Testing a prognostic model in patients with metastatic castrate resistant prostate cancer (MCRPC) treated prechemotherapy with abiraterone.
283 Background: There is a need for prognostic models in metastatic prostate cancer to determine optimal treatment pathways on development of castrate resistance. Abiraterone is an androgen biosynthesis inhibitor. Efficacy was proven in patients with MCRPC prechemotherapy in the COU-AA-302 trial with PSA PFS 11.1 months (Raj et al New Eng J med 2013). We have previously investigated prognostic factors in MCRPC (Charnley et al J Clin Oncol 2014). We reported a prognostic model which predicts that if alk phos rise ≥100 in the preceding 3 months and/or Hb<11 there is no response. The model was correct in 78.9% of responders and 55% of non responders (overall 66.7%). We test the model in an independent group of 81 patients prechemotherapy, some from a different treating centre. Methods: 81 patients with metastatic CRPC received abiraterone 1g daily prechemotherapy. Response to abiraterone was defined as a 50% reduction in PSA. Alk phos rise ≥100 in the previous 3 months and Hb<11 were related to PSA response. Results: 19 patients had both alk phos rise ≥100 in the preceding 3 months, and Hb<11. None of these patients had a PSA response. For the 31 patients with no risk factors 66% of cases responded. The predictive model was correct in 66% of true responders, and 85.3% of non responders, 74.1% overall. Conclusions: In this group of 81 patients with castrate resistant metastatic prostate cancer a previously defined model, which incorporates alk phos rise ≥100 in the preceding 3 months and/or Hb<11 is 74.1 % accurate in an independent group of patients. No patients with both alk phos rise>100 in the preceding 3 months, and Hb<11, responded to abiraterone. The model has potential in guiding treatment pathways on development of castrate resistance in prostate cancer, but needs to be tested in a greater number of patients.