparity presented in PRS studies. To compensate for the lack of diversity in GWASs, a study scaled ancestry-specific PRS distributions that, when considered separately in each ethnic group, can help identify individuals with higher PCa risk in each group36. Another study conducted a cross-validated search on a dataset that comprised only men with African genetic ancestry and identified three SNPs that significantly improved the performance of PRS in the studied population37. While more efforts are underway to improve diversity in the field of GWAS and improve PRS performance in minority groups, substantial gaps remain in our understanding.