Did Corticosteroids Worsen Survival in Patients with Metastatic Prostate Cancer Enrolled in the AFFIRM Trial? 

Get Permission

A post hoc analysis of the AFFIRM trial found that on-study use of corticosteroids led to worse outcomes in metastatic castration-resistant prostate cancer regardless of whether patients were randomly assigned to enzalutamide (Xtandi) or placebo.1 On-study corticosteroid use was associated with reduced survival and higher rates of grade 3 or 4 adverse events. Regardless of on-study corticosteroid use, enzalutamide remained consistently superior to placebo.

Presenting author Howard I. Scher, MD, Chief, Genitourinary Oncology Service, Memorial Sloan-Kettering Cancer Center, New York, told listeners at the Genitourinary Cancers Symposium that AFFIRM was not designed to assess the efficacy of corticosteroids on survival. “Nevertheless, those patients taking corticosteroids had reduced survival times and higher grades of serious side effects than those who did not in this post hoc analysis. These findings require prospective validation,” Dr. Scher stated.

Study Background

Enzalutamide is a novel oral androgen receptor inhibitor that blocks nuclear translocation of the androgen receptor as well as DNA binding. The randomized phase III AFFIRM trial found that baseline use of corticosteroids was associated with inferior overall survival. Without corticosteroids, median overall survival was 7.5 months longer and risk of death was reduced by 53%.2

“Corticosteroids have a range of favorable effects in prostate cancer management. These include antitumor effects, palliation of symptoms, and reduction of adverse events associated with the administration of approved therapies. They are used in both the pre- and postchemotherapy settings, but now a body of evidence suggests that corticosteroids may stimulate prostate cancer growth by a range of mechanisms including the activation of promiscuous androgen receptors,” Dr. Scher explained.

In the AFFIRM registration trial, enzalutamide prolonged survival in post–chemotherapy-treated metastatic castration-resistant prostate cancer by a median of 4.8 months, with a 37% reduction in risk of death. Dr. Scher and coauthors performed an unplanned post hoc analysis to determine if on-study use of corticosteroids compromised outcomes.

In the study, patients were allowed to enter the trial on corticosteroids, but use of these agents varied widely, he continued. “Some patients did not take them at all, about 30% were taking them at the start of therapy, and others received them during the study,” he said.

Overall, corticosteroids were used at the start of therapy and on study in 48% of enzalutamide recipients and 45% of the placebo group. Patients taking corticosteroids were generally sicker and had more advanced disease; they had poorer survival and more rapid disease progression.

Key Data

Median survival for on-study corticosteroid users was 12.8 months for the enzalutamide arm vs 9.6 months for placebo (P < .001). For those not taking corticosteroids, the median survival was not yet reached in the enzalutamide-treated patients vs 18.8 months for placebo (P < .001).

A similar pattern was seen for progression-free survival and time to prostate-specific antigen (PSA) progression. Corticosteroid-treated patients had inferior outcomes relative to those not taking corticosteroids, while the benefit of enzalutamide relative to placebo was seen in both corticosteroid- and noncorticosteroid-treated patients. Median progression-free survival was 5.6 vs 2.9 months, respectively, in on-study corticosteroid users, and 11.1 vs 3 months, respectively, in patients who did not use corticosteroids (P < .001 for both comparisons). Median time to PSA progression in on-study corticosteroid users was 5.6 months for enzalutamide vs 3.1 months for placebo (P < .001), and among nonusers, 8.6 vs 2.9 months, respectively (P < .001).

Use of on-study corticosteroids was associated with higher rates of treatment-emergent grade 3 and 4 adverse events, including anemia, fatigue, spinal cord compression, and back pain. Both the benefit of enzalutamide and the adverse effect of corticosteroids on outcomes remained in a multivariate analysis addressing imbalances in the groups.

“Importantly, these findings are not definitive and require prospective validation,” Dr. Scher commented. ■

Disclosure: Dr. Scher has received research funding and other remuneration from Medivation.


1. Scher HI, Fizazi K, Saad F, et al: Impact of on-study corticosteroid use on efficacy and safety in the phase III trial of enzalutamide (ENZA), an androgen receptor inhibitor. 2013 Genitourinary Cancers Symposium. Abstract 6. Presented February 14, 2013.

2. Scher HI, Fizazi K, Saad F, et al: Association of baseline corticosteroid with outcomes in a multivariate analysis of the phase 3 AFFIRM trial of enzalutamide versus placebo. 2012 ESMO Congress. Abstract 899PD. Presented September 30, 2012.

Related Articles

Expert Point of View: William K. Oh, MD

Formal discussant William K. Oh, MD, Chief of the Division of Hematology and Medical Oncology, Professor of Medicine and Urology, and the Ezra M. Greenspan, MD, Professor in Clinical Cancer Therapeutics at Icahn School of Medicine at Mount Sinai, New York, was not convinced that corticosteroids...