Metabolic Syndrome Is Associated with Aggressive Prostate Cancer Risk Regardless of Race

Purpose A recent meta-analysis suggested a link between Metabolic Syndrome (MS) and high-grade prostate cancer (PC), though few black men were included. We tested the link between MS and PC risk in a population of black and white men undergoing prostate biopsy. We hypothesized MS would be linked with aggressive PC, regardless of race. Among men undergoing prostate biopsy at the Durham Hospital, we abstracted history of or treatment for hypertension ( ≥ 130/85 mmHg), dyslipidemia (HDL<40 mg/dL), hypertriglyceridemia ( ≥ 150 mg/dL), diabetes/impaired fasting glucose (fasting glucose ≥ 100ml/dL), and obesity (waist circumference ≥ 40 inches) in the year prior to biopsy. Biopsy grade group (GG) was categorized as low (GG1) or high-grade (GG2-5). Multinomial logistic regression was used to examine MS (3-5 components) vs. no MS (0-2 components) and risk of high-grade and low-grade vs. no PC adjusting for key confounders. Interactions between race and MS were tested. an of group 2-5). between (categorical, 0-2 vs. yes: 3-5 components), and clinical variables chi-squared categorical variables and Wilcoxon rank sum for continuous variables. Variables included age (continuous), year (continuous), race (black vs. non-black), baseline PSA (continuous), DRE (normal vs. abnormal) and pre-study prostate volume (continuous). In an exploratory analysis, examined by both MS and race using chi-squared or Kruskal-Wallis tests.


Introduction
Metabolic Syndrome (MS) prevalence is increasing and is a serious public health problem worldwide. [1,2] In the US, MS affects more than one-third of adults. [1] MS consists of a diagnosis of at least three of the following ve risk factors: increased abdominal obesity, dyslipidemia (high triglycerides and high HDL cholesterol), hypertension, and abnormal glycemic status. [3] Given obesity, a MS component, is linked with multiple cancers, [4] it is plausible MS may also be associated with cancer development and progression. Indeed, MS has been linked with numerous cancers, including increased PC risk. [5] A recent meta-analysis of 130,000 men found that MS is associated with increased PC incidence, [6] though the overall increased risk was modest. However, not all individual studies were consistent with some nding MS was associated with lower PC risk [7,8] Nonetheless, the meta-analysis found MS was strongly linked with high-grade PC (OR=1.89, p<0.0001). [6] Importantly, nearly all studies evaluating MS and PC risk focused on White, European, or Asian men. [9][10][11] Whether results differ by race and speci cally whether MS is linked with PC among black men, who have one of the highest PC incidence and mortality rates in the world, is limited. [12] To date, few studies examined the link between MS and PC in Black men. [12,13] To evaluate the link between MS and PC risk across different races, we analyzed a case-control study of black and non-black (predominantly white) men undergoing prostate biopsies at the Durham Veterans Affairs (VA) Medical Center (DVAMC) in Durham, North Carolina. We choose an equal access healthcare setting to minimize potential effects that access to care may have on our results. Given obesity, which is one of the components of MS, is linked with lower PSA, [14] but larger prostate sizes, [15] it is noteworthy that all men in this biopsy cohort had data available on PSA and prostate volume to adjust for in the analyses to account for potential differences between men with and without MS.
Furthermore, obesity and insulin resistance have been recognized as the leading contributors in MS, which in turn is associated with the production of various pro-in ammatory cytokines which may promote genomic instability and a greater risk of cancer development. [16] Likewise, in the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) clinical trial, high cholesterol levels, another MS component, were associated with high-grade PC but not overall nor low-risk PC. [17] Based upon the recent meta-analysis [6] and our understanding we were unlikely to be powered to detect a very small increase in total PC incidence, we a priori hypothesized that MS would be linked with aggressive PC regardless of race, but not overall PC incidence.

Study Design
Data were obtained from an ongoing case-control study of veterans undergoing prostate biopsy for concerns about PC at the DVAMC. The study was approved by the institutional review board and written informed consent was obtained from all subjects. Subjects were recruited between January 2007 and July 2018 from the urology clinic. Eligible subjects were men with no prior PC history undergoing a prostate biopsy because of abnormal PSA and/or suspicious digital rectal exam (DRE) as clinically indicated. For men with multiple biopsies performed at the DVAMC, we only used data from their initial biopsy at the DVAMC. Of 2279 eligible subjects, 1322 consented to participate (58% response rate). We excluded 249 subjects due to missing PSA, DRE, prostate volume, MS data, or grade group, and 22 active surveillance biopsies as these subjects would have been diagnosed with PC, resulting in 1051 men.
MS was de ned as three or more of the following: dyslipidemia, hypertriglyceridemia, diabetes/impaired fasting glucose, hypertension, and abdominal obesity. Dyslipidemia was de ned as HDL <40 mg/dL in the year prior to biopsy. Hypertriglyceridemia was de ned as triglycerides ≥150 mg/dL or prescription drugs to treat high triglycerides in the year prior to biopsy. Diabetes was de ned as fasting glucose >100ml/dL, prescription of an anti-diabetic agent, or diagnosis of diabetes in the medical record in the year prior to biopsy. Hypertension was de ned as two blood pressure measurements where systolic blood pressure >130 mmHg or diastolic blood pressure >85 mmHg within the year prior to biopsy. Waist circumference was measured by trained personnel and obesity was de ned as waist circumference ≥40 inches. Race was selfreported as black, white, Asian/Paci c Islander, or American Indian/Alaska Native. As <1% of patients were races other than black or white, race was grouped as black or non-black with 99% of the non-black men being white. Prostate volume was measured at biopsy via transrectal ultrasound. Biopsy grade was categorized as low-grade (grade group 1) or high-grade (grade group 2-5).
We tested the association between MS (categorical, no: 0-2 vs. yes: 3-5 components), demographic and clinical variables using chi-squared for categorical variables and Wilcoxon rank sum for continuous variables. Variables included age (continuous), year (continuous), race (black vs. non-black), baseline PSA (continuous), DRE (normal vs. abnormal) and pre-study prostate volume (continuous). In an exploratory analysis, we examined variables by both MS and race using chi-squared or Kruskal-Wallis tests.
Logistic regression was used to examine the association between MS and PC risk versus no PC. Multinomial logistic regression was used to examine the association between MS and risk of low-grade PC (GG 1) versus no PC and high-grade PC (GG 2-5) versus no PC. The number of MS components was also tested in the above models as a continuous variable. Multivariable models were adjusted for age, year of consent, race, log-transformed baseline PSA, DRE, and log-transformed prostate volume. The interaction between race and MS in adjusted analyses was also tested and models were strati ed by race. In secondary analyses, the association between each individual MS component and PC risk and grade was examined in separate univariable and multivariable models, and interactions were tested between race and each component. As a sensitivity analysis, we then repeated all analyses de ning low-grade PC as GG 1-2 and high-grade PC as GG 2-5.
Signi cance was de ned as P<0.05. All analyses were performed using SAS 9.4 (SAS Institute; Cary, NC).

Baseline characteristics
Among 1051 men, 532 (51%) had MS and 519 (49%) did not. Having MS was associated with older age, non-black race, and larger prostate volume (Table 1, all p≤0.011). There was no association between MS with PSA or DRE.
Black men were younger at biopsy, had higher PSA, fewer suspicious DREs, and higher rates of overall cancer (all p≤0.015; Supplementary Table 1). Non-black men with MS had the largest prostate volumes, while black men without MS had the lowest prostate volumes (p<0.001).

MS and PC outcomes
While MS was not associated with overall PC (52% vs. 55%, p=0.27), MS was associated with PC grade (p=0.005, Table 1

Discussion
MS and PC are both common problems. [2] Despite a recent meta-analysis nding MS was associated with increased PC risk, [6] individual studies are mixed. [7][8][9][10][11] Importantly, the association between MS and PC risk in black men has not been well studied. Given black men are at increased risk of some MS components [18,19] such as hypertension, obesity, diabetes, and aggressive PC, [20] understanding whether race modi es this association is of utmost importance. We tested the link between MS and PC risk in a black and non-black case-control study of men undergoing prostate biopsy. In the REDUCE study, where most of the participants were Caucasians, all men underwent mandated biopsies regardless of PSA levels, MS was associated with increased risk of high-grade PC but not with overall or low-grade PC, [21] therefore, we hypothesized that MS would not be linked with overall PC risk but would be linked with aggressive PC and results would be similar in both black and non-black men.
Consistent with our hypothesis, MS and individual components were not associated with overall PC risk regardless of race. While on the surface these results differ from the meta-analysis and several papers that found a positive association between MS and total PC risk, [6,9,11] it is noteworthy that in the present study the OR=1.17, identical to the OR=1.17 (95%CI 1.00-1.36) from the meta-analysis. [6] Thus, while our ndings for PC risk were not signi cant, they are consistent with modest increased risk and in-line with prior studies. Given the modest link between MS and overall PC risk, this association is of unclear clinical signi cance.
Given practice trends away from trying to diagnose low-grade PC, [22] it is important to also focus on high-grade PC. On unadjusted analyses, there was no association between MS and high-grade PC. However, after adjusting for clinical characteristics, MS was signi cantly linked with nearly twofold increased odds of high-grade PC. Our results were nearly identical to the recent meta-analysis, [6] that found an OR=1.89 (95%CI 1.50-2.38, p<0.0001) for high-grade PC. Likewise, other large studies have also found a link between MS and high-grade PC. [9,21] When we examined individual MS components separately, each component was modestly linked with increased PC risk, but none reached signi cance.
Thus, there was no single MS component that drove the association with high-grade PC, but rather the greater number of components, the greater the risk. Collectively, our results and those of prior studies suggest that MS is a signi cant risk factor for high-grade PC. [9,21] A major gap in the literature, however, is that nearly all prior studies evaluating MS and PC risk focused on either white or Asian men. [10,11] The association between MS and PC in black men is under-studied. [12] This is a major unmet need as black men are at increased risk for some MS components, [18,19] and for aggressive PC. [20,23] We found no evidence for an interaction between MS and race, suggesting results apply equally to black and non-black men, regardless of how we de ned high-grade disease. One prior study examined interactions between race and MS for PC risk and found that MS was suggestively predictive of increased PC risk in black men, but not in white men. [12] However, this study only included 378 black men (vs. 613 in the current) and included controls that did not undergo biopsy. Moreover, no formal interaction testing was done in that study, and thus, whether the results truly differed by race is unknown. Clinically, our results are important as the known risk factors for PC are family history, age, and race. Our data support the hypothesis that regardless of race, MS should be considered a risk factor for high-grade PC. Ultimately, whether managing MS via drugs (statins, metformin, etc.) or lifestyle interventions can reduce the risk of high-grade PC warrants further study. [24,25] While the biological mechanisms why MS is differentially associated with high-grade PC are not clear, one possible mechanism is the MSassociated pro-in ammatory state. [26] Obesity-linked in ammation leads to altered metabolic signaling with activation of cytokines, imbalance between adipokines along with insulin growth factor-1 (IGF-1) axis expression. [27] This environment induces the production of reactive oxygen and free radicals with subsequent genomic alterations as DNA breaks promoting the transmembrane protease, serine 2, erythroblast transformation-speci c-related gene (TMPRSS2: ERG) gene fusion [28] present in 50% of PCs. [29] Intriguingly, in tumors with TMPRSS2: ERG gene fusion, obesity is a particularly strong PC risk factor. [30] Unfortunately, as TMPRSS2: ERG gene fusion status is not known in this study, we are unable to verify this. While the TMPRSS2: ERG gene fusion is less common in black men [31] and we found no interaction by race, the strength of the association between MS and high-grade PC was slightly weaker in black men. Whether this association is indeed weaker in black men certainly requires further study, but if true, it is intriguing to speculate that this may, in part, relate to less TMPRSS2: ERG gene fusion positive tumors in black men.
Another MS component, dyslipidemia, appears to be related to PC, [32] with several mechanisms proposed, including accumulation of cholesterol in PC cells membranes allowing pro-carcinogenic cell signaling. [33] Consistent with MS being selectively linked with high-grade PC, we showed in the REDUCE study that high cholesterol was unrelated to overall PC risk, but was associated with high-grade PC. [32] Our study builds on prior studies supporting the need for PC prevention clinical trials to modify MS factors to reduce high-grade PC including both white and black men.
Our study has some limitations. First, our MS de nition may differ from the rigorous NCP ATP III de nition [3] but is consistent with classic MS de nitions in prior studies. [6] Second, our observed null associations for overall PC risk may due to modest sample size and we cannot exclude a modest association between MS and overall PC risk. Third, our ndings were based on Veterans in the VA system; whether results apply to men outside the VA system requires further study, though our results were consistent with the meta-analysis [6] which included nearly only men outside the VA system, suggesting that our results do apply outside the VA. Fourth, we lacked data on other key factors associated with MS such as glycemic control, distribution of diabetes type 1 vs type 2, diabetes duration, testosterone, in ammatory markers, physical activity, and diet, known confounders of MS [2] possibly attenuating the ORs in our results. In addition, the de nition of PC aggressiveness was grade on biopsy. Though this correlates with long-term PC progression risk, [34] future studies with alternatives de nitions (i.e., metastases, stage, and PC death) are recommended. Lastly, the study subjects were men referred for a biopsy due to elevated PSA, abnormal digital exam or both. Therefore, the prevalence of the exposure of MS in the "healthy" control may be different and this impact remains to be determined. Notwithstanding these limitations, our key strength was the use of a contemporary prospective collection of data from men all undergoing biopsy.

Conclusion
In our study of black and white men undergoing prostate biopsy, we found MS was not associated with overall PC risk but was linked to highgrade PC. Importantly, results were similar in black and non-black men. Given the strength of the data linking MS and high-grade PC, future studies should test whether interventions including lifestyle modi cations can reduce the impact of MS on high-grade PC, and merits further investigation in large diverse multiethnic populations.