The Evidence for Hereditary Prostate Cancer

Prostate cancer is an umbrella term for a clinically heterogenous group of diseases with distinct natural histories. Despite strong evidence for the existence of prostate cancer susceptibility genes, family-based linkage studies have been unable to identify compelling candidates. This raises the possibility that the clinical heterogeneity may be further complicated by an underlying genetic heterogeneity that the studies performed to date have been underpowered to detect.

Age, ethnicity, and family history are the principal risk factors for prostate cancer. Family history has been confirmed as a risk factor for prostate cancer irrespective of age, race, or ethnicity, with a reported odds ratio of 2.5 (95% confidence interval [CI], 1.9–3.3) after adjusting for age and ethnicity. In Sweden, a population-based study suggested that approximately 11.6% of all prostate cancer can be accounted for by familial factors alone. A meta-analysis of 33 studies investigating familial clustering suggested that risk was greater for men with affected brothers (risk ratio [RR], 3.4; 95% CI, 1.8–5.7) than for men with affected fathers (RR, 2.2; 95% CI, 1.9–2.5), and risk increased with the number of close relatives affected. Second-degree relatives (RR, 1.7; 95% CI, 1.1–2.6) conferred a lower risk than first-degree relatives, and 2 or more first-degree relatives (RR, 5.1; 95% CI, 2.6–4.2) conferred increased risk compared with 1 (RR, 3.3; 95% CI, 2.6–4.2). Twin studies have further investigated this familial clustering and confirmed a strong genetic component, with a concordance among monozygotic twins of 27% compared with 7.1% between dizygotic twins, suggesting that 42% (CI, 29%–50%) of prostate cancer risk is caused by hereditary factors.

The incidence of prostate cancer is 24 to 60 times higher in Western compared with Asian countries, and in the United States, incidence rates are 60% higher for African American men than for whites. Although the availability of screening may bias ethnic variations in disease incidence, the magnitude of the difference supports the credibility of this association. Dissecting the interplay of genetic, environmental, and social factors that contribute to these differences is challenging, but would undoubtedly be helped by an improved understanding of the genetic factors involved.

The later median age of onset of prostate cancer may support a polygenic inheritance pattern. With increasing age, somatic mutations accumulate in cells. Many predisposition syndromes, for example hereditary breast-ovarian cancer syndrome caused by germline BRCA mutations, manifest clinically when a second mutation occurs in a specific gene (the Knudson 2-hit hypothesis). Affected individuals inherit 1 abnormal copy of the gene, but the second normal copy of the gene must be knocked out, so that gene function is lost and the disease phenotype develops. If prostate cancer inheritance followed a similar pattern, and several different genes had to be lost before disease occurred, this may explain why early-onset prostate cancer is unusual and why individuals with inherited predisposition still develop late-onset disease.

Many segregation studies have been performed in an effort to explain the pattern of prostate cancer inheritance. Data supporting autosomal recessive, X-linked, and multifactorial inheritance have been reported but the strongest evidence is for a rare autosomal dominantly inherited gene with an allele frequency of 0.3% to 1.7%. However, such a gene would only explain a minority of prostate cancer inheritance and it remains likely that many prostate cancer susceptibility genes exist.

Prostate Cancer Candidate Genes

Summary of Prostate Cancer Predisposition

Phenotypic variation and genetic heterogeneity have complicated the search for prostate cancer susceptibility genes. Many candidates have been identified but few results have been consistently replicated. BRCA2 is one exception that has repeatedly been associated with prostate cancer, and, more recently, with an aggressive subset of the disease, but accounts for a small percentage (<1%) of overall prostate cancer susceptibility. Recent GWAS have identified common variants associated with prostate cancer and, as this technology advances and our understanding of the genome improves, the role these loci play in prostate cancer susceptibility may emerge. It remains possible that rare, single-gene disorders explain familial predisposition to cancer, but more accurate clinical and/or molecular characterization of disease within families will be required to utilize recent technological advances and to identify these genes.

The Evidence for Hereditary Bladder Cancer

Bladder cancer is the fifth commonest malignancy in the United States and, because of the cost of screening, is the most expensive to manage, making it an important public health issue. Transitional cell carcinoma represents more than 90% of bladder cancer. It most commonly occurs in the bladder but also arises in the renal pelvis or ureter (upper urinary tract), and, irrespective of its site of origin, is collectively termed urothelial cancer (UC). Many environmental causes of UC have been reported, among them smoking and exposure to aromatic amines, but epidemiologic data suggest that poorly defined genetic factors also play a role.

A population-based study in Utah reported that first-degree relatives of individuals with bladder cancer have an increased risk (RR, 1.51; 95% CI, 0.98–2.2) of developing the disease, and this increased to 5.07 (95% CI, 0.97–12.5) when the proband was less than 60 years of age. There are 16 multiple-case UC reports in the literature, detailing 32 families with 86 affected individuals. The early age of onset and co-occurrence of other malignancies seems to suggest a genetic predisposition, but it is difficult to separate this from shared environmental exposure. A large Dutch case-control study of 1193 UC cases and 853 controls reported a smoking-adjusted relative risk of 1.8 (95% CI, 1.3–2.7) for a positive family history of UC, suggesting that smoking history and genetic susceptibility are independent risk factors for the disease. Other case-control and cohort studies completed to date support a modest hereditary component to UC, with most risk ratios lying between 1.4 and 1.9.

High-penetrance UC Genes

The most common hereditary cancer syndrome with UC as a characteristic is LS. This is an autosomal dominantly inherited condition resulting from germline mutation in mismatch repair (MMR) genes. It is the most common form of hereditary colorectal cancer and extracolonic manifestations include many malignancies, including UC. The revised Bethesda Guidelines, which were developed to help identify individuals and families who may be at risk for LS, include UC of the upper tract but not bladder among the syndrome-defining malignancies. Although several reports of bladder cancer in LS exist, the association remains unclear. A Dutch LS study reported an RR of 14 (95% CI, 6.7–29.5) for upper-tract UC, but the risk for developing UC of the bladder was not increased in this cohort. A discussion of screening for UC is beyond the scope of this manuscript, but guidelines for individuals with UC presently recommend considering annual urinary cytology.

UC has been reported as a manifestation of other rare, high-penetrance syndromes. An association with hereditary RB was initially believed to be related to treatment effects from primary therapy for the RB ; however, survivors who did not receive radiation or chemotherapy had a considerably higher bladder cancer–specific mortality (standardized mortality ratio, 26.3; 95% CI, 8.5–61.4) than is seen in the general population, suggesting that UC of the bladder is a manifestation of hereditary RB. This is consistent with data demonstrating somatic mutations in RB or p53 in 50% of high-grade UC. Somatic mutations in HRAS and FGFR3 are commonly found in superficial UC. Costello syndrome is a hereditary condition with a large number of characteristic features, including UC of the bladder for adolescents and adults. HRAS is the only known mutation associated with this syndrome and, again, this germline association is consistent with somatic alterations identified in superficial UC. Germline FGFR3 mutations cause achondroplasia that is not associated with an increased risk of UC, but Apert syndrome, which results from mutations in FGFR2, has been associated with UC.

These high-penetrance predisposition syndromes explain a small minority of familial UC. Linkage studies have failed to identify other genes, and, although they have been small and underpowered, they seem to suggest a polygenic inheritance for the familial UC.

Low-penetrance UC Genes

Two candidate genes involved in carcinogen metabolism have been associated with an increased risk of bladder cancer. N -acetyltransferase 2 ( NAT2 ) is a phase II enzyme that detoxifies aromatic amines, 1 family of carcinogens found in tobacco smoke, and glutathione S-transferase ( GSTM1 ) is a phase II enzyme that detoxifies carcinogenic polycyclic aromatic hydrocarbons, such as benzopyrene. The NAT2 slow acetylator phenotype and the GSTM1 null genotype increase bladder cancer risk by 1.4- and 1.5-fold respectively, but, because of their high prevalence in the white population (40%–60% and 40%–50% of whites respectively), have been estimated to account for 31% of bladder cancer in whites.

Further support for the role of common genomic variation in bladder cancer predisposition was provided by recent GWAS. In a study of 4000 cases and 38,000 controls, Kiemeney and colleagues identified 3 new susceptibility loci at 8q24, 3q28, and 5p15 that increase the risk of bladder cancer by 22%, 19%, and 16%, respectively. The same 8q24 locus has been associated with prostate cancer as discussed earlier, as well as colorectal and breast cancer, suggesting that this poorly understood region may contain a common carcinogenic pathway. A second GWAS in 969 bladder cancer cases and 957 controls identified a second, independent SNP on 8q24 located in the promoter region of PSCA.

Summary of Bladder Cancer Predisposition

Individuals with LS are screened for UC, but most familial UC remains unexplained. Candidate genes involved in carcinogen metabolism have been associated with increased risk of the disease, and recent GWAS have identified further common variants that confer increased risk of UC. Like prostate cancer, UC seems to be a polygenic disorder, although rare familial single-gene disorders may exist. More is known about environmental causes of bladder cancer than the other genitourinary malignancies discussed, and future studies investigating genetic susceptibility should consider gene-environment interaction in the analysis.

The Evidence for Hereditary Bladder Cancer

Bladder cancer is the fifth commonest malignancy in the United States and, because of the cost of screening, is the most expensive to manage, making it an important public health issue. Transitional cell carcinoma represents more than 90% of bladder cancer. It most commonly occurs in the bladder but also arises in the renal pelvis or ureter (upper urinary tract), and, irrespective of its site of origin, is collectively termed urothelial cancer (UC). Many environmental causes of UC have been reported, among them smoking and exposure to aromatic amines, but epidemiologic data suggest that poorly defined genetic factors also play a role.

A population-based study in Utah reported that first-degree relatives of individuals with bladder cancer have an increased risk (RR, 1.51; 95% CI, 0.98–2.2) of developing the disease, and this increased to 5.07 (95% CI, 0.97–12.5) when the proband was less than 60 years of age. There are 16 multiple-case UC reports in the literature, detailing 32 families with 86 affected individuals. The early age of onset and co-occurrence of other malignancies seems to suggest a genetic predisposition, but it is difficult to separate this from shared environmental exposure. A large Dutch case-control study of 1193 UC cases and 853 controls reported a smoking-adjusted relative risk of 1.8 (95% CI, 1.3–2.7) for a positive family history of UC, suggesting that smoking history and genetic susceptibility are independent risk factors for the disease. Other case-control and cohort studies completed to date support a modest hereditary component to UC, with most risk ratios lying between 1.4 and 1.9.

High-penetrance UC Genes

The most common hereditary cancer syndrome with UC as a characteristic is LS. This is an autosomal dominantly inherited condition resulting from germline mutation in mismatch repair (MMR) genes. It is the most common form of hereditary colorectal cancer and extracolonic manifestations include many malignancies, including UC. The revised Bethesda Guidelines, which were developed to help identify individuals and families who may be at risk for LS, include UC of the upper tract but not bladder among the syndrome-defining malignancies. Although several reports of bladder cancer in LS exist, the association remains unclear. A Dutch LS study reported an RR of 14 (95% CI, 6.7–29.5) for upper-tract UC, but the risk for developing UC of the bladder was not increased in this cohort. A discussion of screening for UC is beyond the scope of this manuscript, but guidelines for individuals with UC presently recommend considering annual urinary cytology.

UC has been reported as a manifestation of other rare, high-penetrance syndromes. An association with hereditary RB was initially believed to be related to treatment effects from primary therapy for the RB ; however, survivors who did not receive radiation or chemotherapy had a considerably higher bladder cancer–specific mortality (standardized mortality ratio, 26.3; 95% CI, 8.5–61.4) than is seen in the general population, suggesting that UC of the bladder is a manifestation of hereditary RB. This is consistent with data demonstrating somatic mutations in RB or p53 in 50% of high-grade UC. Somatic mutations in HRAS and FGFR3 are commonly found in superficial UC. Costello syndrome is a hereditary condition with a large number of characteristic features, including UC of the bladder for adolescents and adults. HRAS is the only known mutation associated with this syndrome and, again, this germline association is consistent with somatic alterations identified in superficial UC. Germline FGFR3 mutations cause achondroplasia that is not associated with an increased risk of UC, but Apert syndrome, which results from mutations in FGFR2, has been associated with UC.

These high-penetrance predisposition syndromes explain a small minority of familial UC. Linkage studies have failed to identify other genes, and, although they have been small and underpowered, they seem to suggest a polygenic inheritance for the familial UC.

Low-penetrance UC Genes

Two candidate genes involved in carcinogen metabolism have been associated with an increased risk of bladder cancer. N -acetyltransferase 2 ( NAT2 ) is a phase II enzyme that detoxifies aromatic amines, 1 family of carcinogens found in tobacco smoke, and glutathione S-transferase ( GSTM1 ) is a phase II enzyme that detoxifies carcinogenic polycyclic aromatic hydrocarbons, such as benzopyrene. The NAT2 slow acetylator phenotype and the GSTM1 null genotype increase bladder cancer risk by 1.4- and 1.5-fold respectively, but, because of their high prevalence in the white population (40%–60% and 40%–50% of whites respectively), have been estimated to account for 31% of bladder cancer in whites.

Further support for the role of common genomic variation in bladder cancer predisposition was provided by recent GWAS. In a study of 4000 cases and 38,000 controls, Kiemeney and colleagues identified 3 new susceptibility loci at 8q24, 3q28, and 5p15 that increase the risk of bladder cancer by 22%, 19%, and 16%, respectively. The same 8q24 locus has been associated with prostate cancer as discussed earlier, as well as colorectal and breast cancer, suggesting that this poorly understood region may contain a common carcinogenic pathway. A second GWAS in 969 bladder cancer cases and 957 controls identified a second, independent SNP on 8q24 located in the promoter region of PSCA.

Summary of Bladder Cancer Predisposition

Individuals with LS are screened for UC, but most familial UC remains unexplained. Candidate genes involved in carcinogen metabolism have been associated with increased risk of the disease, and recent GWAS have identified further common variants that confer increased risk of UC. Like prostate cancer, UC seems to be a polygenic disorder, although rare familial single-gene disorders may exist. More is known about environmental causes of bladder cancer than the other genitourinary malignancies discussed, and future studies investigating genetic susceptibility should consider gene-environment interaction in the analysis.

Hereditary Kidney Cancer Syndromes

Conservative estimates suggest that 3% to 5% of kidney tumors are inherited. Explanation of the genetic causes of hereditary renal cancer (HRC) identified important molecular pathways in renal cell cancer pathogenesis. This understanding has proved instrumental in the development of clinical trials with targeted molecular therapies, which have become the new standards for medical therapies. Several HRC syndromes have been characterized. These include VHL disease, BHD syndrome, hereditary papillary (type I) renal cancer (HPRC), HLRCC, and tuberous sclerosis (TS) ( Table 3 ). Others, including familial renal oncocytoma and the association between lymphoma and kidney cancer, have been reported, although genetic evidence is lacking.

Management Summary for HRCs

A therapeutic and preventive strategy in patients with HRC requires coordinated multimodality care, which serves to minimize treatment related morbidity and preserve renal function. The general approach to HRC includes screening tests and physical examinations, as well as a regimented schedule of follow-up appointments. Recent advances in nephron-sparing techniques that prioritize functional kidney preservation and expectant management of lesions have diminished kidney-related mortality and chronic renal dysfunction. Retrospective studies have demonstrated the safety of observation of solid renal masses to a diameter of 3 cm. Minimally invasive surgical procedures for multiple partial nephrectomies have also been used satisfactorily at experienced centers, including tumor ablation performed in prospective clinical trials. Long-term outcomes with regard to efficacy and renal function have not been reported.

Syndrome-specific management strategies have also been developed. VHL mutation carriers require neurologic, otologic, ophthalmologic, and endocrine follow-up involving magnetic resonance imaging (MRI)/CT scanning. The development of pheochromocytomas can complicate patient management, increasing the risk of intracranial bleeding in patients at risk for CNS or ocular hemangiomas. Annual functional adrenal studies, and secondary meta-iodobenzylguanidine immunoscintigraphy are often indicated. Medical management before surgical resection is the most frequent management strategy. Partial adrenalectomy for a solitary adrenal gland or bilateral lesions can be safely performed to avoid steroid replacement therapy.

HLRCC management differs from other HRC syndromes because of the aggressive nature of renal tumors. Among affected individuals, rapid distant disease progression, even when the primary tumors are small, is usually the case. This is also evident in pediatric patients. As a result, thorough and early imaging assessment, close follow-up, and rapid intervention is the management paradigm for these patients. Nephron-sparing surgery is less well established in this setting, and complete wide excision, including lymph node dissection, should be attempted. Preoperative positron emission tomography scans may prove beneficial in cases where lymph node or nonlocalized disease is suspected.