Imatinib mesylate (IM) therapy is effective in patients with chronic myeloid leukemia (CML). However, its discontinuation or dose variation in patients who experience sustained molecular response is debated. The possibility of treating patients with an intermittent therapy could also be applied to the second-generation TK-inhibitors. We describe our single institute experience in patients with undetectable levels or in major molecular response of BCR-ABL transcript who reduced or were discontinued from IM therapy. We applied a model for the analysis of longitudinal data to study the BCR/ABL variation according to dose change or discontinuation.
One hundred forty CML patients came to our observation between 1985 and 2009. Among these, 89 patients in chronic phase were eventually treated with IM. Fifty-five patients were treated with IM as naive patients. Each patient's treatment history was subdivided into time periods at constant dosage. Fifty-nine patients were followed-up for a total of 288 periods at constant dosage. At the end of each period, cytogenetic and/or molecular responses were evaluated. Thirty-eight progressions were recorded: 22 molecular, 6 cytogenetic and 10 of both types. The association between progression (molecular, cytogenetic or either) and treatment dose was assessed with the aid of generalized estimating equations (GEE) models, i.e. regression models designed to account for correlation due to repeated measurements over time on the same subject. Ten patients discontinued IM therapy for a period which ranged from three to 60 months, after the patient's individual request and informed consent.
We found no association between dose and progression, not even after accounting for period length. Discontinuation of treatment was not associated to an increased risk of progression. No association with a higher risk of progression was found for periods at reduced dosage (<400mg). Four of the 10 patients who discontinued IM are still in complete molecular response and off treatment.
It is unclear whether IM can “cure” chronic myeloid leukemia but according to our data this therapy can be safely stopped or its dose varied in patients with complete cytogenetic and major molecular response of up to 18 months. Our experience suggests that withdrawal of IM therapy in CP-CML patients after achievement of a complete molecular response may result in divergent molecular outcomes. The prompt improvement seen after the restart of therapy argues against the development of resistance. The selection of resistant clones after IM exposure, and the emergence of Ph negative clones with secondary cytogenetic abnormalities, are matters of concern, particularly in patients receiving long-term IM. The improved quality of life while off therapy, and the prompt response to restart of IM therapy, suggest that the subset of patients who have sustained complete molecular response may be candidates for a tailored approach to intermittent therapy. We suggest that the same statistical analysis can also be used for other TK-inhibitors that are under study for both retrospective or prospective trials in CML.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.