Results of a phase II study of carboplatin and etoposide in patients with progressive metastatic castration refractory prostate cancer (mCRPC) and neuro-endocrine differentiation

Abstract

e16073 Background: Neuro-endocrine differentiation is often observed in the evolution of mCRPC. We assessed the efficacy and toxicity of a platin-based chemotherapy regimen in patients with mCPRC and neuro-endocrine differentiation defined by: either increased circulating neuro-endocrine markers (chromogranin A: CgA, Neuron Specific Enolase: NSE) more than 1.5 X normal serum values and/or visceral metastases confirmed with immunihistochemical proof of neuro-endocrine differentiation on pathological sample. Methods: Sixty patients were included (Simon optimal two stages design with P0 = 20%, P1 = 40%, α = 5% and power 90% for a total of 54 evaluable patients) and were treated by the combination of carboplatin AUC 4 d1 IV and etoposide 100 mg/m2/d d1–3 IV every 3 weeks for a maximum of 6 cycles. Efficacy endpoints included PSA and neuro-endocrine marker response rate (defined as a decrease of 50% or greater of the baseline serum value), objective response rate (according to RECIST criteria), and toxicity. Results: Sixty patients were included between April 2005 to January 2008, median age was 67 (range: 45–80). Sixty-seven per cent patients received prior chemotherapy. Patients had bone metastases (78%), lymph nodes involvement (49%), lung metastases (35%), hepatic involvement (33%) and other localizations (17%). The objective response rate was 33% in the 48 assessable patients. A neuro-endocrine response was observed in 28% of 32 evaluable patients for neuro-endocrine marker level (CgA 6%, NSE 25% and both 3%). The PSA response rate was 9%. The most common grade 3–4 treatment-related toxicities were neutropenia (67%), thrombocytopenia (31%), anemia (27%), asthenia (14%) nausea and vomiting grade (12%). There were 6% febrile neutropenia, with one related toxic death. The median follow-up is 9 months. The median response duration was 1.8 months (range: 0.2–13.4 months). The median overall survival is 10 months. Conclusions: Despite an absolute response rate in accordance with the study assumptions, the benefit-risk ratio of this regimen seems unfavourable due to observed toxicities. Another trials must be conducted in order to define a group of patients which may benefit of this regimen. No significant financial relationships to disclose.