Prostate cancer is the most common cancer and the second most common cause of cancer death in North American men. Population prostate cancer risk is approximately 0.5 % by the age of 65 years and 2 % by the age of 75 years. The etiology of prostate cancer is unknown, but androgen stimulation is implicated by observations of a low incidence in males castrated before the age of 40 years, and a high-fat diet has also been implicated. Evidence for a genetic contribution was provided by genetic epidemiological studies in Mormon families that suggested that the heritability of prostate cancer was greater than that of breast or colorectal cancer. A case-control study by Steinberg et al. (1990) found 15 % of prostate cancer patients had an affected father or brother, compared to 8 % of controls (Table 1).
Table 1 Cumulative risks of prostate cancer according to ethnicity, age, and family history
|Cumulative risk of prostate cancer by age|
|Group||50 years||60 years||70 years|
|No family history in (%)||0.2||2||7.5|
|One FDR in (%)||0.4||5.2||19|
|Two or more FDR in (%)||0.8||10||38|
|No family history in (%)||0.4||3.6||10.6|
|One FDR in (%)||1.1||9.2||27.1|
|Two or more FDR in (%)||2.1||18||34|
Adapted from Nieder et al. (2003)
FDR first-degree relative
Furthermore, the RR increased: (1) with the number of affected relatives such that men with one, two, and three affected first-degree relatives were at twofold, fivefold, and 11-fold increased risk for developing prostate cancer, respectively, and (2) the younger the age at onset of prostate cancer in the proband. Meikle and Smith (1990) reported. A 17-fold increase in RR of prostate cancer has been reported in brothers of men developing prostate cancer between 45 and 50 years of age. Segregation analysis of the family histories of 740 prostate cancer patients suggested that familial clustering of the disease could be caused by a rare, highly penetrant, dominantly inherited predisposition gene. Under the most likely genetic model, 43 % of early-onset prostate cancer (in men <55 years) would occur in gene carriers, but only 9 % of cases in men aged <85 years. This proposed model of a rare dominant predisposing gene (or genes) was similar to that suggested for breast cancer. Johns and Houlston (2003) undertook a meta-analysis of 13 studies of prostate cancer risk in first-degree relatives. They found that the pooled RR in first-degree relatives was 2.5 and 3.5 in men with one or two affected relatives. Risks were also increased if the affected relative had early-onset disease (§60 years). Nieder et al. (2003) provided cumulative risks by age according to family history details and ethnicity (i.e., Table 6.5). The clinical course of familial and sporadic prostate cancer appears to be similar, although the disease may be more aggressive in African-Americans.
Multiple prostate cancer susceptibility loci have been mapped by family linkage studies (e.g., CAPB, HPC1, HPC2, HPX, MSR1, PCAP, HPC20, RNASEL), but none of these equates with major high-penetrance susceptibility genes present in many populations (as for BRCA1 and BRCA2 in familial breast cancer). Although mutations in three candidate prostate cancer susceptibility genes have been reported, RNASEL (HPC1, 1q24–q25), MSR1 (8p22–p23), and ELAC2 (HPC2, 17p11), in most cases these genes do not appear to represent rare highly penetrant loci and familial risks of prostate cancer may be explained better by a model of multiple interacting low-risk genetic variants. On the other hand, germline BRCA2 mutations do account for a small but significant group (2–5 %) of familial prostate cancer clusters or early-onset cases (Edwards et al. 2002), and the relative risk of prostate cancer in carriers is increased significantly, higher at younger ages, in mutation carriers (RR 4.71) (Kirchhoff et al. 2004). The risk of prostate cancer is also increased in males with Lynch syndrome, highest in MSH2 mutation carriers. Kaplan–Meier analysis suggested that cumulative risk by 70 years in MMR mutation carriers may be as high as 30 % (SE, 0.088) as compared to 8.0 % in the general population (Grindedal et al. 2009, Barrow et al. 2013).
The prognosis of prostate cancer in men with germline mutations in BRCA2 appears to be worse than in the general population. Men who carry BRCA2 mutations are more likely to develop early-onset prostate cancer (Grönberg et al. 2001; Willems et al. 2008; Tryggvadóttir et al. 2007; Mitra et al. 2008) and have a shorter survival time compared to BRCA1 mutation carriers and to men in the general population (Narod et al. 2008; Edwards et al. 2010). In the latter study, the median survival in men with germline BRCA2 mutation was 4.8 years, compared 8.5 years in controls (P = 0.002). Loss of heterozygosity was found in the majority of Tumours of BRCA2 mutation carriers and multivariate analysis confirmed that the poorer survival of prostate cancer in BRCA2 mutation carriers is associated with the germline BRCA2 mutation per se. Others have also suggested that the decreased survival in these men may be explained by the finding of an aggressive phenotype (Gleason score ≥8 and high T stage, ≥pT3) in the majority of Tumours in the BRCA2 carriers who die from prostate cancer (Thorne et al. 2011a, b). A strikingly worse outcome for prostate cancer has been reported in carriers of founder BRCA2 mutations in Icelandic population (Sigurdsson et al. 1997; Tryggvadóttir et al. 2007). In addition, the Ashkenazi Jewish BRCA2 founder mutation, c.5946delT (p. Ser1982fs), confers a 3-fold risk for prostate cancer, and carriers are more likely to develop high-grade prostate cancer than noncarrier (Gallagher et al. 2010).
Grönberg et al. (2001) reported a large family with a truncating BRCA2 mutation which was responsible for hereditary prostate cancer in the father and four of his sons who developed the early-onset disease. This mutation was also detected in three daughters diagnosed with breast cancer. All of the affected men with prostate cancer died of metastatic disease. Thus, more aggressive screening and treatment for prostate cancer may be justified, and an international study to this effect, known as IMPACT, is underway.
Fine-mapping by massively parallel sequencing of a previously identified susceptibility loci at 17q21–22 (Zuhlke et al. 2004) revealed a single mutation, HOXB13 G84E, that co-segregated with the disease in four unrelated families and confirmed to be a risk allele in a large case-control population (Ewing et al. 2012). Independent analysis in an international sample of prostate cancer families validated G84E to be a disease-predisposing mutation, present among 5 % of prostate cancer families with an estimated odds ratio of 4.42 (Xu et al. 2013). Although in most Western populations the allele frequency is less than 1 per 1,000 men, the G84E variant has a much higher prevalence in Nordic populations and is found on a common haplotype block, suggesting presence of a founder effect originating in Finland; the missense variant appears to be very rare or absent among African-Americans and Asians (Gudmundsson et al. 2012). While it is
unknown how the G84E variant predisposes to prostate cancer, the recurrent nature of the mutation and the lack of inactivating mutations suggest a gain-of-function mutation. Functional studies of the G84E variant are needed to clarify its role in cancer predisposition.
Men at increased risk of prostate cancer can be offered screening by annual measurement of prostate-specific antigen (PSA) and digital rectal examination from age 40 years. Male BRCA1/BRCA2 mutation carriers are eligible for screening by annual PSA between 40 and 69 years with prostate biopsy if PSA > 3 mg/ml, under a research trial, IMPACT (Mitra et al. 2011).
Genome-wide association studies have discovered at least 50 separate loci that are associated with an increased risk for prostate cancer. All odds ratios are individually less than 1.5, but it is possible that combinations of alleles, and combining alleles with other established risk factors such as body mass index and family history could be of value n refining prostate cancer risk estimates. They could also be useful in modifying the risk of those carrying more highly penetrant alleles, such as HOXB13 and BRCA2 mutations.