Abstract
Cholesterol metabolism has been implicated in prostate cancer pathogenesis. Here, we assessed the association of intratumoral mRNA expression of cholesterol synthesis enzymes, transporters, and regulators in tumor specimen at diagnosis and lethal prostate cancer, defined as mortality or metastases from prostate cancer in contrast to non-lethal disease without evidence of metastases after at least eight years of follow-up. We analyzed the prospective prostate cancer cohorts within the Health Professionals Follow-up Study (n = 249) and the Physicians’ Health Study (n = 153) as well as expectantly managed patients in the Swedish Watchful Waiting Study (n = 338). The expression of squalene monooxygenase (SQLE) was associated with lethal cancer in all three cohorts. Men with high SQLE expression (>1 standard deviation above the mean) were 8.3 times (95% confidence interval, 3.5 to 19.7) more likely to have lethal cancer despite therapy compared to men with the mean level of SQLE expression. Absolute SQLE expression was associated with lethal cancer independently from Gleason grade and stage, as was a SQLE expression ratio in tumor versus surrounding benign prostate tissue. Higher SQLE expression was tightly associated with increased histologic markers of angiogenesis. Collectively, this study establishes the prognostic value of intratumoral cholesterol synthesis as measured via SQLE, its second rate-limiting enzyme. SQLE expression at cancer diagnosis is prognostic for lethal prostate cancer both after curative-intent prostatectomy and in a watchful waiting setting, possibly by facilitating micrometastatic disease.
Keywords: Prostate cancer, cholesterol, squalene monooxygenase, lethal cancer, angiogenesis
Introduction
Cholesterol is required for proliferation in all animal cells and is especially important for membrane formation (1). In prostate cancer, cholesterol may also act as a substrate for intratumoral androgen biosynthesis even after androgen deprivation therapy via the CPY17A1 enzyme, which is expressed by castration resistant prostate cancer cells that de-novo synthesize androgens (2). Several lines of evidence link cholesterol metabolism and prostate cancer progression. First, a positive association between serum cholesterol levels and high-grade prostate cancer has been described (3). Second, in-vivo xenograft and in-vitro cancer progression models have identified numerous aberrations in regulators of cholesterol metabolism (4–7). Third, statin use as a cholesterol-lowering therapy has been associated with a lower risk of prostate cancer diagnosis (8), advanced disease (9), and mortality (10).
In light of these laboratory and observational data, randomized-controlled trials are currently assessing the effect of statin use in prostate cancer. However, it is unknown whether and how local tumor cholesterol metabolism is related to clinical outcomes. Understanding of such molecular differences is paramount, since it may help distinguish indolent prostate cancers from those that progress to lethal disease. In this study, we investigated the association of cholesterol metabolism with lethal prostate cancer and putative mechanisms in three prospective patient cohorts with both predominantly primary prostatectomy and watchful waiting approaches. We hypothesized that mRNA expression of cholesterol synthesis enzymes, transporters, and regulators in tumor specimen is associated with lethal prostate cancer.
Methods
Patients
The analysis included Caucasian prostate cancer patients sampled from three well-described prospective cohort studies.
The Health Professionals Follow-up Study (HPFS) prostate cancer cohort is comprised of those participants of the HPFS, a prospective observational study of initially 51,529 U.S. male health professionals aged 40 to 75 years, who developed prostate cancer during follow-up (11). Data were obtained through biennial questionnaires, systematic medical record review, and in-depth ascertainment of death causes (98% complete).
The Physicians’ Health Study (PHS) prostate cancer cohort includes patients who developed prostate cancer during follow-up of the PHS, initially a randomized, controlled trial (RCT) of acetylsalicylic acid and micronutrients for the primary prevention of cardiovascular disease and cancer among 29,071 healthy U.S. male physicians initially aged ≥ 40 years with follow-up beginning in 1982 (12) or 1999 (13). Data were obtained similarly to HPFS with 99% complete follow-up for cause-specific mortality. Both HPFS and PHS were partly conducted primarily in the era of prostate-specific antigen (PSA) screening. We included men for whom archival tumor tissue was available either from radical prostatectomy (92%) or transurethral resection of the prostate (TURP, 8%).
The Swedish Watchful Waiting Study (SWWS) was population-sampled from hospital catchment areas in Southeastern and central Sweden (14). Patients were incidentally diagnosed with localized prostate cancer when treated for symptomatic prostatic hyperplasia with TURP without PSA screening in place. All men were initially untreated at diagnosis, but symptomatic progression was treated with medical or surgical castration. Deaths were ascertained via medical record review, autopsies, and the national death registry.
Tissue specimens and biomarker data
For patients in all cohorts, tumor tissue from prostatectomy or the initial diagnostic transurethral resection of the prostate (TURP) underwent standardized pathological re-review (15). This review included standardized Gleason grading, assessment of high-grade prostatic intraepithelial neoplasia, and other histological features. In addition, areas of tumor tissue were identified for construction of tissue microarrays.
For patients from all three cohorts, mRNA expression profiling was performed from archived specimen consisting of high-density tumor areas and, if applicable, adjacent prostate tissue without cancer, with processing and quality control as previously reported (14, 16, 17). Whole transcriptome expression profiling was performed using the GeneChip Human Gene 1.0 ST array (Affymetrix, Santa Clara, CA), which resulted in 20,254 unique named genes in HPFS and PHS specimens (Gene Expression Omnibus accession number GSE62872). For SWWS, the DASL platform (Illumina, San Diego, CA) was used, resulting in 6,144 unique named genes (GSE8402).
We examined associations of cholesterol pathways with markers of prostate tumor progression defined by angiogenesis and cell proliferation. Tumor angiogenesis, which facilitates micrometastases, was measured as previously described in the HPFS cohort (18), using CD34 staining and semi-automated histologic analysis. Vessel area and irregularity were markers of interest since both are associated with lethal cancer (18).
Cell proliferation was measured with tissue microarrays for the Ki-67 antigen of proliferating cells, as previously described (19).
Statistical analysis
Within all three prospective tissue cohorts, case–control studies were designed. Cases included men who died from prostate cancer or developed metastases (lethal cancer). Controls were men with prostate cancer who survived for more than eight years without evidence of metastasis (nonlethal cancer).
The exposure of interest was mRNA expression of all genes for cholesterol synthesis (Reactome cholesterol biosynthesis, M16227) and transport (Reactome ABCA transporters in lipid homeostasis, M524) as defined by Reactomes (20, 21). We added all further genes implicated in cholesterol regulation in prostate cancer from prior literature (regulators) (4–7).
The HPFS cohort was used as a discovery set to test whether tumor mRNA expression levels were associated with lethal cancer. In total, 43 genes were available in the gene expression array in this cohort (Supplementary Table S1). Univariable logistic regression models were fit for each gene and P-values for the tests for non-zero coefficients were adjusted using Bonferroni correction for 43 comparisons. Genes found significant after Bonferroni correction were tested for validation in PHS and SWWS.
In a secondary analysis, combining HPFS and PHS, we determined whether mRNA expression in normal adjacent prostate tissue was associated with lethal cancer in univariable logistic regression models with adjustment for multiple comparisons as above. Similarly, we determined if a difference or a ratio of gene expression in tumor versus in adjacent normal tissue was associated with lethal cancer; using a likelihood-ratio test, we also tested a nested model that included normal tissue in addition to tumor expression levels. We also assessed which cholesterol metabolism genes were differently expressed between tumor and normal tissue of the same patient, adjusting for multiple comparisons as above.
Subsequently, we further characterized the expression of the identified gene, squalene monooxygenase (SQLE) using regression models to estimates risks and 95% confidence intervals (95% CIs). Tests for trend across categories of SQLE expression were done by including ordinal category indices as continuous variables; the reference category was mean ± 1 standard deviation (SD). Multivariable models for lethal cancer with cohort-specific intercepts included known predictors for mortality after prostate cancer diagnosis. Given that Gleason grade, stage, prostate-specific antigen (PSA), and Ki-67 expression are likely intermediates between mRNA expression (exposure) and cause-specific mortality (outcome), these variables were not included in our pre-specified multivariable model to estimate causal effects. However, we did adjust a model for Gleason grade (5–6, 3+4, 4+3, 8, 9–10), stage (T1/T2, T3, T4/N1/M1), and PSA (<4, 4–10, ≥10 ng/ml) to evaluate SQLE as a prognostic marker in the combined HPFS and PHS cohorts. PSA at diagnosis was not measured in SWWS patients. Associations between SQLE expression and Gleason grade were assessed using linear regression.
Associations between statin use at diagnosis and hypercholesterolemia (binary exposures) and SQLE mRNA expression (outcome) were assessed using univariable linear regression. In HPFS, patients reported a diagnosis of hypercholesterolemia; in PHS, total serum cholesterol above 240 mg/dl was defined as high (22). For the latter cohort, also Pearson correlation coefficients r for the association of SQLE and total cholesterol levels were calculated. To test whether the association of SQLE expression with lethal disease differed according to statin use at diagnosis or presence of hypercholesterolemia, multiplicative interaction terms of these with SQLE expression were constructed. Associations between angiogenesis indices, Ki-67 (expressed as logarithm of the ratio of Ki-67-positive nuclei over the total number of tumor nuclei), and SQLE expression were assessed using Pearson correlation coefficients r and linear regression.
All analyses were conducted in Stata 12.1 (StataCorp LP, College Station, TX) and followed the REMARK protocol for oncologic biomarkers (23). The research protocol was approved by the institutional review board at Harvard T.H. Chan School of Public Health and Partners Healthcare.
Results
Participants
Characteristics of the three study populations are presented in Table 1. The 249, 153, and 338 men in the HPFS, PHS, and SWWS cohorts had median ages at diagnosis of 66, 66, and 73 years, respectively. All patients were Caucasians. Most patients in the HPFS and PHS cohorts were diagnosed in the PSA screening era, while SWWS patients were diagnosed incidentally without PSA screening.
Table 1.
Baseline characteristics of men with prostate cancer and transcriptome data in the Health Professionals Follow-up Study (HPFS), Physicians’ Health Study (PHS), and Swedish Watchful Waiting Study (SWWS).
| Cohort | HPFS | PHS | SWWS | |||||
|---|---|---|---|---|---|---|---|---|
| Lethal cancer | Nonlethal | Lethal cancer | Nonlethal | Lethal cancer | Nonlethal | |||
| n | 81 | 168 | 31 | 122 | 153 | 185 | ||
| Age at diagnosis, median (range) |
67 (47–80) | 66 (49–77) | 69 (58–81) | 65 (51–80) | 74 (54–91) | 73 (51–89) | ||
| Year of diagnosis, median (range) |
1992 (1986–2004) |
1996 (1986–2003) |
1993 (1983–2005) |
1994.5 (1982–2001) |
1991 (1977–1998) |
1992 (1977–1998) |
||
| Diagnosis before 1993, n (%) |
44 (54) | 33 (20) | 14 (45) | 44 (36) | No PSA testing | |||
| Stage, n (%) | ||||||||
| pT1a | 24 (16) | 64 (35) | ||||||
| pT1b | 129 (84) | 121 (65) | ||||||
| pT1/T2 | 27 (33) | 119 (71) | 13 (42) | 77 (63) | ||||
| pT3 | 37 (46) | 46 (27) | 9 (29) | 41 (34) | ||||
| pT4/N1/M1 | 17 (21) | 3 (2) | 9 (29) | 4 (3) | ||||
| Gleason, n (%) | ||||||||
| 5–6 | 1 (1) | 22 (13) | 0 (0) | 35 (29) | 18 (12) | 87 (47) | ||
| 7 (3+4) | 10 (12) | 80 (48) | 2 (6) | 46 (38) | 28 (18) | 52 (28) | ||
| 7 (4+3) | 28 (35) | 46 (27) | 8 (26) | 20 (16) | 35 (23) | 33 (18) | ||
| 8 | 11 (14) | 8 (5) | 7 (23) | 15 (12) | 12 (8) | 8 (4) | ||
| 9–10 | 31 (38) | 12 (7) | 14 (45) | 6 (5) | 60 (39) | 5 (3) | ||
| PSA, n (%)a | ||||||||
| <4 | 1 (2) | 17 (11) | 3 (19) | 12 (10) | ||||
| 4–10 | 28 (53) | 90 (58) | 8 (50) | 73 (63) | ||||
| ≥10 | 24 (45) | 49 (31) | 5 (31) | 30 (26) | ||||
| Tissue, n (%) | ||||||||
| TURPb | 15 (19) | 2 (1) | 12 (39) | 5 (4)b | 153 (100) | 185 (100) | ||
| Prostatectomy | 66 (81) | 166 (99) | 19 (61) | 117 (96) | 0 (0) | 0 (0) | ||
| BMI, mean (SD)c | 26.0 (3.4) | 25.3 (2.8) | 25.9 (2.9) | 24.9 (2.8) | 25.3 (3.2) | 26.0 (3.4) | ||
| Family history, n (%) |
26 (32) | 35 (21) | 2 (6) | 37 (30) | ||||
| Smoker, n (%)d | 15 (19) | 16 (10) | 2 (7) | 4 (3) | ||||
| Diabetes, n (%) | 4 (5) | 6 (4) | 1 (3) | 3 (2) | ||||
| High cholesterol, n (%) |
32 (40) | 68 (40) | 4 (13) | 11 (9) | ||||
| Statin use at diagnosis, n (%) |
8 (10) | 25 (15) | 2 (6) | 8 (7) | ||||
Prostate-specific antigen in nl/ml. Values were only available for a subset of patients.
Transurethral resection of the prostate. Includes one patient with tumor sample from a lymph node in PHS.
Body mass index, with standard deviation. In SWWS, BMI only available for 268 patients (120 cases and 148 controls).
Current tobacco use at diagnosis.
Cholesterol metabolism gene expression in normal and tumor tissue
Using HPFS as the discovery cohort, only tumor mRNA expression of squalene monooxygenase (SQLE, previously squalene epoxidase) was significantly associated with lethal prostate cancer after adjustment for multiple testing (Padjusted = 0.001). Associations with all other genes are shown in Supplementary Table S1.
For 204 patients (118 from HPFS and 86 from PHS), adjacent normal prostate tissue in addition to tumor areas was available for a secondary analysis of gene expression profiling. Ten genes (ABCA2, EBP, TM7SF2, DHCR24, ABCA1, HSD17B7, ABCA10, GGPS1, ABCA3, LSS) were significantly differentially expressed between tumor and normal tissue (all Padjusted ≤ 0.022). However, none of these differences in expression between tumor and normal tissue were associated with lethal cancer. Expression levels in normal tissue were not associated with lethal cancer (Padjusted = 1 for all genes, including SQLE).
Squalene monooxygenase expression and lethal prostate cancer
We aimed to validate the association of SQLE and lethal cancer in the PHS cohort, which also consisted of predominantly surgically treated men, as well as the initially untreated SWWS cohort. Higher SQLE expression was also associated with an increased risk of lethal cancer in PHS (Tab. 2). Combining the HPFS and PHS cohorts, men with a SQLE expression one SD or more above the mean had an 8.3 times higher odds of lethal cancer (95% CI, 3.5 to 19.7; Ptrend < 0.001) than men with the mean level of expression. In SWWS, SQLE expression had weaker associations with lethal cancer (odds ratio 1.23 for SQLE >1 SD v. mean; 95% CI, 0.66 to 2.29; Ptrend = 0.047; Tab. 2).
Table 2.
Odds ratios for lethal prostate cancer associated with squalene monooxygenase (SQLE) expression, by cohort.
| HPFS | PHS | SWWS | |
|---|---|---|---|
| Unadjusted | |||
| Linear, per 1 SD | 2.06 (1.49–2.86) | 2.32 (1.42–3.79) | 1.22 (0.98–1.52) |
| P | < 0.001 | 0.001 | 0.075 |
| < 1 SD | 0.75 (0.27–2.14) | 0.41 (0.05–3.31) | 0.48 (0.26–0.91) |
| Mean ± 1 SD | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| > 1 SD | 9.73 (3.47–27.3) | 3.91 (1.29–11.9) | 1.25 (0.69–2.28) |
| Ptrend | < 0.001 | 0.011 | 0.018 |
| Adjusteda | |||
| Linear, per 1 SD | 2.19 (1.48–3.24) | 2.51 (1.37–4.59) | 1.17 (0.93–1.47) |
| P | < 0.001 | 0.003 | 0.174 |
| < 1 SD | 0.91 (0.27–3.08) | 0.61 (0.06–5.95) | 0.54 (0.27–1.03) |
| Mean ± 1 SD | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| > 1 SD | 16.1 (3.99–65.0) | 3.57 (0.94–13.6) | 1.23 (0.66–2.29) |
| Ptrend | < 0.001 | 0.067 | 0.047 |
Adjusted for age (continuous), year of diagnosis (quintiles), body mass index (< 25, 25–30, >30), current smoking at diagnosis (binary), family history of prostate cancer in brother or father (binary), diabetes mellitus (binary), and statin use at diagnosis (binary) in HPFS and PHS. Adjusted for age and year of diagnosis (both continuous) in SWWS.
Abbreviations: HPFS, Health Professionals Follow-up Study. PHS, Physicians’ Health Study. Ref., reference category. SD, standard deviation. SWWS, Swedish Watchful Waiting Study.
Tumor tissue of lethal cancers in HPFS and PHS had on average an 8.1% higher SQLE expression (95% CI, 3.9 to 12.3%) than surrounding normal prostate tissue of the same patient. SQLE expression in normal prostate tissue did not carry prognostic information in addition to tumor expression (P = 0.43). However, in this within-patient comparison, a 10% higher SQLE expression in tumor than surrounding normal tissue was associated with 40% higher odds of lethal cancer (95% CI, 15 to 72%). Adjusting for Gleason grade, age, and year of diagnosis, this odds ratio was 1.29 (95% CI, 1.00 to 1.65; P = 0.047).
Higher tumor SQLE expression was associated with higher Gleason grades in all three cohorts (HPFS/PHS, Ptrend < 0.001; SWWS, Ptrend = 0.019; Supplementary Figure S1). SQLE expression was associated with lethal cancer independently from Gleason grade and stage in the combined HPFS and PHS cohorts (P < 0.001). The risk associated with SQLE did not differ significantly by Gleason grade (Pinteraction = 0.79). Further adjusting for Gleason grade, pathological stage, and PSA in the combined HPFS and PHS subsets that included PSA values (n = 289) to evaluate SQLE for prognostication, men with a SQLE expression 1 SD or more above the mean had a 9.0 times higher odds of lethal cancer (95% CI, 2.8 to 28.7) than men with the mean level of expression. All associations in HPFS and PHS were similar when analyses were restricted to men who received primary prostatectomy (92% of patients).
SQLE expression did not differ between users v. non-users of statins before prostate cancer diagnosis (P = 0.41). The association between SQLE expression and lethal cancer was not modified by statin use at diagnosis (Pinteraction ≥ 0.74 for both categorical and linear SQLE expression). Hypercholesterolemia was not associated with SQLE expression (P = 0.17); there was no correlation between total cholesterol levels and SQLE expression in PHS (r = 0.04; P = 0.70). Hypercholesterolemia was also not associated with lethal cancer and did not interact with the association of SQLE expression and lethal cancer.
Angiogenesis and cell proliferation
To study whether tumor angiogenesis could be driven by SQLE expression, we assessed angiogenesis histologically among 169 patients in the HPFS (46 lethal and 123 nonlethal cancers). Tumors with higher SQLE expression were more angiogenic, as evidenced by increased vessel irregularity (r = 0.25; P = 0.001) and decreased mean vessel diameter (r = −0.31; P < 0.001). SQLE expression (P = 0.013) and Gleason grade (P = 0.027) were independently associated with vessel irregularity. When adjusting for SQLE expression (P = 0.001), Gleason grade (P = 0.11) did not remain an independent predictor of vessel diameter.
SQLE expression was associated with cell proliferation as measured by Ki-67 positivity (r = 0.18; P = 0.004). However, Ki-67 expression was not associated with angiogenesis indices and adjustment for Ki-67 expression consequently did not alter their association with SQLE. SQLE was associated with lethal cancer (P < 0.001) independently from Ki-67 expression (P = 0.005).
Discussion
In this prospective investigation of prostate tissue gene expression within three well defined cohorts, we identified and validated a key aberration in the cholesterol metabolism of prostate cancer that is associated with lethal disease. Expression of SQLE, the second rate-limiting enzyme of cholesterol metabolism, was not elevated in prostate cancers in general compared to normal prostatic tissue but distinguished tumors at high risk of metastasis. Specifically, men with prostate tumors characterized by increased SQLE expression at diagnosis had a substantially increased risk of lethal cancer. Moreover, these tumors were more angiogenic. Not only absolute SQLE expression values across the cohorts but also a ratio of expression in tumor versus normal prostate specimen of the same patient was associated with lethal cancer; thus, the availability of normal tissue expression as an internal control may be advantageous for translation of this prognostic marker to clinical management. In a watchful waiting setting, SQLE was less tightly associated with lethal cancer. To our knowledge, this is the first study to characterize how the tumor cholesterol metabolism transcriptome is related to lethal disease in cancer patients.
Previous studies have focused on differences in cholesterol metabolism between normal prostate and tumor tissue (4, 5). One such gene, the transporter ABCA1, was also differently regulated in normal versus tumor tissue in our study, but this difference was not associated with lethal cancer. In a gene expression analysis, using data from heterogeneous clinical sources, a ratio of SQLE expression to a second gene (TPD52L2) was associated with tumor progression after prostatectomy (24). In early-stage breast cancer patients, high SQLE expression was associated with decreased distant-metastasis free survival (25).
In our study, 92% of men in HPFS and PHS underwent primary prostatectomy, and associations of SQLE with lethal disease in this subgroup were similar to the overall cohorts. A biomarker measured in the tumor and associated with lethal disease in men who underwent surgical removal of the prostate can be conceptualized as a biomarker of micrometastases; the mere 44% reduction in prostate cancer-specific mortality associated with prostatectomy in men with clinically diagnosed localized prostate cancer suggests micrometastases in this population are not uncommon (26). Given that we found SQLE expression to be a strong predictor of angiogenesis, a requirement for tumor metastasis and an established predictor of prostate cancer progression and mortality (18), increased angiogenesis may represent the mechanism linking high SQLE expression to lethal disease despite prostatectomy. In fact, SQLE more strongly predicted angiogenesis than Gleason score. Factors that indicate presence of micrometastases are conceivably less predictive of outcomes when studied in absence of primary prostatectomy, as untreated men who succumb to their disease also include men whose tumors do not yet have established micrometastases but have metastatic potential. In men with incidentally diagnosed, early-stage tumors managed expectantly in SWWS, SQLE was a weak predictor of lethal cancer. The possible causality of an association between cholesterol synthesis and angiogenesis is further supported by seminal xenograft studies of prostate and breast cancer in mice (27, 28). Angiogenesis was higher when serum cholesterol levels were experimentally increased and was decreased by the cholesterol uptake blocker ezetimibe. However, given the multifaceted roles of cholesterol in proliferating cells, higher cholesterol synthesis likely contributes to cancer progression through multiple, synergistic biological mechanisms. Furthermore, SQLE is also involved in the synthesis of sterols other than cholesterol, such as 24,25-epoxycholesterol (29).
Our analysis was neither a pharmacoepidemiologic study of statins nor powered to test whether men with increased SQLE expression had different outcomes based on statin use. In addition, the statin target 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) and SQLE are the two rate-limiting steps of cholesterol synthesis and as such likely regulated partly independently. Therefore, SQLE enzyme activity might be an important modifiable risk factor for prostate cancer outcomes. Pharmacologic SQLE inhibitors have not been studied clinically due to toxicity concerns (30). However, polyphenols from green tea (Camellia sinensis) have been identified as potent inhibitors of SQLE (31). The Japan Public Health Center-based Prospective Study demonstrated a dose-dependent association between green tea intake and risk of advanced prostate cancer (32).
This study has several strengths, including the analysis of three carefully designed prospective cohorts representing different biology, the systematic discovery approach, its extensive follow-up, and utilization of lethal cancer as primary endpoint. Future studies might address whether SQLE protein expression, for example using immunohistochemistry, is also associated with prostate cancer outcomes. However, there are several limitations. First, we cannot distinguish temporally whether high SQLE expression was cause or effect of cancer progression. Second, since our outcome was lethal cancer, we cannot exclude bias from competing risks from other causes of death. However, SQLE expression was not associated with overall survival in patients with indolent cancer in SWWS (data not shown). Third, the generalizability of this study may be limited as only white men were included; other racial groups have significant and mostly less well studied burden of prostate cancer.
In summary, our study described and validated a novel and potentially modifiable molecular risk factor for mortality from prostate cancer. Assessing SQLE expression may help identifying high-risk patients, especially since Gleason grading of prostate biopsies in clinical practice is less predictive of outcomes than the rigorously re-reviewed histology used in our study (15). Beyond justifying RCTs of statins in prostate cancer, our findings call for designing adjuvant therapy for high-risk patients that is targeted at tumor cholesterol synthesis.
Supplementary Material
Acknowledgments
Grant support
The Health Professionals Follow-up Study was supported by National Institutes of Health grants P01 CA055075 and UM1 CA167552. The Physicians' Health Study was supported by grants CA097193, CA34944, CA40360, HL26490, and HL34595. Research was further supported by the Dana-Farber/Harvard Cancer Center Specialized Programs of Research Excellence program in Prostate Cancer (5P50CA090381-08) and the National Cancer Institute (CA141298; T32 CA009001, T.A. Gerke; CA133891, E.L. Giovannucci). J.R. Rider, K.L. Penney, and L.A. Mucci are Prostate Cancer Foundation Young Investigators.
We are grateful to the participants and staff of the Physicians’ Health Study and Health Professionals Follow-Up Study for their valuable contributions. In addition, we would like to thank the following state cancer registries for their cooperation and assistance: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WY. The authors assume full responsibility for analyses and interpretation of these data.
Footnotes
Conflicts of interest: The authors disclose no potential conflicts of interest.
References
- 1.Pelton K, Freeman MR, Solomon KR. Cholesterol and prostate cancer. Curr Opin Pharmacol. 2012;12:751–759. doi: 10.1016/j.coph.2012.07.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.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;68:4447–4454. doi: 10.1158/0008-5472.CAN-08-0249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Platz EA, Clinton SK, Giovannucci E. Association between plasma cholesterol and prostate cancer in the PSA era. Int J Cancer. 2008;123:1693–1698. doi: 10.1002/ijc.23715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Leon CG, Locke JA, Adomat HH, Etinger SL, Twiddy AL, Neumann RD, et al. Alterations in cholesterol regulation contribute to the production of intratumoral androgens during progression to castration-resistant prostate cancer in a mouse xenograft model. Prostate. 2010;70:390–400. doi: 10.1002/pros.21072. [DOI] [PubMed] [Google Scholar]
- 5.Lee BH, Taylor MG, Robinet P, Smith JD, Schweitzer J, Sehayek E, et al. Dysregulation of cholesterol homeostasis in human prostate cancer through loss of ABCA1. Cancer Res. 2013;73:1211–1218. doi: 10.1158/0008-5472.CAN-12-3128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Twiddy AL, Cox ME, Wasan KM. Knockdown of scavenger receptor class B type I reduces prostate specific antigen secretion and viability of prostate cancer cells. Prostate. 2012;72:955–965. doi: 10.1002/pros.21499. [DOI] [PubMed] [Google Scholar]
- 7.Murtola TJ, Syvala H, Pennanen P, Blauer M, Solakivi T, Ylikomi T, et al. The importance of LDL and cholesterol metabolism for prostate epithelial cell growth. PLoS One. 2012;7:e39445. doi: 10.1371/journal.pone.0039445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Farwell WR, D'Avolio LW, Scranton RE, Lawler EV, Gaziano JM. Statins and prostate cancer diagnosis and grade in a veterans population. J Natl Cancer Inst. 2011;103:885–892. doi: 10.1093/jnci/djr108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Platz EA, Leitzmann MF, Visvanathan K, Rimm EB, Stampfer MJ, Willett WC, et al. Statin drugs and risk of advanced prostate cancer. J Natl Cancer Inst. 2006;98:1819–1825. doi: 10.1093/jnci/djj499. [DOI] [PubMed] [Google Scholar]
- 10.Yu O, Eberg M, Benayoun S, Aprikian A, Batist G, Suissa S, et al. Use of statins and the risk of death in patients with prostate cancer. J Clin Oncol. 2014;32:5–11. doi: 10.1200/JCO.2013.49.4757. [DOI] [PubMed] [Google Scholar]
- 11.Giovannucci E, Liu Y, Platz EA, Stampfer MJ, Willett WC. Risk factors for prostate cancer incidence and progression in the health professionals follow-up study. Int J Cancer. 2007;121:1571–1578. doi: 10.1002/ijc.22788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook NR, et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med. 1996;334:1145–1149. doi: 10.1056/NEJM199605023341801. [DOI] [PubMed] [Google Scholar]
- 13.Christen WG, Gaziano JM, Hennekens CH. Design of Physicians' Health Study II--a randomized trial of beta-carotene, vitamins E and C, and multivitamins, in prevention of cancer, cardiovascular disease, and eye disease, and review of results of completed trials. Ann Epidemiol. 2000;10:125–134. doi: 10.1016/s1047-2797(99)00042-3. [DOI] [PubMed] [Google Scholar]
- 14.Setlur SR, Mertz KD, Hoshida Y, Demichelis F, Lupien M, Perner S, et al. Estrogen-dependent signaling in a molecularly distinct subclass of aggressive prostate cancer. J Natl Cancer Inst. 2008;100:815–825. doi: 10.1093/jnci/djn150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Stark JR, Perner S, Stampfer MJ, Sinnott JA, Finn S, Eisenstein AS, et al. Gleason score and lethal prostate cancer: does 3 + 4 = 4 + 3? J Clin Oncol. 2009;27:3459–3464. doi: 10.1200/JCO.2008.20.4669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hoshida Y, Villanueva A, Kobayashi M, Peix J, Chiang DY, Camargo A, et al. Gene expression in fixed tissues and outcome in hepatocellular carcinoma. N Engl J Med. 2008;359:1995–2004. doi: 10.1056/NEJMoa0804525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Penney KL, Sinnott JA, Tyekucheva S, Gerke T, Shui IM, Kraft P, et al. Association of prostate cancer risk variants with gene expression in normal and tumor tissue. Cancer Epidemiol Biomarkers Prev. 2015;24:255–260. doi: 10.1158/1055-9965.EPI-14-0694-T. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mucci LA, Powolny A, Giovannucci E, Liao Z, Kenfield SA, Shen R, et al. Prospective study of prostate tumor angiogenesis and cancer-specific mortality in the health professionals follow-up study. J Clin Oncol. 2009;27:5627–5633. doi: 10.1200/JCO.2008.20.8876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rider JR, Fiorentino M, Kelly R, Gerke T, Jordahl K, Sinnott JA, et al. Tumor expression of adiponectin receptor 2 and lethal prostate cancer. Carcinogenesis. 2015;36:639–647. doi: 10.1093/carcin/bgv048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Croft D, O'Kelly G, Wu G, Haw R, Gillespie M, Matthews L, et al. Reactome: a database of reactions, pathways and biological processes. Nucleic Acids Res. 2011;39:D691–D697. doi: 10.1093/nar/gkq1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–15550. doi: 10.1073/pnas.0506580102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.ATPIII. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–3421. [PubMed] [Google Scholar]
- 23.McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM, et al. Reporting recommendations for tumor marker prognostic studies. J Clin Oncol. 2005;23:9067–9072. doi: 10.1200/JCO.2004.01.0454. [DOI] [PubMed] [Google Scholar]
- 24.Zhao H, Logothetis CJ, Gorlov IP. Usefulness of the top-scoring pairs of genes for prediction of prostate cancer progression. Prostate Cancer Prostatic Dis. 2010;13:252–259. doi: 10.1038/pcan.2010.9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Helms MW, Kemming D, Pospisil H, Vogt U, Buerger H, Korsching E, et al. Squalene epoxidase, located on chromosome 8q24.1, is upregulated in 8q+ breast cancer and indicates poor clinical outcome in stage I and II disease. Br J Cancer. 2008;99:774–780. doi: 10.1038/sj.bjc.6604556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bill-Axelson A, Holmberg L, Garmo H, Rider JR, Taari K, Busch C, et al. Radical prostatectomy or watchful waiting in early prostate cancer. N Engl J Med. 2014;370:932–942. doi: 10.1056/NEJMoa1311593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Solomon KR, Pelton K, Boucher K, Joo J, Tully C, Zurakowski D, et al. Ezetimibe is an inhibitor of tumor angiogenesis. Am J Pathol. 2009;174:1017–1026. doi: 10.2353/ajpath.2009.080551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pelton K, Coticchia CM, Curatolo AS, Schaffner CP, Zurakowski D, Solomon KR, et al. Hypercholesterolemia induces angiogenesis and accelerates growth of breast tumors in vivo. Am J Pathol. 2014;184:2099–2110. doi: 10.1016/j.ajpath.2014.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Nelson JA, Steckbeck SR, Spencer TA. Biosynthesis of 24,25-epoxycholesterol from squalene 2,3;22,23-dioxide. J Biol Chem. 1981;256:1067–1068. [PubMed] [Google Scholar]
- 30.Seiki S, Frishman WH. Pharmacologic inhibition of squalene synthase and other downstream enzymes of the cholesterol synthesis pathway: a new therapeutic approach to treatment of hypercholesterolemia. Cardiol Rev. 2009;17:70–76. doi: 10.1097/CRD.0b013e3181885905. [DOI] [PubMed] [Google Scholar]
- 31.Abe I, Seki T, Umehara K, Miyase T, Noguchi H, Sakakibara J, et al. Green tea polyphenols: novel and potent inhibitors of squalene epoxidase. Biochem Biophys Res Commun. 2000;268:767–771. doi: 10.1006/bbrc.2000.2217. [DOI] [PubMed] [Google Scholar]
- 32.Kurahashi N, Sasazuki S, Iwasaki M, Inoue M, Tsugane S, Group JS. Green tea consumption and prostate cancer risk in Japanese men: a prospective study. Am J Epidemiol. 2008;167:71–77. doi: 10.1093/aje/kwm249. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.