Survival Advantage for Patients (pts) with Chronic Myeloid Leukemia (CML) in Accelerated Phase (AP) Treated with Imatinib.

Abstract

When treated with imatinib, pts with CML in AP have a hematologic response rate of over 80%, and 25% achieve a major cytogenetic (CG) response. These results are superior to any other therapy available and have established imatinib as the standard of care for AP CML. However, a survival benefit was not defined in these single-arm studies. With longer follow-up of pts in AP treated with imatinib, we investigated whether imatinib therapy led to a survival advantage over other therapies that have been used for this disease. For this analysis, AP was defined as the presence of any of the following: blasts ≥15%, blasts + promyelocytes ≥30%, basophils ≥20%, platelets (plts) <100 x109/L unrelated to therapy, or clonal evolution. Since 1982, 395 pts with AP have been treated at MDACC: 177 (45%) with imatinib and 218 (55%) with other therapies including IFN-α-based therapy (n=102), homoharringtonine (n=36), decitabine (n=47), daunorubicin + ara −C (n=24), or others (n=9). The rate of complete hematologic response was 82% with imatinib and 38% with others (p < 0.0001), and major cytogenetic response was 48% and 13%, respectively (p <0.0001). The median follow-up is 41 months (mo) (range, 3 – 206): 38 mo (3 to 63) for the imatinib group and 82 mo (3 to 206) for the others. A total of 249 pts have died: 67 in the imatinib cohort and 182 in the control group. The median survival was 21 months for the control group, and has not been reached for the imatinib cohort (estimated rate 53% at 3 yrs). Clinical characteristics adversely affecting survival (p<0.05) included time from diagnosis to treatment >36 mo, splenomegaly, Hgb <10 g/dl, plts <100 x109/L, peripheral or marrow blasts ≥5%, marrow basophils >5%, blasts + promyelocytes >5%, clonal evolution, and treatment with imatinib. Pts treated with imatinib were older, but had fewer pts with splenomegaly, high blasts or clonal evolution than the control group. By multivariate analysis, the independent pre-treatment characteristics associated with survival (p<0.05) were splenomegaly, anemia, presence of peripheral blood blasts, and disease duration >1 yr. After adjusting for these differences in a multivariate analysis, treatment with imatinib was the most significant prognostic factor associated with improved survival (p<0.0001). We conclude that treatment with imatinib has changed the natural history of AP CML with an expected median survival in excess of 3 years.