With cancer currently the second leading cause of death in the United States, cancer prevention is a top research priority. A limitation of the conduct of cancer prevention trials is that they are very costly as they require a large cohort of participants and several years of follow-up to adequately assess primary outcomes. However, the use of secondary data sources to augment clinical trial data collection could alleviate some of the cost burden and substantially increase follow-up time after trials end. In a recently published study in the Journal of the National Cancer Institute, Dr. Joseph Unger and researchers from the Public Health Sciences Division utilized Medicare claims to facilitate the assessment of long-term prostate cancer risk in participants of the Prostate Cancer Prevention Trial (PCPT). “These secondary data sources are emerging as a new paradigm for long-term follow up for cancer clinical trials. It’s an exciting new avenue of research,” said Dr. Unger.
Prostate cancer is common in men and it is estimated that approximately one in seven will be diagnosed with it in his lifetime. The PCPT was a randomized, placebo-controlled trial designed to test whether finasteride, a drug prescribed for conditions such as an enlarged prostate and male pattern hair loss, reduces the risk of developing prostate cancer. The trial began in 1993 and over the course of the study a total of 18,880 men 55 years of age or older were randomized to placebo or finasteride daily for up to seven years.
The original results from the PCPT revealed that the relative risk for prostate cancer was reduced by 25% in men on the finasteride intervention arm. This result was based on analysis of seven years of follow-up from the time of randomization. However, analyses of the original clinical trial data alone could not address a number of related questions, such as whether the protective effect of finasteride endures long term or whether it only delays the onset of prostate cancer rather than prevents it.
In the new study, the authors assessed long-term development of prostate cancer in PCPT participants beyond the seven years of follow-up that formed the basis of the original trial. Dr. Unger explained the methodology for the new analyses, “We used an innovative linkage between the clinical records of the PCPT and Medicare claims data to examine outcomes for trial participants many years after the original clinical trial had ended.” More than 14,000 participants (75% of both placebo and finasteride arms) had Medicare claims that met study criteria for linkage to the trial records. By combining the two data resources, the authors were able to examine prostate cancer diagnoses for a median of 16 years of participant follow-up from the time of randomization (see figure).
Over the 16 years of follow-up, a total of 3,244 prostate cancer cases were diagnosed with more in the placebo arm (1,805) than in the finasteride arm (1,439). Thus, the cumulative incidence was 22.3% for the placebo group and 18.1% for finasteride group over the extended period. Additional analyses revealed that once protocol-directed finasteride use was stopped at seven years, the protective effect largely did not diminish over time. This is supported by the finding that overall, participants on the finasteride arm had a relative reduction in the development of prostate cancer of 21.1% over the 16-year window. However, there was also no additional benefit of finasteride beyond the seven years of use.
Dr. Unger indicated that the results suggest that “seven years of finasteride can reduce prostate cancer diagnoses over a much longer period than was previously known.” This study also demonstrates an advantage of incorporating health indices obtained from secondary data sources into datasets from trials, which “can enhance the ability to detect outcomes over the long term from prospective clinical studies,” noted Dr. Unger. This is supported by the current study, in which 42% more cases of prostate cancer were identified by the use of the Medicare claims. This result also emphasizes the effectiveness and impact of using Medicare claims specifically in the study of diseases of older individuals.
The combination of the two data sources for assessment of PCPT outcomes provide a rich resource that can also be utilized to address additional follow-up questions. When asked about the direction of future studies, Dr. Unger said, “Our next objective is to examine the extent to which finasteride use limited the total burden of treatments from prostate cancer and the complications of those treatments for participants on the PCPT, and to map the potential impact on reduced treatments and complications onto the U.S. population of older men.”
This research was supported by the National Cancer Institute.
Unger JM, Hershman DL, Till C, Tangen CM, Barlow WE, Ramsey SD, Goodman PJ, Thompson IM, Jr. 2018. Using Medicare claims to examine long-term prostate cancer risk of finasteride in the prostate cancer prevention trial. Journal of the National Cancer Institute. doi:10.1093/jnci/djy035.
Basic Sciences Division
Human Biology Division
Maggie Burhans, Ph.D.
Public Health Sciences Division
Vaccine and Infectious Disease Division
Clinical Research Division
Julian Simon, Ph.D.
Clinical Research Division
and Human Biology Division
Arnold Digital Library