Clinicopathologic correlates and natural history of atypical chronic myeloid leukemia.

2021 
Background There are limited data on the clonal mechanisms underlying leukemogenesis, prognostic factors, and optimal therapy for atypical chronic myeloid leukemia (aCML). Methods The authors evaluated the clinicopathologic features, outcomes, and responses to therapy of 65 patients with aCML. The median age was 67 years (range, 46-89 years). Results The most frequently mutated genes included ASXL1 (83%), SRSF2 (68%), and SETBP1 (58%). Mutations in SETBP1, SRSF2, TET2, and GATA2 appeared at variant allele frequencies (VAFs) greater than 40%, whereas other RAS pathway mutations were more likely to appear at low VAFs. The acquisition of new, previously undetectable mutations at transformation was observed in 63% of the evaluable patients, with the most common involving signaling pathway mutations. Hypomethylating agents (HMAs) were associated with the highest response rates but with a short duration of response (median, 2.7 months). Therapy with ruxolitinib was not associated with clinically significant responses as a single agent or in combination with an HMA. Allogeneic stem cell transplantation was the only therapy associated with improved outcomes (hazard ratio, 0.144; 95% CI, 0.035-0.593; P = .007). Age, platelet counts, bone marrow blast percentages, and serum lactate dehydrogenase (LDH) levels were independent predictors of survival and were integrated in a multivariable model that allowed the prediction of 1- and 3-year survival. Conclusions aCML is characterized by high frequencies of ASXL1, SRSF2, and SETBP1 mutations and is associated with a high risk of acute myeloid leukemia transformation. Response and survival outcomes with current therapies remain poor. The incorporation of age, platelet counts, bone marrow blast percentages, and LDH levels can allow survival prediction, and allogeneic stem cell transplantation should be considered for all eligible patients.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    33
    References
    1
    Citations
    NaN
    KQI
    []