Hereditary Prostate Cancer

Introduction

The McKusick catalogue (1990) lists 66 Mendelian disorders that involve cancer or a predisposition for cancer as a component of the phenotype. The genetic loci of 21 of these 66 traits have been mapped by genetic epidemiological investigations. Prostate cancer has a well-established familial clustering, including both dominant and recessive syndromes similar to those found in breast and ovarian cancer syndrome families. However, extensive research in families affected by prostate cancer has so far failed to identify major genes as risk factors for prostate cancer. More recently, research has been directed to a “common disease common variant” approach employing a very large number of cases and controls in genome-wide association studies (GWASs). These association studies produced a large number of genetic variants that are common in the population, and are associated with prostate cancer, but their effect on prostate cancer risk was only modest. The most recent development in the field is the increasing utilization of high-throughput sequencing technology that is providing us with an unprecedented opportunity to thoroughly investigate the genome at single base resolution, and thus opening the door to understanding the role of genetic variants, common and rare, in the development of prostate cancer.

In this chapter, we summarize the evidence from family-based and GWASs. We also briefly discuss the clinical implications of our developing knowledge of the genetic basis of prostate cancer, and possible future directions for inherited prostate cancer research.

Epidemiologic studies of family history and prostate cancer risk

Studies of familial clustering of cancer provided the rudimentary basis for the investigation of a heritable component to prostate cancer. Morganti and Woolf were the first to demonstrate an increased risk of prostate cancer and prostate cancer death in relatives as compared to the general population. These findings were corroborated by later studies using a database of Utah Mormons. Steinbert and Walsh reported a case-control study to estimate the relative risk of developing prostate cancer in men with a positive family history. They found a 2-, 5-, or 11-fold increase in men with 1, 2, or 3 affected first-degree relatives. This implied that an increased number of affected relatives increased the genetic disposition for prostate cancer. These findings were substantiated in a more contemporary study by Lesko et al., who studied a case-control cohort of 1266 men from Massachusetts in the 1990s. The authors found an odds ratio (OR) of 2.3 in men who had a brother or a father with prostate cancer. The OR was 2.2 for men with a single relative with prostate cancer and 3.9 for men with two or more relatives with prostate cancer.

The relationship between family history of prostate cancer and age of onset was also examined by Lesko et al. who found that the association between family history and the risk of developing prostate cancer was greatest when prostate cancer was diagnosed in a family member younger than age 65 with an OR of 4.1. The OR for that association dropped to 0.76 if the affected family member was 75 years or older at the time of diagnosis establishing an inverse relationship between risk of prostate cancer and age of affected relative at time of diagnosis.

Many studies have attempted to determine if familial prostate cancer is related to biologic aggressiveness or risk of recurrence with mixed results. Kupelian et al. showed 3-year freedom from biochemical recurrence rates of 52.5% versus 72% for patients with and without family history, respectively. The 5-year rates showed a similar trend with biochemical freedom from recurrence of 29% versus 52%. When the authors stratified the cohort into two groups based on PSA <10, Gleason score <7, and clinical stage T1-2, they found that patients with a family history of prostate cancer had a 5-year biochemical recurrence free survival rate of 49% versus 80% for those with no family history. Patients showed a similar difference in biochemical recurrence free survival of 20% versus 35% (with and without family history, respectively).

The results from Kupelian’s study hinted to the possibility of increased aggressiveness of familial prostate cancer. Bova et al. attempted to characterize familial prostate cancer aggressiveness by matching a cohort of men who met the criteria for HPC to others by Gleason score, pathologic stage, and time interval of prostatectomy. The cohort was followed for 5 years and did not show a significant difference in the likelihood of freedom from biochemical recurrence.

In an effort to quantify the survival of patients with familial prostate cancer, Siddiqui et al. examined a cohort of 3560 patients who underwent prostatectomy for cancer from 1987 to 1997. A total of 865 patients were found to have either familial or hereditary prostate cancer. The authors found that although preoperative PSA was higher in patients with hereditary prostate cancer, no significant difference was found in the pathological stage, Gleason score, seminal vesicle involvement, margin positivity, or DNA ploidy. In addition, postoperative adjuvant therapy was equivalent in all groups. Importantly, the 10-year data from this cohort showed no difference in biochemical recurrence free survival, systemic progression free survival, or cancer-specific survival.

One major limitation of familial studies in prostate cancer is the potential for detection bias. Almost all of the authorities on screening for prostate cancer, including the American Urologic Association, recommend early screening for prostate cancer in males with family history of the disease. In addition, men with family history of prostate cancer have been shown in previous studies to be more likely to ask for prostate cancer screening. This higher prevalence of screening in males with family history of prostate cancer would obviously lead to a higher detection bias. A familial history of prostate cancer could also indicate a shared environmental exposure among members of the same family. However, twin studies were able to isolate genetic from environmental risk factors by comparing the risk of prostate cancer between monozygotic (MZ) and dizygotic (DZ) twins. Grönberg et al. studied 4840 male twin pairs, in which 458 prostate cancers were identified between 1959 and 1989. Among these, 16 MZ and 6 DZ twin pairs were concordant for prostate cancer. The rate of concordance for MZ versus DZ, respectively was 0.192 and 0.043, and the correlation of liability for the twin pairs was found to be 0.40 and −0.05. The authors concluded that the differences in concordance rates between the two groups were notable and implicate genetic factors in the development of prostate cancer. Along the same lines, a study of World War II veteran twins examined concordance rates to estimate the risk of a twin developing cancer given that the other twin had cancer. This study found that MZ versus DZ twins had a concordance rate of 27.1% versus 7.1%, which calculated to a risk ratio of 3.83. They concluded that concordance rates were higher in MZ pairs and estimated 50% of the variability in liability in this prostate cancer cohort was due to genetics.

Segregation and linkage studies

Twin and familial studies of prostate cancer were the impetus for further investigation into the possibility of a genetic cause of prostate cancer. Once a genetic cause is suspected, a segregation analysis can be performed to determine if a Mendelian mode of inheritance exists. Mendelian inheritance denotes the possibility of a specific gene involved in prostate cancer, a gene that could be theoretically shared by familial and sporadic forms of prostate cancer. While some of the early segregation studies have suggested a possible autosomal dominant transmission of a major rare risk factor for prostate cancer, other studies suggest more of a recessive, or polygenic, genetic basis for this disease.

With the distinct possibility of a genetic component indicated by segregation analyses, the next step in further qualifying the genetic causes for prostate cancer was to perform linkage analyses to specific genetic landmarks, and measure the extent of the linkage. The results from these studies were given in terms of a logarithm of odds (LOD) score (logarithm of the OR in favor of linkage vs. no linkage) as a metric for the likelihood of linkage. Table 17.1 demonstrates many of the contemporary studies attempting to establish genetic linkages to prostate cancer in different populations.

While many putative prostate cancer genes have been found and examined, evidence for the elusive high-risk allele is still inconclusive. The absence of a major finding from these linkage studies is strong evidence to the fact that prostate cancer risk is determined by a group of genes with small or modest risk effects. Having said that, the research for a major gene in prostate cancer should continue in families and homogeneous populations as previous research in similar groups have provided us with keys to understanding the biology of important cancers (Von Hippel–Lindau in kidney cancer, and the MSH2 gene in colon cancer).

Genome-wide association studies in prostate cancer

The development of commercially available and inexpensive arrays capable of genotyping up to one million single nucleotide polymorphisms (SNPs) in parallel, in combination with the cataloging human SNPs through the Human Genome Project and the International HapMap Project, has made it possible for prostate cancer research to move from familial and linkage studies to investigating the entire human genome for genetic variants associated with prostate cancer risk. These GWASs are based on the concept of linkage disequilibrium, which occurs when a certain allele combination (haplotype) is preserved over subsequent generations. What follows is that the likelihood of recombination between two alleles is a function of the genetic distance between them. Thus, tagging SNPs that are correlated with neighboring SNPs can be used to infer the human haplotype. This linkage phenomenon would therefore allow us to examine genetic variants in the entire genome without the need to tag every single SNP. These haplotypes are then compared between cases and controls, and are proving to be a very powerful tool to identify genetic variants associated with different diseases.

The large number of GWAS performed in prostate cancer research since 2007 is a testimony to the high success rates of such studies in discovering genetic variants associated with prostate cancer. The first GWAS in prostate cancer revealed a risk locus at 8q24, rs6983267 to be associated with prostate cancer. Ever since, GWAS have identified nearly 100 common variants associated with susceptibility to prostate cancer. Not only are these GWAS carried out to investigate the association of genetic changes with prostate cancer, but they are also helping to understand the genetic basis for disease aggressiveness, treatment complications, and outcomes. Herein, we summarize the findings of GWAS and examine the functional relevance of some of the SNPs that have been reported in multiple studies. This discussion is not all-inclusive of the studied SNPs, and the interested reader should review the original studies dealing with the subject.