Responses to Induction Chemotherapy and High Dose Melphalan Can Be Incorporated within a Risk Stratification Model and Help to Identify Myeloma Patients Who May Benefit from Novel Therapies.

We propose a multi-parametric risk stratification model for the identification of patients with myeloma who may benefit from novel therapies. This model is based on an intention to treat analysis of 383 patients with previously untreated myeloma, their responses to induction chemotherapy and to intensive high dose melphalan (HDM). Using this strategy we outline a proposal for the incorporation of new agents such as bortezomib and other targeted therapies into treatment plans for new patients with myeloma. Patients in this study, who were collected prospectively into a comprehensive clinical data base, received induction therapy with CVAMP followed by HDM (200mg/m²) or intermediate dose melphalan IDM (140mg/m²) and had a median follow-up of 8 years from presentation. The overall response rate to CVAMP was 66% (15% CR, 51% PR), with 29 (8%) early deaths, and 101 (26%) non responders, defined using European Bone Marrow Transplant criteria. Response to induction therapy was highly predictive of response to HDM (p<0.00005) and of overall survival (OS) and progression free survival (PFS) after HDM. On multivariate analysis, B2M (>3.65mg/l) was the most significant independent prognostic factor to predict death, event free survival (EFS), and relapse (p<0.004). A Kaplan- Meier analysis following 3 cycles of CVAMP demonstrated that this was the earliest time point during induction treatment at which a response could significantly predict a better outcome for responders versus non-responders (p=0.035) and was an independent prognostic factor in the model. Following CVAMP, 253 patients proceeded to HDM and 29 patients received IDM (140mg/m²). Overall survival was superior for patients receiving HDM compared to IDM, median survival 7.1 years versus 2.6 years; p<0.00001. A complete response (CR) at 3 months post HDM was achieved in 50% of the 282 patients and was significantly associated with an improved OS and PFS (median OS 7.4 years CR group versus 5.3 years non-CR group; p=0.023). The benefit of attaining a CR post HDM was independent of prior response. These results identify two therapeutic windows: the first following the third cycle of induction chemotherapy and the second 3 months after HDM, at which outcome can be assessed and used to predict subsequent disease behaviour. We have integrated these responses into a novel predictive strategy. This prognostic information collected sequentially over time could then be used to direct the use of novel therapeutic agents, with the aim of improving response status at these time points. Such a strategy, aimed at achieving the maximal numbers of patients in CR after transplantation with minimal toxicity, is likely to be crucial for improving outcomes in newly diagnosed patients with myeloma.