Prognostic significance of pronormoblasts in erythrocyte predominant myelodysplastic patients.

Recent studies of acute erythroleukemias have reaffirmed DiGuglielmo's syndrome (M6a, myeloblast-predominant) and disease (M6b, pronormoblast-predominant). M6c (mixed myeloblast/pronormoblast) has also been described. However, MDS is still defined according to the percentage of myeloblasts (% myeloblasts) without including the pronormoblast count. A 20-year retrospective study was performed to identify cases demonstrating >or=50% erythrocytic component and <30% calculated blasts (FAB exclusion criteria) without underlying cause (96 cases). Pronormoblast and myeloblast counts and other variables were analyzed as possible explanatory variables of the variations in survival. Considered alone, increasing % myeloblasts and/or percentage of pronormoblasts (% pronormoblasts) were significant predictors of decreasing survival. When all variables were considered as a multivariate group, the best fitting statistical model for predicting survival was a function of age, % pronormoblasts, IPSS cytopenias, platelet count, and percentage erythrocytic component. Of these, % pronormoblasts was by far the most significant. Nonappearance of % myeloblasts in this model is indicative of high correlations of this count with other variables.