Germline mutation landscape of DNA damage repair genes in African Americans with prostate cancer highlights potentially targetable RAD genes

In prostate cancer, emerging data highlight the role of DNA damage repair genes (DDRGs) in aggressive forms of the disease. However, DDRG mutations in African American men are not yet fully defined. Here, we profile germline mutations in all known DDRGs (N = 276) using whole genome sequences from blood DNA of a matched cohort of patients with primary prostate cancer comprising of 300 African American and 300 European Ancestry prostate cancer patients, to determine whether the mutation status can enhance patient stratification for specific targeted therapies. Here, we show that only 13 of the 46 DDRGs identified with pathogenic/likely pathogenic mutations are present in both African American and European ancestry patients. Importantly, RAD family genes (RAD51, RAD54L, RAD54B), which are potentially targetable, as well as PMS2 and BRCA1, are among the most frequently mutated DDRGs in African American, but not in European Ancestry patients.

The authors state that the results of their study could enhance patient stratification for specific targeted therapies. However, I don't find convincing evidence of this claim in my review of this manuscript. The authors could have spent more effort taking the significant gene variant findings and conducted other secondary analyses (e.g. pathway analysis) in order to gain a more informed sense of known and novel specific therapeutic targets that could be studied in the future by this research group or others who research prostate cancer in health disparate populations, most notably in African American men.
I do feel that this paper will influence thinking in the field because of its focus on a health disparate population, African Americans with prostate cancer.
The following are my concerns and questions about this manuscript: Lines 143-148: How did the authors conclude that a test using this 8-gene panel would detect 11.9% (31 of 259) of AA CaP patients and only 5.8% (16 of 272) of CA CaP patients with potentially targetable mutations (P = 0.0017)? I don't recall reading how this was determined in my review of the manuscript so I think that additional clarification is necessary.
For Supplementary Figure 1. Principle Component Analysis (PCA) of Ancestry using the PEDDY program. Although I agree with the populations included, why was the 1K Genomes reference population "ASW" population not included in the principle component analysis of ancestry? The ASW reference cohort is comprised of African Americans from the United States Southwest. This cohort would likely have the same percentage of European and African admixture as the African Americans in your study. I think the PCA analysis should be repeated after also including the 1K Genomes ASW reference population.
Lines 97-98: Of the 11.5% (69/600) excluded from the analysis, how many were due to of the patients due to various sequencing QC criteria and how many were excluded due to a mismatch between genomic ancestry and self-reported race? Could more African American subjects in particular have been included if the ASW cohort had been included in the PCA analysis?
Lines 136-138: Why all significant RAD mutations were grouped together? The rationale for this was unclear to me. Please provide additional clarification/justification for doing this.
The statistical analyses appear to be appropriate and valid. I was encouraged to learn that the authors used the Benjamini-Hochberg procedure to control for FDR. With the exception of the comments made in the manuscript about patient stratification for specific targeted therapies, I think that a different researcher could reproduce the work described in this manuscript given the level of detail provided.
Overall, the attention to the differences in DDRG genes involved in hereditary prostate cancer between AA and CA populations is noteworthy as it helps to address the disparities that exist for AA afflicted with prostate cancer who despite having the highest morbidity and mortality of the disease, are not well-represented in prostate cancer genomics research. Thank you for the opportunity to review this submission.
Reviewer #2 (Remarks to the Author): Expert in cancer genetics and epidemiology, biomarkers, and ancestry

Overview
The present study seeks to characterize the mutational landscape of DNA-damage repair genes (DDRGs) in a matched retrospective cohort study of Black and White men treated in an equal access healthcare setting. The authors demonstrate that 46 of 276 known DDRGs were mutated among study participants. Of the 46, there was mutational overlap for only 13 of the 46 genes between Black and White men.

Novel Findings
This study makes a significant contribution to the literature by using a whole-genome sequencing platform to identify mutations in the 276 DDRGs. Black men were more likely to harbor mutations with potentially therapeutic targets. This finding may be clinically important as currently available genetic testing platforms do not capture the full spectrum of DDRG mutations.

Interpretation
The authors may wish to consider reframing their discussion of the health disparity. The inequity present is that Black men are an understudied population. This study addresses that inequity and identifies additional mutations that may have clinical relevance that have not been previously described. The genetic variation in DDRGs may help to explain the observed differences between Black and White men as it pertains to prostate cancer incidence and outcomes. However, the observed genetic variation between races alone does not indicate a disparity.
The authors note that a limitation of current genetic testing is that prognostic variants were identified among cohorts of men who were predominantly White and who had advanced stage disease. While the current study provides greater breadth of the mutational landscape of DDRGs, DDRG mutations are most likely to occur at advanced disease stage. A more informative approach for a future study may include characterizing DDRG variants among Black men with metastatic/advanced stage disease.

Reviewer #3 (Remarks to the Author): Expert in prostate cancer genomics
The authors present a paper on the frequency of germline mutations in African American vs Caucasian American prostate cancer patients. Indeed the paper is well written, concise and of interest to the community.
My comments: 1) The main assumption that is used here concerns the categorization of mutations as pathogenic/ likely pathogenic. The use of Clinvar and other databases that are used by geneticists is an accepatible way to circumvent this problem, however it would be stronger if indeed the authors would present evidence that the mutation they find are indeed pathogenic. This could be done by providing family histories of cases or analysis of actual cancer genomes for mutations in the same genes (second hit).
2) The main strength of the paper is that the authors represent in an improved manner (WGS, which is by definition comprehensive) the frequency of germline alterations in DDRG genes. It would help the impact of the paper if a table could be added where the results of this analysis and the previous analyses would be summarized: showing which genes were missed in previous analyses and how the frequencies of the genes that concur differ in the different datasets.
3) finally the outcome analysis is somewhat flawed as it is retrospective and could suffer from a number of biases. Here again validation would be great. There have been cohorts of local prostate cancer that have been characterized using WGS, interrogating those with the question whether these mutations indeed confer a poor prognosis would be helpful.

Reviewer #1
The authors' study objective was to profile germline mutations in all known DDRGs (N =276) using whole genome sequences from blood DNA of a matched cohort of patients (300 African Americans and 300 Caucasian Americans) with primary prostate cancer to determine if knowledge of DDRG mutation status can enhance patient stratification for specific targeted therapies. The authors found only 13 pathogenic or likely pathogenic DDRG variants shared among AA and CA patients. Additionally, RAD1, RAD54L, RAD54B, PMS2 and BRCA genes were more frequently mutated in AA patients compared to CA patients.
While the genomic analysis methods implemented in this paper are standard (not novel), I think the results of this paper would be of high interest to investigators who study prostate cancer. Particularly, prostate cancer in men of African descent where significant health disparities in diagnosis, treatment and research currently exists.
The authors state that the results of their study could enhance patient stratification for specific targeted therapies. However, I don't find convincing evidence of this claim in my review of this manuscript. The authors could have spent more effort taking the significant gene variant findings and conducted other secondary analyses (e.g. pathway analysis) in order to gain a more informed sense of known and novel specific therapeutic targets that could be studied in the future by this research group or others who research prostate cancer in health disparate populations, most notably in African American men.

Response:
We thank the reviewer for the kind feedback and the important points raised. Based on the reviewer's suggestions, we have incorporated information on significant gene findings, pathway analysis and added relevant references in the manuscript (pages #7 and #8).
Considering significant gene findings in relation to clinical utility, we found that several RAD genes (RAD51, RAD54L, RAD54B), as well as PMS2 and BRCA1, were among the most frequently mutated DDRGs in AA patients, but not in CA patients. These genes are part of targetable DDRG pathway, specifically the homologous recombination (HR) pathway, suggesting potential benefit for AA men.
As a secondary analysis suggested by the reviewer, we performed pathway analysis based on published DNA Damage Repair pathways (Knijnenburg et al, Cell report 2018). The analysis showed that out of known 14 DDR pathways, the HR pathway (with 15 mutations) is associated with disease progression to BCR (p 0.018), while the NHEJ pathway (with 3 mutations) is associated with metastasis (p 0.045). We have added these new results to the new Figure 3; referred to on page #8, Line 157) and the new Supplementary Following the reviewer's advice, in an additional secondary analysis, we also functionally scored and validated the germline mutations using an in silico tool, enGenome. These new results, which further supports our findings, has been incorporated in page #7 (line 134-137) I do feel that this paper will influence thinking in the field because of its focus on a health disparate population, African Americans with prostate cancer.

Response:
We thank the reviewer for this comment.
The following are my concerns and questions about this manuscript: 1. Lines 143-148: How did the authors conclude that a test using this 8-gene panel would detect 11.9% (31 of 259) of AA CaP patients and only 5.8% (16 of 272) of CA CaP patients with potentially targetable mutations (P = 0.0017)? I don't recall reading how this was determined in my review of the manuscript, so I think that additional clarification is necessary.
Response: Based on the reviewer's question, we have added clarifications in the manuscript. We know from the sequence and ddPCR validation that 11.9% of the AA CaP patients had germline mutation in any of the 8 genes. Therefore, we assume that a test for these 8 genes will detect the same 11.9% of the cases with germline mutation. The same logic applies for the CA patients. Supplementary Table 6 provides the summary of these 8 candidate genes, which were selected based on their potential clinical utility (being part of potentially targetable DDR pathways), and mutation carrier frequency of over 1%. This information with relevant references has been included in page #8 (lines 165-166). Figure 1. Principle Component Analysis (PCA) of Ancestry using the PEDDY program. Although I agree with the populations included, why was the 1K Genomes reference population "ASW" population not included in the principle component analysis of ancestry? The ASW reference cohort is comprised of African Americans from the United States Southwest. This cohort would likely have the same percentage of European and African admixture as the African Americans in your study. I think the PCA analysis should be repeated after also including the 1K Genomes ASW reference population.

Response:
We thank the reviewer for the comment on inclusion of ASW in PCA ancestry analysis. The ASW subpopulation group (N=61) within the AFR major population was incorporated in the ancestry analysis but was not specifically indicated. To address this point, we have now expanded the description of the PCA ancestry analysis by better defining the major populations from the 1000 Genome project that was used in our analysis, in the Supplemental Figure 1 legend (Page 1, lines 6-12 of Supplementary Information).
3. Lines 97-98: Of the 11.5% (69/600) excluded from the analysis, how many were due to of the patients due to various sequencing QC criteria and how many were excluded due to a mismatch between genomic ancestry and self-reported race? Could more African American subjects in particular have been included if the ASW cohort had been included in the PCA analysis?
Response: Thank you for this comment. We note that we had already included ASW in the PCA analysis as described in point #2 above, so our method of genomic-ancestry inference is not an explanation for sample exclusion. To further respond to this point, we have added detail on the reason for sample exclusion, on page 5 (lines 95-97). Specifically, 26 samples were excluded because of low quality sequencing results due to DNA yields, fragment size and contamination, and 33 samples were further excluded considering differences in self-reported and genomicallyinferred ancestry. Since the aim of our study is to compare AA and CA patients, we decided to exclude samples with a lack of agreement between self-reported race and genomically-inferred ancestry. This new information is now added to page 5 on lines 95-97.
4. Lines 136-138: Why all significant RAD mutations were grouped together? The rationale for this was unclear to me. Please provide additional clarification/justification for doing this.
Response: Based on the reviewer's suggestions, we have provided additional justifications in the manuscript (page # 7), on lines 152-154. In brief, one of the major findings from our study is that several RAD family genes (RAD51, RAD54L, RAD54B), in addition to PMS2 and BRCA1, were among the most frequently mutated DDRGs in AA patients, but not in CA patients, when compared to the relevant control datasets (please see Figure 2). RAD51 was found to be the most frequently mutated HR pathway gene in AA cohort (please see new Figure 3). Because RAD genes are functionally related, we grouped all RAD mutations together. We observed a greater mutation rate in AA (6.95%) than in CA patients (1.10%) (P = 0.001; OR 6.68) (Please see Supplementary Table 4). Additionally, literature supporting potential clinical utility of the mentioned RAD genes has been included in page #7 (lines 144, 145, 149, 150, 152).
5. The statistical analyses appear to be appropriate and valid. I was encouraged to learn that the authors used the Benjamini-Hochberg procedure to control for FDR. With the exception of the comments made in the manuscript about patient stratification for specific targeted therapies, I think that a different researcher could reproduce the work described in this manuscript given the level of detail provided.

Response:
We thank the reviewer for this comment regarding the methodological rigor in our manuscript. We have added detailed description and new references about the targeted therapies on page #7 (lines 144, 145, 149, 150, 152) and page #9 (line 191-192) Reviewer #2 (Remarks to the Author): Expert in cancer genetics and epidemiology, biomarkers, and ancestry 1. The authors may wish to consider reframing their discussion of the health disparity. The inequity present is that Black men are an understudied population. This study addresses that inequity and identifies additional mutations that may have clinical relevance that have not been previously described. The genetic variation in DDRGs may help to explain the observed differences between Black and White men as it pertains to prostate cancer incidence and outcomes. However, the observed genetic variation between races alone does not indicate a disparity.

Response:
We agree and thank the reviewer for pointing out the correct usage of the racial disparity term. We have corrected the manuscript text by using the term "racial differences" for "racial disparity" [page# 3 (line 50), #4 (line 67) , #6 (lines 118, 119) and #9 (line 200)]. Some