Prostate Biopsy Samples Show Evidence of Differential Gene Expression in Patients After Supplement Use

Earlier observational studies had reported that selenium, vitamin E, or both, might reduce the risk of prostate cancer.

To determine whether gene expression patterns change in patients after supplement use, Jeri Kim, M.D., of the University of Texas MD Anderson Cancer Center in Houston, and colleagues enrolled 39 patients in a phase IIA trial and randomly assigned them to receive selenium, vitamin E, both supplements, or placebo for three to six weeks before prostatectomy. Researchers then isolated and studied tumor cells, stromal cells, and normal cells from biopsy samples removed from the prostatectomy specimens.

The investigators found that gene expression patterns differed between the treatment groups. Expression of several molecular pathways associated with cancer were altered in the supplement groups compared with the placebo group.

"We have demonstrated the feasibility and efficiency of the preoperative model and its power as a hypothesis generating engine," the authors conclude.

In an accompanying editorial, Eric A. Klein, M.D., of the Glickman Urological and Kidney Institute in Ohio, reviewed the history of selenium and vitamin E in prostate cancer chemoprevention efforts and put the new data into context. "…the study is noteworthy for demonstrating that even short-term exposure (i.e., 3 – 6 weeks) to these agents can affect expression of a majority of the genes interrogated and, in the robust demonstration of the utility of the preprostatectomy model, for deriving information on modulation of biomarkers…"

These data are consistent with data from past observational studies. However, two large randomized trials failed to show statistically significant benefit from the use of vitamin E and/or selenium in the prevention of prostate or other cancers or cardiovascular disease. The lack of consistency between epidemiological and in vitro studies and randomized trials suggest that randomized controlled trials "do not always validate what we believe biology indicates and that our model systems are imperfect measures of clinical outcomes in the real world," the editorialist concludes.