Search

Figure S1. Cross-validation error in ADMIXTURE analyses. The best fit to data occurs at K = 3.

Successfully called markers on the MADCaP Array. This tab-delimited file includes genomic positions, inclusion criteria, and overlap with other arrays.

Although prostate cancer is the leading cause of cancer mortality for African men, the vast majority of known disease associations have been detected in European study cohorts. Furthermore, most genome-wide association studies have used genotyping arrays that are hindered by SNP ascertainment bias. To overcome these disparities in genomic medicine, the Men of African Descent and Carcinoma of the Prostate (MADCaP) Network has developed a genotyping array that is optimized for African populations. The MADCaP Array contains more than 1.5 million markers and an imputation backbone that successfully tags over 94% of common genetic variants in African populations. This array also has a high density of markers in genomic regions associated with cancer susceptibility, including 8q24. We assessed the effectiveness of the MADCaP Array by genotyping 399 prostate cancer cases and 403 controls from seven urban study sites in sub-Saharan Africa. Samples from Ghana and Nigeria clustered together, whereas samples from Senegal and South Africa yielded distinct ancestry clusters. Using the MADCaP array, we identified cancer-associated loci that have large allele frequency differences across African populations. Polygenic risk scores for prostate cancer were higher in Nigeria than in Senegal. In summary, individual and population-level differences in prostate cancer risk were revealed using a novel genotyping array.Significance:This study presents an Africa-specific genotyping array, which enables investigators to identify novel disease associations and to fine-map genetic loci that are associated with prostate and other cancers.

Clonal hematopoiesis of indeterminate potential (CHIP) has been associated with inflammation, which is a risk factor for cancer, including prostate cancer. We previously reported weak evidence of an association between CHIP and prostate cancer risk in men of European ancestry. However, little is known for African ancestry populations. We investigated the association of age-related CHIP with overall and aggressive prostate cancer risk in a large whole-exome sequencing study of 12,049 African ancestry men, including 7,176 prostate cancer cases (of which 3,283 had aggressive disease and 1,074 had metastatic disease) and 4,873 controls. Somatic variant calling was carried out using GATK Mutect2, and only variants with minor allele frequencies (MAF) 5% were included. Variants with a MAF ≥0.1% in gnomAD were excluded. CHIP variants were identified from a list of pre-specified mutations in 74 genes. Associations were tested using regression models adjusting for age, sub-study, and top 10 principal components, with statistical significance tested by the likelihood ratio test and applying a Bonferroni correction to account for multiple testing. In total, 998 variants in 57 CHIP genes were identified. Consistent with previous results, we observed a strong association between CHIP and age at blood draw. CHIP genes in aggregate were not statistically significantly associated with risks of total (OR=1.12, 95% CI=0.97-1.28), aggressive (OR=1.14, 95% CI=0.92-1.43) or metastatic (OR=1.17, 95% CI=0.91-1.49) prostate cancer. We observed that carriers of variants in DNMT3A, which is the gene that harbors the most CHIP driver mutations, had a nominally elevated risk of prostate cancer compared to non-carriers (OR=1.35, 95% CI=1.08-1.68, p=0.007). Additionally, carriers of variants in EZH2, which is implicated in cancer progression, showed a suggestive association with aggressive prostate cancer (OR=7.33, 95% CI=1.01-53.21, p=0.029). After adjusting for age at blood draw, CHIP genes in aggregate were not associated with age at prostate cancer diagnosis. However, we found that EZH2 variants carriers were diagnosed 12.9 years earlier on average than non-carriers (95% CI=6.1-19.7, adjusted p=0.01). A prostate cancer polygenic risk score (PRS) constructed using 269 risk variants was not associated with CHIP carrier status in aggregate (OR=0.99, 95% CI=0.92-1.06, p=0.70) or with any individual gene (all adjusted p>0.05). In summary, overall CHIP is not likely to be a risk factor of prostate cancer or aggressive disease in men of African ancestry. However, our results do confirm the association of CHIP in DNMT3A with prostate cancer risk as reported in previous studies in men of European ancestry. Future work will be needed to evaluate the biological causality of DNMT3A- and EZH2- related CHIP on prostate cancer. Citation Format: Anqi Wang, Yili Xu, Xin Sheng, Raymond Hughley, Ben Adusei, Mohamed Jalloh, Serigne Magueye Gueye, Andrew A Adjei, James Mensah, Pedro W. Fernandez, Akin Olupelumi Adebiyi, Oseremen Inokhoife Aisuodionoe-Shadrach, Lindsay Petersen, Maureen Joffe, Jeannette T. Bensen, James L. Mohler, Jack A. Taylor, Eboneé N. Butler, Sue A. Ingles, Benjamin A. Rybicki, Janet L. Stanford, Wei Zheng, Sonja I. Berndt, Chad D. Huff, Joseph Lachance, Luc Multigner, Caroline Andrews, Timothy R. Rebbeck, Laurent Brureau, Stephen J. Chanock, Adam de Smith, Fei Chen, Burcu F. Darst, David V. Conti, Christopher A. Haiman. Association between clonal hematopoiesis and risk of prostate cancer in a large sample of African ancestry men [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3508.

Background: There is a growing body of evidence supporting the contributions of germline rare variants to the susceptibility of prostate cancer (PCa), especially aggressive PCa. Our previous exome sequencing analysis highlighted 36 aggressive PCa candidate genes in populations of European ancestry. Here we investigated whether rare germline pathogenic, likely pathogenic, or deleterious (P/LD/D) variants in these genes were associated with overall and aggressive PCa risk in men of African ancestry. Methods: This exome sequencing analysis consists of 7,176 prostate cancer cases and 4,873 controls from the Research on Prostate Cancer in Men of African Ancestry (RESPOND) study. Among the PCa cases, 3,283 are aggressive cases (tumor stage T3/T4, regional lymph node involvement, metastatic disease, Gleason score >= 8.0, prostate-specific antigen [PSA] level >= 20 ng/mL or PCa as the underlying cause of death) including 1,074 metastatic cases, and 1,752 are non-aggressive cases (Gleason score ⇐ 7.0, PSA < 20 ng/mL, and tumor stage T1/T2). P/LP/D variants analyzed were rare (minor allele frequency < 1% in controls) and had either a Variant Effect Predictor impact score of “high” or a pathogenic or likely pathogenic ClinVar classification. The association between P/LP/D carrier status with risk of overall PCa, aggressive PCa, and metastatic PCa was evaluated in logistic regression models, adjusting for age and the top ten principal components. All statistical tests are two-sided. Results: Of the 36 PCa candidate genes, BRCA2 was the most frequently affected gene, with 1.7% of cases and 1.1% of controls harboring a germline P/LP/D variant, followed by MUTYH (1.5%/1.3%) ATM (0.93%/0.49%), MSH5 (0.70%/0.51%) and HOXB13 (0.70%/0.35%). Nominally significant associations with overall PCa were observed for ATM (OR=1.83, 95% CI=1.14-2.92, P=0.012), BRCA2 (OR=1.52, 95% CI=1.10-2.10, P=0.011), HOXB13 (OR=2.10, 95% CI=1.12-3.66, P=0.008), and PALB2 (OR=3.46, 95% CI=1.18-10.1, P=0.02). In case-case analyses (aggressive vs. non-aggressive cases), the association with aggressive PCa was nominally significant for ATM (OR=5.10, 95% CI=1.96-13.3, P=8.7 × 10−4) and BRCA2 (OR=2.00, 95% CI=1.19-3.38, P=0.009) and was suggestive for PALB2 (OR=2.99, 95% CI=0.83-10.7, P=0.09). Similar associations with metastatic PCa were also observed for these three genes. Conclusion: The associations of BRCA2, ATM, and PALB2 with overall PCa and aggressive PCa observed in men of African ancestry are consistent with findings from our previous study in men of European ancestry. These findings further support the importance of these genes in the consideration of screening and active surveillance for high-risk and advanced disease. Citation Format: Fei Chen, Burcu F. Darst, Xin Sheng, Anqi Wang, Yili Xu, Raymond Hughley, Ben Adusei, Mohamed Jalloh, Serigne Magueye Gueye, Andrew A. Adjei, James Mensah, Pedro W. Fernandez, Akindele O. Adebiyi, Oseremen Aisuodionoe-Shadrach, Lindsay Petersen, Maureen Joffe, Jo McBride, Jeannette T. Bensen, James L. Mohler, Jack A. Taylor, Eboneé N. Butler, Sue A. Ingles, Benjamin A. Rybicki, Janet L. Stanford, Wei Zheng, Sonja I. Berndt, Chad D. Huff, Joseph Lachance, Luc Multigner, Caroline Andrews, Timothy R. Rebbeck, Laurent Brureau, Stephen J. Chanock, David V. Conti, Christopher A. Haiman. Association of prostate cancer candidate genes with overall and aggressive prostate cancer in men of African ancestry [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1182.