Search

BACKGROUND Azacitidine (AZA) is the standard treatment for myelodysplastic syndromes (MDS); however, many patients prematurely stop therapy and have a dismal outcome. METHODS The authors analyzed outcomes after AZA treatment for 402 MDS patients consecutively enrolled in the Italian MDS Registry of the Fondazione Italiana Sindromi Mielodisplastiche, and they evaluated the North American MDS Consortium scoring system in a clinical practice setting. RESULTS At treatment discontinuation, 20.3% of the patients were still responding to AZA, 35.4% of the cases had primary resistance, and 44.3% developed adaptive resistance. Overall survival (OS) was better for patients who discontinued treatment while in response because of planned allogeneic hematopoietic stem cell transplantation (HSCT; median OS, not reached) in comparison with patients with primary resistance (median OS, 4 months) or adaptive resistance (median OS, 5 months) or patients responsive but noncompliant/intolerant to AZA (median OS, 4 months; P = .004). After AZA discontinuation, 309 patients (77%) received best supportive care (BSC), 60 (15%) received active treatments, and 33 (8%) received HSCT. HSCT was associated with a significant survival advantage, regardless of the response to AZA. The North American MDS Consortium scoring system was evaluable in 278 of the 402 cases: patients at high risk had worse OS than patients at low risk (3 and 7 months, respectively; P < .001). The score was predictive of survival both in patients receiving BSC (median OS, 2 months for high-risk patients vs 5 months for low-risk patients) and in patients being actively treated (median OS, 8 months for high-risk patients vs 16 months for low-risk patients; P < .001), including transplant patients. CONCLUSIONS Real-life data confirm that this prognostic scoring system for MDS patients failing a hypomethylating agent seems to be a useful tool for optimal prognostic stratification and for choosing a second-line treatment after AZA discontinuation.

OBJECTIVES Morphology and cytogenetics are currently used to define prognosis in myelodysplastic syndromes (MDS). However, these parameters have some limits. Flow cytometry has been recently included in the diagnostic panel for MDS, and its prognostic significance is under evaluation. METHODS Marrow aspirates from 424 MDS patients were analyzed by flow cytometry to evaluate the impact of bone marrow cell immunophenotype on overall survival (OS) and leukemia-free survival (LFS). The immature compartment of myeloblasts was analyzed by the quantitative expression of CD34 (

Background. MDS may be characterized by hypocellular marrow, irrespective of their WHO classification or molecular characteristics. Their prognostic weight must still be completely evaluated. MDS with hypocellular marrow tend to be considered an aplastic anemia "overlap syndrome" or a pre-aplastic anemia stage. There are no strong specific therapy recommendations, and large studies analyzing the outcome of hypocellular MDS are lacking. While selective sensitivity to immunosuppressive therapy is suggested, evidence in this sense is controversial. Aims. We wanted to evaluate the clinical characteristics, outcome and choice of therapy of patients with hypocellular MDS and compare them with normocellular MDS. Methods. We analyzed 2559 MDS cases with complete clinical annotations and with evaluable bone trephine biopsy, enrolled in our Italian National Registry FISMonlus. In this cohort of patients, 438 had a bone marrow cellularity Results. In FISM cohort median age was 72.5 yrs in the hypocellular group and 72,3 yrs in the normocellular group; M/F were 53.2%/46.8% for hypocellular MDS vs 62.6%/37.4% in normocellular MDS. IPSS-R risk categories were distributed as follows: Hypocellular MDS Very Low 15.5%, Low 35.1%, Intermediate 30.1%, High 11.3%, Very high 8%; Normocellular MDS Very Low 12.8%, Low 37.2%, intermediate 23.7%, High 15.5%, Very High 11.4%. Global median overall survival (OS) was 77 months for hypocellular MDS and 56 months for normocellular MDS. When OS was evaluated in the different IPSS-R risk groups, Lower risk MDS cases with hypocellular BM had a median OS of 125 mos while normocellular had a median OS of 74 mos (p< .001). Higher risk MDS with hypocellular BM had 19 mos median OS vs 20 mos OS in normocellular MDS. Regarding the first line therapy, the comparison of hypocellular MDS with normocellular ones yielded the following: watch and wait 33.8% vs 31.6% for IPSS-R lower risk, 12.1% vs 16% for higher risk cases; AML like chemotherapy and HSCT were chosen for< 1% of lower risk cases overall, and in 1.7% of higher risk hypocellular MDS, while higher risk MDS with normocellular marrow received it in 6.2% of the cases. Azacitidine was first line treatment for 36.2% of the higher risk MDS patients with hypocellular BM and 25% of normocellular BM. Immunosuppressive treatments were emploied for < 1% and 1.5% respectively in lower risk cases only. Erythroid stimulating agents were administered in 42.6% and 41.2% MDS IPSS-R lower risk, hypo- and normocellular, respectively. We then focused on the validation cohort (REL registry). Median age was 74 yrs in the hypocellular group IPSS-R risk categories were distributed as follows: Very Low risk 9 %, Low 36%, Intermediate 39%, High 17 %, Very high 9%. Global median overall survival (OS) was 79 months for hypocellular MDS and 64 months for normocellular MDS. The difference was significant in very low/low IPSS-R risk groups, cases with hypocellular BM having a median OS of 103 mos vs. 69 mos of normocellular cases (p=.011). No significant differences were present in higher disease risk categories. No significant difference was noticed on first line treatment of choice between hypocellular an normocellular MDS. Immunosuppressive treatments were used in a very low proportion of cases (2% and 3% respectively). Conclusion. Our results are based on an unbiased analysis of "real life" MDS patients with hypocellular BM compared to normocellular ones. Clinical characteristics between the two groups were not significantly different in terms of age, gender, and distribution in the various IPSS-R risk categories. The outcome of the hypocellular marrow-MDS cases was better in comparison with that of normocellular MDS, with significantly longer OS in IPSS-R lower risk cases. The advantage in OS for hypoplastic MDS wasn't present for IPSS-R higher risk cases. Finally, the choice of first line therapy doesn't seem to be influenced by the BM cellularity, with a surprising very low proportion of patients receiving immunosuppressive agents, despite several guideline recommend of this treatment for hypoplastic MDS. Disclosures Santini: Novartis: Honoraria; Otsuka: Consultancy; Celgene: Honoraria, Research Funding; AbbVie: Membership on an entity's Board of Directors or advisory committees; Amgen: Membership on an entity's Board of Directors or advisory committees; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees. Finelli:Novartis: Consultancy, Speakers Bureau; Celgene: Research Funding, Speakers Bureau; Janssen: Consultancy, Speakers Bureau. Gaidano:Janssen: Consultancy, Honoraria; AbbVie: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Morphosys: Honoraria; Roche: Consultancy, Honoraria. Oliva:Celgene: Consultancy, Other: Royalties, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Amgen: Consultancy, Speakers Bureau; Janssen: Consultancy, Speakers Bureau; La Jolla: Consultancy; Sanofi: Consultancy, Speakers Bureau. Cilloni:celgene: Membership on an entity's Board of Directors or advisory committees; Janssen: Membership on an entity's Board of Directors or advisory committees.

Introduction Azacytidine(AZA) is the current standard of care for patients with high-risk myelodysplastic syndrome (MDS) in Europe. AZA has shown a survival advantage when compared with conventional therapies and has also shown activity in IPSS lower-risk patients. However, about 40% of patients do not respond and most patients show a loss of response within 2 years. Treatment options for MDS patients who progressed while on, failed to respond to, or became intolerant tohypomethylatingagents are scarce and it has been shown that overall survival (OS) is extremely short. Objectives of this study were to describe in a cohort of real life MDS patients treated with AZA, the reasons to discontinue treatment, and to evaluate their clinical outcomes. Methods We present here the results of a real life study from a large group ofnon selectedpatients recorded in the MDS registry of FondazioneItalianaSindromiMielodisplastiche(FISM). Only patients treated with AZA from January 2009 to June 2014 were considered for the analysis. All types of conventional published schedules of AZA were allowed. The primary end point was the evaluation of OS from start of AZA treatment to the date of death from any cause. Secondary end points were clinical response, cause of discontinuation, salvage treatments and OS from discontinuation of the drug. Results Between January 2009 to June 2014 1799 newly diagnosed MDS patients were enrolled , and 420 received AZA; 271 patients received AZA as 1st line treatment, 115 patients as 2nd line treatment, 34 as a line ≥3rd. Median age was 73 years (IQR: 67-77); 261 patients (62%) were male. WHO diagnosis was RA or RARS (n=27, 6%), RCMD with or without RS (n=62, 15%) AREB-1 (n=126, 30%), AREB 2 (n=190, 45%), other subtypes (n=15, 4%). At start of AZA therapy IPSS score was low in 11 (3%), int-1 in 80 (19%), int-2 in 143 (34%), high in 54 patients (13%), and not available in 132 patients (31%). Forty-three (47%) low and int-1 risk MDS patients had a transfusion-dependent anemia. Patients received a median of 7 courses of treatment (range 3-12). Twenty-four patients (6%) received concomitant erythropoietin therapy. OS at 1 year from beginning of AZA therapy was 73% for the whole cohort (420 pts)(95%CI: 0.69-0.78), and median OS was 23 months, 25 months for patients with IPSS lower-risk MDS and 21 months for those with IPSS higher risk MDS (log-rank test: 0.72). OS after discontinuation of AZA was 8 months. Clinical response was reported according to IWG criteria only in 288/420 patients (69%); 94 patients (33%) achieved a hematological response, that was complete in 35 patients (12%) and partial in 59 (20%), 78 patients (27%) had stable disease while 116 (40%) patients did not respond. Response was achieved after a median of 6 cycles (IQR: 4-11), in both lower and higher risk MDS patients. After a median follow up of 16 months (IQR:7-35) in 97 patients (23%) AZA therapy was still ongoing and in 323 has been discontinued (77%). Interruption of treatment was due to loss of response/worsening of hematological parameters in 24 patients who achieved complete or partial response (7%) and in 20 patients with stable disease (6%). AML evolution was the cause for interruption in 105 cases (32%), death in 28 (9%), toxicity or poor compliance in only 26 (8%), while clinical decision of the treating hematologist determined interruption in 22 cases (7%). In additional 98 patients AZA was discontinued early for reasons not reported by the treating physician (30%). Of the 323 patients who discontinued AZA 10 (3%) were managed with intensive AML-like chemotherapy, 17 (5%) underwent an allogeneic HSCT, 18 (6%) received low-dose chemotherapy, 42 (14%) other treatments and 236 patients (73%) received only transfusions and other supportive therapy. Conclusions Our data confirm that AZA therapy is effective for MDS patients, both with higher and lower IPSS risk disease. Response rate is consistent with what previously reported and median OS is 23 months. The interesting observation is that at 16 months, 77% of patients had discontinued treatment, either for progression or loss of response (45% of cases) and only in 8% of cases for reported toxicity. As there are few treatment options after AZA interruption, it is important to establish the reasons other than progression yielding to a stop in therapy, in order to avoid too early discontinuation and loss of survival advantage. Disclosures Finelli: Celgene: Research Funding; Celgene: Other: Speaker fees; Novartis: Other: Speaker fees. Angelucci:Novartis oncology, celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees. Santini:Celgene: Consultancy, Honoraria, Research Funding; Janssen: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Astex: Consultancy; Amgen: Consultancy; Onconova: Consultancy.

Morphology and cytogenetics are currently used to define prognosis in myelodysplastic syndromes (MDS). However, these parameters have some limits. Flow cytometry has been recently included in the diagnostic panel for MDS, and its prognostic significance is under evaluation.Marrow aspirates from 424 MDS patients were analyzed by flow cytometry to evaluate the impact of bone marrow cell immunophenotype on overall survival (OS) and leukemia-free survival (LFS). The immature compartment of myeloblasts was analyzed by the quantitative expression of CD34 (3% vs. ≥3%), CD117, and CD11b(-) /CD66b(-) (5% vs. ≥5%); myeloid maturation was analyzed by the expression of CD11b(+) /CD66b(++) (15% vs. ≥15%) and CD11b(+) /CD66b(+) (25% vs. ≥25%).In univariate analysis, the expression of immaturity markers (CD34(+) , CD117(+) , and CD11b(-) /CD66b(-) ) was associated with shorter LFS and OS (P 0.0001); higher expression of differentiation markers (CD11b(+) /CD66b(++) and CD11b(+) /CD66b(+) ) was associated with longer LFS (P 0.0001 and P = 0.0002, respectively) and OS (P 0.0001). In multivariate analysis, expression of CD34(+) (P = 0.007), CD117(+) (P = 0.013), and CD11b(+) /CD66b(++) (P = 0.023) retained independent prognostic value for OS, while only the expression of CD34(+) was a prognostic factor for LFS (P = 0.0003). Two different risk groups were defined according to the presence of 0-1 or ≥2 of these factors with significant different LFS and OS (P 0.0001). This score showed prognostic value in predicting survival even in subanalysis according to IPSS and WHO subgroups.Flow cytometric analysis in MDS may provide meaningful prognostic information. Blast percentage expressed as CD117(+) or CD34(+) cells and the quantitative assessment of myeloid maturation showed prognostic value for survival.

Abstract 943 Introduction: Arsenic trioxide (ATO) inhibits angiogenesis, induces apoptosis and differentiation in variety of malignant cells. Two multicenter phase II studies have documented that ATO has as a moderate activity both in low and high risk myelodysplasia with 19-21% haematological improvement. Ascorbic acid in “vitro” increases in tumor cells the activity of ATO by depleting GSH and inhibiting NF-kb. High EVI-1 levels have been reported to predicted the response to arsenic trioxide and thalidomide combination therapy. We report the results of a multicenter phase II study with combination of arsenic trioxide and acid ascorbic in myelodysplastic patients. Patient eligibility included:a) pts with 3q26 rearrangement or high EVI-1 levels; b) non-RAEB pts at low-intermediate/1 risk who failed a previous therapy; c) RAEB or high-intermediate-2 risk patients not elegible to chemotherapy and/or bone marrow transplant. Study design: Arsenic trioxide was administrated intravenously over 1 hour at the loading dose of 0.30 mg/kg/day for 5 consecutive days, followed by 0.25 mg/kg/day twice weekly for 15 weeks. Ascorbic acid 1000 mg was given IV within 30 minutes after each arsenic trioxide infusion. Response evaluation was scheduled after two and four months of treatment according to the IWG criteria (Cheson 2000) Patients: 44 pts were enrolled .Median age was 71 years (range 47–80). WHO categories were non-RAEB (15 pts), RAEB-1 (12 pts), RAEB-2 (17 pts). 23 patients had a low/intermediate-1 IPSS score; 18 had a high/intermediate-2 risk. Risk category, could not be assessed in 3 pts lacking of cytogenetic data. Before starting therapy EVI-1 was highly expressed in 13 out of 34 evaluated pts (38%); high WT-1 levels was documented in 25 out of 35 pts (71%). 31 pts (57%) were transfusion dependent at baseline. Results: Ten out of 44 evaluable patients obtained a response (23%). They included 1 complete remission, 2 major erythroid hematologic improvement (HI-E major), 3 minor HI-E, 2 major neutrophil improvemet (HI-N) and 2 major trilineage responses. Response rate was 35% in lower risk IPSS patients and 6% in higher risk pts (p 0.05).Only 3 pts with high EVI-1 expression achieved a response. In 8 out of ten responders, the response was evident within the first 8 weeks of treatment. Twenty three (52%) patients discontinued treatment because disease progression (11%), severe adverse events (32%), drug unrelated adverse events (5%), withdrawal of consent (5%). Severe neutropenia and thrombocytopenia were observed respectively in 45% and 23% pts. The other G3-4 adverse events observed were: ATRA-like syndrome (9%), cardiotoxicity (11%) infection (11%), hepatotoxicity (9%). No toxic death was observed. Summary and Conclusions: The combination of ATO and ascorbic acid is tolerable and it is active in about 25% of MDS patients. The addition of ascorbic acid to ATO does not increase neither the toxicity nor the response rate to ATO. The tolerability of this regimen is reduced in elderly and high risk patients. In our study high EVI-1 transcript levels were not confirmed to predict the response to ATO. Disclosures: Off Label Use: Arsenic trioxide is yet approved for the treatment of MDS, even if its efficacy has been shown in other phase II studies.

Introduction Erythropoiesis stimulating agents (ESAs) have been demonstrated to reduce the transfusion requirements in International Prognostic Scoring System (IPSS) low/Int-1 risk MDS patients and response to ESAs is associated with an improved quality of life. Response to ESAs varies from 20 to 60% according to the different studies and an association between ESAs response and a better overall survival (OS), without any disease evolution to AML has been reported. Aim of our study is to analyze the real life use of ESAs in Italy and to assess which group of lower risk MDS patients can benefit more by this treatment. Methods From 1999 to 30 Jun 2013, 2487 patients affected by MDS were included in the Network of the Italian MDS Registries. Morphological MDS diagnosis was confirmed. Among them, 1411 IPSS low/int-1 patients with follow-up data available were selected. Furthermore we excluded the following cases: 87 patients who were treated at any time by azacytidine (n=61) or lenalidomide (n=26), 95 patients with WHO 2002 diagnosis of CMML and aCML, 96 patients with incomplete data, 84 patients unhomogeneously treated with ESAs other than EPO alpha. We obtained a final data base of 1049 cases. OS and leukemic evolution (LE) according to EPO treatment were analyzed using the Kaplan-Meier product-limit survival curve estimates and log-rank tests. Results Median age was 74 years, 59% of the patients were male and 41 % female. WHO 2002 classification was as follows: 296 (28 %) Refractory Anemia, 347 (33 %) refractory cytopenia with multilineage dysplasia, 94 (9%) refractory anemia with ringed sideroblasts, 46 (4 %) Del 5q, 188 (18 %) refractory anemia with excess of blasts and 78 (7 %) MDS unclassifiable. Median hemoglobin level at baseline was 10.0 g/dL. One hundred and twenty patients (11%) were previously transfused with a Hb baseline level less than 8.0 g/dL; 448 not yet transfused patients (43%) presented a mild anaemia with Hb ranging from 8 to 10 g/dL and 481 patients (46%) had a Hb level higher than 10 g/dL. Patients classification according to IPSS was as follows: low 55% (n=575); int-1 45% (n=474). 335 (32%) out of 1049 patients received EPO alpha treatment while the remaining 714 patients (68%) were supported without any ESAs treatment. The rate of erythroid response to EPO therapy was 61%. Median duration of response was 82 weeks. The response rate to EPO was significantly higher in non transfused than in transfused patients: 69% versus 14% respectively (p value < 0.001). Only 52 out of 1049 patients showed a LE and a higher incidence of LE in EPO treated patients was not observed. Considering the whole group of patients, the OS of EPO treated patients was not statistically different from untreated ones. However EPO responders patients showed a better OS in comparison to both the non responders and non-EPO treated patients (p value < 0.05). When we stratified patients according to the basal hemoglobin level (lower than 8 g/dL, from 8 to 10 g/dL and higher than 10 g/dL) in patients with Hb ranging from 8 to 10 g/dL we observed a significant improvement in OS following successful EPO treatment (median survival: EPO vs. non-EPO 64 vs. 43 months respectively; p 10 g/dL. Even if adverse events were not prospectively recorded in the Registry, no frequent thromboembolic events was reported in EPO group in comparison to non treated patients. Conclusions Our survey can suffer from selection bias due to the limits typical of a voluntary-based registry, but it can be considered a good real life picture of EPO alpha therapy in lower risk Italian MDS patients. Our data confirm that EPOa treatment is globally safe and that patients who can benefit more are the mild anemic untransfused subjects, with baseline Hb level ranging from 8 to 10 g/dL. Disclosures: Santini: Celgene: Honoraria; Novartis: Honoraria; GSK: Honoraria; Janssen: Honoraria.

Background: Erythroid stimulating agents (ESAs) are a common treatment for lower risk anaemic MDS patients (pts). The response rate varies from 20 to 60% according to the different published studies and ESAs are recommended in International Guidelines. Many clinical and biological parameters have been proposed as predictors of response. Little is known about ESAs therapy in real life. Purpose:We performed an analysis based on the network of regional Italian MDS Registries in order to assess which group of patients could benefit more of ESAs treatment. Materials and Methods: From 2487 pts enrolled in the Registry from 1999 to November 2012, we selected a group of 1411 low or intermediate-1 cases with follow-up data available. No differences in leukemic evolution rate have been observed by ESAs treatment (p=0,279). Subsequently, in order to study the effect of ESAs on survival, we excluded the 92 pts who experienced a leukemic evolution and the cases whose evolution was influenced by lenalidomide and/or azacytidine treatments. Cases treated with ESAs other than erythropoietin alpha (EPOa) were excluded because of the low number of observations and/or incomplete information and/or inhomogeneous dosages. Finally, we obtained a data base of 1110 pts treated by EPOa or observation only. We evaluated Overall Survival (OS) according to EPOa treatment considering the degree of anaemia at baseline. Results: The 1110 pts were subdivided according to Hb level as follows: 123 (11%) with Hb 10 g/dL. Three hundred and fifty six pts were treated by EPOa. The difference in OS by EPOa treatment in the whole cohort was not statistically significant. However, when considering patients not yet transfusion dependent with Hb levels ranging from 8 g/dL to 10 g/dL, there was a clear statistically significant difference in OS: median time 216 months in EPOa treated group vs 99 months in non treated pts (p=0,002). The influence of EPOa on OS in pts with Hb 10 g/dL did not reached a significant level. Additionally, no statistically significant difference was observed in deaths for thrombotic events. Conclusions: Our real life experience confirms that EPOa treatment is safe inMDS patients and it is particularly useful in prolonging OS of the selected group of lower risk pts with mild anaemia not yet transfusion dependent.

Abstract 1729 Background: Management of neoplastic patients is strongly influenced by comorbidities, especially in more advanced ages. Due to that, comorbidities evaluation is a critical issue in the global assessment of patients (pts) affected by myelodysplastic syndromes (MDS). Until now, there is no agreement on which comorbidity index (CI) is more suitable in this setting and different CI has been proposed. Recently a new MDS-specific score (MDS-CI) has been published by Della Porta et al., while the majority of CI in use has been developed in geriatric oncology setting. One of the most useful is Cumulative Illness Rating Scale of Geriatrics (CIRS-G). Aim of our study: Aim of our study was to test the usefulness of the conventional and easy to apply CIRS-G score among a cohort of MDS pts enrolled in the MDS Piedmont Registry from 1999 to 2010 in predicting OS and leukemic progression. Materials and methods: 788 patients from the MDS Piedmont Registry with CIRS-G evaluation at diagnosis were included in our statistical analysis. 78% of the patients were low and Int-1 IPSS risk, the remaining 22% were Int-2-high risk. The majority of patients (69%) carried an histological diagnosis of non RAEB MDS according to WHO classification, the remaining 31% were RAEB I-II. Age stratification was as follows: 10% up to 60, 23% from 61 to 70, 43% from 71 to 80, 24% over 80 years. Comorbidities with score up to 2 were considered mild while the ones with values higher than 2 were considered severe. We evaluated the global impairment of each patient creating two comorbidity scores based on the number of mild comorbidities (mild comorbidities score, MCS) and severe comorbidities (severe comorbidities score, SCS). Results: The majority of our patients showed only mild comorbidities and the comorbidities with the greater number of patients carrying severe grade of impairment are the cardiac (25%), hypertensive (30%) and endocrinological (20%) ones. COX analysis did not show an impact of comorbidities on leukemic progression risk while there is a statistically significant impact on overall survival of respiratory, renal, urological and osteo-muscular comorbidities (HR respectively of 1,18; 1,3; 1,3; 1,16). There is a trend of increased risk of non MDS related death in patients with severe grade of each comorbidity. Then we set up a Fine and Gray regression model in order to evaluate the global impact of comorbidities on leukemic progression and overall survival according to SCS and MCS. Neither SCS nor MCS showed an impact on the leukemic progression risk. Considering overall survival (OS), MCS showed a HR of 1,12 (p= 0,009) and moreover SCS has a strong impact on the risk of death (HR 1,59; p= 0,000). MCS remains statistically significant in low IPSS risk patients (p Disclosures: No relevant conflicts of interest to declare.

Abstract 3807 Background: The prognostic role of serum ferritin (SF) evaluation at baseline in patients (pts) affected by myelodysplastic syndrome (MDS) is still controversial. In fact, increased SF mainly due to transfusion requirement during disease history has a clear negative impact on overall survival (OS) (Malcovati et al., 2006) and also on leukemic evolution (LE) (Sanz, ASH 2008, de Swart ASH 2010) while contrasting data about its role at baseline on OS has been published. Park and colleagues (ASH 2010) failed to identify a negative prognostic impact of SF higher than 300 ng/mL in a cohort of low risk untransfused MDS patients while data from the European LeukemiaNet MDS registry identified SF as an independent prognostic factor for OS and progression-free survival in low- and int-1 MDS (de Swart, Edimburgh 2011). SF can be a marker of iron overload but also of inflammation and little is known about the impact on survival of other iron parameters such as transferrin saturation (TS) or inflammation such as C reactive protein (CRP) in MDS pts at diagnosis. Aim: Aim of our study was to evaluate the prognostic role of iron parameters and inflammation at diagnosis in MDS patients analyzing data collected in the MDS Piedmont Registry. Materials and methods: 1360 patients enrolled in the MDS Piedmont Registry (1999–2010) were analyzed. Patients with information on OS and LE and available baseline SF (n=670), TS (n=299), CRP (n=287) were included in the analysis. Survival analysis was performed using Kaplan-Meier method. Patients were stratified according to a cut off value of 800 ng/mL for SF, 40% for TS and values within or higher the normal range (0,8 mg/dL) for CRP. In order to compare survival curves, log-rank test was used. Cumulative incidence of LE, according to SF, TS and CRP levels, was calculated accounting for death from any causes as a competing event. Results: In the population with SF baseline values, 3-years OS in pts with SF < 800 ng/mL was 80.7 (95%CI: 75.8–84.8) and in pts with SF >800 ng/mL was 66.1 (95%CI: 46.7–79.8) (p= 0,006). The result seems to be confirmed in the low risk MDS subgroup in toto (n=226) and considering only the untransfused pts (136 cases) (p=0,0073 and p=0,0038 respectively) but no statistically significance in OS of high risk pts (n=108) has been observed. In subjects with available data on TS, 3-years OS for pts with TS lower than 40% was 75.0 (95% CI: 64.2–83.0) while in pts higher than 40% was 72.1 (95%CI: 59.5–81.4) (p=0,1). Finally, in pts with CRP values 3-years OS was 80.8 (95%CI: 69.7–88.2) for patients < 0,8 mg/dL and 47.2 (95%CI: 34.5–58.9) for pts > 0,8 mg/dL. Also 3-year cumulative incidence of LE was higher in pts with SF > 800 ng/mL [35.8 (95%CI: 20.3–51.2) vs 18.5 (95%CI: 14.4–22.5); p=0,002 ] and in those with CPR > 0,8 mg/dL [35.3 (95%CI: 23.9–46.7) vs 12.7 (95%CI: 5.7–19.7); p40%; p=0.172]. Conclusions: although the limits of missing data, our results suggest that high levels of SF and CRP above the normal range at baseline should have a prognostic role in MDS pts, while TS seems to have little impact on OS. Moreover SF and CRP seem to have both a negative impact on LE. Our data suggest a more important prognostic role of chronic inflammation parameters than iron overload or oxidative stress in MDS patients at diagnosis. Further prospective evaluation of more specific parameter of oxidative stress and inflammation need to be analyzed in order to confirm our preliminary observation. Disclosures: No relevant conflicts of interest to declare.

Abstract 4028 BACKGROUND. The corner stone of the WHO classification and prognostic scores of myelodysplastic syndromes (MDS) is the blast count in bone marrow. The standard cytology evaluation of at least 500 bone marrow cells is easy to perform, but some concerns arise about reproducibility of this method. Nowadays bone marrow trephine biopsy and flow cytometry are frequently considered for the diagnosis of MDS. However there is so far paucity of data comparing cytology, histology and flow cytometry in quantifying bone marrow blasts in order to differentiate non RAEB from BAEB-I and RAEB-II cases. AIM OF THE WORK. The Aim of the work was to analyse the differences and the prognostic impact of cytology, histology and flow cytometry in differentiating non RAEB from BAEB-I and RAEB-II. PATIENTS AND METHODS. Since 1999, clinical and laboratory data from 1256 new cases of MDS were prospectively recorded into the Piemonte MDS Registry. Blast count could be performed with the three different methods: BMC (bone marrow cytology) has been performed in 844 cases, BMH (bone marrow histology) in 874 cases, and BMF (bone marrow flow cytometry) in 636. In order to quantify blasts, immune-histochemistry evaluation of CD34+ cells was used in BMH, while both CD34+ and CD117+ cells were considered in BMF. Out of the total of the 636 patients analysed by BMF only 420 had an accurate and complete registration of CD34 and CD117 positivity and were considered for the present analysis. In two hundred and thirty six cases all three evaluations were contemporary available. The concordance of each diagnostic method with the others and their prognostic value were evaluated in both univariate and multivariate analyses. A comparison between BMC and BMH was available in 571 cases, between BMC and BMF in 228 cases, and between BMH and BMF in 279 cases. RESULTS. The disagreement in classifying patients as non-RAEB or RAEB-I or RAEB-II between BMC and BMH was 156/571 (27%), with BMH over-evaluating blasts in 114/571 cases (20%) and under-evaluating blasts in 42/571 cases (7%). The disagreement between BMC and BMF was 80/228 (35%), with BMF over-evaluating and under-evaluating blast percentage in comparison to BMC in 53/228 (23%) and in 27/228 (12%) cases respectively. The disagreement between BMH and BMF was present in 113/279 (41%), with BMF over-evaluating and under-evaluating blast percentage in comparison to BMH in 44/279 (16%) and in 69/279 (25%) cases respectively. In univariate analysis all three methods of quantifing blasts and differentiating non-RAEB from RAEB-I and RAEB-II retained an important prognostic value for both leukemic evolution and survival. However when the three models were tested in multivariate analysis in order to define the best predictor of leukemic evolution, BMC retained the best predictive value. CONCLUSIONS. When BMH or BMF are used instead of BMC in order differentiate non-RAEB from RAEB-I and RAEB-II, the shift to a different WHO category is evident in at least 30% of patients and BMH and BMF do not play the same role as BMC. BMC still remain the standard method to quantify blasts for classification and prognostic evaluation of MDS. Disclosures: Off Label Use: Lenalidomide in Mantle Cell Lymphoma. Boccadoro:Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen-Cilag: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding. Saglio:Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Bristol Myers Squibb: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

BACKGROUND. The International Prognostic Scoring System (IPSS) of myelodysplastic syndromes (MDS) is based on bone marrow blast count, cytogenetic features and number of peripheral cytopenias. Transfusion requirement has been recently proposed as an important risk factor (Malcovati et al JCO23, 7594, 2005), and a new prognostic system (WPSS) has been suggested, based on WHO diagnostic subgroups, cytogenetic abnormalities and transfusion requirements (2005 ASH meeting abs.789). AIM OF THE WORK. To confirm the prognostic role of transfusion requirement and to compare the prognostic value of WPSS to that of IPSS. PATIENTS AND METHODS. The Piedmont MDS register is active since 1999. Clinical and laboratory data of patients were centrally recorded through our web site. Transfusion requirement data were retrospectively obtained from blood banks when possible. IPSS was calculated according to Greenberg et al. (Blood 89, 2079Blood 89, 1997). WPSS was calculated according to Malcovati et al. (2005 ASH meeting abst 789) by summing the score values of the following three variables: cytogenetic abnormalities scored according to the IPSS: 0 for good; 1 for intermediate; 2 for poor; transfusion requirement: 0 for absent; 1 for regular; WHO category: 0 for RA, RARS, 5q-, and unclassifiable; 1 for RCMD and RCMD-RS; 2 for RAEB-I; 3 for RAEB-II. WPSS score values were stratified into the proposed five risk groups: very low (score 0), low (1); intermediate (2); high (3–4); very high (5–6). RESULTS. From June 1999 to December 2004, 762 MDS patients were registered from 37 different institutions. Transfusion information was available for 376 patients. Data on both cytogenetics and transfusion requirements were available for 202 patients who are the object of the present analysis. 132 patients (65 %) needed regular transfusions, while the remaining 70 ones (35 %) did not. In univariate analyses all variables of both IPSS and WPSS scoring systems statistically influenced overall survival (OS). The time from diagnosis to the first transfusion was variable: less than six months in 97 cases (48 %); 6 to 12 months in 15 (7 %); more than 12 months in 20 (10 %). The OS of the 132 transfused patients was significantly worse than that of the 70 non-transfused ones, but the OS of patients who began to require transfusions more than 12 months after diagnosis was not significantly worse than that of not transfused patients. Only patients with an early transfusion requirement (within the first 12 month) had a poor outcome. When considering the prognosis of WHO subgroups the difference in OS between the subgroups 0 and 1 did not reach a significant level (p>0.05). In stepwise multivariate analysis the best prognostic factors were blast count, peripheral cytopenias and cytogenetic abnormalities. When the IPSS variables were included into the Cox model, neither WHO subgroups nor transfusion requirement retained an independent prognostic value. CONCLUSIONS. The prognostic value of transfusion requirement was confirmed, but only patients requiring transfusion within the first 12 months since diagnosis showed a poor outcome. In our experience IPSS model fits prognosis better than WPSS.

Background. A leukaemic evolution is evident in less than 50% of myelodysplastic patients. They can often die of age-related problems which are independent of myelodysplastic syndrome (MDS) itself, and the best therapy is difficult to define for this group of old patients, in which aggressive strategies are at high risk and supportive care can constitute the most useful option. A systematic analysis of causes of death is so far lacking. Aim of the work. To analyse the prognosis of a large group of myelodysplastic syndromes with particular reference to the causes of death. Patients and methods. From January 1999 to June 2005, data from 783 new cases of MDS were prospectively recorded into the Piedmont MDS register through our web site. Thirty two and 68 cases were excluded because RAEB-t and CMMoL respectively. The remaining 680 patients, who are the object of the present analysis, can be subdivided according to the WHO classification as follows: 99 RAEB-II; 160 RAEB-I; 104 RCMD; 317 MDS other than RAEB and RCMD. Data regarding co-morbidity and IPSS score are available for 457 and 404 patients respectively. At the moment of the analysis, 157 deaths were recorded and causes of death were registered for 153 patients. Results. Median age was 73 (range 27–95), with 151 patients (22%) older than 80. One or more co-morbidities were present at diagnosis in 399/457 (87%). The prognostic role of both IPSS scoring system and WHO classification were confirmed. The causes of death were subdivided as follows: complications due to cytopenia and/or leukaemic transformation in 57 patients (37%); infections in 20 patients (13%); other age or co-morbidity related causes in the remaining 76 patients (50%). No significant differences of causes of death were seen according to sex, while deaths from unrelated causes increased with increasing age from 29% under 60 years up to of 61% over 80 years (test for linear trend: p=0.02). Deaths due to cytopenia, and/or leukaemic transformations, and/or infections were more frequent in patients with no co-morbidities (75%), while no differences were seen according to the number of co-morbidities: 44%, 39% and 55% for patients with respectively one, two and three associated diseases. A significant relationship was evident between diagnostic subgroups and deaths from unrelated causes: 21% for RAEB-II; 51% for RAEB-I; 53% for RCMD; 76% for MDS other than RAEB and RCMD (p Conclusions. The prognostic analysis of this group of MDS patients with attention to the causes of dearth suggest that the majority of patients die of unrelated causes. Age and co-morbidities should play a major role in defining the treatment strategy of this group of patients. Anti-leukaemic treatments should therefore be limited to a small group of patients with diagnosis of RAEB and high IPSS score. An improvement in supportive treatment should be useful for the majority of patients.