Response to therapy is independently associated with survival prolongation in chronic myelogenous leukemia in the blastic phase

BACKGROUND. Achievement of minimal tumor burden, such as complete response, has been accepted as a surrogate marker for improved survival in many solid and hematologic carcinomas. Several new agents have been approved recently for orphan disease indications or unmet medical needs, based on response analyses. This has not been the case for chronic myelogenous leukemia in the blastic phase (CML-BP), despite its poor prognosis, because response has not been proven to be a valuable endpoint for survival prolongation. The purpose of this study was to analyze the effect of response in CML-BP on survival prolongation. STUDY GROUP AND METHODS. In total, 328 patients with CML-BP referred from 1989 to 1999 were studied; 311 patients received therapy for CML-BP, and 275 were evaluable for response. Blastic phase CML was defined by the presence of 30% or more blasts in the blood or bone marrow, or extramedullary disease. Treatment responses were reviewed and categorized as proposed in previous large studies. Four categoric response groups were defined further based on significant differences in outcome: cytogenetic response, hematologic response, bone marrow improvement, and failure. The association of treatment response with survival was evaluated by multivariate and landmark analyses. RESULTS. The association of response with survival was analyzed among the 275 patients who had evaluable responses, and follow-up information was documented. Univariate analysis of pretreatment characteristics found performance status, hemoglobin levels, platelet counts, peripheral blasts, additional chromosomal abnormalities, and blastic phase morphology as showing significant associations with survival (P < 0.1). A multivariate analysis found platelet counts and blastic morphology as independent significant factors associated with survival (P < 0.05). A landmark analysis conducted at 8 weeks from start of therapy showed the beneficial effect of achieving response on survival prolongation (P < 0.001). A repeat multivariate at the 8-week landmark time in the 240 patients alive at that time, which included pretreatment characteristics and treatment response, confirmed the independent significant association of morphology (P = 0.003), platelet counts (P = 0.04), and response (P < 0.001) with survival. CONCLUSIONS. Response to therapy is a significant independent factor associated with survival prolongation and maybe an acceptable therapeutic endpoint for approving new treatments in CML-BP.