Long-Term Outcome of Anemic Non Del 5q Lower-Risk MDS Refractory to or Relapsing After Erythropoiesis Stimulating Agents (ESAs)

Abstract 442 Background. ESAs are usually the first line tx of anemia in non del 5q lower risk MDS. However, not all pts respond to ESAs and median response duration is only about 2 years (Park, Blood 2008;111:574). Long-term outcome of pts who do not respond to or relapse after response to ESAs is incompletely known. We analyzed this outcome by updating a previously reported lower-risk MDS cohort of 403 pts treated with ESA in centers of the GFM (Blood 2008;111:574). Methods. We analyzed in that cohort low and int-1 (lower risk) IPSS pts with Hb<10g/dL, requiring or not RBC transfusions, who did not respond to or relapsed after ESAs, according to IWG 2000 criteria. Pts with MPN/MDS, unclassified MDS, therapy related-MDS, del 5q or 3q, IPSS int-2/high and pts who, at the time of relapse of ESAs, had progressed to AML or higher risk MDS were excluded. Pts had started ESA tx between 1998 and 2006, and data were reanalyzed 4 years after the last pt inclusion (at the reference date of 1 July 2010). 93 pts of the cohort who responded to ESAs but had not relapsed at the end of this follow up period were used for comparisons. Results. 177 pts, including 94 with primary resistance to ESAs and 83 with relapse after an initial response were analyzed. In the 94 pts with primary resistance to ESAs, M/F was 2, median age 75, WHO classification at tx onset RA, RCMD, RARS, RAEB-1 in 17%, 27%, 28%, 27% of cases, respectively (resp), karyotype fav, intermediate (int) in 86% and 14% pts, resp, IPSS low, int-1 or assignable to low/int-1 in 35%, 60% and 5% pts, resp. Median serum ferritin was 658 ng/mL and median serum EPO level (sEPO) 125 IU/L. 63% of the pts were RBC transfusion dependent (TD) (median 2 RBC units/month). Median overall survival (OS) and 3-y cumulative incidence (CI) of AML from tx onset were 43 months (mo) and 18%, resp. Among pre-tx characteristics, age >75 was associated with shorter survival (median OS 31 mo vs. not reached for age <75, P=0.01) along with int karyotype (median 26 vs 56 mo in pts with fav karyotype, P=0.005). Using multivariate analysis, only cytogenetics maintained prognostic significance for OS. No prognostic factor of AML progression was found. 21% of the 94 pts were aged <65. Their 3-y CI of AML was 19.4% vs 7.4% in pts aged >65 (P=0.93), and their median OS was not reached vs 41 mo in pts aged >65 (P=0.03). 83 pts relapsed after an initial response (IWG 2000 major and minor in 60.2% and 39.8% pts, resp) of 16.5 mo median duration (range 3–74 mo). At tx onset, M/F was 1.35, median age 74.3, WHO classification RA, RCMD, RARS, RAEB-1 in 14%, 38%, 32%, 16% of cases, resp, karyotype fav, int in 92% and 8% pts, resp, IPSS low, int-1 in 51% and 49% of pts. Median serum ferritin was 695 ng/mL and median sEPO 64 IU/L. 45% of the pts were TD (median 2 RBC units/mo). Median OS and 3-y CI of AML after relapse were 53 mo and 9.7%, resp. Median OS after relapse was 26 mo in RAEB-1 and not reached in other WHO subtypes (P=0.06) and was not influenced by the presence of multilineage dysplasia. Pts who relapsed after 24 mo had a 4.1% 3-y CI of AML vs 12.8% in pts who relapsed before 24 mo (P=0.40). Median OS was not reached in pts who relapsed after 24 mo vs 53 mo in those relapsing before 24 mo (P=0.90). No pre-tx characteristic was predictive of relapse before or after 24 mo. 16% of the 83 pts were aged <65. Their median OS after relapse was not reached at 4 years vs also not reached in pts aged >65 (P=0.17), and their 3- CI of AML after relapse was 0% vs 12% in pts aged >65 (P=0.31). In the overall pt population (ie pts with primary resistance, pts with relapse and pts with sustained response), univariate competing risk modeling found CI of death from cardiovascular causes to be correlated with TD and older age at tx onset but not with response status, while only age remained significant in multivariate analysis (HR=1.12 [1.014-1.24], P=0.02). Both older age and early failure (ie primary failure or relapse <24 mo) were associated with increased CI of death from MDS-related causes (AML, hemorrhage, infection) (HR=1.05 [1.00-1.10], P=0.04; and HR=5.64 [1.85-17.22], P=0.002, resp). Conclusions. Failure to respond or loss of response to ESAs in the absence of frank disease progression to AML or higher risk MDS was not associated with poor outcome in lower-risk MDS, except in some pt subgroups (pts with intermediate karyotype and with a diagnosis of RAEB-1). Those figures have to be taken into account for therapeutic decisions, especially in pts aged <65 years, where median survival was not reached with relatively long term follow up. Disclosures: Off Label Use: ESAs for anemia in MDS. Fenaux:CELGENE, JANSSEN CILAG, ROCHE, AMGEN, MERCK, GSK, NOVARTIS, CEPHALON: Honoraria, Research Funding.