Height, Obesity, and the Risk of TMPRSS2:ERG-Defined Prostate Cancer

Abstract

The largest molecular subtype of primary prostate cancer is defined by the gene fusion. Few studies, however, have investigated etiologic differences by status. Because the fusion is hormone-regulated and a man's hormonal milieu varies by height and obesity status, we hypothesized that both may be differentially associated with risk of -defined disease. Our study included 49,372 men from the prospective Health Professionals Follow-up Study. Participants reported height and weight at baseline in 1986 and updated weight biennially thereafter through 2009. Tumor ERG protein expression (a marker) was immunohistochemically assessed. We used multivariable competing risks models to calculate HRs and 95% confidence intervals (CIs) for the risk of ERG-positive and ERG-negative prostate cancer. During 23 years of follow-up, we identified 5,847 incident prostate cancers, among which 913 were ERG-assayed. Taller height was associated with an increased risk of ERG-positive disease only [per 5 inches HR 1.24; 95% confidence interval (CI), 1.03-1.50; = 0.07]. Higher body mass index (BMI) at baseline (per 5 kg/m HR 0.75; 95% CI, 0.61-0.91; = 0.02) and updated BMI over time (per 5 kg/m HR 0.86; 95% CI, 0.74-1.00; = 0.07) were associated with a reduced risk of ERG-positive disease only. Our results indicate that anthropometrics may be uniquely associated with -positive prostate cancer; taller height may be associated with greater risk, whereas obesity may be associated with lower risk. Our study provides strong rationale for further investigations of other prostate cancer risk factors that may be distinctly associated with subtypes. .

Family history of prostate cancer and the incidence of ERG‐ and phosphatase and tensin homolog‐defined prostate cancer

“…Men assayed for a molecular marker were more likely to be diagnosed in earlier years, to be younger at diagnosis, and to have more information regarding their clinical characteristics. Previous studies from our group suggest that statistically accounting for differences between cases with and without molecular marker information does not materially change results …”

“…Lifestyle may also play a role in the development of PCa defined by TMPRSS2:ERG . For example, lycopene consumption from tomato products has been shown to be inversely associated with ERG‐positive but not ERG‐negative disease, and two studies have shown obesity to be associated with a reduced risk of developing TMPRSS2:ERG ‐positive PCa . Given that families often share lifestyle exposures, the association between a family history of PCa and molecular subtypes of PCa could operate through environment.…”

“…It is also reassuring that the prevalence of family history was similar among those with and without tissue biomarker data. In addition, two previous studies showed that the use of inverse probability weighting to account for differences between men with and without ERG status available did not materially change results . ERG and PTEN status were not available for family members diagnosed with PCa, as these data are not collected in clinical practice.…”

“…In the most common known molecular subtype (i.e., roughly half of primary PCa), the oncogene ERG fuses with androgen‐regulated promoter genes, most often TMPRSS2 . Although ERG status is unlikely prognostic by itself, several studies have shown that various risk factors are differentially associated with ERG‐defined PCa . In particular, preliminary evidence suggests that there are distinct inherited genetic factors associated with the risk of ERG‐positive versus ERG‐negative disease, lending plausibility to the hypothesis that family history could be differentially associated with the risk of ERG‐defined disease.…”

“…6 Although ERG status is unlikely prognostic by itself, 7 several studies have shown that various risk factors are differentially associated with ERG-defined PCa. [8][9][10][11][12][13][14][15][16] In particular, preliminary evidence suggests that there are distinct inherited genetic factors associated with the risk of ERG-positive versus ERG-negative disease, [8][9][10] lending plausibility to the hypothesis that family history could be differentially associated with the risk of ERG-defined disease.…”