Background: The unmet need for predictive biomarkers emerged from the unpredictable pattern of response to androgen signalling inhibition in metastatic castration-resistant prostate cancer (mCRPC). Here, we report on the testing of a previously identified candidate androgen signalling signature associated with response to androgen signalling inhibition.
Patients and methods: We report on the outcome of the first module of a phase II trial on abiraterone acetate (AA) followed by combination with dasatinib or sunitinib. Bone marrow biopsies (BMBs) with matched bone marrow aspirate and blood samples were collected at baseline and upon progression. End-points included assessment of a prespecified molecular signature consisting of nuclear androgen receptor (AR) overexpression, cytochrome P450, family 17, subfamily A, polypeptide 1 (CYP17) expression, and AR-C-/N terminal expression ratio of ≥0.8 by immunohistochemistry (IHC) in patients with benefit versus primary resistance to AA (i.e. progression within 4 months). Tumour markers also included v-ets avian erythroblastosis virus E26 oncogene homologue (ERG), androgen receptor splice variant (ARV7) by IHC and steroids by liquid chromatography-tandem mass spectrometry.
Results: Of 170 patients accrued from 03/2011 to 02/2015, 44 (26%) were primary resistant to AA. Forty-eight patients had tumour infiltrated BMB at baseline. Pretreatment androgen signalling signature was linked to benefit from AA (p < 0.001). Presence of ERG was associated with benefit (p = 0.05), whereas nuclear ARV7 presence and 20 or more bone lesions at baseline with primary resistance (p = 0.04 and p = 0.0006, respectively).
Conclusion: Testing of a prespecified androgen signalling signature was highly supportive of its predictive value in maximal androgen deprivation strategies in mCRPC. Further validation is under way.
Trial registration: ClinicalTrials.gov NCT01254864.
Keywords: Abiraterone acetate; Androgen receptor; Bone metastasis; Castration-resistant prostate cancer; Predictive biomarkers.
Copyright © 2020 Elsevier Ltd. All rights reserved.