Results of a retrospective analysis, including data from a multi-institutional cohort of more than 1,200 men, suggest that obesity and race may be important modifiers in determining the association between diabetes and prostate cancer aggressiveness.
Durham, NC-Results of a retrospective analysis, including data from a multi-institutional cohort of more than 1,200 men, suggest that obesity and race may be important modifiers in determining the association between diabetes and prostate cancer aggressiveness.
In the recently published study, lead author Stephen J. Freedland, MD, and colleagues reported the findings from analyses exploring the association of diabetes and outcomes after radical prostatectomy (Cancer Epidemiol Biomarkers Prev 2010; 19:9-17). In the primary analysis, no correlations were found between diabetes and either risk of biochemical recurrence or PSA doubling time (PSADT). However, secondary analyses performed with patients stratified into subgroups by race and obesity showed diabetes was associated with a significantly increased risk of biochemical recurrence and a trend toward a shorter PSADT among obese Caucasian men.
Diabetes had no significant effects on the risk of biochemical recurrence in any other race-obesity subset and even showed a trend for improving that outcome as well as an association with a significantly longer PSADT.
"Although diabetes has been shown to be associated with a decreased risk of being diagnosed with prostate cancer, our findings agree with previous studies that have shown diabetes is not correlated with prostate cancer outcomes overall," said Dr. Freedland, associate professor of urology and pathology at Duke University, Durham, NC.
"However, we took the investigation of this issue further to see if there may be specific subgroups of men where diabetes may play a role, and we considered race and obesity, as both of these demographic features are related to both aggressive prostate cancer and diabetes. To our knowledge, ours is the first study to explore the contributions of race and body mass index to the interaction of diabetes and prostate cancer outcomes.
"The results of this initial study should be considered hypothesis-generating and require further investigation. While we believe diabetes along with race and obesity influence prostate cancer aggressiveness, the directions in which they do so are clearly complex, and a lot more research needs to be done."
The retrospective study included men from the Shared Equal-Access Regional Cancer Hospital Database who had undergone radical prostatectomy between 1998 and 2008 at one of four Veterans Affairs Medical Centers. Of the 1,262 men comprising the multi-institutional cohort, 233 (19%) were diabetic, 670 (53%) were Caucasian, and 368 (29%) were obese (BMI ≥30 kg/m2 ). Diabetes prevalence was found to be significantly higher among African-American men compared with Caucasian men and among obese versus non-obese men.
In multivariate analysis of how diabetes correlated with pathologic characteristics, diabetes was associated with significantly increased risks of high-grade disease and seminal vesicle invasion but was not associated with extracapsular extension or positive margins. There were no statistically significant interactions between diabetes and race or obesity for the association between diabetes and tumor characteristics.
High recurrence in obese Caucasians
During follow-up, biochemical recurrence occurred in 401 men (32%). While in multivariate analysis there was no significant association between diabetes and biochemical recurrence, in secondary analysis with men stratified by race, there was a trend for an increased risk of biochemical recurrence among Caucasian men with diabetes and a decreased risk among African-American men with diabetes.
With further stratification by obesity, obese Caucasian men with diabetes had a statistically significant, 2.5-fold increased risk of a biochemical recurrence compared with Caucasian obese men without diabetes. The risk of biochemical recurrence was decreased among non-obese Caucasians, non-obese African-Americans, and obese African-Americans with diabetes, although none of the associations was statistically significant.
"The chances for finding spurious statistically significant associations are increased when performing a multiple stratified analysis. We addressed this issue with rigorous statistical methods, although ours are initial findings requiring external validation," said Dr. Freedland.
Delineating associations between diabetes and prostate cancer risk and outcomes is of particular interest because diabetes is accompanied not only by increased glucose levels but also by decreases in insulin, testosterone, and in Caucasian men, decreased insulin-like growth factor-1 (IGF-1), Dr. Freedland noted.
"Identifying effects of diabetes on outcomes after radical prostatectomy may give us insight into the roles of insulin, IGF-1, and testosterone in driving prostate cancer growth," he said.
While it has been theorized that lower levels of testosterone and insulin may create a growth factor-poor environment that explains why diabetes is associated with a decreased risk of prostate cancer, Dr. Freedland and colleagues postulate that the specific chemical environment in obese Caucasian men with diabetes, which also includes low levels of IGF-1, may select for more aggressive tumors.
In order to further investigate interactions between diabetes and prostate cancer aggressiveness, Dr. Freedland and colleagues are currently conducting a prospective study of men undergoing biopsy for prostate cancer in which they will be analyzing tumor characteristics and looking for associations with various serum laboratory values and diabetes medication history.