Abstract 3819Transfusion-dependent anemia is a characteristic complication of myelodysplastic syndromes (MDS) with del(5q) chromosome abnormality. Background appears to be a haplo-insufficiency of the RPS-14 gene located within the CDR. Lenalidomide has improved the treatment of MDS with 5q deletion inducing a hematologic (erythroid) and cytogenetic response in the majority of patients. The exact mechanisms of action of lenalidomide explaining its effects on erythropoiesis and their possible prognostic relevance are not clear so far. PATIENTS:We present the first data on in-vivo changes at the level of hematopoietic stem cells (SC), erythroid committed SC (SC-e), maturation of SC-e to erythroblastic islands and normoblasts observed during lenalidomide therapy within the bone marrow from a total of 39 transfusion-dependent patients with low- or intermediate-1-risk MDS with del(5q) chromosome aberration (MDS.5q-) who were recruited and treated with lenalidomide (10 mg / d) at the European site of the MDS-003 study and whose bone marrow samples (aspirates and biopsies) were taken prospectively at 6-month intervals after start of treatment. Mature erythroid precursors were marked by anti-hemoglobin-A-antibody (HbA); immature, HbA- erythroid precursors by anti-glycophorin-C-antibody (GPC), and the SC compartment by anti-CD34 antibody. Due to their low number, the numerical densities of SC-e were estimated indirectly by the numerical density of erythroblastic islands within bone marrow whereas the total number of SC was estimated by a statistical approach assuming that SC give rise to clusters of CD34+ precursors. The results were correlated with lenalidomide dose, RPS-14, SPARC, and glycophorin A (GPA) gene expression, degree of anemia, transfusion dependence, cytogenetic and molecular (FISH) response, clonal evolution, AML-free and overall survival of patients. RESULTS:Lenalidomide therapy resulted in a marked increase of SC-e, differentiation of SC-e to GPC+HbA- erythroid precursors, pronounced production of HbA+ precursors and erythrocytes resulting in transfusion independence in 54 % of patients. Lenalidomide-induced increase of SC-e did not correlate with a rise in the total number of SC. It was more pronounced in patients with a cytogenetic response, but it occurred also in patients without a cytogenetic remission. 17 / 39 patients with a minimal therapy effect on erythropoiesis achieving the 1% percentile of normal healthy marrow and continuing for > 12 months showed an excellent prognosis with a probability of survival > 90 % 60 months after start of treatment and 120 months after diagnosis of disease irrespective of their hematologic response whereas the median survival time of the patients without minimal effect on erythroid precursors (n = 22) was shorter than 20 months after start of treatment (P < 0.0002). In multivariate survival analysis, changes of erythropoiesis occurring during lenalidomide treatment, clonal evolution of disease, and age of patients were the only variables providing independent prognostic information (P < 0.00005). Classification of disease (FAB, WHO, IPSS, WPSS), changes of peripheral blood cell counts, transfusion independence, cytogenetic and molecular response, RPS-13, SPARC, or GPA gene expression did not provide additional independent prognostic information (P > 0.05). CONCLUSIONS:Lenalidomide promotes erythroid differentiation at stem cell level as well as at more mature stages of erythropoiesis affecting both, the neoplastic clone as well as non-neoplastic hematopoiesis. The anti-neoplastic effect of lenalidomide does not seem to be relevant for the prognosis of the patients; (1) genetic instability of neoplastic SC / precursor resulting in clonal evolution on the one and (2) the capacity of neoplastic SC / precursors to respond to lenalidomide therapy with the production of SC-e and more mature erythroid precursors on the other hand appear to be significant. By combining both variables, three groups can be discriminated allowing for a therapy-specific prognostic scoring: one group with excellent prognosis (100 % survival 5 years after start of therapy) comprising ∼ 33 % of patients, another with poor prognosis (0 % survival 5 years after start of therapy) comprising ∼ 28 % of patients, and a third group with intermediate prognosis (53 % dying during this period of observation) comprising ∼ 39 % of patients (P < 0.000001). Disclosures:Buesche:Celgene Corp.: reference pathologist of the MDS-004 and the MDS-Le-Mon-5 studies. Giagounidis:Celgene Corp: Consultancy, Honoraria, Membership on an entity’s Board of Directors or advisory committees. Göhring:Celgene Corp.: reference cytogeneticist of the MDS-004 study. Schlegelberger:Celgene Corp.: reference cytogeneticist of the MDS-004 study. Knight:Celgene Corp.: Employment.