AR-V7 Predicts Response in Castration-Resistant Prostate Cancer

Emmanuel S. Antonarakis, MD

The presence of androgen receptor splice variant-7 (AR-V7) mRNA in circulating tumor cells (CTCs) predicted poorer outcomes for men with castration-resistant prostate cancer assigned to novel hormonal therapies (NHT), according to results recently published in the Journal of Clinical Oncology.

In an update on results first presented in 2014, researchers observed prostate-specific antigen (PSA) response rates of 75.5% in CTC-negative patients, 52.2% in CTC-positive/AR-V7—negative patients, and just 13.9% in CTC-positive/AR-V7–positive patients among 202 patients starting treatment with enzalutamide (Xtandi) or abiraterone acetate (Zytiga).

The study was not powered to assess overall survival (OS), but those outcomes also favored CTC-negative patients. Median OS in that group was 28.7 months (95% CI, 28.4-not reached) compared with 29.5 months (95% CI, 18.4-not reached) in the CTC-positive/AR-V7—negative group and 11.2 months (95% CI, 8.3-17.1) in the CTC-positive/AR-V7–positive group.

“The outcomes were exactly as we would have predicted,” study author Emmanuel S. Antonarakis, MD, an associate professor of urology and oncology at Johns Hopkins School of Medicine, said in an interview with OncLive. “The best outcomes were for those who didn’t have any circulating tumor cells detected. The very worst were in men who had CTC detected and AR-V7 detected.”

Antonarakis added that AR-V7 appeared to be a predictor only for NHT, though AR-V7 has not been evaluated in immunotherapy or radiotherapy settings.

“We did not include the updated chemotherapy results in this analysis—that will be part of a future paper,” he said. “The overwhelming data seems to suggest that this biomarker is only relevant, or most relevant, in the setting of hormone therapy-resistance but not chemotherapy-resistance.”

Researchers enrolled men assigned to enzalutamide (n = 107) or abiraterone (n = 95) from December 2012 through November 2015. At baseline, 26.2% were CTC-negative, 56% were CTC-positive/AR-V7—negative, and 17.8% were CTC-positive/AR-V7–positive. Patients in the latter group were more likely to have Gleason scores ≥8, present with metastatic disease, have higher levels of PSA and alkaline phosphatase, have used abiraterone or enzalutamide previously, have used taxanes, experience bone pain, and have ECOG status ≥1.

Enzalutamide was administered at 160 mg once daily. Abiraterone was administered at 1000 mg once daily, along with 5 mg of prednisone twice daily.

The first-line NHT cohort was made up of 124 men who had never received enzalutamide or abiraterone. In the second-line cohort, 78 men had received one of those drugs previously. There were fewer CTC-negative patients and more CTC-positive/AR-V7—positive patients in the second-line group.

In the first-line NHT cohort, median OS was 29.7 months (95% CI, 28.7-to not reached) for CTC-negative patients, 30.7 months (95% CI, 29.5-not reached) for CTC-positive/AR-V7—negative patients, and 21.5 months (95% CI, 10.4-not reached) for CTC-positive/AR-V7–positive patients (P = .003).

In the second-line NHT cohort, median OS was 18.8 months (95% CI, 12.5-to not reached) for CTC-negative patients, 13.0 months (95% CI, 10.0-22.6) for CTC-positive/AR-V7—negative patients, and 8.5 months (95% CI, 4.9-15.6) for CTC-positive/AR-V7–positive patients (P <.001).

In the first-line cohort, PSA response rates were 52.9% for CTC-negative patients, 27.5% for CTC-positive/AR-V7—negative patients, and 4.8% for CTC-positive/AR-V7–positive patients.

PSA response rates were 86.1% for CTC-negative patients, 65.8% for CTC-positive/AR-V7—negative patients, and 26.7% for CTC-positive/AR-V7–positive patients in the second-line cohort.

Biomarker status was an independent predictor of PSA response for both groups on multivariable logistic regression analysis.

Biomarker status was also a predictor for PFS with enzalutamide or abiraterone. Overall, median PFS was 11.3 months (95% CI, 8.7-13.8) for CTC-negative patients, 6.2 months (95% CI, 5.8-7.3) for CTC-positive/AR-V7—negative patients, and 2.1 months (95% CI, 1.9-3.1) in CTC-positive/AR-V7–positive patients.

In the first-line NHT cohort, median PFS was 12.7 months (95% CI, 11.7-23.9) in CTC-negative patients, 7.3 months (95% CI, 6.2-2.0) in CTC-positive/AR-V7─negative patients, and 2.9 months (95% CI, 2.0-not reached) in CTC-positive/AR-V7—positive patients.

Among second-line patients, median PFS was 6.4 months (95% CI, 5.1-not reached for CTC-negative patients, 4.4 months (95% CI, 3.2-6.0) for CTC-positive/AR-V7-negative patients, and 1.1 months (95% CI, 1.0-3.1) for CTC-positive/AR-V7-positive patients.

Though these data are encouraging, Antonarakis said AR-V7 needs to be evaluated in a prospective, clinical trial before physicians should use the marker to make treatment decisions. He is currently working with Andrew Armstrong, MD, associate professor of medicine at Duke Cancer Institute, on a multicenter study comparing the prognostic value of several different AR-V7 assays. That trial has reached its target enrollment of 120 patients with castration-resistant prostate cancer, and the investigators expect to have preliminary results in 6 to 12 months.

Jorge Garcia, MD, director of the advanced prostate cancer program at the Cleveland Clinic Taussig Cancer Institute reviewed the data for OncLive. He said these results were “powerful enough” that he would be willing to use AR-V7 as a biomarker when making treatment decisions, provided he had an assay that could reliably detect CTCs.

“Until now, outside of PSA, we did not have a biomarker that we could utilize to define the appropriate sequence and the appropriate patient for the treatments that we have in castration-resistant disease,” Garcia said. “If you look at the NCCN guidelines, we have radium-223, docetaxel, sipuleucel-T [Provenge], abiraterone, and enzalutamide as treatment options for the same patient phenotype. The biggest question is, can we tease out who are the patients who are unlikely to benefit from oral therapy and should move to something else? That is the relevance of this data.”

ES Antonarakis, Lu C, Luber B, et al. Clinical significance of androgen receptor splice variant-7 mRNA detection in circulating tumor cells of men with metastatic castration-resistant prostate cancer treated with first- and second-line abiraterone and enzalutamide [published online April 6, 2016]. J Clin Oncol. doi: 10.1200/JCO.2016.70.1961.