Continuous Therapy in Prostate Cancer Patients

Chicago—A phase 3, international, randomized trial of patients with hormone-sensitive metastatic prostate cancer found that overall survival was inferior in those who had intermittent androgen deprivation therapy compared with those who underwent continuous therapy. A secondary analysis found intermittent androgen deprivation therapy was noninferior in patients with extensive disease, but it was statistically inferior in patients with minimal disease.

Maha Hussain, MD, professor of medicine and urology at the University of Michigan’s Comprehensive Cancer Center and the study’s lead author, said continuous therapy should remain the standard of care in this patient population.

“These observations suggest inherent potential biological differences and certainly warrant further mechanistic evaluations,” said Dr. Hussain, who presented the results at the ASCO meeting in a plenary session.

The National Cancer Institute sponsored the study, which began on May 15, 1995, and was closed on September 1, 2008. Dr. Hussain said that earlier clinical trials, including phase 2 studies, found that intermittent androgen deprivation could prolong the duration of disease response and may be associated with improvement in quality of life.

Patients were recruited from the United States, Canada, and Europe, and were included if they had newly diagnosed metastatic prostate cancer, prostate-specific antigen (PSA) level ≥5 ng/mL prior to initiating androgen deprivation, and Southwest Oncology Group performance status between 0 and 2.

During the induction phase, patients were treated with hormone therapy (goserelin and bicalutamide) for 7 months. If patients’ PSA levels were ≤4 ng/mL during months 6 and 7 and showed a declining trend, they were randomized to continuous or intermittent therapy. Men in the intermittent group discontinued therapy and were monitored with monthly PSA tests. They resumed therapy based on prespecified criteria.

There were 770 patients randomized in the intermittent arm and 765 patients randomized in the continuous arm. The groups were well balanced, and the median age was approximately 70 years. They had similar PSA levels at randomization, performance status, presence of visceral disease, bone pain, and exposure to prior hormone therapy.

Dr. Hussain said that when the researchers designed the study, they assumed a median survival of 3 years after randomization for the control group, which was too short a time period. Thus, it took them longer to observe the number of deaths needed for a final analysis.

Median overall survival was 5.8 years in the continuous group and 5.1 years in the intermittent group (hazard ratio [HR], 1.09; 95% confidence interval [CI], 0.95-1.24), representing a 9% relative increase in the rate of death with intermittent therapy.

Dr. Hussain noted that the upper limit of the 95% CI was higher than the prespecified threshold based on the statistical design, which deemed intermittent therapy would be inferior if the 95% CI exceeded 1.20.

“Thus, we conclude intermittent therapy is in fact inferior in this patient population overall,” she said.

A secondary analysis found the overall treatment effect was consistent across most subgroups, although Dr. Hussain said intermittent therapy was favored in patients with extensive disease. The median overall survival for patients with extensive disease was 4.4 years in the continuous therapy group compared with 5 years in the intermittent therapy group (HR, 0.96; 95% CI, 0.80-1.16). The median overall survival for patients with minimal disease was 7.1 years in the continuous therapy group and 5.2 years in the intermittent therapy group (HR, 1.23; 95% CI, 1.02-1.49; P=.034).

Dr. Hussain said disease extent was a prespecified stratification factor, but the study was not designed to evaluate survival based on the extent of disease.

“Nonetheless, this finding is unexpected and quite striking,” she said.

The groups had no significant differences in treatment-related adverse events, including cardiovascular events. There were a low number of grade 4 adverse events, and no grade 5 adverse events were reported.