Finasteride Use Appears Unlikely to Induce High-Grade Prostate Tumors

Roxanne Nelson

September 17, 2007 — The chemopreventive agent finasteride (Proscar, Merck & Co) does not appear to increase the rate of histologically high-grade prostate carcinomas, a possibility that had been suggested by a previous study showing also that the drug reduces the risk for prostate cancer. According to 2 new studies, published online in the September 11 issue of the Journal of the National Cancer Institute, the elevated rate of high-grade tumors observed in men who received finasteride in the Prostate Cancer Prevention Trial (PCPT) is most likely due to an increased rate of cancer detection.
The PCPT was the first phase 3 clinical trial to evaluate the chemoprevention of prostate cancer. Men who used finasteride had a lower incidence of prostate cancer, but among those who did develop the disease while taking finasteride, a higher incidence of high-grade tumors was observed. This finding raised the possibility that finasteride induces the development of a higher-grade cancer.
"The primary finding of the PCPT was that prostate cancer, overall, was reduced by 25% in men in the finasteride group of the study," said first author M. Scott Lucia, MD, of the University of Colorado Health Sciences Center in Denver. "A secondary finding from the trial was that there was a higher incidence of high-grade cancer in the finasteride group."
Dr. Lucia and colleagues looked at possible explanations for the latter finding, and histologically examined high-grade prostate biopsies with a Gleason score of 8 to 10 for hormonal effects; 90 specimens were from men treated with finasteride and 52 were from the placebo group. They also assessed biopsies with a Gleason score of 7 to 10 (n = 282 for finasteride; n = 244 for placebo) for pathologic surrogates of disease extent.
They found that prostate volumes were lower among men who had received finasteride, but degenerative hormonal changes seen in the high-grade biopsies were similar in both groups. In the finasteride group, pathologic surrogates for tumor extent were also lower with finasteride than with placebo. Among patients who had undergone a prostatectomy, the increased rate of high-grade disease that was initially observed with diagnostic needle biopsy in the finasteride group (42.7% for finasteride vs 25.4% for placebo) was less pronounced at prostatectomy (46.4% for finasteride vs 38.6% for placebo).
A larger proportion of the men who received finasteride were found to have high-grade disease at prostatectomy (69.7% vs 50.5%), but pathologic staging and the proportion of tumors that were low grade at biopsy and upgraded to high-grade cancer at prostatectomy were similar in the 2 groups.
"We found that there was no pathologic evidence that high-grade cancer was more advanced in the finasteride group, and there was no evidence that finasteride induced a hormonal artefact that would have interfered with proper grading of the cancers," Dr. Lucia told Medscape. "There is evidence that finasteride reduced gland volume in men taking the drug, such that high-grade cancer was better detected by biopsy."
The pathologic evidence indicates that it is unlikely that finasteride induces a high-grade cancer, although it cannot be completely ruled out. Hpwever,"the risk of high-grade cancer in men taking finasteride or other 5-alpha reductase inhibitors (5ARIs) is lower than originally believed," Dr. Lucia said. "Physicians should consider discussing prostate-cancer prevention using 5ARIs with their patients."
In the second study, Peter Gann, MD, ScD, director of the division of pathology research at the University of Illinois at Chicago, and colleagues hypothesized that the increase in high-grade tumors seen among finasteride users who participated in the PCPT was caused by the drug's ability to reduce prostate volume, thus allowing a greater detection of high-grade tumors because of sampling density. The researchers used 2 logistic models to assess the role of detection bias as a possible reason for the larger number of high-grade tumors observed in men who received finasteride.
The median prostate volume was 25% lower among patients receiving finasteride than among those receiving placebo, and the logistic model that was developed for the placebo group showed that detection of high-grade tumors would likely decrease as prostate volume increased. Using this model, it was predicted that a total of 239 high-grade tumors would be found in the finasteride group. The actual number of high-grade cancers detected was 243, and the difference was not statistically significant.
"The analyses presented here show that adjustment for changes in gland volume due to the drug could account for essentially all of the observed association between finasteride assignment and high-grade cancer," write the researchers.
In an accompanying editorial, Gerald Andriole, MD, of Washington University School of Medicine in St. Louis, Missouri, and colleagues point out that the results of these 2 studies strengthen the evidence that the increased proportion of high-grade prostate tumors observed in the PCPT is not likely to be clinically relevant.
"In the future, molecular testing of these PCPT tissue samples may be useful [in determining] whether the high-grade carcinomas seen in the finasteride arm have DNA ploidy and gene-expression profiles similar to those of untreated high-grade carcinoma," they write. "Ultimately, analysis of radical prostatectomy specimens and long-term clinical follow-up of the finasteride-treated patients will be necessary to completely ascertain the clinical significance of the increased high-grade cancer on biopsy in the PCPT."
Dr. Lucia's study was supported by National Cancer Institute Public Health Service grants through the Southwest Oncology Group. Dr. Lucia has been a scientific consultant for GlaxoSmithKline. All of Dr. Gann's coauthors except for one are paid employees of Merck & Co Inc, which manufactures finasteride.
J Natl Cancer Inst. 2007;99:1375-1383, 1366-1374, 1355-1356.

Finasteride (Proscar) Associated with Boost in Detection of High-Grade Prostate Cancer

DENVER, Sept. 13 -- The increase in high-grade prostate cancer seen among men taking finasteride (Proscar) is probably caused by increased detection, two groups of researchers said.
Although a seven-year study reported in 2003 showed that finasteride reduced the occurrence of prostate cancer by 24.8%, compared with placebo, it also showed that finasteride patients had a significant increase in the rate of lesions with Gleason scores of 7 or higher.
Now the researchers have reported online in the Journal of the National Cancer Institute that the finding of high-grade lesions, which alarmed many physicians and patients, was mainly a result of better detection among finasteride patients.
Both studies used data from the original study, expanded to include several hundred participants whose outcome status wasn't known at the time of the original report.
According to M. Scott Lucia, M.D., of the University of Colorado Health Sciences Center and colleagues, degenerative hormonal changes in high-grade biopsies were equivalent between the groups, but prostate volumes were significantly smaller (at P<0.001) in the finasteride group - 25.1 cc versus 34.4.
The smaller volume would make detection of high-grade tumor cells more likely on biopsy. Dr. Lucia and colleagues said.
Also, they noted that among finasteride patients who had a prostatectomy during the study, the rate of high-grade disease - which was significantly elevated at biopsy compared to those taking placebo - was equivalent to placebo when tumors were graded after surgery.
Although it's not possible to rule out the alarming possibility that finasteride induces high-grade tumors, Dr. Lucia and colleagues said, their results "suggest that high-grade cancer was detected earlier and was less extensive in the finasteride group than in the placebo group."
A similar conclusion was reached by Peter Gann, M.D., Sc.D., of the University of Illinois at Chicago and colleagues with Merck, which makes the drug.
Dr. Gann and colleagues performed an "observational re-analysis" of the data from the original study, asking whether the likelihood of detection of high-grade disease at biopsy varied depending on the size of the gland.
In fact, they said, a logistic model based on the placebo group showed that the likelihood of finding high-grade cancer decreased as volume increased.
Specifically, for each increase of 10 cc, the chance of detecting high-grade cancer by biopsy fell by 19%. The odds ratio was 0.81, with a 95% confidence interval from 0.74 to 0.90.
On the basis of the model, Dr. Gann and colleagues said 239 high-grade lesions would be expected among the finasteride patients - not significantly different from the 243 that were observed.
When prostate volume was used as a covariate, the odds ratio for high-grade cancer in the finasteride group, compared to placebo, fell from a significant 1.27 to a non-significant 1.03, they found.
Both groups of researchers cautioned that their conclusions are based on post hoc analyses of the data, and should be evaluated carefully. Dr. Gann and colleagues said their findings "must be interpreted cautiously and (do) not allow us to recommend definitive changes in clinical practice."
Nevertheless, the studies taken together seem to show that the increased rate of high-grade cancer "is not likely to be clinically relevant," commented Gerald Andriole, M.D., of Washington University School of Medicine in St. Louis, and colleagues.
Writing in an accompanying editorial, they noted that despite the "impressive" drop in cancer occurrence among finasteride patients, the drug "has not been embraced in clinical practice."
That was largely the result of disquiet about the possibility that the drug was somehow inducing high-grade cancer, they said - a worry that still can't be completely dispelled despite the "substantial reassurance" provided by the two new studies.
Dr. Andriole and colleagues suggested that the results of a study of a sister drug to finasteride - dutasteride - are due in the next few years and should provide better information on the safety of this class of medications.
All authors of the Chicago study except Dr. Gann were paid employees of Merck, which makes finasteride. The Denver study was supported by the National Cancer Institute and Merck. Dr. Lucia has been a scientific consultant for GlaxoSmithKline, and coauthors Michael Kattan, Ph.D., and Jonathan Epstein, M.D., have been consultants for Merck. Primary source: Journal of the National Cancer InstituteSource reference: Cohen YC et al. "Detection Bias Due to the Effect of Finasteride on Prostate Volume: A Modeling Approach for Analysis of the Prostate Cancer Prevention Trial." J Natl Cancer Inst 2007; 99: 1366-74. Additional source: Journal of the National Cancer Institute Source reference: Lucia MS et al. "Finasteride and High-Grade Prostate Cancer in the Prostate Cancer Prevention Trial." J Natl Cancer Inst 2007; 99: 1375-83. Additional source: Journal of the National Cancer Institute
Source reference: Andriole GL et al. "High-Grade Prostate Cancer in the Prostate Cancer Prevention Trial: Fact or Artifact." J Natl Cancer Inst 2007; 99: 1355-56.