PCPT data on finasteride, high-grade prostate cancer debunked

"If it's true that we improved detection by using finasteride, then the benefits are that we are picking up higher-grade tumors at an earlier stage when they are curable," Dr. Lucia said.

The PCPT randomized more than 18,000 men to receive either finasteride, 5 mg daily, or placebo over a 7-year period. The goal was to assess the ability of finasteride to reduce the incidence of prostate cancer during the evaluation period.

"Three years ago, we learned finasteride could reduce the rate of prostate cancer by 25%, but the concern that it induced high-grade cancer overshadowed the results," said E. David Crawford, MD, professor of surgery/radiation oncology and head of the section of urologic oncology, University of Colorado Health Sciences Center. "It has taken a few years, but we now know that concern was unfounded."

For the analysis, three expert pathologists convened for a blinded review of high-grade tumors from both the finasteride and placebo groups. They evaluated the slides and graded them for individual criteria associated with hormone effects. When the review was unblinded, they realized that no criteria could distinguish tumors in the placebo group tumor from those in the finasteride group. They concluded that the hormone effect could occur in isolated cases, but that the effect wasn't large enough to explain the differences between the two arms.

Coincident with that analysis, Dr. Lucia reviewed the same biopsies for pathologic prognostic factors, such as the number of biopsy cores that were positive for cancer and tumor length. He hypothesized that if finasteride were potentiating tumors, more cores would test positive for cancer, and tumor size would increase with the length of time the men were exposed to finasteride. But the data didn't support that hypothesis. Instead, the cancers were smaller in the finasteride arm compared with those in the placebo arm. Epidemiologic data also showed that the hazard ratio for high-grade disease did not increase in proportion to the length of time the men were on the drug.

Dr. Lucia and his colleagues then examined more than 500 prostatectomy specimens from patients diagnosed with prostate cancer during the study. Prostate volume was reduced by more than 25% in men in the finasteride arm, compared with men in the placebo arm. Upon further examination, researchers determined that the sensitivity of needle biopsy for detecting high-grade prostate cancer was 20% higher among specimens from the finasteride arm than it was in specimens from the placebo arm.

"When you stick needles into a smaller prostate, not only are you more likely to hit tumor, but if there is a minor high-grade component, you are more likely to pick it up," Dr. Lucia explained.

The final analysis examined the performance of PSA as a marker for the diagnosis of high-grade cancer. The area under the receiver operating characteristic curve-the sensitivity and specificity-was higher in the finasteride group than it was in the placebo group.

"The bottom line is, we haven't proven that the possibility that finasteride potentiated growth of high-grade tumor is absolutely false," said Dr. Lucia. "All we've done is added some supportive data of a possible detection bias with finasteride that individuals can look at and make up their own minds."

Dr. Lucia has received grants or research support from GlaxoSmithKline.