Prostate cancer (PCa) is the second most common cancer in men worldwide (1). Despite PCa incidence in China remains far lower than that in western countries, it has increased rapidly in recent years due to lifestyle changes and screenings (2). Genomic and molecular complexity in prostate cancer have been tremendously profiled (3–7). Although many correlations between different genetic variants and PCa clinical outcomes are yet to be revealed, patients with pathogenic variants in DNA damage repair genes, especially those from homologous recombinant repair (HRR) such as ATM, BRCA2, and BRCA1, are found to be closely associated with younger onset, increased risks, and poorer prognosis (8, 9). Furthermore, platin based regimens and poly (ADP-ribose) polymerase inhibitors have been proved to have additional benefits in this specific subgroup (10, 11).

However, most of these findings came from studies enriched with non-Asian population, and research data on domestic Chinese patients are still inadequate. Previous studies have found unique genomic and epigenomic features in Chinese PCa patients, implying the importance of taking population-specific contexts into consideration (12). There are a series of studies addressing the HRR genes in Chinese population (13–19). These studies have provided outlines of mutation landscape in HRR genes in Chinese PCa patients. However, most of these studies came from south part of China. Furthermore, these studies focused mostly on germline variants and metastatic cases. Integrative analysis of somatic variants somatic and germline alterations in Chinese PCa patients is still limited.

In this study, we sequenced 714 fresh blood samples and 200 formalin-fixed paraffin-embedded (FFPE) PCa tissues from a total of 720 PCa patients across all disease states at Peking University First Hospital (PUFH). The majority of patients in this PUFH cohort came from north China. We also analysed clinical characteristics and obtained the surgical results of these patients. Thus, PUFH cohort would serve as a unique supplementation to the current bank of HRR gene alterations in Chinese population.

Through somatic sequencing of 200 cases and germline sequencing of 714 cases, we identified 53 patients harboring either somatic or germline mutations ( Figure 4 ). Unlike previous studies in Chinese population, we profiled PCa in a range of clinical states—from locoregional to metastatic, and, thus, enabled comparison of genomic landscape across disease states using a single cohort. Furthermore, our study enriched PCa patients from north China, which shared a distinct genomic background but unequally represented before. Similar to previous studies, our results showed ATM and BRCA2 germline mutations were of great importance in Chinese population as were in western population. In addition, we echoed previous findings that TP53 accounted for the most common somatic mutation across all population, followed by PTEN and ATM.

The somatic mutational rates of HRR genes in this study was 3% (6/200). By contrast, 8.68% of the cases in TCGA PARD and 9.4% in MSK-IMPACT harbored somatic aberrations in HRR genes (6). All the three studies included PCa from locoregional to metastatic states. The lower somatic mutational rates in our PUFH cohort confirmed the observation that mutational rates were generally lower in Asian patients compared to Caucasian patients. However, this 3% mutational rate was also lower than 7.89% in the CPGEA cohort which only included Chinese patients (15). This might be due to the fact that our study included a smaller proportion of metastatic (2.5% VS 8.65%) and GS 9-10 (14% VS 28.8%) patients. Furthermore, 17.8% patients in SU2C/PCF harbored somatic HRR gene mutations (4). As SU2C/PCF only included individuals at metastatic castration resistant state, the lower mutational rates in our cohort confirmed that aberrations in HRR genes were less common in early state diseases.

When it came to germline alterations, 3.9% (28/714) patients carried HRR gene mutations in PUFH cohort. The germline mutational rate was 1.8% for BRCA2 and 0.6% for ATM. This was lower compared to the results reported by Nicolosi et al., SU2C/PCF, and MSK-IMPACT in Caucasian population and by Zhu et al. (4.3% for BRCA2 and 1.04% for ATM) in south Chinese population (4, 5, 7, 15). Of note, in the study by Zhu et al, 6% PCa patients had reginal lymph node metastasis and 65% had remote metastasis. However, only 9.2% and 17.6% patients had reginal lymph node and remote metastasis in our cohort. Therefore, it was once again proved that the size of the dataset, stage of disease, and patient diversity—even for the same ethnicity from different regions—were of great importance when interpreting prostate cancer genomic data.

Due to the low mutational rates for HRR genes in patients from north China, a smarter protocol to select potential candidates for genetic testing is necessary. Broadly speaking, as HRR gene mutations were associated with poor outcomes, any clinical feature potentially linked to aggressive disease could serve as a candidate marker. Currently, most guidelines addressed the use of clinical characteristics including age of onset, family history, and personal cancer history to enhance the positive rate of genetic testing (39). Nicolosi et al. argued clinical factors frequently used to identify patients who qualified germline testing, including age, race, and family history, did not correlate with positive test results in their data (7). So was GS. However, their data revealed a higher likelihood of germline mutations in HRR genes with higher stages of diseases. These findings were not echoed by those of Wu et al. In their study, Wu et al. did not observe a relationship between the presence of germline HRR gene mutations and any clinical characteristics except age at diagnosis (14). In our study, Gleason grades at diagnostic biopsy were found to be correlated with somatic aberrations in general and in TP53, while age of onset, PSA at diagnosis, and GS at biopsy were factors correlated with germline ATM and BRCA2 mutations. Due to these conflicting evidences, clinicians should be cautious using these clinical factors when prescribing tailored genetic tests for PCa patients. Indeed, it would be a better idea to expand genetic tests to include wider range of men diagnosed with PCa as much as possible due to the fact that these mutation carriers demanded radical treatment as early as possible.

As a needle biopsy and corresponding RP might not always get the same GS, we also examined the results of RP GS. Our results showed a higher incidence for GS upgrading for most groups except GS 9-10 when compared to the benchmark from Epstein et al (40). Thus, it reflected a potential bias in using biopsy GS alone to pick patients with aggressive patients. However, some pathological features including Intraductal/ductal histology and lymphovascular invasion appeared to be associated with pathogenic germline DNA-repair gene mutations in men with prostate cancer (41). Therefore, it would be reasonable for future studies to investigate the underlying morphology in these specimens possessing HRR gene mutations, and, thus, could provide extra help to smarter screening.

There were puzzles waiting to be answered. We found no BRCA1 mutation in our cohort. Interestingly, in the only one previous study focused on north China population but with a smaller size, researchers found no BRCA1 mutation neither (19). This could possibly be a distinctive genomic feature for north China individuals. Future studies should validate this observation with multiple center studies and more powerful epidemiology tools.

Our study had some limitations. First, we treated VUS as negative alterations in this study. However, preliminary yet increasing data implies the pathogenicity of some of the current VUS. Second, our panel only tested variants including single-nucleotide polymorphisms and indels, but not copy number variants, which also could result in loss of function of proteins. Third, we interpreted the somatic variants with a variety of certified knowledge bases. Of note, there was currently no ‘gold standard’ somatic mutation database that was similar to what existed for germline. These concerns warranted future efforts and advancement in clinical genetics.

In summary, we investigated the genomic landscape of both somatic and germline HRR gene alterations in the PCa patients across all clinical states from north Chinese population. The lower HRR gene mutational rate in this PUFH cohort underlined the need for a more efficient pre-testing candidate-selection protocol in genetic testing.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

The study was conducted in accordance with the Declaration of Helsinki. The studies involving human participants were reviewed and approved by Committee for Ethics at Peking University First Hospital (protocol code: 2022-199-001; date of approval: 2022-04-22). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

HXL and WY set up the experimental design and supervised the work. YL and BJ collected and selected the data. CS, XG, XQ, MM, HZL, HH, QT, KY, YM, and ZH organized the database. BJ and JG carried out the genetic testing. YL and XF performed the statistical analysis and wrote the manuscript. All authors reviewed the manuscript and agree with its contents. All authors contributed to the article and approved the submitted version.