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.

The effect of recombinant interleukin 2 (IL2) on the in vitro and in vivo proliferation and growth of human acute leukaemia cells of both myeloid and lymphoid origin was investigated. In none of the 25 primary samples tested could a continuously in vitro growing cell line be obtained by adding IL2 to the culture medium. Although IL2 induced a proliferative signal in three of the 31 acute leukaemias analysed, the overall 3H-thymidine uptake of the neoplastic cells was significantly reduced (P less than 0.05) in the presence of IL2. The unlikelihood of an important proliferative signal triggered by IL2 was confirmed in a semisolid clonogenic assay, which failed to document an increased colony growth in the 26 samples studied. Furthermore, using a colorimetric assay as a test for cell proliferation and survival, in seven of the 11 fresh acute leukaemia samples tested a 22-40% reduction in viability was observed in the presence of IL2, while in the remaining four, IL2 was ineffective. In order to investigate the effect of IL2 in an in vivo setting, an experimental model in heavily immunosuppressed nu/nu mice was established. In no case did IL2 promote the in vivo proliferation and growth of human myeloid and lymphoid acute leukaemia cells injected in the mice. On the contrary, with seven of the eight leukaemic cell lines which gave rise spontaneously to leukaemic masses, this could be prevented when the mice received locally 300 U of IL2 three times daily for 90 d. IL2 also blocked the growth in vivo of three fresh acute leukaemia samples (two myeloid and one lymphoid). Co-culture experiments using leukaemic cell lines and increasing numbers of normal lymphocytes suggest that the inhibitory effect of IL2 is probably exerted via an indirect mechanism. These findings, coupled to the well-documented ability of IL2 to generate lymphokine activated killer cells cytolytic against leukaemic blasts, further point to the potential role of immunotherapy with IL2 in the management of patients with haematological malignancies.

In 26 myeloid and lymphoid acute leukemia patients at presentation the capacity to generate interleukin-2 (IL-2)-induced lymphokine-activated killer (LAK) cells effective against the natural killer (NK)-resistant Raji cell line, as well as the susceptibility of the blasts to normal peripheral blood (PB) LAK cells and to autologous LAK effectors was analyzed. The overall PB LAK activity against Raji cells was significantly lower in acute leukemia patients compared with normal controls (mean, 1,473 +/- 971 SD LU/10(8) LAK effectors v 3,340 +/- 1,862; P less than .001). The sensitivity of the blasts to autologous LAK cells was also significantly lower than to normal LAK effectors (517 +/- 593 LU/10(8) LAK effectors v 1,304 +/- 1,066; P less than .01). When the data were analyzed independently, four patterns of behavior could be recognized. The relatively largest group (9 of 26) included patients in whom effective LAK cells could be generated against the Raji line, but in whom the blasts were resistant to autologous PB-LAK effectors while being susceptible to normal LAK cells (defective specific LAK activity). In 5 of 26 cases, an incapacity to generate LAK activity against both allogeneic and autologous target cells was observed (defective LAK generation). In six further cases, the blasts were resistant to both allogeneic and autologous LAK populations, though the latter were effective against the Raji line (resistant blasts). The same defects could also be shown with bone marrow-derived LAK cells. Only in six cases did the leukemic blasts appear susceptible to autologous and allogeneic LAK cells. In four patients the analysis could be repeated at remission, and in three a restoration of the LAK function against the primary blasts was recorded. In the 10 cases studied at relapse, the blasts were resistant to autologous LAK effectors in nine and to normal LAK in seven. These data demonstrate that in most acute leukemia patients with active disease, a defect of the LAK machinery, either a deficient generation of LAK cells or the resistance of the blasts to LAK effectors, may be documented, pointing therefore to a possible contributory role of the LAK system in the control of leukemic cell growth. In view of the frequent normalization of the autologous LAK activity at the time of remission, immunotherapy with IL-2/LAK cells should be primarily aimed to patients with minimal residual disease.

611 Background: Improvements in hematologic parameters have been associated with iron chelation therapy (ICT) in transfusion-dependent patients with chronic anemia associated with hematologic malignancies. Data from a significant cohort of myelodysplastic syndromes (MDS) patients enrolled in the EPIC study and treated with deferasirox reported a percentage of 22.6% of erythroid responses. Several sporadic reports showed hematologic improvement in patients treated with deferoxamine or deferasirox in patients affected by myelofibrosis (PMF) and Aplastic Anemia (AA). The aim of this study was to retrospectively evaluate the hematologic response in the entire cohort of chronic anemias with iron overload receiving ICT with both deferasirox (DFX) or deferoxamine (DFO) in 6 hematological Italian centers from 1993 to 2011. Methods: 105 patients received ICT for at least 3 months. Sixteen were PMF, 8 AA, 75 MDS, 4 Chronic Myelomonocytic Leukemia (CMML), 2 Acute Myeloid Leukemia (AML). 30 patients received deferoxamine (6 PMF, 3 AA, 1 CMML, 2 AML, 18 MDS), and 68 deferasirox (9 PMF,5 AA, 3 CMML, 51 MDS), and 7 received deferasirox after a prior treatment with deferoxamine (1 PMF, 6 MDS). The median serum ferritin levels at the time of ICT was 1983 ng/ml and it was not significantly different between the two cohorts (p=0,8). Patients, at the time of ICT, had transfused a median of 30 Units of RBC (40 in the DFO cohort and 24.5 in DFX cohort, p=0.001). 25 out of 105 were receiving EPO therapy at the time of chelation, started at least 6 months before ICT, without a significant clinical improvement and three were receiving a JAK2 inhibitor started at least 1 year before ICT. Patients receiving any kind of therapy able to modify the erythroid response including azacitidine were excluded as well as patients receiving EPO started less than 6 months before ICT or JAK2 inhibitors or immunosuppressive therapy less than 12 months before. Hematological response (HR) was evaluated as follow: Achieving a RBC transfusion independency (complete HR) or Hematological improvement (HI-e) for patients showing a Hb increase of 1.5 g/dL or a reduction of 4 RBC transfusions/8 weeks (IWG 2006). Results: We retrospectively analyzed an unselected cohort of patients with transfusion dependent iron overload affected by different hematologic malignancies who received ICT outside clinical trials thus allowing the inclusion of high risk MDS/AML. 13 patients were not evaluable because they were receiving ICT for less than 3 months. 41 patients out of 92 (42.7%) evaluable patients achieved a hematologic response. In details: 18 (19,5%) became completely RBC transfusion independent. Six (1 AA, 3 RARS, 1 RCMD, 1 AML) were under DFO treatment and 12 (3 AA, 2 RA, 3 RARS, 1 RAEBII, 1 CMML, 2 PMF) under DFX. In addition, all 4 AA patients who achieved transfusion independency significantly increased the number of platelets ( median 17.000/mm3 before ICT and 35.000 and 55.000 after 6 and 12 months of ICT). Median time to response was 15 months for DFO and 3 months for DFX. 16 patients (17.3%) (6 RA, 4 RARS,1 RCMD, 1 RAEB, 4 PMF) obtained HI-e defined as a reduction of 4 U/8 weeks (5 in DFO and 11 in DFX cohorts) after a median of 6 months for both DFO and DFX. HI-e defined as an increased of 1.5 g/dL was observed in 7 patients (7.6%) ( 4 RA, 1 RARS, 1 RCMD, 1 PMF) after a median of 6 months for DFO and 3 for DFX. The hematologic improvement is not strictly related to an effective reduction of serum ferritin (p=0,4). Conclusions: Our data show a high rate of complete responses, mainly in AA and RARS but also in high risk MDS/AML representing 11% of those achieving complete transfusion independency. Notably 50% of AA achieved RBC and platelet transfusion independency. Despite the limitation due to the retrospective collection of data we suggest the ICT could result in hematologic improvement in a wide population including patients who are, at present, outside the published ICT guidelines. This study warrants further investigation on the mechanism of action of ICT in inducing erythroid response. Disclosures: Saglio: Novartis, Brystol Myers: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau.

In 26 myeloid and lymphoid acute leukemia patients at presentation the capacity to generate interleukin-2 (IL-2)-induced lymphokine-activated killer (LAK) cells effective against the natural killer (NK)-resistant Raji cell line, as well as the susceptibility of the blasts to normal peripheral blood (PB) LAK cells and to autologous LAK effectors was analyzed. The overall PB LAK activity against Raji cells was significantly lower in acute leukemia patients compared with normal controls (mean, 1,473 +/- 971 SD LU/10(8) LAK effectors v 3,340 +/- 1,862; P less than .001). The sensitivity of the blasts to autologous LAK cells was also significantly lower than to normal LAK effectors (517 +/- 593 LU/10(8) LAK effectors v 1,304 +/- 1,066; P less than .01). When the data were analyzed independently, four patterns of behavior could be recognized. The relatively largest group (9 of 26) included patients in whom effective LAK cells could be generated against the Raji line, but in whom the blasts were resistant to autologous PB-LAK effectors while being susceptible to normal LAK cells (defective specific LAK activity). In 5 of 26 cases, an incapacity to generate LAK activity against both allogeneic and autologous target cells was observed (defective LAK generation). In six further cases, the blasts were resistant to both allogeneic and autologous LAK populations, though the latter were effective against the Raji line (resistant blasts). The same defects could also be shown with bone marrow-derived LAK cells. Only in six cases did the leukemic blasts appear susceptible to autologous and allogeneic LAK cells. In four patients the analysis could be repeated at remission, and in three a restoration of the LAK function against the primary blasts was recorded. In the 10 cases studied at relapse, the blasts were resistant to autologous LAK effectors in nine and to normal LAK in seven. These data demonstrate that in most acute leukemia patients with active disease, a defect of the LAK machinery, either a deficient generation of LAK cells or the resistance of the blasts to LAK effectors, may be documented, pointing therefore to a possible contributory role of the LAK system in the control of leukemic cell growth. In view of the frequent normalization of the autologous LAK activity at the time of remission, immunotherapy with IL-2/LAK cells should be primarily aimed to patients with minimal residual disease.