Aplicability Of The Predictive Model Of Response To Erythropoetic Stimulating Agents (ESA) From Myelodysplastic Syndromes (MDS) and Analysis Of Response and Overall Survival (OS) In a Series Of 99 Patients (Pts) With Chronic Myelomonocytic Leukemia (CMML) From The Spanish Registry Of MDS and The Düsseldorf-MDS Registry, Germany

Background There is scarce information about the efficacy of ESA in CMML although their use is common in clinical practice. The objective of this study was to analyze the response and OS in a series of 99 pts with CMML treated with ESA and to evaluate the feasibility of the predictive model of response to ESA used in MDS (Hellstrom-Lindberg, et al. Br J Haematol. 1997; 99: 344-51). Method Between January 1997 and March 2013 99 pts with CMML from the Spanish Registry of MDS and the Dusseldorf-MDS registry were studied. Clinical characteristics, response and OS were analyzed. Predictive model of response to ESA (0 good, 1 intermediate, and 2 poor) based on erythropoetin (EPO) level (< or ≥ 500 U/L) and red blood cell (RBC) transfusion need (< 2 or ≥ 2 RBC/month) was applied. Results 66 (67%) pts were males. Median age (range) was 75 (52-93) years. CMML subtype: myelodysplastic 58 (59%), myeloproliferative 41 (41%), CMML-I 84/98 (86%), CMML-II 14/98 (14%). CPSS score: Low/Int-1 65/90 (72%), Int-2/High 25/90 (28%). Transfusion dependence on initiation of ESA 24/86 (28%). Score based on predicted model of response to ESA: 0 43/62 (69%), 1 15/62 (24%), 2 4/62 (7%). ESA type: EPO alfa 22/94 (24%), EPO beta 16/94 (17%), EPO theta 3/94 (3%) darbepoetin 53/94 (56%). Concomitant medication: hydroxyurea 19 (39%), iron 18 (37%), steroids 4 (8%), azacitidine 3 (6%), etoposide 2 (4%), G-CSF 1 (2%), romiplostim 1 (2%) and oral chelation 1 (2%). Four pts were excluded for response analysis because they received azacitidine (3) and oral chelation (1). Response: Erythroid response (ER) 55/86 (64%), transfusion independence 5/22 (23%). ER according to CPSS (Low/Int-1 vs. Int-2/High): 71% vs. 43%, p=0,032. Median (min,max) time of ER was 4 months (0,88). Pts with 0 score according to predictive model of response to ESA presented significantly higher ER than pts with 1-2 score (77% vs. 24%, p<0.001) ([Table 1][1]). Median (range) follow-up was 2.1 years (0.1-10.5) years and median OS was 3.3 years (95%CI 2.7-4). OS of pts of Low/Int-1 risk group with ER (n= 40) was significantly higher than that of non-responding pts (n=16) (median in years (4.4, 95%CI (0.4-8.3) vs. 2.3, 95%CI (1.5-3), p<0.001). | ER according to predictive model of response to ESA (MDS) | | --------------------------------------------------------- | --------- | | | 0 | 1-2 | P | | ER n (%) | Yes | 30/39 (77) | 4/17 (24) | <0.001 | | No | 9/39 (23) | 13/17 (76) | Conclusions A high frequency of ER was observed in this series of pts with CMML, the majority belonging to Low-risk CPSS. Predictive model of response to ESA from MDS was feasible, with a similar ER than that of MDS pts. Pts with low-risk CPSS with ER to ESA had a better OS than non-ER patients Supported by RD12/0036/0029 from RTICC-Instituto Carlos III, Spain Disclosures: Germing: Celgene: Honoraria, Research Funding; Jansen-Cilag: Honoraria; Novartis: Research Funding; GSK: Research Funding; Amgen: Research Funding. [1]: #T1