Sipuleucel-T significantly improves OS in African-Americans vs. Caucasians

September 5, 2019

Analyses of data collected in PROCEED, a large real-world registry, corroborate phase III study findings demonstrating that sipuleucel-T (Provenge) treatment for metastatic castrate-resistant prostate cancer has a particular benefit for improving overall survival in African-American men.

Analyses of data collected in PROCEED, a large real-world registry, corroborate phase III study findings demonstrating that sipuleucel-T (Provenge) treatment for metastatic castrate-resistant prostate cancer (mCRPC) has a particular benefit for improving overall survival (OS) in African-American men.

In a PSA-matched cohort that included 438 Caucasians and 219 African-Americans, there was a statistically significant 9.5-month difference in median OS favoring African-American patients over Caucasian patients (35.3 vs. 25.8 months; hazard ratio [HR]: 0.70; p<.001). When the men were split into two groups according to median baseline PSA (≤29.48 ng/mL and >29.48 ng/mL), a significant difference in median OS favoring the African-Americans was achieved only in the subgroup with a baseline PSA ≤29.48 ng/mL, but the difference between groups increased to almost 21 months. Within the lower baseline PSA subgroup, median OS was 54.3 months for the African-American men and 33.4 months for the Caucasians (HR: 0.52, p<.001), reported first author Oliver Sartor, MD, at the American Society of Clinical Oncology annual meeting in Chicago.

“The difference in median OS in the lower PSA subgroup is quite remarkable and represents the largest known racial difference in OS in response to any therapy for mCRPC,” said Dr. Sartor, medical director of Tulane Cancer Center, and C. E. and Bernadine Laborde Professor of Cancer Research at Tulane University School of Medicine, New Orleans.

“Therefore, in my practice, I am trying to ensure that African-Americans, and particularly those with PSA below 30 ng/mL, have the opportunity to receive sipuleucel-T if it is feasible based on their insurance and other issues.”

Review of the registry enrollees identified 1,902 men who had received at least one sipuleucel-T infusion, of whom 221 were African American and 1,649 were Caucasian. Because PSA differed significantly at baseline in the two racial groups, a PSA-matched cohort was created, matching two Caucasian patients to each African-American based on having a baseline PSA ±10%. The matched cohort included only 219 African-Americans after excluding one man who lacked baseline PSA data and another who had a very high PSA that could not be matched according to the selection criterion.

In both racial groups, the median patient age was 71 years, about two-thirds of the men had an ECOG performance status of 0, and slightly less than one-half had a Gleason sum ≥8. Median PSA at baseline was 28.7 ng/mL in the Caucasians and 32.9 ng/mL in the African-Americans. The mean number of life-prolonging anti-cancer therapies was 1.6 in both groups; the median number received was 2.0 in the Caucasians and 1.0 in the African-Americans.

Next: Analysis divides patients into PSA quartilesAnalysis divides patients into PSA quartiles

Another analysis of OS divided patients into PSA quartiles (≤8, 8 to ≤29.48 ng/mL, >29.48 to ≤82.4 ng/mL, and >82.4 ng/mL). A significant improvement in median OS among African-Americans compared with Caucasians was found only in the two lowest quartiles (HR: 0.49 for men with PSA ≤8 ng/mL, 0.54 for men with PSA >8 ng/mL to 29.48 ng/mL).

Analyses were also performed to identify baseline variables that were predictors of OS. In univariate analysis, age, weight, race, ECOG performance status, PSA, alkaline phosphatase, hemoglobin, lactate dehydrogenase, lymph node-only metastases, prior prostatectomy, prior abiraterone (ZYTIGA)/enzalutamide (XTANDI), and prior docetaxel (Taxotere)/cabazitaxel (Jevtana) had a statistically significant association with OS.

Also see: Prostate-sparing treatment using nanoparticles shows promise

In a Cox regression model adjusting for the significant covariates, African-American race was independently associated with improved OS and had the strongest predictive value of all associated covariates (HR: 1.67).

Dr. Sartor noted that the explanation for racial difference in OS is unknown.

“It may involve differences in pathophysiology or immune responsiveness,” he said.

Dr. Sartor is a consultant to Dendreon and to other companies that market or are developing treatments for prostate cancer. For full disclosures, see bit.ly/proceeddisclosures.

download issueDownload Issue : Vol 47 No 9