Risk stratification of smoldering multiple myeloma (SMM): Predictive value of free light chains and group-based trajectory modeling (GBTM).

Abstract

8031 Background: A serum free light chain ratio (FLCR) ≥ 100 and bone marrow plasma cell (BMPC) ≥ 60% have recently been classified as myeloma defining events to identify a group of SMM patients at high risk (ie. > 50% risk at 2 years) of organ damage. However rates of progression for FLCr range from 30% to 98%. Reasons for the discrepancy include selection/referral bias, patient heterogeneity, and lack of consideration of disease evolution. The objective of this study was to determine the predictive value of baseline FLCR > 100 and BMPC > 60% in our patient population and also to determine the significance of m spike and FLC evolution using GBTM. Methods: We retrospectively investigated the predictive value of these events, including involved–uninvolved FLC (dFLC), in 273 consecutive SMM patients seen at our institution between 2010 and 2015. GBTM uses the outcomes of groups of individuals with similar trajectories over time (vs apriori assumptions of % change) to model population trajectories. Results: In patients with FLCR ≥ 100 at diagnosis, the median time to progression was 23 mos with 2-year progression of 52%. In patients with BMPC ≥ 60% at diagnosis, median time to progression was 25 months with 2-year progression of 47%. For 111 patients available for analysis, GBTM two distinct trajectories in dFLC during the 1st year. 18% of patients fell into a high risk group experiencing a 171% increase in dFLC at 1 year vs the remaining patients only had a 16% increase. The high risk group had a median TTP of 13.7 mos versus not reached for the remaining (log-rank p = 0.0063). Similarly the 25% of patients who had a 62% increase in mean m spike within 1 year had TTP of 27 mos versus the remaining 75% who had no increase had a median TTP of 84 mos (log rank p = 0.0943). Conclusions: Our results not only confirm a more modest 52% 2 year risk of PD with an FLCR > 100 and dFLC > 100, but also suggest that a high risk dFLC trajectory evolution may help identify a SMM high risk group, with a median TTP of only 13.7 mos. Results of multivariate analysis and sensitivity/specificity comparisons of baseline as well as evolving biomarkers will be presented at the meeting.