Predictors of Survival in Multiple Myeloma Patients after Relapse from a Delayed Autologous Stem Cell Transplant

Background: Early (<12 months from diagnosis) autologous stem cell transplantation (ASCT) remains a standard treatment for eligible patients with multiple myeloma (MM). However studies suggest comparable overall survival (OS) outcomes in patients undergoing delayed (>12 months from diagnosis) ASCT.  Little is known about the outcomes upon relapse from a delayed ASCT and hence we describe the natural course of the disease and identify factors that predict post-ASCT relapse survival in this group of patients.

Patients and Methods: We examined the outcomes of 207 MM patients who underwent a delayed ASCT (12 or more months from their initial diagnosis) at the Mayo Clinic, Rochester and then relapsed between 2000 and 2012. Patients who received a tandem or an allogeneic stem cell transplant post first ASCT were not included in this study. Multivariate analysis was performed using a Cox proportional hazards model.

Results: Median (range) age at ASCT and at relapse was 62 yrs (40-75) and 63 yrs (40-76) respectively. The median time to ASCT from MM diagnosis was 30 mos (12-202) and the median time to relapse post-ASCT was 11 mos (2-100). The median follow-up (95% CI) from relapse was 78 mos (60-99). The median (range) number of regimens prior to ASCT was 2 (1-6). Of the 112 patients who had FISH testing at diagnosis, 30% were categorized as high risk. The median (95% CI) LDH and plasma cell labeling index at diagnosis was 169 (178 – 215) and 0.6 (1 – 1.5) respectively. The median (95% CI) post-ASCT relapse survival (PRS) and overall survival from diagnosis was 30 (22-36) and 77 (69-87) mos, respectively. The distribution (N, %) of the post-relapse treatment regimens were as follows: immunomodulator-based (103, 50%), proteasome inhibitor-based (47, 23%), non-novel agent chemotherapy (31, 15%), novel investigational therapy (22, 11%) and repeat ASCT (4, 1%). In a multivariate analysis, only shorter time to relapse, presence of CRAB symptoms or extramedullary disease at relapse, higher pre-ASCT plasma cell labeling index and LDH predicted for a shorter PRS.  The presence of high-risk cytogenetics by FISH, and choice of salvage therapy upon relapse post-ASCT did not affect PRS in the multivariable model. 

Conclusion: MM patients who undergo a delayed first ASCT experience improved PRS if they have a longer duration of response to ASCT, lack CRAB symptoms or extramedullary disease at relapse (only biochemical progression) and have a lower pre-ASCT plasma cell labeling index and serum LDH.