Basophil-lineage commitment in acute promyelocytic leukemia predicts for severe bleeding after starting therapy

Severe hemorrhagic events occur in a significant fraction of acute promyelocytic leukemia patients, either at presentation and/or early after starting therapy, leading to treatment failure and early deaths. However, identification of independent predictors for high-risk of severe bleeding at diagnosis, remains a challenge. Here, we investigated the immunophenotype of bone marrow leukemic cells from 109 newly diagnosed acute promyelocytic leukemia patients, particularly focusing on the identification of basophil-related features, and their potential association with severe bleeding episodes and patient overall survival. From all phenotypes investigated on leukemic cells, expression of the CD203c and/or CD22 basophil-associated markers showed the strongest association with the occurrence and severity of bleeding (p ≤ 0.007); moreover, aberrant expression of CD7, coexpression of CD34+/CD7+ and lack of CD71 was also more frequently found among patients with (mild and severe) bleeding at baseline and/or after starting treatment (p ≤ 0.009). Multivariate analysis showed that CD203c expression (hazard ratio: 26.4; p = 0.003) and older age (hazard ratio: 5.4; p = 0.03) were the best independent predictors for cumulative incidence of severe bleeding after starting therapy. In addition, CD203c expression on leukemic cells (hazard ratio: 4.4; p = 0.01), low fibrinogen levels (hazard ratio: 8.8; p = 0.001), older age (hazard ratio: 9.0; p = 0.002), and high leukocyte count (hazard ratio: 5.6; p = 0.02) were the most informative independent predictors for overall survival. In summary, our results show that the presence of basophil-associated phenotypic characteristics on leukemic cells from acute promyelocytic leukemia patients at diagnosis is a powerful independent predictor for severe bleeding and overall survival, which might contribute in the future to (early) risk-adapted therapy decisions.
This work was supported by the Fundación Científica de la Asociación Española Contra el Cáncer (AECC, Madrid, Spain) and the Fundación Rafael del Pino (Madrid, Spain) and both CIBERONC (CB16/12/00400, CB16/12/00233, CB16/12/00480) and grant PI16/00787 from Instituto de Salud Carlos III (Ministerio de Economía y Competitividad, Madrid, Spain).