However, in most places, enzalutamide and abiraterone (+prednisone) have become first-line therapies for metastatic castration-resistant prostate cancer (mCRPC), once patients have failed ADT – as they are oral agents and tolerated well, they are favored over chemotherapy. Yet, the decision to choose one over the other is limited to physician judgment, comfort, and adverse effect profiles. Sequencing these expensive agents has become the next big challenge, especially as they may share resistance pathways.
In this first prospective trial to compare sequencing, the authors initiated a multicenter trial of ABI+P (abiraterone + prednisone) followed by enzalutamide (ENZ) at PSA progression (arm A) versus ENZ followed by ABI+P (arm B). Primary endpoints were PSA decline > 50% (PSA50) on 2nd-line therapy and time to 2nd PSA progression (TT2P) (from start of 1st-line). Deep-targeted sequencing of serial samples of circulating tumour DNA (ctDNA) was performed to potentially identify a biomarker of efficacy.
The study design is below:
A total of 202 were accrued with median follow-up 22.3 months (m). 101 patients were randomized to each arm. At baseline:
1. Patients treated with enza first were older (median age 78 vs. 73, p<0.05)
2. 80-90% in both groups were ECOG 0-1
3. Median PSA in both groups was 35-37
4. Lab values and presence of visceral/bone mets were similar between the two groups at baseline
In terms of progression on 1st line therapy and cross-over to second line therapy during this time frame, 65 pts from arm A and 71 from arm B crossed-over to 2nd-line treatment and 30 pts (15/arm) stopped treatment without cross-over.
At the time of advance to 2nd line therapy, baseline characteristics at time of 2nd-line therapy were balanced: median age 75 (range 50-93), PSA 14.5 (6.6-62.1), alkaline phosphatase > upper limit of normal (ULN) in 39% of pts and bone/liver metastasis (mets) in 89%/9%. ECOG PS was 0-1 in 89% vs 76% of pts for arm A vs arm B (p = 0.044) and LDH was > ULN in 25% vs 8% (p = 0.013).
In terms on oncologic response to 2nd line therapy, PSA50 for 2nd-line therapy for arm A vs arm B was 34% vs 4% (p˂0.001) and median time to PSA progression on 2nd-line therapy (TTPP) was 2.7 m vs 1.3 m (HR 0.38, 95% CI 0.26-0.56). Waterfall plot demonstrates a better response to Arm A (AbiEnza):
For the intention-to-treat population (including those that did not go on 2nd line therapy), TT2P was 13.6 vs 11.9 m (HR 0.75, 95% CI 0.53-1.06).
Median overall survival (OS) was not reached (NR) vs 24.3 m (HR 0.82, 95% CI 0.53-1.27).
On multivariable analysis, factors associated with 2nd-line TTPP were: bone mets (HR 2.22, 95% CI 1.08-4.54), liver mets (HR 3.18, 95% CI 1.21-8.41) and treatment arm A vs B (HR 0.27, 95% CI 0.17-0.40).
In terms of biomarkers, at progression, AR gene copy number increased in 14% of evaluable pts (7/49) and AR L702H/T878A(S) mutations were present in 8% of pts. ctDNA fraction ≥2% at baseline was associated with worse TT2P (HR 2.04, 95% CI 1.43 - 2.90) and OS (HR 4.07, 95% CI 2.40-6.91).
In terms of adverse event profile, the proportion of grade 3-4 AEs was similar in both groups (23/65 in Arm 1 and 24/71 in Arm 2).
Based on this interesting data, it would appear (similar to other stages of PCa management), the sequence of ABI+P followed by ENZ was associated with superior PSA50 and TTPP on 2nd-line therapy. Further work is needed to validate these findings, but they may help clarify a better treatment paradigm in the future.
Presented by: Daniel Khalaf, MD
Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, Twitter: @tchandra_uromd at the 2018 ASCO Annual Meeting - June 1-5, 2018 – Chicago, IL USA