Current risk stratification strategies for prostate cancer (PC) and selection criteria for active surveillance (AS) rely primarily on clinical and pathological factors and do not consider family history or germline genetic risk. It has long been established that family history of prostate cancer increases the risk of prostate cancer diagnosis and mortality 2- to 4-fold.¹ More recently, there is growing evidence that men with a family history of other cancers, such as breast, ovarian, and pancreatic cancer, also have an increased risk for lethal prostate cancer.^2-4 While both genetic and environmental risk factors are shared among family members, twin studies attribute >50% of the variation of PC risk to genetic factors.⁵ The established genetic risk factors for PC range from rare pathogenic mutations in DNA repair genes (eg, BRCA2) to common single-nucleotide polymorphisms (SNPs) associated with PC risk. While >260 PC-associated SNPs have been identified to date through large-scale genome-wide association studies and used to generate polygenic risk scores (PRSs), most of these variants are associated with risk of developing any PC (not aggressive or high-grade PC). Studies investigating genetic association with aggressive PC have identified specific variants in the 8q24 and 19q13 regions associated with aggressive disease and/or PC-specific mortality.^6-9 To our knowledge, no prior studies have examined the combined prognostic value of family history and germline risk variants among patients with favorable-risk PC eligible for AS. In this study, we evaluated the combined impact of a positive expanded family history (including not only PC but other genetically related cancers) and germline risk SNPs on long-term outcomes of AS-eligible patients.

In a prospective cohort of 51,529 U.S. men in the Health Professionals Follow-up Study, 1,367 were diagnosed with clinically favorable-risk (National Comprehensive Cancer Network [NCCN] low- or favorable intermediate–risk) PC from 1986 to 2016 and underwent genome-wide SNP genotyping. Positive expanded family history was defined by having any first-degree relatives with prostate, breast, and/or pancreatic cancer. A previously validated PRS derived from 269 PC-associated SNPs was calculated for each patient. Multivariable Cox proportional hazards models were used to estimate the association between expanded family history, germline risk variants, and PC recurrence or death through 2017, adjusting for age, PSA, grade group, clinical T stage, and primary treatment.

The median follow-up time from diagnosis was 14.1 years. Median age at diagnosis was 69 years (IQR 64-74). At diagnosis median PSA was 5.9 ng/mL (IQR 4.5-8.1), and 80.8% of participants had grade group 1 PC. By NCCN risk stratification, 896 participants (65.5%) had low-risk disease and 471 (34.5%) had favorable intermediate–risk disease. During 30 years of follow-up, 263 (19.2%) patients had PC recurrence and 52 patients (3.8%) died of PC. A positive expanded family history was observed in 489 (36%) patients (Fig. 1, A). Men with a positive expanded family history had an 83% greater risk of PC-specific death (multivariable-adjusted HR 1.83, 95%CI 1.06-3.17, P = .03; Fig. 1, B). Notably, a model including positive expanded family history performed significantly better compared to the model including family history of PC alone (ANOVA, P < .001, Fig. 1, B).

We next investigated whether the 269-SNP PRS was associated with either PC recurrence or death and found the PRS was not associated with either outcome (Fig. 2). We then focused on specific variants with previously validated associations with PC aggressiveness and found rs2735839 on 19q13 to be significantly associated with PC-specific death (multivariable–adjusted HR 1.77, 95%CI 1.02-3.10, P = .04).

Finally, we evaluated the combined effect of both family history and common genetic variants. Our modeling demonstrated that family history and the 19q13 risk SNP rs2735839 were each independently associated with PC-specific death. We generated a heritable risk score giving 1 point each to a positive expanded family history and rs2735839 risk allele. Each point of the heritable risk score was associated with a 73% increased risk of PC-specific death (multivariable–adjusted HR 1.73, 95%CI 1.16-2.57, P = .007; Fig. 3, A). This association was even more pronounced for the subgroup of patients with NCCN low-risk disease (multivariable–adjusted HR 2.86, 95%CI 1.53-5.33, P < .001; Fig. 3, B).

Our study represents one of the largest cohorts, to our knowledge, in which the combined impact of family history and germline genetic risk on outcomes on AS-eligible PC has been investigated. Unique to our study is the long-term follow-up allowing investigation of PC death as a primary outcome and our broader definition of family history inclusive of other malignancies with shared genetic underpinnings with PC. While previous studies have concluded that family history of PC is not a risk factor for patients eligible for or on AS,¹⁰ these studies relied on a narrower definition of family history of PC alone, short-term outcomes such as adverse pathology at surgery, or upgrading on subsequent surveillance biopsy, and shorter follow-up time.

In summary, our study found that patients with AS-eligible prostate cancer with a family history of prostate, breast, or pancreatic cancer and/or a 19q13 germline risk allele had an elevated risk of PC-specific death. These findings have implications for how familial and germline genetic risk should be factored into clinical decision making around AS.