Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Clinical Cancer Research Imaging, Diagnosis, Prognosis Molecular Markers in Key Steroidogenic Pathways, Circulating Steroid Levels, and Prostate Cancer Progression Le vesque1,2, Shu-Pin Huang3,4, Etienne Audet-Walsh1, Louis Lacombe2, Bo-Ying Bao6,7, Yves Fradet2, Eric 1 1 lanie Rouleau , Chao-Yuan Huang8, Chia-Cheng Yu5, Patrick Caron1, and re , Me Isabelle Laverdie 1 Chantal Guillemette Abstract Purpose: Prostate cancer is a heterogeneous genetic disease, and molecular methods for predicting prognosis in patients with aggressive form of the disease are urgently needed to better personalize treatment approaches. The objective was to identify host genetic variations in candidate steroidogenic genes affecting hormone levels and prostate cancer progression. Experimental Design: The study examined two independent cohorts composed of 526 Caucasian men with organ-confined prostate cancer and 601 Taiwanese men on androgen–deprivation therapy. Caucasians were genotyped for 109 haplotype-tagging single-nucleotide polymorphisms (SNP) in CYP17A1, ESR1, CYP19A1, and HSD3B1, and their prognostic significance on disease progression was assessed using Kaplan–Meier survival curves and Cox regression models. Positive findings, including previously identified SRD5A1, SRD5A2, HSD17B2, HSD17B3, and HSD17B12 polymorphisms, were then explored in Taiwanese men (n ¼ 32 SNPs). The influence of positive markers on the circulating hormonal levels was then appraised in Caucasians using specific and sensitive mass spectrometry–based methods. Results: After adjusting for known risk factors, variants of CYP17A1 (rs6162), HSD17B2 (rs4243229 and rs7201637), and ESR1 (rs1062577) were associated with progressive disease in both cohorts. Indeed, the presence of these variations was significantly associated with progression in Caucasians (HR, 2.29–4.10; P ¼ 0.0014–2 107) and survival in Taiwanese patients [HR ¼ 3.74; 95% confidence interval (CI): 1.71–8.19, P ¼ 0.009]. Remarkably, the CYP17A1 rs6162 polymorphism was linked to plasma dehydroepiandrosterone-sulfate (DHEA-S) levels (P ¼ 0.03), HSD17B2 rs7201637 with levels of dihydrotestosterone (P ¼ 0.03), and ESR1 rs1062577 with levels of estrone-S and androsterone-glucuronide (P 0.05). Conclusion: This study identifies, in different ethnic groups and at different disease stages, CYP17A1, HSD17B2, and ESR1 as attractive prognostic molecular markers of prostate cancer progression. Clin Cancer Res; 1–11. 2012 AACR. Introduction Prostate cancer is clearly a major public health concern because it is the most common cancer in men and the ^ tel-Dieu Authors' Affiliations: 1Pharmacogenomics Laboratory and 2L'Ho bec, Centre Hospitalier Universitaire de Que bec (CHUQ) Research de Que bec, Canada; 3DepartCenter, Faculty of Medicine, Laval University, Que ment of Urology, Faculty of Medicine, College of Medicine, Kaohsiung Medical University; 4Department of Urology, Kaohsiung Medical University Hospital; 5Division of Urology, Department of Surgery, Kaohsiung Veterans General Hospital, Kaohsiung; 6Department of Pharmacy, China Medical University; 7Sex Hormone Research Center, China Medical University Hospital, Taichung; and 8Department of Urology, National Taiwan University Hospital, College of Medicine, National Taiwan University, Taipei, Taiwan Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). Corresponding Author: Chantal Guillemette, CHUQ Research Center, T3bec, Canada, G1V 4G2. Phone: 418-654-2296; 48, 2705 Boul. Laurier, Que Fax: 418-654-2298; E-mail: [email protected] doi: 10.1158/1078-0432.CCR-12-2812 2012 American Association for Cancer Research. second leading cause of cancer-related death in North America (1). It is well established that androgen hormones play a central role in prostate cancer development and progression. Indeed, androgen deprivation therapy (ADT) is the cornerstone therapy for recurrent and metastatic prostate cancer (2, 3). Furthermore, castration-resistant prostate cancer (CRPC), once thought to be hormone refractory, remains clearly driven by sex steroid hormones (4). Even in a low-circulating testosterone environment, cancer cells generate potent intracellular hormones to support cancer growth and proliferation (5). In addition, several studies sustain the role of sex steroid formation by prostate cancer cells that possess the enzymatic machinery to convert precursors into more potent hormones (5–7). This reinforces the need to block intracellular steroidogenesis of prostatic cells. In metastatic settings, this concept is further reinforced with the advent of new compounds such as abiraterone acetate and MDV-3100, which targets, respectively, the CYP17 androgens’ biosynthesis pathway and the androgen receptor, both effective for CRPC (8–12). www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF1 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 vesque et al. Le Translational Relevance There are significant limitations in our ability to predict the progression and metastatic behavior of a patient’s prostate cancer. Risk of disease progression differs greatly between individuals, and the variability in clinical outcome further emphasizes the need to find novel markers of progression. Sex steroid hormones are essential for the development and progression of prostate cancer. The purpose of this study was to assess the prognostic value of common genetic variants in key steroidogenic genes. Our findings support, in different ethnic groups and at different disease stages, the importance of CYP17A1, HSD17B2, and ESR1 as attractive prognostic molecular markers of prostate cancer progression. The use of understanding germline determinants of progression is that this could be used, along with known and established clinical factors, to prognosticate clinical course. Poor prognostic genetic factors could be used to select patients for future clinical trials and identify subgroups of patients that would most benefit from endocrine treatment. Therefore, our findings could lead to a more personalized approach to guide patient therapy based on molecular markers. However, despite the improvement in survival of some patients with advanced prostate cancer, there are no available clinical prognostic molecular biomarkers of progression and mortality after treatment modalities such as radical prostatectomy, radiotherapy, hormonal manipulation, and chemotherapy. In addition to tumor markers, common germline variations might determine tumor behavior or serve as informative biomarkers. In advanced disease state, studies have shown the impact of HSD17B4 genetic variations (13) and expression (14) relative to disease progression, and germline variations in SLCO2B1 and SLCO1B3 in response to palliative ADT therapy (15). However, there is still a lack of molecular markers better defining aggressive and lethal prostate cancer. Recently, germline polymorphisms in SRD5A and HSD17B genes as well as deletions of UGT2B genes were shown to be associated with biochemical recurrence (BCR) in Caucasian and Taiwanese prostate cancer patients after radical prostatectomy (16, 17). Such results are in agreement with a meaningful effect of inherited genetic variations in the androgenic pathway on prostate cancer recurrence and progression, regardless of race. Here, we sought to investigate the link between common germline variations in additional candidate steroidogenic genes on cancer progression and mortality, and to evaluate their relationships with circulating endogenous sex steroids. Patients and Methods Clinical data The study included 526 Caucasians with localized prostate cancer who underwent radical prostatectomy as OF2 Clin Cancer Res; 2012 initial therapy and 601 Taiwanese men on ADT, described previously (16, 18). Briefly, in the Caucasian cohort, PSA failure (n ¼ 130) after radical prostatectomy was defined as values of 0.3 mg/L or more (16; Table 1). In the Taiwanese ADT cohort (18), treatment modalities are indicated in Table 2. PSA nadir was defined as the lowest PSA value achieved during ADT treatment. The cause of death was obtained by matching patients’ personal identification number with the official cause of death registry provided by the Department of Health, Executive Yuan (Taiwan, Republic of China). Prostate cancer-specific mortality (PCSM) was defined as the interval from the initiation of ADT to death from prostate cancer. The ACM was defined as the period from the initiation of ADT to death from any cause. Overall, 145 deaths were identified, and 101 of them died from prostate cancer. Participants provided written informed consent for the analysis of their genome and the Institutional Review Board approved the research protocol. Genetic analysis For genotyping, PCR amplifications were conducted on genomic DNA and products were analyzed by Sequenom iPLEX matrix-assisted laser desorption/ionization-time-offlight mass spectrometry. Polymorphisms in selected genes were chosen to explain most of the haplotype diversity. Indeed, a region covering all the exons, introns, and 5 kb of each of the 50 and 30 untranslated regions of selected genes was screened using a haplotype tagging SNPs (htSNP) strategy to maximize coverage (16, 17). A total of 112 SNPs were genotyped in Caucasians. Two SNPs were excluded because of a missing genotype frequency more than 5% (rs8041933 in CYP19A1 and rs9340941 in ESR1), and one SNP was excluded because of lack of genetic diversity (rs11968373 in ESR1). All SNPs were in Hardy– Weinberg equilibrium, except rs12900487, rs4441215, and rs2470152 in CYP19A1 and rs2982894 in ESR1. SNPs significantly associated with BCR, including those in SRD5A and HSD17B genes (17, 19) were successfully genotyped in Taiwanese men. There were not enough minor allele homozygotes to evaluate the impact of HSD17B2 rs1119933, rs1364287, rs2955162, and rs8059915 SNPs (17). For both populations, the genotype call rate was in average 98% or more, and SNPs with a missing call rate 5% or more were excluded (n ¼ 6). Negative controls were present for every analysis and quality controls included 5% or more blind duplicates. Measurement of steroid levels Plasma samples were available from the Caucasian cohort collected on the morning of the surgical procedure. We excluded patients that received neoadjuvant hormonal treatment and those with missing genotype information for studied SNPs. Steroids were measured by validated gas chromatography/mass spectrometry or liquid chromatography/tandem mass spectrometry methods (16, 20). Deuterated steroids were added to samples and quality controls were included in each Clinical Cancer Research Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Biomarkers of Prostate Cancer Progression Table 1. Clinical and pathologic characteristics of the Caucasian population (n ¼ 526) Characteristics Localized prostate cancer Age at diagnosis, y Mean SD Range Follow-up median (months) 63.3 6.8 43.5–80.7 88.8 Biochemical recurrence (%) PSA at diagnosis (ng/mL) 10 >10–20 >20 Pathologic Gleason score 6 7 8 Pathologic T stage pT ¼ T2 pT ¼ T3 pT ¼ T4 Nodal invasion N0 Nþ Neoadjuvant hormonotherapy Yes No Adjuvant hormonotherapy Yes No Margin status Negative Positive N (%) 130 (24.7) 362 (69) 103 (20) 56 (11) 158 (31) 244 (48) 107 (21) 313 (60) 195 (37) 13 (3) 481 (92) 44 (8) 31 (6) 495 (94) 30 (6) 496 (94) 368 (70) 154 (30) Abbreviations: PSA, prostate-specific antigen; T, tumor; N, node. run. The lower limit of quantification is as follows: testosterone (0.03 ng/mL), dihydrotestosterone (DHT; 0.005 ng/mL), ADT (0.025 ng/mL), ADT-glucuronide (ADT-G; 1 ng/mL), and 3a-diol-3G and 3a-diol17G (0.25 ng/mL), DHEA (0.1 ng/mL), DHEA-S (0.075 ug/mL), E1-S (0.075 ng/mL), E1 (0.005 ng/mL), E2 (0.001 ng/mL), androstenedione (4-dione; 0.05 ng/ mL), and androstenediol (5-diol; 0.05 ng/mL). Coefficients of variation for intra- and interassays for these methods were 10.0% or below, and accuracies for steroid hormones were as follows: testosterone (100.5%), DHT (103.5%), ADT (96.5%), ADT-G (99.5%), 3a-diol-3G (89.2%), 3a-diol-17G (100.2%), DHEA (96.13%), DHEA-S (99.9%), E1-S (100.45%), E1 (106.9%), E2 (103.15%), 4-dione (95.5%), and 5-diol (95.33%). www.aacrjournals.org Table 2. Clinical and pathologic characteristics of the Asian population (n ¼ 601) Characteristics Age at diagnosis, y Median (IQR) 73 >73 Follow-up median (months) Mean (range) Advanced prostate cancer 73 (67–79) 320 (53.2) 281 (46.8) 39 (3–125) N (%) Disease progression No ADT failure ADT failure PCSM Alive Dead of disease ACM Alive Dead of any cause Clinical stage at diagnosis T1/T2 T3/T4/N1 M1 Gleason score at diagnosis 6 7 8 PSA at ADT initiation (ng/mL) Median (IQR) <34.6 34.6 PSA nadir (ng/mL) Median (IQR) <0.2 0.2 Time to PSA nadir, months Median (IQR) <10 10 Treatment modality ADT as primary treatment ADT for postradical prostatectomy PSA failure ADT for post RT PSA failure Neoadjuvant/adjuvant ADT with RT Others 184 (30.7) 415 (69.3) 499 (83.2) 101 (16.8) 455 (75.8) 145 (24.2) 189 (31.7) 184 (30.8) 224 (37.5) 194 (33.0) 180 (30.6) 214 (36.4) 35 (11–129) 287 (49.6) 292 (50.4) 0.18 (0.01–1.33) 301 (50.8) 292 (49.2) 10 (5–18) 293 (49.4) 300 (50.6) 333 (55.7) 68 (11.4) 18 (3.0) 125 (20.9) 54 (9.0) Abbreviation: IQR, interquartile range. Statistical analysis For association of SNPs with progression and PCSM and overall survival, individual htSNP was first considered using a model based on 3 categories (genomic model), namely Clin Cancer Res; 2012 Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF3 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 vesque et al. Le major allele homozygote, heterozygote, and minor allele homozygote, because the function of most htSNPs remains unknown. For rare homozygotes (frequency < 2%), minor allele homozygotes were combined with the heterozygotes. Cox regression was conducted on each SNP as an isolated covariate, with adjustment for clinicopathologic variables, as previously described (17, 19). In Caucasians, multiple explanatory variables included known prognostic factors: age at diagnosis, smoking status, pathologic stage, Gleason score, PSA values, nodal and margins status, and neoadjuvant and adjuvant ADT (17, 19). In Asians, multiple explanatory variables included known prognostic factors: age at diagnosis, clinical stage, Gleason score, PSA at ADT initiation, PSA nadir, time to PSA nadir, and treatment modality, as defined (18). After analyses with the genomic model, a secondary model was developed (recessive or dominant) for each SNP. All covariables were treated as categorical with 4% or less missing values for all covariables. To estimate the cumulative impact of variations, the 3 categories subdivision was considered as a continuous variable. P values were considered significant if <0.05 and false discovery rates (q values) were calculated to determine the degree to which the tests were prone to false positives, using the R QVALUE package (http://genomics.princeton.edu/ storeylab/qvalue/; ref. 21). To adjust for differences in the absolute levels of sex steroids, we calculated residuals of the natural logarithm of the hormone level regressed on age at blood donation and smoking status. The association with SNPs was conducted by regression of hormone residuals on each SNP independently for 4 models: recessive, dominant, codominant, and additive with 1 degree of freedom for the first 3 models, and 2 degrees of freedom for the additive model. We considered the association of a SNP with variation in hormone levels to be significant if the P value was <0.05. To facilitate the comparison between groups, we displayed the hormone level as untransformed data by using geometric mean and SEM. Statistical analyses were conducted using SAS Statistical Software version 9.2 (SAS Institute) and using PASW statistics version 17 (SPSS Inc.). Results A total of 109 htSNPs distributed across CYP17A1, CYP19A1, HSD3B, and ESR1 were studied in Caucasians (n ¼ 526) based on their role in hormone action, biosynthesis, and metabolism. Positive findings in these 4 genes, combined with previously identified markers in SRD5A1, SRD5A2, HSD17B2, HSD17B3, and HSD17B12 (17, 19) were subsequently tested in the Taiwanese ADT cohort to evaluate their potential impact on survival (n ¼ 32 SNPs). Genetic analyses in Caucasians Relative frequencies of SNPs in patients with cancer and their corresponding HRs (95% CI) are shown in Table 3 and Table 3. Association between htSNPs in candidate steroidogenic pathways and BCR in Caucasians (n ¼ 526) Cox regression analysesb Heterozygous SNP CYP17A1 rs1004467 rs2486758 rs6162 rs743572 CYP19A1 rs1870050 rs2446404 ESR1 rs1062577 rs2982683 rs3003922 rs488133 rs9341016 Homozygous Best-fitting genetic models Major Minor MAF BCRa No-BCRa HR (95% CI) P HR (95% CI) P Model HR (95% CI) P TT TT AA GG CC CC GG AA 7% 22% 59% 65% 1/11/118 9/55/65 65/49/16 66/47/16 1/58/332 19/123/250 143/204/46 150/199/43 0.50 1.82 0.71 0.69 0.052 0.002 0.27 0.24 0.69 1.50 1.19 1.17 0.72 0.29 0.60 0.60 Dom Dom Rec Rec 0.52 1.77 1.54 1.56 0.051 0.003 0.022 0.020 AA AA CC GG 6% 6% 0/20/110 0/18/112 2/40/351 1/39/345 Dom Dom 2.03 (1.22–3.38) 0.006 2.04 (1.20–3.47) 0.008 TT CC GG CC TT AA TT CC TT CC 11% 30% 20% 33% 4% 1/33/96 13/61/54 2/41/87 17/56/57 1/15/114 2/75/314 30/162/198 1.07 (0.72–1.58) 0.75 14/133/247 0.92 (0.62–1.37) 0.68 40/170/181 0.87 (0.58–1.29) 0.49 1/24/360 (0.25–1.01) (1.24–2.68) (0.39–1.31) (0.38–1.27) (0.09–5.15) (0.71–3.19) (0.66–2.13) (0.65–2.11) Dom 1.85 (0.98–3.51) 0.06 Rec 0.26 (0.06–1.09) 0.06 Rec 1.73 (0.97–3.09) 0.06 Rec Dom 2.10 1.80 0.27 1.86 1.98 (0.27–1.00) (1.22–2.57) (1.07–2.24) (1.07–2.26) (1.38–3.22) (0.98–3.29) (0.06–1.11) (1.08–3.20) (1.15–3.39) 0.0006 0.06 0.07 0.026 0.013 NOTE: If there were too few individuals being minor-allele homozygotes, illustrated by shades of gray, they were combined with heterozygotes. Significant associations in HSD17B2 and HSD17B3 were reported elsewhere (17). Bold, significant P values. Abbreviations: Dom, dominant model; Rec, recessive model. a The number represents minor allele homozygotes, heterozygotes, and major allele homozygotes, respectively. b Multivariate models included PSA at diagnosis, pathologic Gleason score, pathologic T stage, age at diagnosis, neoadjuvant therapy, smoking status, adjuvant therapy, surgical margins status, and nodal invasion status. The major allele homozygotes were considered as the reference group, with a fixed HR ¼ 1.00. OF4 Clin Cancer Res; 2012 Clinical Cancer Research Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Biomarkers of Prostate Cancer Progression Supplementary Table S1. After adjustments for known clinicopathologic variables, 4 of the 6 htSNPs in CYP17A1 were significantly associated with progression, 4 of 37 in CYP19A1, 3 of 57 in ESR1, and none in HSD3B1. Of note, significant associations were observed for CYP17A1, involved in one of the initial steps in hormone biosynthesis immediately downstream the action of the P450 side chain cleavage (P450scc) gene, which transforms cholesterol into C21-steroids. The CYP17A1 variant rs6162G and rs2486758C were associated with progression with HR values of 1.54 (95% CI, 1.07–2.24; P ¼ 0.022; q ¼ 0.18) and 1.77 (95%CI, 1.22–2.57; P ¼ 0.003; q ¼ 0.065), respectively. Two additional variants in genes involved in the estrogen biosynthetic pathway merit attention. The rs1870050C variant in CYP19A1, encoding the aromatase enzyme and being the rate-limiting step involved in the estrogens biosynthesis from androgens, was associated with a higher risk of progression (HR 2.03; 95%CI, 1.22– 3.38; P ¼ 0.006; q ¼ 0.10). In addition, the rs1062577A minor allele of the ESR1 gene, which encodes for the estrogen receptor alpha, was also associated with an increased risk of recurrence (HR ¼ 2.10; 95% CI, 1.38– 3.22; P ¼ 0.0006; q ¼ 0.02). Genetic analyses in Taiwanese ADT cohort In agreement with the observations obtained in the Caucasian cohort, positive findings were also revealed for CYP17A1 in the Taiwanese ADT cohort. The CYP17A1 rs6162G variant was associated with mortality with a HR value of 1.69 (95% CI, 1.03–2.78; P ¼ 0.037; q ¼ 0.17). Interestingly, the rs4243229A in the HSD17B2 gene, the product of which inactivates sex steroids, was notably associated with an increased risk of death (HR 1.75; 95% CI, 1.22–2.51; P ¼ 0.002; q ¼ 0.03). Moreover, the ESR1 rs1062577A minor allele was similarly associated with worse outcome with an increased risk of mortality with a HR of 1.45 (95% CI, 1.02–2.07; P ¼ 0.038; q ¼ 0.17). Interestingly, the variant rs2257157C of HSD17B3, involved in hormone bioactivation, was also associated with an increased risk of death (HR 1.60, 95% CI: 1.11– Table 4. Association between htSNPs in candidate steroidogenic pathways and all-cause mortality following ADT in the Asian cohort (n ¼ 601) All-cause mortality ACM SNP CYP17A1 rs1004467 rs2486758 rs6162 rs743572 CYP19A1 rs1870050 rs2446404 ESR1 rs1062577 rs2982683 rs3003922 rs488133 rs9341016 HSD17B2 rs4243229 rs7201637 rs9934209 HSD17B3 rs10739847 rs1810711 rs2257157 Heterozygotes Homozygotes Best-fitting genetic models Major Minor MAF Dead Alive HR (95% CI) P HR (95% CI) P Model HR (95% CI) P TT TT AA GG CC CC GG AA 33% 22% 47% 47% 15/59/59 7/47/80 34/66/34 32/65/36 40/196/177 17/148/246 85/207/120 84/206/117 0.94 0.96 1.24 1.13 0.75 0.81 0.35 0.59 0.78 1.87 1.69 1.53 0.42 0.12 0.037 0.10 DOM REC REC REC 0.91 1.90 1.49 1.42 0.58 0.10 0.06 0.10 AA AA CC GG 29% 3% 7/51/76 0/7/127 39/174/200 0.72 (0.50–1.04) 0.08 1/29/382 0.57 (0.26–1.25) 0.16 DOM DOM 0.70 (0.49–1.00) 0.048 0.55 (0.24–1.28) 0.17 TT CC GG CC TT AA TT CC TT CC 24% 18% 46% 7% 14% 9/52/72 2/41/91 24/69/38 1/16/117 2/36/93 23/144/249 11/127/275 75/226/107 0/54/356 12/85/310 1.42 (0.98–2.06) 0.06 0.90 (0.61–1.32) 0.58 1.11 (0.73–1.68) 0.63 1.68 (0.82–3.41) 0.15 0.99 (0.24–4.08) 0.99 1.10 (0.64–1.88) 0.73 1.00 (0.67–1.49) 0.99 1.52 (0.37–6.34) 0.56 DOM REC DOM DOM REC 1.45 1.03 1.11 0.66 1.52 GG TT GG AA AA CC 20% 18% 44% 9/48/76 15/123/273 1.73 (1.19–2.51) 0.004 1.96 (0.87–4.38) 0.10 4/33/97 13/130/270 0.65 (0.43–0.98) 0.041 1.13 (0.41–3.10) 0.82 25/68/41 77/211/126 1.06 (0.70–1.59) 0.79 1.21 (0.72–2.03) 0.47 DOM DOM REC 1.75 (1.22–2.51) 0.002 0.68 (0.46–1.02) 0.06 1.17 (0.74–1.84) 0.50 AA GG TT TT TT CC 41% 41% 35% 22/60/50 73/191/144 1.03 (0.70–1.52) 0.89 0.95 (0.56–1.62) 0.85 21/66/47 60/214/137 0.87 (0.59–1.29) 0.50 1.08 (0.64–1.83) 0.76 14/76/44 52/180/181 1.72 (1.17–2.52) 0.005 1.14 (0.61–2.15) 0.68 REC REC DOM 0.94 (0.58–1.51) 0.79 1.18 (0.73–1.88) 0.50 1.60 (1.11–2.33) 0.013 (0.65–1.37) (0.66–1.38) (0.80–1.92) (0.73–1.74) (0.43–1.42) (0.85–4.11) (1.03–2.78) (0.93–2.52) (0.63–1.29) (0.87–4.13) (0.99–2.25) (0.93–2.17) (1.02–2.07) (0.25–4.21) (0.74–1.64) (0.38–1.14) (0.37–6.32) 0.038 0.97 0.62 0.14 0.56 NOTE: For SNPs in HSD17B2 (rs1119933, rs1364287, rs2955162, and rs8059915), frequency of minor-allele homozygotes was too low to compute HRs in the recessive model as previously described in Caucasians (17). P < 0.05 are indicated in bold. HRs were adjusted for age, TNM staging, Gleason score, PSA at ADT initiation, PSA nadir, time to PSA nadir, and treatment modality as described in the Materials and Methods section. When the frequency of rare minor-allele homozygotes was 2% or less, they were combined with heterozygotes. For these SNPs, values were not computed for heterozygotes or minor-allele homozygotes, and corresponding boxes are in gray. Bold, significant P values. Abbreviations: DOM, dominant model; REC, recessive model. www.aacrjournals.org Clin Cancer Res; 2012 Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF5 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 vesque et al. Le 2.33; P ¼ 0.013). Conversely, the variant rs1870050C in CYP19A1 was linked with a lower risk of death (HR 0.70; 95% CI, 0.49–1.00; P ¼ 0.048; q ¼ 0.17; Table 4). Most of these associations were also significant for PCSM in Taiwanese men (Supplementary Table S2). Cumulative effects of adverse genotypes When the cumulative effects of the adverse genotypes were investigated, significant associations with progression and overall survival were observed for patients with unfavorable genotypes in CYP17A1, ESR1, and HSD17B2 in both Caucasian and Taiwanese men (Fig. 1). In Caucasians, the cumulative model includes CYP17 rs6162, the presence of one HSD17B2 risk allele (rs1364287, rs8059915, rs4243229, rs2955162, and rs1119933) and ESR1 rs1062577. In Taiwanese men, in addition to CYP17A1 and ESR1 variants, only the HSD17B2 alleles (rs4243229 and rs7201637) were included in the model because of very low frequency of the other SNPs in this population. Relationship with endogenous sex-steroid hormones levels The relationship of prognostic markers positive in both cohorts with circulating sex steroids was then assessed in the Caucasian cohort. Remarkably, CYP17A1 (rs6162), HSD17B2 (rs7201637), and ESR1 (rs1062577) variations were associated with significant changes in plasma steroid levels. The CYP17A1 rs6162 variation is associated with a 20% difference in DHEA-S levels (P ¼ 0.031). Although homozygotes for the minor allele are rare in Caucasians for the HSD17B2 rs7201637 variation (n ¼ 2), these patients display 50% lower 5-DIOL and DHT levels. The ESR1 rs1062577 variant correlated with reduced levels of E1-S and ADT-G. Additional hormone data in relation to genetic polymorphisms in CYP17A1, HSD17B2, and ESR1 are provided in Table 5. Discussion There are clear unmet oncological needs to better predict prostate cancer progression, especially lethal prostate cancer, and so far, no prognostic markers fill this gap in knowledge. Host genetic variations in the well-characterized hormone biosynthetic and degradation pathways have not been studied systematically in terms of molecular markers associated with prostate cancer progression and none have been validated at different stages of the disease (13, 18, 22). Here, we exposed several common inherited variations in CYP17A1, HSD17B2, and ESR1 associated with prostate cancer progression in 2 independent cohorts of patients and further showed their biologic association with plasma hormone levels. The HSD17B2 gene is involved in steroids interconversion and controls C21, C19-, and C18-steroid bioavailability for nuclear receptor bindings and actions (Fig. 2; refs. 23, 24). Remarkably, variants in this gene, although being at different frequencies in the 2 ethnic groups studied herein, persist as prognostic markers associated with BCR, disease OF6 Clin Cancer Res; 2012 Figure 1. Impact of germline variations in CYP17A1, ESR1, and HSD17B2 with prostate cancer progression. Prognostic SNPs located in ESR1, CYP17A1, and HSD17B2 were combined in relation to BCR in Caucasians (A) and all-cause mortality in Asians under ADT (B). Log rank (LR) P values are shown in each frame and HRs for each category are shown under both panels. progression, and mortality following ADT [herein and ref. (17)]. In Taiwanese men, the HSD17B2 rs4243229A [a variant more frequent in Asians than Caucasians; MAF; Clinical Cancer Research Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. www.aacrjournals.org Hormone SNP A/G G/A T/C T/C A/G G/A T/C T/C C/T G/A G/A T/A G/C C/T G/A G/A T/A G/C C/T G/A G/A T/A G/C C/T G/A G/A T/A G/C C/T T/A T/C C/T T/A T/C 54/223/189 51/217/196 404/59/2 289/147/28 54/226/187 51/220/194 404/60/2 289/149/27 251/185/26 276/160/26 453/12 403/56/2 177/213/74 231/173/25 254/149/26 420/12 376/50/2 169/196/66 251/187/26 276/162/26 455/12/ 405/56/2 178/215/73 252/187/26 277/162/26 456/12 406/56/2 178/215/74 213/204/48 366/97/2 425/32/2 214/202/47 364/96/3 423/32/2 Homozygote minor Homozygote major 1 minor allele 1 minor allele Homozygote minor Secondary model (dominant) Secondary model (recessive) 1.83 0.15 1.82 0.16 1.61 0.07 1.66 0.07 0.97 0.09 0.94 0.08 0.80 þ 0.03 0.85 0.03 527.27 18.51 521.75 17.54 519.42 14.08 511.71 15.22 509.41 22.82 0.63 0.04 0.63 0.04 0.63 0.03 0.63 þ 0.03 0.63 0.03 309.07 9.47 306.61 8.93 307.24 7.16 303.08 7.56 295.38 11.32 17.68 1.27 17.39 1.16 17.92 0.75 18.03 þ 0.83 18.21 1.81 30.81 2.75 31.27 1.73 30.72 1.52 0.42 0.03 0.42 0.02 0.40 0.02 1.55 0.09 1.57 0.09 1.63 0.16 1.51 0.11 0.78 0.04 0.78 0.04 0.81 þ 0.09 0.71 0.05 503.22 21.81 513.68 23.52 454.39 81.06 569.64 34.01 518.69 19.84 0.62 0.03 0.63 0.03 0.66 0.10 0.61 þ 0.05 0.65 0.05 303.47 11.23 301.46 12.42 326.29 36.17 j 347.53 19.12 315.20 10.70 18.07 0.48 18.80 0.70 20.39 2.21 17.32 þ 0.93 17.65 0.45 29.92 1.30 27.40 1.62 27.04 3.38 0.38 0.03 0.34 0.03 0.35 0.07 1.83 0.15 1.82 0.16 1.61 0.07 1.66 0.07 0.97 0.09 0.94 0.08 0.80 þ 0.03 0.85 0.03 527.27 18.51 521.75 17.54 519.42 14.08 199.24 122.38 511.71 15.22 533.07 37.28 509.41 22.82 0.69 0.08 0.63 0.04 0.67 0.08 0.63 0.04 0.63 0.03 0.78 þ 0.09 0.63 þ 0.03 0.58 0.04 0.63 0.03 321.31 32.76 309.07 9.47 360.06 30.57 306.61 8.93 307.24 7.16 157.65 33.73 303.08 7.56 313.98 16.10 295.38 11.32 19.46 2.86 17.68 1.27 19.21 1.81 17.39 1.16 17.92 0.75 12.33 þ 0.24 18.03 þ 0.83 18.34 0.85 18.21 1.81 30.98 2.50 30.81 2.75 27.39 7.18 31.27 1.73 17.57 5.38 30.72 1.52 0.40 0.05 0.42 0.03 0.33 0.08 0.42 0.02 0.82 0.15 0.40 0.02 1.63 0.10 1.63 0.09 0.70 0.06 1.78 0.36 0.79 0.06 0.79 0.05 0.44 þ 0.03 0.87 0.25 524.48 76.08 510.49 75.01 1.59 0.06 1.59 0.07 1.58 0.16 1.55 0.11 0.79 0.03 0.79 0.03 0.79 þ 0.09 0.74 0.06 505.79 21.23 513.23 22.70 454.39 81.06 549.38 34.26 522.36 17.55 0.63 0.03 0.64 0.03 0.66 0.10 0.61 þ 0.04 0.63 0.04 305.59 10.62 308.96 11.56 326.29 36.17 338.19 19.17 314.89 8.96 18.24 0.55 18.86 0.65 20.39 2.21 17.11 þ 0.91 17.82 0.40 30.12 1.16 27.40 1.59 26.37 3.23 0.38 0.02 0.34 0.03 0.37 0.06 1.60 0.08 1.61 0.08 1.61 0.06 1.61 0.06 0.82 0.04 0.81 0.04 0.80 þ 0.03 0.80 0.03 516.93 14.10 518.77 14.05 517.63 13.87 518.45 13.99 514.46 14.97 0.63 0.03 0.63 0.03 0.63 0.02 0.63 þ 0.03 0.64 0.03 306.67 7.23 304.69 7.25 307.71 7.03 308.16 7.05 306.06 7.79 17.85 0.76 17.90 0.77 17.98 0.73 17.94 þ 0.74 17.90 0.86 30.37 1.54 30.42 1.41 30.45 1.43 0.40 0.02 0.40 0.02 0.40 0.02 Pb 0.166 0.168 0.157 0.280 0.094 0.133 0.433 0.029 0.629 0.984 0.350 0.028 0.938 0.638 0.772 0.804 0.704 0.131 0.824 0.159 0.709 0.007 0.580 0.521 0.085 0.273 0.417 0.590 0.834 0.088 0.173 0.452 0.054 0.278 0.089 0.071 0.759 0.322 0.031 0.045 0.927 0.027 0.378 0.888 0.350 0.325 0.807 0.880 0.611 0.804 0.805 0.779 0.769 0.722 0.709 0.062 0.297 0.415 0.027 0.273 0.441 0.453 0.935 0.027 0.123 0.241 0.017 0.565 0.233 0.688 0.588 0.794 0.164 0.944 0.439 0.188 0.030 0.725 0.322 0.406 0.458 0.048 0.653 0.057 0.023 0.742 0.390 0.521 0.779 0.990 0.054 0.392 0.710 0.672 0.198 0.495 0.892 0.882 Codomin. Add. Dom. Reces. 0.170 0.262 0.947 0.141 0.309 0.403 0.875 0.008 0.336 0.942 0.350 199.24 122.38 0.145 533.07 37.28 0.997 0.69 0.08 0.571 0.67 0.08 0.840 0.804 0.78 þ 0.09 0.682 0.58 0.04 0.248 321.31 32.76 0.609 360.06 30.57 0.578 0.709 157.65 33.73 0.020 313.98 16.10 0.446 19.46 2.86 0.717 19.21 1.81 0.060 0.273 12.33 þ 0.24 0.588 18.34 0.85 0.304 30.98 2.50 0.678 27.39 7.18 0.029 17.57 5.38 0.220 0.40 0.05 0.221 0.33 0.08 0.024 0.82 0.15 0.351 1.63 0.10 1.63 0.09 0.70 0.06 1.78 0.36 0.79 0.06 0.79 0.05 0.44 þ 0.03 0.87 0.25 524.48 76.08 510.49 75.01 Mean SEMa Mean SEMa Mean SEMa Mean SEMa Mean SEMa Mean SEMa Mean SEMa Heterozygote Genomic model NOTE: Data are presented for individuals on whom we had genotyping, and plasma assay data. Bold, significant P values. Abbreviations: Codomin, codominant model; Add, additive model; Dom, dominant model; Reces, recessive model. a Hormone levels are expressed as geometric mean and SEM. b P values from linear regression of the residuals on the SNP under each model. rs6162 rs743572 rs 1004467 rs2486758 DHEA-S (mg/mL) rs6162 rs743572 rs 1004467 rs2486758 HSD17B2 5-diol (pg/mL) rs1119933 rs2955162 rs4243229 rs7201637 rs8059915 4-dione (ng/mL) rs1119933 rs2955162 rs4243229 rs7201637 rs8059915 DHT (pg/mL) rs1119933 rs2955162 rs4243229 rs7201637 rs8059915 E2 (pg/mL) rs 1119933 rs2955162 rs4243229 rs7201637 rs8059915 ESR1 ADT-G (ng/mL) rs488133 rs1062577 rs9341016 E1-S (ng/mL) rs488133 rs1062577 rs9341016 CYP17A1 DHEA (ng/mL) Gene Major/ Minor Homozygote allele N (0, 1, 2) major Table 5. Association between positive htSNP markers and circulating steroid hormone levels in the Caucasian cohort Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Biomarkers of Prostate Cancer Progression Clin Cancer Res; 2012 Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF7 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 vesque et al. Le Figure 2. Simplified schematic representation of sex–steroid biosynthesis pathways. 5-DIOL, androst-5-ene-3b, 17bdiol; A-DIONE, androstanedione; DHEA, dehydroepiandrosterone; DHEA-S, DHEA-sulfate; DHT, dihydrotestosterone; E1, estrone; E2, estradiol; TESTO, testosterone. 20% vs. 1%; ref. (17)] was associated with shorter survival after adjustment for known prognostic factors and clearly defines a subgroup of patients at higher risk of progression and mortality. Although the impact of this polymorphism could not be assessed precisely on the steroid hormonal profile due to its lower frequency in Caucasians (no minor allele homozygote), other HSD17B2 markers influenced steroid hormone levels in patients with cancer, further reinforcing their potential role in cancer progression. On the basis of our preliminary data and those of others (24–27), we can speculate that genetic variations associated with progression in the HSD17B2 gene might lead to a loss-of-function in favor of the reductive pathway and accumulation of potent androgens and estrogens, thereby potentially modifying the course of the disease. Inversely, the observation of an association for the protective HSD17B2 rs7201637 with lower DHT in available homozygotes, suggests a more efficient inactivation of potent hormone by the 17bHSD type 2 enzyme, thereby plausibly reducing progression risk in these patients. Therefore, these positive markers, or any functional polymorphism in strong linkage, may significantly impair gene expression/function with subsequent modification in hormone levels that favor cell proliferation, and ultimately the survival of the patient. Together, data sustain a critical role of the HSD17B2 pathway as an important determinant in prostate cancer aggressiveness, progression, and survival (24–28), and suggest that targeting this pathway and/or potentially the reverse HSD17B3/5 pathway (29, 30) might be beneficial for anticancer therapy (Fig. 2). Significant associations were also observed in both cohorts for SNPs in CYP17A1 and ESR1. CYP17A1 is one of the early steps in the biosynthesis of sex steroids and the molecular target of abiraterone that improves survival after docetaxel treatment in patients with CRPC (8). The CYP17A1 rs6162 variation was associated with prostate cancer progression and mortality herein, further reinforcing OF8 Clin Cancer Res; 2012 the importance of this critical pathway in cancer progression at numerous disease stages and in different ethnic groups. In support, 3 previous studies have looked at the impact of CYP17A1 on prostate cancer outcomes (31–33). In agreement with our findings, in these studies, the presence of polymorphisms in CYP17A1 was also associated with disease progression and mortality in Caucasians and Asians (31–33; Table 6). Overall, our data combined with previous observations (31–33) clearly support a major role of CYP17A1 on cancer progression after diagnosis. Moreover, we showed that the presence of CYP17A1 SNPs is further associated with altered levels of circulating DHEA-S in Caucasians, which likely modify steroid precursor levels available for intracrine conversion to more potent hormones in tissues and prostatic cells. It is unclear whether this modification in DHEA-S levels is caused by an altered enzyme activity or expression in target cells and/or feedback mechanism on the hypothalamic–pituitary–adrenal axis or any other unknown physiologic process. Interestingly, germline variations of CYP17A1, associated with outcomes herein and in previous studies (31, 32), are mostly located in the promoter region and exon 1 raising the possibility of a functional impact on transcriptional activity of the gene. Further study is clearly needed to decipher the molecular mechanism(s) underlying this pathophysiologic association. Ryan and colleagues showed that baseline levels of androstenedione might be predictive of ketoconazole response in metastatic patients (34). Therefore, in view of data presented here, the high frequency of the CYP17A1 rs6162 (50%) and its association with circulating DHEA-S levels, it would be relevant to evaluate, in future studies, the predictive and prognostic impact of CYP17A1 variations on abiraterone response and outcome. Moreover, the ESR1 rs1062577 variant associated with progression also seems to affect the circulating hormone profile, with carriers of the minor allele showing reduced Clinical Cancer Research Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Biomarkers of Prostate Cancer Progression Table 6. Summary of current data supporting CYP17A1 SNPs as germline prognostic molecular markers of prostate cancer progression Sample size (n) Ethnic background Disease stage Treatment CYP17A1 rs6162(G) rs743572 (A) rs10883783 (T) Wright and colleagues (32) Hamada and colleagues (31) Yamada and colleagues (30) This study 598 Caucasian Localized (RP/radiation/ADT) 222 Caucasian Advanced (CRPC) 214 Asian Japanese Advanced (ADT) 526 Caucasian Localized (RP) 601 Asian Taiwanese Advanced (ADT) — — " — " " n/s " " n/s " " n/s a " NOTE: Bold, significant P values. Abbreviations: RP, radical prostatectomy; n/s, nonsignificant; —, not evaluated. a Trend. systemic E1-S and ADT-G levels. A previous report suggests that E1-S is associated with prostate cancer aggressiveness, being positively associated with margin status, stage, and PSA levels (35). Accordingly, we also observed a correlation between E1-S and margin status, with higher levels being associated positive margins (P ¼ 0.007, not shown). A previous extensive evaluation of ESR1 genetic variations was also conducted on cancer aggressiveness and efficacy of ADT; however, to our knowledge, the rs1062577 variant was not specifically evaluated in previous studies (28, 36). Another group reported that this common SNP (rs1062557) might affect the binding capacity of miR-186 based on in silico analysis (37). Interestingly, the HSD17B3 rs2257157 SNP was associated with cancer-related death in Taiwanese men on ADT, whereas the CYP19A1 rs1870050 SNP was associated with better survival. This variation is located in the promoter region of CYP19A1 encoding the aromatase enzyme, a ratelimiting step in estrogens production from androgens. Remarkably, the CYP19A1 rs1870050 was associated with progression in Caucasians in agreement with previous association displaying a shorter time to progression in a cohort of men on ADT composed mainly of Caucasians (13). Differences in outcome between populations are perhaps linked to different haplotypes tagged by these SNPs in diverse ethnic groups. Nevertheless, validation of these 2 latter markers in cohorts with similar clinical and pathologic characteristics is still required to fully ascertain their role in cancer progression. The importance of inherited genetic variations in the steroidogenic pathways is not surprising because they contribute to the maintenance and bioavailability of active hormones in multiple target organs despite hormonal therapy (5, 6). Nevertheless, most of these htSNPs, or their associated SNPs, have not been biochemically studied and their functionality still remains undefined. As such, limitations of the study are mainly related to the scarcity of the functional data for positive markers and lack of additional cohorts for definitive validation at each disease stage. www.aacrjournals.org Strengths of the study include a significant number of patients, a candidate gene approach, the high plausibility of the association based on the biologic function of selected candidate pathways, repeated associations in 2 genetically divergent ethnic groups, an impact on circulating steroid levels associated with positive markers, and clinical endpoints including both progression and survival. Finally, results further underscore the need to examine the potential relationship between prognostic markers and tissue hormone levels to better understand their physiologic consequence and contribution to cancer progression. Data clearly emphasizes the biologic significance of the HSD17B2, CYP17A, and ESR1 pathways in prostate cancer progression and provides promising prognostic candidates for explaining differences in clinical outcomes. The assessment of host genetic variations in key steroidogenic pathways, such as those identified herein, represent additional evidence that sex steroids are likely involved throughout all stages of disease progression including a castrate environment. Furthermore, findings support the seed and soil hypothesis in which germline variations in steroidogenic pathways influence the hormonal microenvironment to which cancer cells are exposed and a subsequent impact on prostate cancer proliferation, recurrence, and progression. These molecular markers may improve prognostication of prostate cancer, lead to a better patient stratification in future clinical trials, targeting therapeutic interventions to optimize hormonal manipulation in patients with a high risk of recurrence, and more likely to benefit from treatment. However, before such translational advance is made, additional investigations are required to fully characterize the underlying biologic mechanisms driving the positive associations of inherited germline variations in those key genes on steroid hormone levels, progression, and survival. In conclusion, in combination with tumor changes, the major challenge will be to decipher precisely which sets of inherited gene combinations are associated with favorable and adverse outcomes to ultimately personalize the management of prostate cancer. Clin Cancer Res; 2012 Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF9 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 vesque et al. Le Disclosure of Potential Conflicts of Interest Levesque, C. Guillemette, L. Lacombe, and Y. Fradet have been named E. inventors on a patent application owned by Laval University on related work that preceded this study. No potential conflicts of interest were disclosed by the other authors. Acknowledgments The authors thank the genomics, particularly Marc-Andre Rodrigue, and the statistical service platforms, particularly Sun Makosso-Kallyth, of the CRCHUQ (Quebec, Canada), as well as the National Center for Genome Medicine [National Science Council (NSC), Taiwan], for technical support. Authors' Contributions Levesque, S. Huang, Y. Fradet, C. Guillemette Conception and design: E. Levesque, C. Guillemette Development of methodology: E. Acquisition of data (provided animals, acquired and managed patients, Levesque, S. Huang, E. Audet-Walsh, L. provided facilities, etc.): E. Lacombe, B. Bao, Y. Fradet, I. Laverdiere, M. Rouleau, C. Huang, P. Caron, C. Guillemette Analysis and interpretation of data (e.g., statistical analysis, biosta Levesque, S. Huang, E. Audet-Walsh, I. tistics, computational analysis): E. Laverdiere, M. Rouleau, C. Guillemette Levesque, E. Writing, review, and/or revision of the manuscript: E. Audet-Walsh, L. Lacombe, B. Bao, Y. Fradet, I. Laverdiere, M. Rouleau, C. Guillemette Administrative, technical, or material support (i.e., reporting or orga Levesque, L. Lacombe, C. Huang, nizing data, constructing databases): E. C. Guillemette Levesque, C. Guillemette Study supervision: E. Grant Support This work was supported by Canadian research grants from the Prostate Cancer Research Foundation of Canada (to E. Levesque), Cancer Research Society (to C. Guillemette), and the Canada Research Chair Program (to C. Guillemette). This work was also supported by the NSC (Taiwan, Republic of China; grants NSC-98-2320-B-039-019-MY3, NSC-99-2314-B-037-018MY3, and NSC-100-2314-B-039-009-MY3), China Medical University (grant CMU99-COL-13), and Kaohsiung Medical University Hospital (grant KMUH99-9R12 and grant KMUH100-0R42). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received August 29, 2012; revised October 30, 2012; accepted November 5, 2012; published OnlineFirst November 27, 2012. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. OF10 Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69–90. Harris WP, Mostaghel EA, Nelson PS, Montgomery B. Androgen deprivation therapy: progress in understanding mechanisms of resistance and optimizing androgen depletion. Nat Clin Pract Urol 2009;6: 76–85. Horwitz EM. Prostate cancer: optimizing the duration of androgen deprivation therapy. Nat Rev Urol 2009;6:527–9. Bianchini D, de Bono JS. Continued targeting of androgen receptor signalling: a rational and efficacious therapeutic strategy in metastatic castration-resistant prostate cancer. Eur J Cancer 2011;47 Suppl 3: S189–94. Montgomery RB, Mostaghel EA, Vessella R, Hess DL, Kalhorn TF, Higano CS, et al. Maintenance of intratumoral androgens in metastatic prostate cancer: a mechanism for castration-resistant tumor growth. Cancer Res 2008;68:4447–54. Chang KH, Li R, Papari-Zareei M, Watumull L, Zhao YD, Auchus RJ, et al. Dihydrotestosterone synthesis bypasses testosterone to drive castration-resistant prostate cancer. Proc Natl Acad Sci U S A 2011;108:13728–33. Sharifi N. The 5alpha-androstanedione pathway to dihydrotestosterone in castration-resistant prostate cancer. J Invest Med 2012;60: 504–7. de Bono JS, Logothetis CJ, Molina A, Fizazi K, North S, Chu L, et al. Abiraterone and increased survival in metastatic prostate cancer. N Engl J Med 2011;364:1995–2005. Miller K. Words of wisdom. Re: antitumour activity of MDV3100 in castration-resistant prostate cancer: a phase 1–2 study. Scher HI, Beer TM, Higano CS, et al. Prostate cancer foundation/department of defense prostate cancer clinical trials consortium. Eur Urol 2010;58: 464–5. Payton S. Prostate cancer: MDV3100 has antitumor activity in castration-resistant disease. Nat Rev Urol 2010;7:300. Scher HI, Beer TM, Higano CS, Anand A, Taplin ME, Efstathiou E, et al. Antitumour activity of MDV3100 in castration-resistant prostate cancer: a phase 1–2 study. Lancet 2010;375:1437–46. Yap TA, Zivi A, Omlin A, de Bono JS. The changing therapeutic landscape of castration-resistant prostate cancer. Nat Rev Clin Oncol 2011;8:597–610. Ross RW, Oh WK, Xie W, Pomerantz M, Nakabayashi M, Sartor O, et al. Inherited variation in the androgen pathway is associated with the efficacy of androgen–deprivation therapy in men with prostate cancer. J Clin Oncol 2008;26:842–7. Clin Cancer Res; 2012 14. Rasiah KK, Gardiner-Garden M, Padilla EJ, Moller G, Kench JG, Alles MC, et al. HSD17B4 overexpression, an independent biomarker of poor patient outcome in prostate cancer. Mol Cell Endocrinol 2009;301:89–96. 15. Yang M, Xie W, Mostaghel E, Nakabayashi M, Werner L, Sun T, et al. SLCO2B1 and SLCO1B3 may determine time to progression for patients receiving androgen deprivation therapy for prostate cancer. J Clin Oncol 2011;29:2565–73. 16. Nadeau G, Bellemare J, Audet-Walsh E, Flageole C, Huang SP, Bao BY, et al. Deletions of the androgen-metabolizing UGT2B genes have an effect on circulating steroid levels and biochemical recurrence after radical prostatectomy in localized prostate cancer. J Clin Endocrinol Metab 2011;96:E1550–7. 17. Audet-Walsh E, Bellemare J, Lacombe L, Fradet Y, Fradet V, Douville P, et al. The impact of germline genetic variations in hydroxysteroid (17-Beta) dehydrogenases on prostate cancer outcomes after prostatectomy. Eur Urol 2012;62:88–96. 18. Huang CN, Huang SP, Pao JB, Chang TY, Lan YH, Lu TL, et al. Genetic polymorphisms in androgen receptor-binding sites predict survival in prostate cancer patients receiving androgen–deprivation therapy. Ann Oncol 2012;23:707–13. 19. Audet-Walsh E, Bellemare J, Nadeau G, Lacombe L, Fradet Y, Fradet V, et al. SRD5A polymorphisms and biochemical failure after radical prostatectomy. Eur Urol 2011;60:1226–34. 20. Lepine J, Audet-Walsh E, Gregoire J, Tetu B, Plante M, Menard V, et al. Circulating estrogens in endometrial cancer cases and their relationship with tissular expression of key estrogen biosynthesis and metabolic pathways. J Clin Endocrinol Metab 2010;95: 2689–98. 21. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci U S A 2003;100:9440–5. 22. Huang CN, Huang SP, Pao JB, Hour TC, Chang TY, Lan YH, et al. Genetic polymorphisms in oestrogen receptor-binding sites affect clinical outcomes in patients with prostate cancer receiving androgen–deprivation therapy. J Intern Med 2012;271:499–509. 23. Labrie F, Luu-The V, Lin SX, Simard J, Labrie C. Role of 17 betahydroxysteroid dehydrogenases in sex steroid formation in peripheral intracrine tissues. Trends Endocrinol Metab 2000;11: 421–7. 24. Harkonen P, Torn S, Kurkela R, Porvari K, Pulkka A, Lindfors A, et al. Sex hormone metabolism in prostate cancer cells during transition to an androgen-independent state. J Clin Endocrinol Metab 2003;88: 705–12. Clinical Cancer Research Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Biomarkers of Prostate Cancer Progression 25. Elo JP, Harkonen P, Kyllonen AP, Lukkarinen O, Poutanen M, Vihko R, et al. Loss of heterozygosity at 16q24.1-q24.2 is significantly associated with metastatic and aggressive behavior of prostate cancer. Cancer Res 1997;57:3356–9. 26. Harkonen P, Kyllonen AP, Nordling S, Vihko P. Loss of heterozygosity in chromosomal region 16q24.3 associated with progression of prostate cancer. Prostate 2005;62:267–74. 27. Friedlander TW, Roy R, Tomlins SA, Ngo VT, Kobayashi Y, Azameera A, et al. Common structural and epigenetic changes in the genome of castration resistant prostate cancer. Cancer Res 2012;72:616–25. 28. Sun T, Oh WK, Jacobus S, Regan M, Pomerantz M, Freedman ML, et al. The impact of common genetic variations in genes of the sex hormone metabolic pathways on steroid hormone levels and prostate cancer aggressiveness. Cancer Prev Res (Phila) 2011;4:2044–50. 29. Adeniji AO, Twenter BM, Byrns MC, Jin Y, Chen M, Winkler JD, et al. Development of potent and selective inhibitors of aldo-keto reductase 1C3 (type 5 17beta-hydroxysteroid dehydrogenase) based on Nphenyl-aminobenzoates and their structure-activity relationships. J Med Chem 2012;55:2311–23. 30. Chen M, Adeniji AO, Twenter BM, Winkler JD, Christianson DW, Penning TM. Crystal structures of AKR1C3 containing an N-(aryl) amino-benzoate inhibitor and a bifunctional AKR1C3 inhibitor and androgen receptor antagonist. Therapeutic leads for castrate resistant prostate cancer. Bioorg Med Chem Lett 2012;22:3492–7. 31. Yamada T, Nakayama M, Shimizu T, Nonen S, Nakai Y, Nishimura K, et al. Genetic polymorphisms of CYP17A1 in steroidogenesis pathway www.aacrjournals.org 32. 33. 34. 35. 36. 37. are associated with risk of progression to castration-resistant prostate cancer in Japanese men receiving androgen deprivation therapy. Int J Clin Oncol 2012 Jun 21. [Epub ahead of print]. Hamada A, Danesi R, Price DK, Sissung T, Chau C, Venzon D, et al. Association of a CYP17 polymorphism with overall survival in Caucasian patients with androgen-independent prostate cancer. Urology 2007;70:217–20. Wright JL, Kwon EM, Lin DW, Kolb S, Koopmeiners JS, Feng Z, et al. CYP17 polymorphisms and prostate cancer outcomes. Prostate 2010;70:1094–101. Ryan CJ, Halabi S, Ou SS, Vogelzang NJ, Kantoff P, Small EJ. Adrenal androgen levels as predictors of outcome in prostate cancer patients treated with ketoconazole plus antiandrogen withdrawal: results from a cancer and leukemia group B study. Clin Cancer Res 2007;13:2030–7. Giton F, de la Taille A, Allory Y, Galons H, Vacherot F, Soyeux P, et al. Estrone sulfate (E1S), a prognosis marker for tumor aggressiveness in prostate cancer (PCa). J Steroid Biochem Mol Biol 2008;109:158–67. Sun T, Lee GS, Werner L, Pomerantz M, Oh WK, Kantoff PW, et al. Inherited variations in AR, ESR1, and ESR2 genes are not associated with prostate cancer aggressiveness or with efficacy of androgen deprivation therapy. Cancer Epidemiol Biomarkers Prev 2010;19: 1871–8. Li N, Dong J, Hu Z, Shen H, Dai M. Potentially functional polymorphisms in ESR1 and breast cancer risk: a meta-analysis. Breast Cancer Res Treat 2010;121:177–84. Clin Cancer Res; 2012 Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research. OF11 Published OnlineFirst November 27, 2012; DOI: 10.1158/1078-0432.CCR-12-2812 Molecular Markers in Key Steroidogenic Pathways, Circulating Steroid Levels, and Prostate Cancer Progression Éric Lévesque, Shu-Pin Huang, Étienne Audet-Walsh, et al. Clin Cancer Res Published OnlineFirst November 27, 2012. Updated version Supplementary Material E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-12-2812 Access the most recent supplemental material at: http://clincancerres.aacrjournals.org/content/suppl/2012/11/27/1078-0432.CCR-12-2812.DC1 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from clincancerres.aacrjournals.org on June 14, 2017. © 2012 American Association for Cancer Research.