Abstract
Background/Aim
Several studies reported that patients with benign prostatic hyperplasia (BPH) experienced a 10% increased incidence of prostate cancer (PCa) after the first 5 years of diagnosis. We investigated the association between single nucleotide polymorphisms (SNPs) in the promoter of Serine Protease Inhibitor Kazal Type 1 (SPINK1) and the increased risk of BPH and PCa.
Materials and Methods
We genotyped three SNPs in a cases-control study, including BPH and PCa cases. Multiple logistic regression models were applied to analyze clinical and genotypic data.
Results
We found an inverse association between SNP rs10035432 and BPH under the log-additive (p=0.007) model. No association was found between these SNPs and PCa risk. However, we observed a possible association between rs1432982 and lower-grade PCa (p=0.05) under the recessive model.
Conclusion
SPINK1 promoter variants are likely to be associated with the risk of BPH.
Keywords: African American, prostate cancer, Gleason, BPH, SPINK1, SNPs
Benign prostatic hyperplasia (BPH) is a commonly diagnosed disease in men above 50 years of age (1–3). It refers to the non-malignant growth of the prostate and is defined as prostate gland enlargement secondary to hyper-proliferation of stromal and granular cells with predominance of mesenchymal cells (4). The etiology and pathogenesis of BPH is still poorly understood (5). Many factors have been reported to play a role in the development of BPH, including inflammatory genes, inflammatory mediators, hormones, dietary factors and oxidative stress (6). Additional risk factors include a positive family history of BPH, genetic polymorphisms and early age of onset of BPH (7–11). Another prostatic disease, prostate cancer (PCa), is the second leading cause of cancer-related deaths among men in the United States (12). Chokkalingam et al. reported that patients with BPH experienced a 10% excess incidence of PCa after the first 5-years of follow-up (13). Several researchers reported a genetic association between BPH development and polymorphisms of specific genes expressed in the prostate (14–16). In a recent study, Väänänen et al. reported that Serine Protease Inhibitor Kazal Type 1 (SPINK1) expression was detected in histologically benign-area from PCa specimens (17). However, there exists limited information about single nucleotide polymorphisms (SNPs) in the SPINK1 gene association with the risk of BPH.
In addition, previous studies have shown that SPINK1 is over-expressed in PCa tissues, especially high-grade tumors (Gleason score >7) (18, 19). SPINK1 over-expression exclusively occurs in a subset of PCa tissues without E26 transformation-specific (ETS) gene fusion suggesting that SPINK1-positive tumors may represent a separate PCa subtype (20–22). In a urine-based analysis, SPINK1 showed a significant association for discriminating between men with or without PCa disease and had a greater specificity for diagnosing PCa than prostate-specific antigen (PSA) (23). Recently, it was reported that SPINK1 over-expression is more common in African American (AA) PCa tissues compared to Caucasian men (24). Although previous studies have implicated SPINK1 as a molecular biomarker for a subset of PCa cases, the exact functional role of SPINK1 in PCa is unclear. However, it has been speculated that SPINK1 over-expression is a consequence of upstream genetic events (20).
SPINK1 is located on chromosome 5q32, encodes a 56-amino acid peptide, which functions as an acute-phase protein (APP) in the pancreas (25). Initially, SPINK1 was thought to prevent premature activation of trypsinogen (inactive form of trypsin), thereby ensuring the integrity of acinar cells to prevent auto-digestion of the pancreas (26–28). Some studies have reported SPINK1 to play a role in tissue repair, apoptosis, reproduction and behave as a growth factor in other human organs (29–33).
The promoter of SPINK1 contains a consensus sequence “TTGNNGNAATG” for the nuclear factor-interleukin six (NF-IL6) binding site, known as the IL-6 response element (IL-6-RE) (25). This sequence is a 40-bp DNA fragment located between −3.84 kb and −3.80 kb upstream from the start site (25). This fragment is conserved among various acute-phase genes and is responsible for both SPINK1 transcriptional activity and IL-6-induced expression of SPINK1 (34, 35). Given that IL-6 expression is associated with aggressive PCa phenotypes, this transcription factor site may be involved in the gain-of-function of SPINK1 in PCa tissue (36–38). Still, the pathological relevance of the known SPINK1 promoter variants is largely unknown. Therefore, we hypothesize that polymorphic variations in the IL-6RE of the SPINK1 gene may be associated with increased risk of BPH and PCa. To approach our hypothesis, we (i) genotyped three known single nucleotide polymorphisms (SNPs) in SPINK1 promoter in PCa patients, sub-group of men diagnosed with BPH and controls, (ii) performed a functional in silico analysis to determine the functionality of the tested SPINK1 promoter variants and (iii) integrated the resulting data with the clinical and pathological characteristics in order to assess the pathological relevance of each SNP.
Materials and Methods
Study population
This is a case-control study comprised of 2 selected cohorts from the “African American (AA) Sporadic PCa” study (AAPCA) and the “Vitamin D and PCa Risk in AA Men” studies. These studies were conducted at the National Human Genome Center (NHGC) and Howard University Cancer Center (HUCC). The Howard University Institutional Review Board (IRB) approved the study protocols (IRB-02-MED-42; IRB-11-MED-22) and written informed consent was obtained from all study subjects. Briefly, participants were recruited from the Division of Urology at the Howard University Hospital (HUH) and/or from ongoing free PCa screening program at the HUCC. From the above mentioned two studies, we selected (i) 307 non-related men of African descent (i.e. self-reported AA, Afro-Caribbean, African Latinos, East and West Africans) from the Washington DC area with histologically diagnosed adenocarcinoma of the prostate, PSA of >4.0 ng/ml and a positive digital rectal examination (DRE); (ii) 96 subjects diagnosed with BPH based on the clinical symptoms, such as scrotal discomfort and urination problems; and (iii) control subjects (n=339) with PSA levels <4.0 ng/ml, normal DRE and with no history of PCa among first-degree relatives. Blood samples were drawn from each participant and clinical characteristics, such as Gleason score (high and low Gleason score were defined as less advanced “if a case’s Gleason score is <7” and more advanced “if their Gleason score is >7”) and PSA level were obtained from the HUH Tumor Registry and pathology records.
SNP selection
The SNPs in the IL-6 response element in the promoter region of SPINK1 were selected using National Center for Biotechnology Information (NCBI) (39) and International Haplotype Map (HapMap) (40). Utilizing data from Utah residents with Northern and Western European ancestry from the CEPH collection (CEU) and Yoruba in Ibadan, Nigeria (YRI) population, a Chi-square test was performed to determine significant differences in minor allele frequencies (MAF) within the populations. Haploview version 4 was utilized for selecting the tagSNPs in IL-6 response element in the SPINK1 promoter region (41). The selection parameters were set to capture MAF of 5% or more and r2 values of 0.8 or more in the promoter region of SPINK1. A total of 56 SPINK1 promoter SNPs have been reported in HapMap. Out of these 56 SNPs, 15 were located in the NF-IL-6 binding site. Out of these 15 SNPs, four had minor allele frequencies of 10% or greater, which was needed for adequate power. Of the four SNPs, three variants, rs10035432, rs11748222 and rs1432982, showed significant differences in MAF between CEU and YRI and had adequate power for the sample set.
Genotyping
Genomic DNA was extracted from lymphocytes using the Blood and Cell Culture DNA Midi Kit (Qiagen, Germantown, MD, USA) in accordance with the manufacturer’s instructions. Twenty ng of DNA was used in each predesigned TaqMan® 5′ allelic discrimination assay (Life Technologies, Grand Island, NY, USA) for genotyping the 3 selected SNPs. For quality control, subjects were genotyped in duplicate and the overall concordance rate was 99.9%. Genotyping assays of the SPINK1 promoter SNPs (Table I) were performed using Applied Biosystems 7900HT Real-Time PCR System with TaqMan SNP assay reagents. The ABI 7900HT Sequence Detection System (SDS) software (version 2.0) was used to analyze real-time, end-point fluorescence and allelic discrimination (Life Technologies, Grand Island, NY, USA).
Table I.
TaqMan® SNP genotyping assays identification (ID) and probe sequences.
| SNP ID | Genotyping Assays ID | * Context Sequence [VIC/FAM] |
|---|---|---|
| rs10035432 | C__30608242_20 | TAATTGAGCTTTAGCTATAAAATCC[A/G]TATATATTCCTCTGAAGCTTAAATA |
| rs11748222 | C___3057756_20 | TCTGATACATCAAAAACACCCAGCA[C/T]GTTGGTTATTATTCTTTTTATTGAT |
| rs1432982 | C___8950584_10 | AAACAAATATAAAGACAATTTAAAA[A/G]TGCATGAGCATAGTAAAGCTTGACG |
Each designed TaqMan® SNP Genotyping Assay includes two allele-specific TaqMan® MGB probes containing distinct 5′ fluorescently-labeled single primers and minor groove binder (MGB) non-fluorescent quencher (NFQ) on the 3′ end, as well as 2 unlabeled PCR primers (forward and reverse) to detect specific SNP targets. One VIC® dye - MGB-labeled probe detects the Allele 1 sequence and one 6FAM™ dye - MGB-labeled probe detects the Allele 2 sequence. FAM (6-carboxyfluorescein)/VIC (4, 7, 2′-trichloro-7′-phenyl-6-carboxyfluorescein).
In silico analysis to study the functional effects of the SNPs
The Alibaba 2.1 software (42) was utilized to determine the functionality of the tested SPINK1 promoter variants and their pathological relevance; and if any of the tested SNPs disrupted or resulted in the appearance of putative transcription factors binding site.
Statistical analyses
The statistical analyses were performed using R (43) and SNPStats (44). Allele frequencies in controls were tested for Hardy-Weinberg equilibrium using chi square analysis and Fisher’s exact test, when appropriate. The association of disease status and clinical phenotypes with genotype and haplotype was analyzed by binary and multivariate logistic regression. For each SNP genotype, tests using five different genetic models: co-dominant, dominant, recessive, over-dominant and log-additive, were performed. Multiple logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (95% CI). Two-sided p-values of ≤0.05 were considered as statistically significant. ORs for associations between SNPs, Gleason score and PCa risk were adjusted for age and percentage of West African ancestry. ORs for associations between SNPs and BPH were adjusted for age. To test the combinatorial effects of the SNPs and haplotypes, SNPStats was utilized (44).
Results
Participants’ demographics
In this study, 307 PCa and 96 BPH patients, as well as 339 control subjects, were investigated in order to explore the association of polymorphic variations in the IL-6RE of SPINK1 gene with BPH and PCa risk. The characteristics of the human subjects are shown in Table II. A student’s t-test reveals significant differences in age between the PCa, BPH and control subjects (p<0.0001). The mean age of patients with PCa (67.84±10.02) was significantly older than that of controls (56.59±10.33) and BPH patients (63±9.6). The mean PSA levels were significantly higher in PCa (74.51 ng/ml) and BPH (8.74 ng/ml) patients than controls (1.20 ng/ml) (p<0.0005). The mean Gleason Score with ranges 2–6 and 7–10 in PCa cases was 4.44±0.07 and 8.80±0.12, respectively.
Table II.
Characteristics of study subjects.
| Controls | Cases | * BPH | **p-Value | |
|---|---|---|---|---|
| Total number of participants | 339 | 307 | 96 | |
| Mean age (±SD) | 56.59±10.33 | 67.84±10.02 | 63±9.64 | 0.0001 |
| Mean PSA level | 1.20±0.87 | 74.51±383.52 | 8.74±20.6 | 0.0005 |
| Mean Gleason Score | ||||
| <7 | - | 5.44±0.07 | - | |
| >7 | - | 8.80±0.12 | - | |
|
| ||||
| History of WAA (%) | 67.33% | 70.77% | 68.2% | 0.02 |
Plus-minus values are means±SD. Prostate-specific antigen (PSA) levels (ng/ml) were obtained at the time of diagnosis for cases and at the time of study enrollment for the controls.
Benign prostatic hyperplasia (BPH).
p-value (t-Test, 2-tailed). WAA, West African ancestry.
The percentage of subjects with West African ancestry was significantly different between PCa (70.77%) subjects and controls (67.33%) (p=0.02) and not in BPH subjects (68.2%).
Associations between SNPs, Gleason score and BPH risk
The studied rs10035432, rs11748222 and rs1432982 SNPs were in the Hardy-Weinberg equilibrium (p>0.05). The allele and genotype frequencies in PCa patients were similar to the controls. The majority of PCa patients were homozygous for the major alleles of rs10035432, rs11748222 and rs1432982 SNPs (Table III). For SNP rs10035432 the G/G genotype was 58% and 58.5% among cases and controls, respectively. The T/T genotype of rs11748222 was 80.8% among cases and 79% among controls. For SNP rs1432982, the A/A genotype was 61.5% among cases compared to 58.5% among controls.
Table III.
Association between SPINK1 SNPs and prostate cancer risk.
| Model | Rs10035432
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | * Cases | ** OR (95% CI) | ***p-Value | |
| Co-dominant | G/G | 65 (58%) | 55 (58.5%) | 1 | 0.44 |
| A/G | 44 (39.3%) | 33 (35.1%) | 0.87 (0.44–1.71) | ||
| A/A | 3 (2.7%) | 6 (6.4%) | 0.35 (0.07–1.81) | ||
| Dominant | G/G | 65 (58%) | 55 (58.5%) | 1 | 0.50 |
| A/G-A/A | 47 (42%) | 39 (41.5%) | 0.80 (0.41–1.54) | ||
| Recessive | G/G-A/G | 109 (97.3%) | 88 (93.6%) | 1 | 0.22 |
| A/A | 3 (2.7%) | 6 (6.4%) | 0.38 (0.08–1.87) | ||
| Over-dominant | G/G-A/A | 68 (60.7%) | 61 (64.9%) | 1 | 0.87 |
| A/G | 44 (39.3%) | 33 (35.1%) | 0.95 (0.49–1.83) | ||
| Log-additive | --- | --- | --- | 0.75 (0.43–1.31) | 0.31 |
| Model | rs11748222
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | * Cases | ** OR (95% CI) | ***p-Value | |
| Co-dominant | T/T | 83 (79%) | 63 (80.8%) | 1 | 0.97 |
| C/T | 15 (14.3%) | 10 (12.8%) | 1.10 (0.40–3.01) | ||
| C/C | 7 (6.7%) | 5 (6.4%) | 0.90 (0.23–3.49) | ||
| Dominant | T/T | 83 (79%) | 63 (80.8%) | 1 | 0.96 |
| C/T-C/C | 22 (20.9%) | 15 (19.2%) | 1.02 (0.44–2.40) | ||
| Recessive | T/T-C/T | 98 (93.3%) | 73 (93.6%) | 1 | 0.86 |
| C/C | 7 (6.7%) | 5 (6.4%) | 0.89 (0.23–3.42) | ||
| Over-dominant | T/T-C/C | 90 (85.7%) | 68 (87.2%) | 1 | 0.85 |
| C/T | 15 (14.3%) | 10 (12.8%) | 1.10 (0.40–3.02) | ||
| Log-additive | --- | --- | --- | 0.99 (0.55–1.79) | 0.97 |
| Model | rs1432982
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | * Cases | ** OR (95% CI) | ***p-Value | |
| Co-dominant | A/A | 69 (58.5%) | 59 (61.5%) | 1 | 0.48 |
| A/G | 42 (35.6%) | 27 (28.1%) | 1.16 (0.59–2.30) | ||
| G/G | 7 (5.9%) | 10 (10.4%) | 0.53 (0.15–1.82) | ||
| Dominant | A/A | 69 (58.5%) | 59 (61.5%) | 1 | 0.99 |
| A/G-G/G | 49 (41.5%) | 37 (38.5%) | 1.00 (0.53–1.90) | ||
| Recessive | A/A-A/G | 111 (94.1%) | 86 (89.6%) | 1 | 0.26 |
| G/G | 7 (5.9%) | 10 (10.4%) | 0.50 (0.15–1.69) | ||
| Over-dominant | A/A-G/G | 76 (64.4%) | 69 (71.9%) | 1 | 0.52 |
| A/G | 42 (35.6%) | 27 (28.1%) | 1.25 (0.64–2.43) | ||
| Log-additive | --- | --- | --- | 0.89 (0.54–1.46) | 0.65 |
Cases, included only PCa cases.
OR (95% CI): odd ratio and 95% confidential interval of SNP response to PCa when adjusted by age and %West African ancestry.
p-Value, adjusted p-value by age.
None of the studied SNPs were found to be associated with PCa risk even after adjusting for age and %WAA using the five genetic models (p>0.05) (Table III). The association between the three SNPs and Gleason score was also determined (Table IV). We found a significant association between the minor allele G of rs1432982 with low Gleason score and PCa risk after adjusting for age under the recessive model (OR=2.96, 95% CI=0.91–9.61; p=0.05). In silico analysis of rs1432982 revealed that when the G allele is present, there is a CCAAT-enhancer-binding protein alpha (c-EBPα) site; however when the A allele is present the c-EBPα site is absent. We found an association between the A allele of rs10035432 and BPH risk before adjusting for age (data not shown) and even after adjusting for age and %WAA under the co-dominant (OR=0.52, 95% CI=0.29–0.93; p=0.02), dominant (OR=0.48, 95% CI=0.28–0.84; p=0.01) and the log-additive (OR=0.55, 95% CI=0.35–0.84; p=0.007) models (Table V). In silico analysis of rs10035432 showed that when the A allele is present there is a TATA Binding Protein (TBP) site; however, when the G allele is present this site is lost.
Table IV.
SPINK1 SNPs association with Gleason score (GS) in prostate cancer cases.
| Model | rs10035432
|
||||
|---|---|---|---|---|---|
| Genotype | * Low GS | * High GS | ** OR (95% CI) | ***p-Value | |
| Co-dominant | G/G | 50 (65.8%) | 42 (60.9%) | 1 | 0.48 |
| A/G | 18 (23.7%) | 22 (31.9%) | 0.69 (0.33–1.45) | ||
| A/A | 8 (10.5%) | 5 (7.2%) | 1.34 (0.41–4.42) | ||
| Dominant | G/G | 50 (65.8%) | 42 (60.9%) | 1 | 0.54 |
| A/G-A/A | 26 (34.2%) | 27 (39.1%) | 0.81 (0.41–1.59) | ||
| Recessive | G/G-A/G | 68 (89.5%) | 64 (92.8%) | 1 | 0.49 |
| A/A | 8 (10.5%) | 5 (7.2%) | 1.51 (0.47–4.84) | ||
| Over-dominant | G/G-A/A | 58 (76.3%) | 47 (68.1%) | 1 | 0.27 |
| A/G | 18 (23.7%) | 22 (31.9%) | 0.66 (0.32–1.38) | ||
| Log-additive | --- | --- | --- | 0.96 (0.58–1.58) | 0.88 |
| Model | rs11748222
|
||||
|---|---|---|---|---|---|
| Genotype | * Low GS | * High GS | ** OR (95% CI) | ***p-Value | |
| Co-dominant | T/T | 67 (80.7%) | 66 (84.6%) | 1 | 0.54 |
| C/T | 11 (13.2%) | 10 (12.8%) | 1.08 (0.43–2.72) | ||
| C/C | 5 (6%) | 2 (2.6%) | 2.46 (0.46–13.14) | ||
| Dominant | T/T | 67 (80.7%) | 66 (84.6%) | 1 | 0.51 |
| C/T-C/C | 16 (19.3%) | 12 (15.4%) | 1.31 (0.58–2.99) | ||
| Recessive | T/T-C/T | 78 (94%) | 76 (97.4%) | 1 | 0.27 |
| C/C | 5 (6%) | 2 (2.6%) | 2.44 (0.46–12.94) | ||
| Over-dominant | T/T-C/C | 72 (86.8%) | 68 (87.2%) | 1 | 0.94 |
| C/T | 11 (13.2%) | 10 (12.8%) | 1.04 (0.41–2.60) | ||
| Log-additive | --- | --- | --- | 1.34 (0.72–2.50) | 0.36 |
| Model | rs1432982
|
||||
|---|---|---|---|---|---|
| Genotype | * Low GS | * High GS | ** OR (95% CI) | ***p-Value | |
| Co-dominant | A/A | 50 (60.2%) | 50 (67.6%) | 1 | 0.16 |
| A/G | 21 (25.3%) | 20 (27%) | 1.05 (0.51–2.17) | ||
| G/G | 12 (14.5%) | 4 (5.4%) | 3.00 (0.91–9.94) | ||
| Dominant | A/A | 50 (60.2%) | 50 (67.6%) | 1 | 0.34 |
| A/G-G/G | 33 (39.8%) | 24 (32.4%) | 1.37 (0.71–2.65) | ||
| Recessive | A/A-A/G | 71 (85.5%) | 70 (94.6%) | 1 | 0.05 |
| G/G | 12 (14.5%) | 4 (5.4%) | 2.96 (0.91–9.61) | ||
| Over-dominant | A/A-G/G | 62 (74.7%) | 54 (73%) | 1 | 0.81 |
| A/G | 21 (25.3%) | 20 (27%) | 0.91 (0.45–1.87) | ||
| Log-additive | --- | --- | --- | 1.45 (0.90–2.34) | 0.13 |
Low and high GS were defined as less advanced “if a case’s GS is <7” and more advanced “if their GS is >7”.
OR (95% CI): odd ratio and 95% confidential interval of SNP response to PCa when adjusted by age.
p-Value, adjusted p-value by age.
Table V.
Associations between SPINK1 SNPs and benign prostatic hyperplasia (BPH) risk.
| Model | Rs10035432
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | BPH | * OR (95% CI) | **p-Value | |
| Co-dominant | G/G | 165 (58.1%) | 30 (44.1%) | 1 | 0.02 |
| A/G | 106 (37.3%) | 31 (45.6%) | 0.52 (0.29–0.93) | ||
| A/A | 13 (4.6%) | 7 (10.3%) | 0.32 (0.11–0.90) | ||
| Dominant | G/G | 165 (58.1%) | 30 (44.1%) | 1 | 0.01 |
| A/G-A/A | 119 (41.9%) | 38 (55.9%) | 0.48 (0.28–0.84) | ||
| Recessive | G/G-A/G | 271 (95.4%) | 61 (89.7%) | 1 | 0.11 |
| A/A | 13 (4.6%) | 7 (10.3%) | 0.43 (0.16–1.16) | ||
| Over-dominant | G/G-A/A | 178 (62.7%) | 37 (54.4%) | 1 | 0.07 |
| A/G | 106 (37.3%) | 31 (45.6%) | 0.60 (0.34–1.05) | ||
| Log-additive | --- | --- | --- | 0.55 (0.35–0.84) | 0.007 |
| Model | rs11748222
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | BPH | * OR (95% CI) | **p-Value | |
| Co-dominant | T/T | 221 (81%) | 72 (84.7%) | 1 | 0.28 |
| C/T | 42 (15.4%) | 8 (9.4%) | 1.60 (0.71–3.63) | ||
| C/C | 10 (3.7%) | 5 (5.9%) | 0.56 (0.18–1.76) | ||
| Dominant | T/T | 221 (81%) | 72 (84.7%) | 1 | 0.6 |
| C/T-C/C | 52 (19.1%) | 13 (15.3%) | 1.20 (0.61–2.37) | ||
| Recessive | T/T-C/T | 263 (96.3%) | 80 (94.1%) | 1 | 0.28 |
| C/C | 10 (3.7%) | 5 (5.9%) | 0.52 (0.17–1.65) | ||
| Over-dominant | T/T-C/C | 231 (84.6%) | 77 (90.6%) | 1 | 0.21 |
| C/T | 42 (15.4%) | 8 (9.4%) | 1.65 (0.73–3.74) | ||
| Log-additive | --- | --- | --- | 0.99 (0.60–1.64) | 0.97 |
| Model | rs1432982
|
||||
|---|---|---|---|---|---|
| Genotype | Controls | BPH | * OR (95% CI) | **p-Value | |
| Co-dominant | A/A | 175 (60.5%) | 58 (69.9%) | 1 | 0.44 |
| A/G | 98 (33.9%) | 22 (26.5%) | 1.38 (0.78–2.42) | ||
| G/G | 16 (5.5%) | 3 (3.6%) | 1.62 (0.45–5.89) | ||
| Dominant | A/A | 175 (60.5%) | 58 (69.9%) | 1 | 0.21 |
| A/G-G/G | 114 (39.5%) | 25 (30.1%) | 1.41 (0.82–2.41) | ||
| Recessive | A/A-A/G | 273 (94.5%) | 80 (96.4%) | 1 | 0.55 |
| G/G | 16 (5.5%) | 3 (3.6%) | 1.46 (0.41–5.27) | ||
| Over-dominant | A/A-G/G | 191 (66.1%) | 61 (73.5%) | 1 | 0.31 |
| A/G | 98 (33.9%) | 22 (26.5%) | 1.33 (0.76–2.33) | ||
| Log-additive | --- | --- | --- | 1.33 (0.85–2.10) | 0.21 |
SNPStats was used to construct the haplotypes.
OR (95% CI): odd ratio and 95% confidential interval of SNP response to BPH when adjusted by age and WAA.
p-Value, adjusted p-value by age and WAA.
A haplotype analysis of the three studied SNPs revealed only one associated haplotype with low Gleason score and PCa risk (OR=2.01, 95% CI=1.00–4.04; p=0.05). This haplotype is composed of the rs10035432 G allele, rs11748222 T allele and rs1432982 G allele (Table VI). Whereas, an inverse association was found between haplotype 2 (rs10035432 A allele, rs11748222 T allele and rs1432982 A allele) and BPH risk (OR=0.63, 95% CI=0.38–1.06; p=0.08) (Table VII).
Table VI.
Haplotype frequencies and their association with high and low Gleason score (GS).
| rs10035432 | rs11748222 | rs1432982 | Total Frequencies | * Low GS | * High GS | ** OR (95% CI) | p-Value | |
|---|---|---|---|---|---|---|---|---|
| 1 | G | T | A | 0.5038 | 0.4392 | 0.582 | 1 | --- |
| 2 | A | T | A | 0.2004 | 0.2218 | 0.1712 | 1.44 (0.80–2.60) | 0.22 |
| 3 | G | T | G | 0.168 | 0.2044 | 0.1211 | 2.01 (1.00–4.04) | 0.05 |
| 4 | G | C | A | 0.0644 | 0.0682 | 0.0604 | 1.56 (0.66–3.70) | 0.32 |
| Rare | * | * | * | 0.0634 | NA | NA | 1.11 (0.42–2.95) | 0.83 |
Low and high GS were defined as less advanced “if a case’s GS is <7” and more advanced “if their GS is >7”.
OR (95% CI): Odd ratio and 95% confidential interval. Global haplotype association p-value: 0.02.
Table VII.
Haplotype frequencies and their association with benign prostatic hyperplasia (BPH).
| rs10035432 | rs11748222 | rs1432982 | Total Frequencies | Controls | BPH | * OR (95% CI) | p-Value | |
|---|---|---|---|---|---|---|---|---|
| 1 | G | T | A | 0.5076 | 0.4896 | 0.5166 | 1 | --- |
| 2 | A | T | A | 0.2282 | 0.2585 | 0.2175 | 0.63 (0.38 – 1.06) | 0.08 |
| 3 | G | T | G | 0.1426 | 0.0947 | 0.1555 | 1.49 (0.71 – 3.14) | 0.29 |
| 4 | G | C | G | 0.0599 | 0.0213 | 0.071 | 2.11 (0.63 – 7.00) | 0.22 |
| 5 | G | C | A | 0.0381 | 0.0697 | 0.0279 | 0.46 (0.18 – 1.20) | 0.12 |
| 6 | A | C | A | 0.0125 | 0.0526 | 0 | 0.64 (0.07 – 6.02) | 0.70 |
| 7 | A | T | G | 0.0112 | 0.0136 | 0.0115 | 0.00 (0.00-NA) | 1.00 |
| Rare | * | * | * | 0 | 0 | 0 | 1 | --- |
SNPStats was used to construct the haplotypes.
OR (95% CI): Odd ratio and 95% confidential interval. NA, Not applicable. Global haplotype association p-value: 0.0043.
Discussion
In the present study, we investigated whether SNPs in SPINK1 are associated with an increased risk of BPH and PCa. We identified and genotyped a sub-set of population frequency discriminant alleles in the IL-6 response element (IL-6 RE) of the SPINK1 gene (rs10035432, rs11748222 and rs1432982) in PCa cases, controls and patients diagnosed with BPH, who were of African descent. None of the investigated SPINK1 SNPs was associated with PCa risk. However, we found a significant association between the minor allele G of rs1432982, low GS under the recessive model. We did observe an inverse association between rs10035432 and BPH under the log additive model.
Though SPINK1 is being targeted as a biomarker for PCa, there exists limited information regarding the association between the genetic variants present in this gene and PCa risk. To our knowledge we are the first to investigate rs10035432, rs11748222 and rs1432982 association with PCa and BPH risk. Our in silico functional analysis showed that TBP and c-EBPα transcriptional factors are present or absent depending on the given allele at their associated loci. We propose that these transcription factor changes can impact SPINK1 gene expression, which may ultimately influence SPINK1’s function in the etiology of BPH and PCa. In speculation, if TBP and c-EBPα were to bind to the promoter region, as a putative effect, SPINK1 will be expressed in order to regulate excessive amount of trypsin that may have secreted into the extracellular fluid. In theory, this mechanism of action could protect the prostate tissue against inflammation and, subsequently, prevent PCa development or an aggressive form of PCa. However, if TBP and c-EBPα lose their affinity to bind to the promoter region of SPINK1 due to sequence changes in IL-6 RE, transcription of SPINK1 may not be initiated and chronic inflammation will occur. This may lead to PCa development and a more aggressive form of PCa. Further functional studies are required to validate the impact of TBP and c-EBPα on the regulation of SPINK1 in BPH and PCa cells.
In this study, we found a possible positive association between rs1432982 and lower Gleason score in men of African descent. Previous studies have found that SPINK1 is associated with high Gleason score (18, 45), while other studies reported an association with low Gleason score (24) or a null association (19, 22, 46, 47). Only one previous study has investigated SPINK1 and Gleason score in an African American cohort (24). In this previous study, AA men who were positive for SPINK1 protein over-expression were more likely to have a lower Gleason score compared to men negative for SPINK1 expression. We also found an inverse association between rs10035432 and BPH. We found that SPINK1 was over-expressed in the BPH cell line (data not shown), which is consistent with previous findings (17). However, a previous study reported that SPINK1 expression was absent in benign prostate tissue (48). There is limited amount of data available about the role SPINK1 plays in BPH etiology. Thus, further investigation is needed to confirm SPINK1 expression in BPH cases. The regulation of SPINK1 in prostate cancer cells is unclear.
A major strength of this study provides insight on genetic variations in the promoter region of SPINK1. However, we could have missed some associations, because we did not include SNPs across SPINK1, especially polymorphisms in the coding region. Also, we are the first to genotype rs11748222 and rs1432982 in controls, PCa and BPH who were African American. These SNPs were not reported as a part of the HapMap project. Given the modest size of our sample size and the limitation of other ethnic groups not included in this study, our findings warrant further validation in a lager cohort, which will include men from different ethnic backgrounds.
Conclusion
Our findings showed no association between SPINK1 promoter variants and PCa risk. However, SPINK1 promoter variants are likely to be associated with the risk of BPH.
Acknowledgments
This work was supported with funded grants from the US Army Medical Research and Materiel Command (USAMRMC) DAMD17-03-1-0069, W81XWH-11-1-0560 and, in part, by NIH Grants S06 GM08016 and G12 RR003048.
Footnotes
Conflicts of Interest
The Authors have no personal or financial conflicts of interest and have not entered into any agreement that could interfere with their access to the data on the research or upon their ability to analyze the data independently, to prepare manuscripts and to publish them.
References
- 1.Ruan L, Zhu J, Pan C, Hua X, Yuan D, Li Z, Zhong W. Association between single nucleotide polymorphism of vitamin D receptor gene FokI polymorphism and clinical progress of benign prostatic hyperplasia. Scientific World Journal. 2015;2015:Article ID: 235895. doi: 10.1155/2015/235895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kumar R, Malla P, Kumar M. Advances in the design and discovery of drugs for the treatment of prostatic hyperplasia. Expert Opin Drug Discov. 2013;8:1013–1027. doi: 10.1517/17460441.2013.797960. [DOI] [PubMed] [Google Scholar]
- 3.Choubey VK, Sankhwar SN, Carlus SJ, Singh AN, Dalela D, Thangaraj K, Rajender S. SRD5A2 Gene Polymorphisms and the Risk of Benign Prostatic Hyperplasia but not Prostate Cancer. Asian Pac J Cancer Prev. 2015;16:1033–1036. doi: 10.7314/apjcp.2015.16.3.1033. [DOI] [PubMed] [Google Scholar]
- 4.Djavan B, Margreiter M, Dianat SS. An algorithm for medical management in male lower urinary tract symptoms. Curr Opin Urol. 2011;21:5–12. doi: 10.1097/MOU.0b013e32834100ef. [DOI] [PubMed] [Google Scholar]
- 5.Ahmad M, Suhail N, Mansoor T, Banu N, Ahmad S. Evaluation of oxidative stress and DNA damage in benign prostatic hyperplasia patients and comparison with controls. Indian J Clin Biochem. 2012;27:385–388. doi: 10.1007/s12291-012-0229-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Minciullo PL, Inferrera A, Navarra M, Calapai G, Magno C, Gangemi S. Oxidative stress in benign prostatic hyperplasia: a systematic review. Urol Int. 2015;94:249–254. doi: 10.1159/000366210. [DOI] [PubMed] [Google Scholar]
- 7.Roehrborn CG. Benign prostatic hyperplasia: an overview. Rev Urol. 2005;7(Suppl 9):S3–S14. [PMC free article] [PubMed] [Google Scholar]
- 8.Sanda MG, Beaty TH, Stutzman RE, Childs B, Walsh PC. Genetic susceptibility of benign prostatic hyperplasia. J Urol. 1994;152:115–119. doi: 10.1016/s0022-5347(17)32831-8. [DOI] [PubMed] [Google Scholar]
- 9.Madigan MP, Gao Y-T, Deng J, Pfeiffer RM, Chang B-L, Zheng S, Meyers DA, Stanczyk FZ, Xu J, Hsing AW. CYP17 polymorphisms in relation to risks of prostate cancer and benign prostatic hyperplasia: a population-based study in China. Int J Cancer. 2003;107:271–275. doi: 10.1002/ijc.11378. [DOI] [PubMed] [Google Scholar]
- 10.Novelli G, Margiotti K, Sangiuolo F, Reichardt JK. Pharmacogenetics of human androgens and prostatic diseases. Pharmacogenomics. 2001;2:65–72. doi: 10.1517/14622416.2.1.65. [DOI] [PubMed] [Google Scholar]
- 11.Ban JY, Yoo KH. Promoter Polymorphism (rs12770170, -184C/T) of Microseminoprotein, Beta as a Risk Factor for Benign Prostatic Hyperplasia in Korean Population. Int Neurourol J. 2014;18:63–67. doi: 10.5213/inj.2014.18.2.63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.U.S. Cancer Statistics Working Group. United States Cancer Statistics: 1999–2011 Incidence and Mortality Web-based Report. Atlanta: US Dep Heal Hum Serv Centers Dis Control Prev Natl Cancer Inst; 2014. [Google Scholar]
- 13.Chokkalingam AP, Nyrén O, Johansson J-E, Gridley G, McLaughlin JK, Adami H-O, Hsing AW. Prostate carcinoma risk subsequent to diagnosis of benign prostatic hyperplasia: a population-based cohort study in Sweden. Cancer. 2003;98:1727–1734. doi: 10.1002/cncr.11710. [DOI] [PubMed] [Google Scholar]
- 14.Seok H, Yoo KH, Kim YO, Chung J-H. Association of a Missense ALDH2 Single Nucleotide Polymorphism (Glu504Lys) With Benign Prostate Hyperplasia in a Korean Population. Int Neurourol J. 2013;17:168–173. doi: 10.5213/inj.2013.17.4.168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.El Ezzi AA, Zaidan WR, El-Saidi MA, Al-Ahmadieh N, Mortenson JB, Kuddus RH. Association of benign prostate hyperplasia with polymorphisms in VDR, CYP17, and SRD5A2 genes among Lebanese men. Asian Pac J Cancer Prev. 2014;15:1255–1262. doi: 10.7314/apjcp.2014.15.3.1255. [DOI] [PubMed] [Google Scholar]
- 16.Yoo KH, Kim SK, Chung J-H, Chang S-G. Nitric oxide synthase 2 gene polymorphisms are associated with prostatic volume in Korean men with benign prostatic hyperplasia. Asian J Androl. 2010;12:690–696. doi: 10.1038/aja.2010.37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Väänänen R-M, Lilja H, Kauko L, Helo P, Kekki H, Cronin AM, Vickers AJ, Nurmi M, Alanen K, Bjartell A, Pettersson K. Cancer-associated changes in the expression of TMPRSS2-ERG, PCA3, and SPINK1 in histologically benign tissue from cancerous vs noncancerous prostatectomy specimens. Urology. 2014;83:511.e1–7. doi: 10.1016/j.urology.2013.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Paju A, Hotakainen K, Cao Y, Laurila T, Gadaleanu V, Hemminki A, Stenman U-H, Bjartell A. Increased expression of tumor-associated trypsin inhibitor, TATI, in prostate cancer and in androgen-independent 22Rv1 cells. Eur Urol. 2007;52:1670–1681. doi: 10.1016/j.eururo.2007.01.096. [DOI] [PubMed] [Google Scholar]
- 19.Lippolis G, Edsjö A, Stenman U-H, Bjartell A. A high-density tissue microarray from patients with clinically localized prostate cancer reveals ERG and TATI exclusivity in tumor cells. Prostate Cancer Prostatic Dis. 2013;16:145–150. doi: 10.1038/pcan.2013.7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tomlins SA, Rhodes DR, Yu J, Varambally S, Mehra R, Perner S, Demichelis F, Helgeson BE, Laxman B, Morris DS, Cao Q, Cao X, Andrén O, Fall K, Johnson L, Wei JT, Shah RB, Al-Ahmadie H, Eastham JA, Eggener SE, Fine SW, Hotakainen K, Stenman U-H, Tsodikov A, Gerald WL, Lilja H, Reuter VE, Kantoff PW, Scardino PT, Rubin MA, Bjartell AS, Chinnaiyan AM. The role of SPINK1 in ETS rearrangement-negative prostate cancers. Cancer Cell. 2008;13:519–528. doi: 10.1016/j.ccr.2008.04.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Han B, Mehra R, Suleman K, Tomlins SA, Wang L, Singhal N, Linetzky KA, Palanisamy N, Zhou M, Chinnaiyan AM, Shah RB. Characterization of ETS gene aberrations in select histologic variants of prostate carcinoma. Mod Pathol. 2009;22:1176–1185. doi: 10.1038/modpathol.2009.79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Palanisamy N, Tsodikov A, Yan W, Suleman K, Rittmaster R, Lucia SM, Chinnaiyan AM, Fleshner N, Kunju LP. Molecular profiling of ETS gene rearrangements in patients with prostate cancer registered in REDEEM clinical trial. Urol Oncol. 2015;33:108.e5–108.e13. doi: 10.1016/j.urolonc.2014.07.015. [DOI] [PubMed] [Google Scholar]
- 23.Laxman B, Morris DS, Yu J, Siddiqui J, Cao J, Mehra R, Lonigro RJ, Tsodikov A, Wei JT, Tomlins SA, Chinnaiyan AM. A first-generation multiplex biomarker analysis of urine for the early detection of prostate cancer. Cancer Res. 2008;68:645–649. doi: 10.1158/0008-5472.CAN-07-3224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Khani F, Mosquera JM, Park K, Blattner M, O’Reilly C, MacDonald TY, Chen Z, Srivastava A, Tewari AK, Barbieri CE, Rubin MA, Robinson BD. Evidence for Molecular Differences in Prostate Cancer between African American and Caucasian Men. Clin Cancer Res. 2014;20:4925–4934. doi: 10.1158/1078-0432.CCR-13-2265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yasuda T, Ogawa M, Murata A, Ohmachi Y, Mori T, Matsubara K. Identification of the IL-6-responsive element in an acute-phase-responsive human pancreatic secretory trypsin inhibitor-encoding gene. Gene. 1993;131:275–280. doi: 10.1016/0378-1119(93)90306-n. [DOI] [PubMed] [Google Scholar]
- 26.Mills JS, Needham M, Parker MG. A secretory protease inhibitor requires androgens for its expression in male sex accessory tissues but is expressed constitutively in pancreas. EMBO J. 1987;6:3711–3717. doi: 10.1002/j.1460-2075.1987.tb02705.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Stenman U-H. Tumor-associated trypsin inhibitor. Clin Chem. 2002;48:1206–1209. [PubMed] [Google Scholar]
- 28.Paju A, Stenman U-H. Biochemistry and clinical role of trypsinogens and pancreatic secretory trypsin inhibitor. Crit Rev Clin Lab Sci. 2006;43:103–142. doi: 10.1080/10408360500523852. [DOI] [PubMed] [Google Scholar]
- 29.Playford RJ, Batten JJ, Freeman TC, Beardshall K, Vesey DA, Fenn GC, Baron JH, Calam J. Gastric output of pancreatic secretory trypsin inhibitor is increased by misoprostol. Gut. 1991;32:1396–1400. doi: 10.1136/gut.32.11.1396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ma L, Yu H, Ni Z, Hu S, Ma W, Chu C, Liu Q, Zhang Y. Spink13, an epididymis-specific gene of the Kazal-type serine protease inhibitor (SPINK) family, is essential for the acrosomal integrity and male fertility. J Biol Chem. 2013;288:10154–10165. doi: 10.1074/jbc.M112.445866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lu F, Lamontagne J, Sun A, Pinkerton M, Block T, Lu X. Role of the inflammatory protein serine protease inhibitor Kazal in preventing cytolytic granule granzyme A-mediated apoptosis. Immunology. 2011;134:398–408. doi: 10.1111/j.1365-2567.2011.03498.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Soon WW, Miller LD, Black MA, Dalmasso C, Bin Chan X, Pang B, Ong CW, Salto-Tellez M, Desai KV, Liu ET. Combined genomic and phenotype screening reveals secretory factor SPINK1 as an invasion and survival factor associated with patient prognosis in breast cancer. EMBO Mol Med. 2011;3:451–464. doi: 10.1002/emmm.201100150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ozaki N, Ohmuraya M, Hirota M, Ida S, Wang J, Takamori H, Higashiyama S, Baba H, Yamamura K-I. Serine protease inhibitor Kazal type 1 promotes proliferation of pancreatic cancer cells through the epidermal growth factor receptor. Mol Cancer Res. 2009;7:1572–1581. doi: 10.1158/1541-7786.MCR-08-0567. [DOI] [PubMed] [Google Scholar]
- 34.Lee Y-C, Pan H-W, Peng S-Y, Lai P-L, Kuo W-S, Ou Y-H, Hsu H-C. Overexpression of tumour-associated trypsin inhibitor (TATI) enhances tumour growth and is associated with portal vein invasion, early recurrence and a stage-independent prognostic factor of hepatocellular carcinoma. Eur J Cancer. 2007;43:736–744. doi: 10.1016/j.ejca.2006.11.020. [DOI] [PubMed] [Google Scholar]
- 35.Wang G-P, Xu C-S. Pancreatic secretory trypsin inhibitor: More than a trypsin inhibitor. World J Gastrointest Pathophysiol. 2010;1:85–90. doi: 10.4291/wjgp.v1.i2.85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Santer FR, Malinowska K, Culig Z, Cavarretta IT. Interleukin-6 trans-signalling differentially regulates proliferation, migration, adhesion and maspin expression in human prostate cancer cells. Endocr Relat Cancer. 2010;17:241–253. doi: 10.1677/ERC-09-0200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Shariat SF, Chromecki TF, Hoefer J, Barbieri CE, Scherr DS, Karakiewicz PI, Roehrborn CG, Montorsi F, Culig Z, Cavarretta IT. Soluble gp130 regulates prostate cancer invasion and progression in an interleukin-6 dependent and independent manner. J Urol. 2011;186:2107–2114. doi: 10.1016/j.juro.2011.06.048. [DOI] [PubMed] [Google Scholar]
- 38.Rojas A, Liu G, Coleman I, Nelson PS, Zhang M, Dash R, Fisher PB, Plymate SR, Wu JD. IL-6 promotes prostate tumorigenesis and progression through autocrine cross-activation of IGF-IR. Oncogene. 2011;30:2345–2355. doi: 10.1038/onc.2010.605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.dbSNP. Available from: http://www.ncbi.nlm.nih.gov/snp.
- 40.HapMap. Available from: http://hapmap.ncbi.nlm.nih.gov/
- 41.Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. doi: 10.1093/bioinformatics/bth457. [DOI] [PubMed] [Google Scholar]
- 42.AliBaBa 2.1. Available from: http://www.gene-regulation.com/pub/programs.html#alibaba2.
- 43.González JR, Armengol L, Solé X, Guinó E, Mercader JM, Estivill X, Moreno V. SNPassoc: an R package to perform whole genome association studies. Bioinformatics. 2007;23:644–655. doi: 10.1093/bioinformatics/btm025. [DOI] [PubMed] [Google Scholar]
- 44.Solé X, Guinó E, Valls J, Iniesta R, Moreno V. SNPStats: a web tool for the analysis of association studies. Bioinformatics. 2006;22:1928–1929. doi: 10.1093/bioinformatics/btl268. [DOI] [PubMed] [Google Scholar]
- 45.Leinonen KA, Saramäki OR, Furusato B, Kimura T, Takahashi H, Egawa S, Suzuki H, Keiger K, Ho Hahm S, Isaacs WB, Tolonen TT, Stenman U-H, Tammela TLJ, Nykter M, Bova GS, Visakorpi T. Loss of PTEN is associated with aggressive behavior in ERG-positive prostate cancer. Cancer Epidemiol Biomarkers Prev. 2013;22:2333–2344. doi: 10.1158/1055-9965.EPI-13-0333-T. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Leinonen KA, Tolonen TT, Bracken H, Stenman U-H, Tammela TLJ, Saramäki OR, Visakorpi T. Association of SPINK1 expression and TMPRSS2:ERG fusion with prognosis in endocrine-treated prostate cancer. Clin Cancer Res. 2010;16:2845–2851. doi: 10.1158/1078-0432.CCR-09-2505. [DOI] [PubMed] [Google Scholar]
- 47.Flavin RJ, Pettersson A, Hendrickson WK, Fiorentino M, Finn SP, Kunz L, Judson G, Lis RT, Bailey D, Fiore C, Nuttall EJ, Martin NE, Stack EC, Penney KL, Rider JR, Sinnott JA, Sweeney CS, Sesso HD, Fall K, Giovannucci EL, Kantoff PW, Stampfer MJ, Loda M, Mucci LA. SPINK1 Protein Expression and Prostate Cancer Progression. Clin Cancer Res. 2014;20:4904–4911. doi: 10.1158/1078-0432.CCR-13-1341. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Grupp K, Diebel F, Sirma H, Simon R, Breitmeyer K, Steurer S, Hube-Magg C, Prien K, Pham T, Weigand P, Michl U, Heinzer H, Kluth M, Minner S, Tsourlakis MC, Izbicki JR, Sauter G, Schlomm T, Wilczak W. SPINK1 expression is tightly linked to 6q15- and 5q21-deleted ERG-fusion negative prostate cancers but unrelated to PSA recurrence. Prostate. 2013;73:1690–1698. doi: 10.1002/pros.22707. [DOI] [PubMed] [Google Scholar]