Paris--Results of a second phase III trial suggest that a still-investigational immunotherapy agent offers a survival benefit in men with asymptomatic metastatic androgen-independent prostate cancer. Survival data from the more recent study support the findings of the first phase III trial of the agent, APC8015, also known as sipuleucel-T (Provenge), researchers reported at the European Cancer Conference here.
Paris-Results of a second phase III trial suggest that a still-investigational immunotherapy agent offers a survival benefit in men with asymptomatic metastatic androgen-independent prostate cancer. Survival data from the more recent study support the findings of the first phase III trial of the agent, APC8015, also known as sipuleucel-T (Provenge), researchers reported at the European Cancer Conference here.
In the second trial, D9902A, no statistically significant improvement in time to progression or overall survival was found, but median survival was improved to 19 months in patients receiving the agent compared with 15.7 months in the placebo group. Further, more patients treated with immunotherapy (32%) were alive at 36 months post-treatment than were those in the placebo group (21%). After adjustment for prognostic variables, which was done retrospectively, a statistical advantage was found for the APC8015 arm, reported lead author Celestia S. Higano, MD, director of the genitourinary oncology clinical research group and associate professor of medicine and surgery at the Seattle Cancer Care Alliance.
Eligibility criteria for D9902A included histologically documented carcinoma of the prostate, at least 25% of cells positive for PAP by immunohistochemical analysis, metastatic disease in soft tissue and/or skeleton, testosterone less than 50 ng/dL, tumor progression, ECOG performance 0 to 1, and asymptomatic status. All patients were followed for survival for 3 years after randomization. Enrollment stopped after accrual of 98 of the 120 patients planned.
An exploratory analysis of D9901 found that time to disease progression was prolonged in the APC8015 arm only in patients with a Gleason score ≤7, which was the most important predictor of response to the agent. Patients with a Gleason score ≤7 were placed in a second trial (D9902B), which has been amended with FDA approval to include patients regardless of their Gleason score and those with minimally symptomatic disease-related pain. Five hundred men will be enrolled in this amended study, which will be known as the IMPACT (Immunotherapy for Prostate Adenocarcinoma Treatment) study.
In the amended IMPACT protocol, the primary endpoint has been changed to overall survival, which will be an event-driven analysis, and time to objective disease progression has been established as the secondary endpoint. The primary statistical model to determine efficacy will be the Cox multivariate regression model, the same model used in the two previous studies.
Pooled analysis shows advantage
Pooled analysis of data from D9901 and D9902A showed a statistical survival advantage for APC8015 compared with placebo, with a statistically significant difference in survival of 4.3 months favoring immunotherapy (23.1 months vs. 18.9 months, p=.011). APC8015 was well tolerated and was safe in both trials, with transient fever, fatigue, and chills being the major adverse events reported.
Dr. Higano explained that the rationale for the pooled analysis was based on the two trials' identical inclusion criteria, treatment schema, and study design.
"The endpoints were measured in an identical fashion, and the trials were fairly contemporaneous with a pre-specified plan to pool for the secondary endpoint," she said.
"Given these data, it would appear that APC8015 might be a reasonable treatment option for asymptomatic men with metastatic androgen-independent prostate cancer," Dr. Higano concluded.
The current data were reviewed by Cora Sternberg, MD, chief of the department of medical oncology at San Camillo-Forlanini Hospital in Rome, who noted that both studies lacked data on patients' immunologic parameters. She also pointed out that the studies were small and were not designed to show a statistical difference in survival, creating what she called a "huge alpha error."