ATLANTA—Radical prostatectomy (RP) is associated with better overall and disease-specific survival compared with external beam radiation therapy (EBRT) for localized prostate cancer (PCa), according to findings presented at the American Urological Association 2012 annual meeting.
In a study of 1,655 men with localized PCa—including 1,164 (70.3%) who underwent RP and 491 (29.7%) who had EBRT—researchers found that EBRT was associated with a
40% and 65% decreased likelihood of overall and disease-specific survival, respectively, compared with RP, after adjusting for multiple confounders.
The study, which was presented at the meeting by David F. Penson, MD, MPH, of Vanderbilt University in Nashville, Tenn., included participants in the Prostate Cancer Outcomes Study, which enrolled subjects diagnosed with PCa between October 1994 and October 1995.
“The poorer overall survival with radiotherapy is largely due to selection bias—healthier men are more likely to get surgery,” said lead investigator Richard M. Hoffman, MD, MPH, Professor of Internal Medicine at the University of New Mexico in Albuquerque.
As for why radiotherapy patients had worse disease-specific survival compared with RP patients, Dr. Hoffman observed that in the mid 1990s, radiation dosages were lower and data had yet to be published showing that men with high-risk PCa (based on high PSA levels and high Gleason scores) benefited from receiving both radiation and androgen deprivation therapy. “Current radiation modalities and treatment regimens could provide better outcomes,” he said. “However, another explanation could be that surgery is a better treatment.”
Dr. Hoffman noted that no randomized controlled trials have compared survival outcomes among men with localized PCa treated with RP or radiotherapy. The study by his group provides the first long-term survival results comparing the two treatments for localized PCa. Previous observational studies have focused either on just one type of treatment, used surrogate endpoints such as PSA rises, or were single-center investigations, Dr. Hoffman pointed out.
“Our results, based on a large-population-based cohort, are much more generalizable, though we recognize that patient selection and treatments have changed since we enrolled patients in the mid 1990s,” Dr. Hoffman said.
He and his colleagues obtained information on medical conditions at the time of diagnosis based on well-accepted comorbidity scales, and they used multivariate statistical techniques to adjust for comorbidity differences between treatment groups. These measures are relatively crude, however, Dr. Hoffman said.