Main

Prostate cancer (PRCa) constitutes a major health issue worldwide. Prostate cancer is one of the most common causes of cancer death in men (Parker et al, 1996; Prior and Waxman, 2000). The incidence of PRCa increases with age and it is estimated that 80% of men would be affected by the age of 80 years (Holund, 1980). The aetiology of PRCa is unclear, although current evidence suggests that PRCa is the result of multiple factors that include ethnicity, environmental, genetics, hormonal and dietary factors (Pienta and Esper, 1993; Whittemore et al, 1995; Wingo et al, 1996; Hsieh et al, 1999; Tzonou et al, 1999; Lichtenstein et al, 2000). Benign prostatic hyperplasia (BPH) is a non-neoplastic enlargement of the prostate. Benign prostatic hyperplasia is extremely common, with a rapid increase in prevalence in the fourth decade of life. According to epidemiological studies, most cancers are associated with BPH elsewhere in the prostate (83.3%; Carter and Coffey, 1990; Bostwick et al, 1992) and approximately 3–20% of patients who have undergone transurethral prostatectomy (TURP) or open prostatectomy for BPH subsequently develop PRCa (Armenian et al, 1974; Schwartz et al, 1986; Bostwick et al, 1992). Compared to men without BPH, those with the condition have a five-fold raised risk of developing PRCa and a four-fold raised risk of death from PRCa (Armenian et al, 1974). A previous study reported that a family history of prostate disease (PRCa or BPH) was more frequently seen in relatives of men with BPH (20%) than in relatives of men with PRCa (12.8%) or in healthy controls (5.1%) (Schuman et al, 1977). In addition, in vitro malignant transformation of BPH tissue has been previously reported (Chen and Heidelberger, 1969; Fraley et al, 1970; Frank and Wilson, 1970). These results suggest that common genetic mechanisms may predispose to benign and malignant prostate disease. Moreover, these results suggest that BPH may be part of a premalignant environmental condition in the prostate gland. With the increasing incidence of PRCa in many populations, there is an urgent need for the identification of molecular markers that can serve as indicators of disease risk to focus chemoprevention and early detection strategies. Many candidate PRCa genes have been suggested, including genes influencing cellular growth and differentiation. The cytochrome P450 3A4 enzyme (CYP3A4) is a member of the human P450 family. CYP3A4 protein is responsible for hydroxylation of testosterone, which results in the deactivation of the hormone function (Waxman et al, 1988; Yamazaki and Shimada, 1996). A single-nucleotide polymorphism (SNP) in the CYP3A4 promoter (−290 A to G) was previously reported with two CYP3A4 alleles; CYP3A4*1A is the wild type (−290A) and CYP3A4-V, now designated CYP3A4*1B (−290G), is the variant (Rebbeck et al, 1998; Sata et al, 2000). The functional influence of this SNP is unclear, but initial in vitro and in vivo studies suggest a role in transcriptional control leading to altered CYP3A4 enzyme activity for a number of substrates, including testosterone (Amirimani et al, 1999; Rebbeck, 2000; Wandel et al, 2000). Genetic epidemiology studies found that the CYP3A4*1B allele was associated with higher clinical stage and grade of PRCa (Rebbeck et al, 1998; Paris et al, 1999; Kittles et al, 2002). The CYP3A4*1B allele frequency has been shown in various studies to vary markedly between ethnic groups and match solely with the incidence of PRCa based on ethnicity (Walker et al, 1998; Ball et al, 1999; Paris et al, 1999; Sata et al, 2000; Tayeb et al, 2000; Kittles et al, 2002). The highest incidence of PRCa was found in African Americans, intermediate in Caucasians, and the lowest in Asians (Pienta and Esper, 1993; Wingo et al, 1996). Vitamin D has been implicated in PRCa, with several epidemiological studies linking low vitamin D levels with increased risk of PRCa (Schwartz and Hylka, 1990; Corder et al, 1993). Calcitriol, the biologically active metabolite of vitamin D, 1,25-dihydroxyvitamin D3, has been shown to inhibit prostate cell growth (Skowronski et al, 1993; Hedlund et al, 1996). The action of vitamin D is mediated through binding to its nuclear receptor (VDR). The inherited TaqI SNP in exon 9 of the VDR 3′UTR regions (C352T) has been demonstrated to affect vitamin D levels (Morrison et al, 1994; Ma et al, 1998). Previous studies observed an association between the TaqI SNP and PRCa risk (Taylor et al, 1996; Correa-Cerro et al, 1999; Hamasaki et al, 2001). Our previous nested case–control association studies found that the frequencies of the CYP3A4*1B and VDR TaqI TT genotypes are higher among BPH patients who subsequently develop PRCa than among BPH control patients (Odds ratio, OR: 5.2 and 5.16, respectively; Tayeb et al, 2002 submitted). Moreover, we found that the frequency of CYP3A4*1B and VDR TT combined genotypes is increased in BPH patients who developed PRCa later on in their life compared with BPH patients who did not, and the risk of developing PRCa was 13-fold higher in BPH patients having the CYP3A4*1B and VDR TT combined genotypes than the control (Tayeb et al, 2002, submitted). The association between CYP3A4 and VDR TaqI SNPs, and the risk of developing PRCa in BPH patients have been investigated further in this study by determining the CYP3A4*1B and VDR TT genotype frequencies in 400 patients with BPH who have been followed up to 11 years.

Materials and methods

Data for BPH patients from years 1989 to 1990 (Northeast Scotland; Grampian region) were collected using the University of Aberdeen Department of Pathology data bases. In total, 1010 samples were identified. Data for PRCa from years 1989 to April 2000 were also collected and 44 of the 1010 BPH patients (4.4%) subsequently developed PRCa in the period 1989–April 2000. The geographic region has very little population migration over generations and is served by a single pathology department. Of the 1010 BPH samples, 400 were randomly selected for further molecular analysis, of which 21 had subsequently developed PRCa (5%). All sections were rereviewed by pathologist to confirm the diagnosis.

Genotyping

DNA was extracted from formalin-fixed, paraffin-embedded tissues. The tissue sections were deparaffinised with xylene and ethanol, and then DNA was isolated by proteinase K digestion (Frank et al, 1996). A 289 bp fragment CYP3A4*1B was amplified by PCR and screened using single-strand conformation polymorphism (SSCP) analysis (Tayeb et al, 2000). Previously described primer set was used to amplify the region of 198 bp around the VDR TaqI polymorphic region (Lundin et al, 1999). Genomic DNA (100–500 ng) was subjected to PCR amplification in a 25 μl reaction mixture containing 10 × PCR buffer (MBI, Sunderland, UK), 1 mM MgCl2 (MBI), 200 μ M dNTP mix (Bioline, London, UK), 10 pmol of each primer, 1 U of Taq polymerase (Roche, Lewes, UK), and sterilised distilled water. The genomic DNA was initially denatured at 94°C for 2 min and thereafter subjected to 35 cycles of PCR amplification with denaturation for 1 min at 94°C, annealing for 2 min at 60°C, extension for 2 min 30 s at 72°C, and a final extension step at 72°C for 10 min. The PCR products were digested with the TaqI restriction endonuclease (Roche, Lewes, UK) at 65°C for 5 h. Genotypes for the SNPs were determined after separation on a 3% agarose gel. Individuals were scored as TT homozygous (absence of TaqI restriction sites), Tt heterozygotes, or tt (presence of TaqI restriction sites).

Statistical analysis

Random selection for cohort samples was made using Minitab software version 12.1. The incidence rate and the relative risk (RR) of developing PRCa with studied markers and the power of the cohort study were calculated using Stata 1.0 software.

Results

CYP3A4*1B frequencies across populations

Our cohort study had 83% power to detect an RR of 4. The overall incidence rate of PRCa in this study was 645 per 100 000 men-year. Genotype frequencies of the CYP3A4 SNP in the cohort population are shown in Table 1. From Table 1, the frequencies of the CYP3A4*1B homozygote and heterozygote genotypes were higher in BPH patients who developed PRCa during the time of follow-up compared to BPH patients who did not. Genotype frequencies of CYP3A4 SNP and incidence rate of PRCa in the cohort study are shown in Table 2. From Table 2, the incidence rate of PRCa was higher in BPH patients with CYP3A4*1B genotype compared to BPH patients with CYP3A4*1A homozygotes. The RR of developing PRCa was 2.7 (95% CI=0.77–7.66) in BPH patients having a CYP3A4*1B genotype.

Table 1 Distribution of CYP3A4 genotype frequencies in the cohort population
Full size table
Table 2 Distribution of CYP3A4 genotype frequencies and incidence rate of PRCa in the cohort population
Full size table

VDR TT genotype frequency across population

The power of the cohort study was predicted to detect an RR of 4 with 82% power. Genotype frequencies of the TaqI SNP in the cohort population are shown in Table 3. From Table 3, the frequency of the TT genotype is similar in BPH patients who developed PRCa to BPH patients who did not (33 and 36%, respectively). The incidence rate of PRCa was lower in BPH patients with TT genotype compared to BPH patients with Tt or tt genotypes (Table 4). The RR of developing PRCa was 0.86 (95% CI=0.29–2.28) in BPH patients having a TT genotype. However, the results were not statistically significant.

Table 3 Distribution of TaqI genotype frequencies in the cohort population
Full size table
Table 4 Distribution of TaqI genotype frequencies and incidence rate of PRCa in the cohort population
Full size table

Combined genotype analysis

The frequencies of the CYP3A4*1B homozygote, heterozygote (A/G and G/G), and VDR TT combined genotypes were higher in BPH patients who developed PRCa during the time of follow-up compared to BPH patients who did not (Table 5). The incidence rate of PRCa was higher in BPH patients with CYP3A4*1B (A/G and G/G) and VDR TT combined genotypes compared to BPH patients with other combined genotypes (Table 6). The RR of developing PRCa was 3.43 (95% CI=0.99–11.77) in BPH patients having a CYP3A4*1B and VDR TT combined genotypes.

Table 5 Distribution of CYP3A4 and VDR TaqI combined genotype frequencies in the cohort population
Full size table
Table 6 Distribution of CYP3A4 and VDR TaqI genotype frequencies and incidence rate of PRCa in the cohort population
Full size table

Discussions

Both PRCa and BPH are common diseases for which the incidence increases with age. Previous studies have defined a significant association between BPH and developing PRCa (Armenian et al, 1974; Bostwick et al, 1992). With the increasing incidence of BPH in the ageing population, there is an urgent need for the identification of molecular markers that can serve as prognostic indicators for developing PRCa in those patients with BPH. Germline and somatic variations in genes directly involved in the regulation of prostate cell growth might be critically important in understanding the carcinogenesis of PRCa, as these variants might be used as diagnostic, prevention, and prognostic markers for PRCa. The primary aim of this study was to identify molecular markers that are important in the development of PRCa in patients with BPH. If molecular markers in patients with BPH are shown to be predictors for eventually developing PRCa, then more intensive surveillance and/or early treatment could be offered to those selected patients. Such an approach is also likely to lead to improvements in survival. In the converse situation, those patients who do not have a high risk of developing PRCa could be offered standard follow-up monitoring. Our previous nested case–control association studies showed that a constitutive CYP3A4 and VDR TaqI SNPs are associated with a group of men with BPH who are at an increased risk of PRCa (Tayeb et al, 2002, submitted). The association between these two SNPs and risk of developing PRCa have been investigated further in this study by determining the CYP3A4*1B and VDR TT genotype frequencies in 400 patients with BPH (1989–1990). The median years of follow-up for these patients were 11 years and during the time of follow-up, 21 BPH patients developed PRCa. The results of this study showed that the incidence rate of PRCa was higher in BPH patients having a CYP3A4*1B genotype compared to those homozygous for CYP3A4*1A, but it was not statistically significant. The RR of BPH patients developing PRCa was 2.7, although results failed to reach significance at the 5% level. Regarding VDR TaqI SNP, the RR of BPH patients developing PRCa was 0.86 in patients with the TT genotype, although results were not statistically significant. This lack of significance could be because of the limited power of the cohort study, as the power of the study was determined to detect an RR of 4 with 83 and 82% for CYP3A4 and VDR TaqI SNPs, respectively. The power of a cohort study depends on several factors: (1) the number of subjects enrolled in the cohort, (2) the time for which each subject is followed-up, (3) the rate at which events (PRCa) occur in the cohort, (4) the frequency of ‘exposure’ to the hypothesised risk factor in the cohort (in this case, the frequency of the CYP3A4 genotypes or/and VDR TaqI genotypes, which had been hypothesised to be associated with increased risk of developing PRCa), (5) the size of RR which the investigator want to detect. As these factors change, the power of the study, or the necessary sample size, also changes. PRCa is likely to be caused by complex interactions between genetic, endocrine, and environmental factors. Ethnic differences in the risk of developing PRCa suggest that in addition to environmental factors, common genetic variants with low penetrance and high population attributable risk may play an essential role in the aetiology of PRCa. This study examined the data for gene–gene interactions between putative risk genotypes, CYP3A4*1B and VDR TT. These genetic variations confer an increased risk for the development of PRCa through their mediation of prostate cell growth and differentiation. The identification of evidence of a significant interaction (patients with both risk genotypes) may not necessarily indicate that the two genes are synergistic. They may instead influence risk via independent mechanisms. Gene–gene interactions might be important for the development of PRCa and this interaction needs to be explored. The results of this study showed that BPH patients who subsequently developed PRCa have significantly different frequency of harbouring CYP3A4, and VDR at risk genotypes than those BPH patients who did not develop PRCa (13-fold, Tayeb et al, 2002, submitted). It is interesting to notice that the ORs obtained from these combined genotypes (A/G and TT) were higher than those ORs obtained from each individual variant: 5.2 and 5.16 for heterozygous CYP3A4*1B and homozygous VDR TT, respectively (Tayeb et al, 2002, submitted). This study observed a borderline significant association between these combined genotypes and PRCa risk (RR=3.43; 95% CI=0.999–11.770). The RRs for these combined genotypes were higher than those obtained for each individual marker: 2.7 and 0.86 for heterozygous and homozygous CYP3A4*1B, and homozygous VDR TT, respectively. Calcitriol has been reported to inhibit PRCa proliferation and to promote a more differentiated phenotype. VDR, TaqI SNP has been demonstrated to affect transcriptional activity and mRNA stability, thus altering the abundance of VDR, and in turn affects vitamin D level (Morrison et al, 1994). In addition, higher levels of calcitriol have been reported in those who are homozygous for the t (TaqI site) allele relative to those who are homozygous for the T (no site) allele (Morrison et al, 1994; Ma et al, 1998). It has been speculated that men with the CYP3A4*1B genotype may have altered testosterone metabolism, promoting androgen-mediated prostate carcinogenesis and the occurrence of PRCa (Rebbeck et al, 1998). It might be that BPH patients harbouring both CYP3A4 and VDR TaqI combined risk genotypes have a higher level of androgen hormones and lower level of calcitriol, which might lead to an increase in prostate cell growth and reduce the level of differentiation and apoptosis, which might result in the occurrence of PRCa. However, larger studies are clearly needed to confirm these assumptions. If confirmed, the genetic risk factors examined in this study (VDR, and CYP3A4) are among the strongest risk factors yet identified for PRCa. The finding of this study is consistent with a multigenic model of PRCa, where PRCa risk is influenced by gene–gene and gene–environment interactions. On the basis of the joint effect of several loci, one might ultimately be able to construct a risk profile that could predict the development of the disease and could allow for a more meaningful decision making regarding optimal treatment strategies.