CD56 expression is an indicator of poor clinical outcome in patients with acute promyelocytic leukemia treated with simultaneous all-trans- retinoic acid and chemotherapy

Purpose: Preliminary reports suggest that leukemic cell expression of CD56, a neural cell adhesion molecule, is associated with adverse clinical outcome in either acute myeloid leukemia with t(8;21) or acute promyelocytic leukemia (APL). We investigated the prognostic relevance of CD56 in a series of patients with APL who were treated homogeneously with all-transretinoic acid (ATRA) and chemotherapy. Patients and Methods: Clinicobiologic presenting features and therapeutic results were analyzed in a series of 100 patients with genetically proven APL who were treated, according to the example of the Gruppo Italiano Malattie Ematologiche Maligne dell'Adulto multicenter trial, with ATRA plus idarubicin (AIDA) and for whom data on CD56 expression were available at diagnosis. Results: Fifteen patients (15%) showed expression of CD56 in greater than or equal to 20% blasts at diagnosis and were considered as CD56⁺. No differences were found regarding age, sex, WBC and platelet counts, incidence of coagulopathy, hemoglobin and fibrinogen levels, promyelocytic leukemia/retinoic acid receptor (PML/RAR) alpha fusion type, or complete remission (CR) rate in the comparison of the CD56⁺ and CD56^- populations. Conversely, compared with patients who were CD56^-, patients with CD56⁺ APL had shorter CR duration (P = .04) and overall survival (P = .002). In the multivariate analysis, CD56 positivity and initial WBC count greater than 10 x 10⁹ cells/L retained statistical significance in overall survival (P = .04 and P = .02, respectively). Conclusion: The expression of CD56 is significantly associated with inferior CR duration and survival in patients with APL who were treated with modern front-line treatment that included ATRA and simultaneous chemotherapy. Combined with other well-established prognostic factors such as WBC count, CD56 expression at diagnosis might be used to build prognostic scores for risk-adapted therapy in APL. (C) 2000 by American Society of Clinical Oncology.