Impact of static and dynamic risk assessment in HMA-treated MDS patients undergoing stem cell transplantation.

Abstract

6568 Background: MDS risk assessed by the IPSS-R and IPSS-M at diagnosis impacts outcomes post-hematopoietic stem cell transplant (HSCT). Recent EBMT data showed no post-HSCT survival benefit from downstaging IPSS-R scores with hypomethylating agent (HMA) therapy. However, the lack of mutational data precluded evaluating dynamic IPSS-M changes. This study aimed to determine whether evaluating IPSS-M at diagnosis or pre-HSCT more accurately predicts post-HSCT outcomes in pts treated with HMA. Additionally, in the absence of consensus on the role of cytoreductive therapy in the pre-HSCT setting for pts with higher-risk MDS, we investigated whether the dynamic application of IPSS-M offers any advantages for such therapy. Methods: We analyzed 176 paired samples from higher-risk MDS pts treated with HMA followed by HSCT at Dana-Farber (n=91) and Moffitt (n=85). Disease risk was assessed by IPSS-M at diagnosis and after HMA therapy pre-HCT. Dynamic assessment was categorized as decrease (improvement), no change, or increase (progression) in IPSS-M risk category from diagnosis to HSCT. The primary outcome was post-HSCT progression-free survival (PFS). Results: In the cohort, 60% were male, with a median age of 66 yrs (range 26-79). At diagnosis, 87.5% had MDS with increased blasts and 9.7% had MDS with low blasts. Pts received a median of 4 cycles of HMA prior to HSCT, with 63.1% having MUD donors and 84.1% receiving RIC. At diagnosis, 80% were higher-risk (MH/H/VH) per IPSS-M. Post-HMA, 61.4% improved in IPSS-M, while 24.4% had no change and 14.2% progressed. The 4y PFS for the cohort was 47%, with no significant differences between centers (48% vs 47%, p=0.75). In MVA, there was no difference in prognostic accuracy between IPSS-M estimated at diagnosis and pre-HSCT (c-index: 0.635 vs. 0.645). Dynamic assessment showed a 4y PFS of 50% for both improved/unchanged IPSS-M vs 31% for progressive IPSS-M (c-index: 0.647, p=0.09). Substantial improvement in IPSS-M (≥2.5 score change) yielded a 4y PFS of 38%, comparable to those with progression (23%) and much worse than those with discreet/evident improvement (53%/56%). Pts with substantial improvement in IPSS-M had a higher proportion of VH risk MDS at diagnosis than those with discreet/evident improvement (77% vs. 28%), indicating that adverse disease biology at diagnosis negatively affected outcomes despite favorable response to HMA. Conclusions: Pre-HSCT IPSS-M assessment did not enhance post-HSCT outcome predictions compared to evaluation at diagnosis. Worsening IPSS-M correlated with worse outcomes, while improvement did not yield better results than unchanged risk. For those with high-risk disease, improvement in IPSS-M achieved through HMA does not appear to mitigate the adverse risk established at diagnosis. Thus, in HMA-treated pts, downstaging of IPSS-M pre-HSCT should not be a therapeutic goal or an endpoint for response evaluation in MDS trials.