Philadelphia-Negative Clonal Hematopoiesis Following Imatinib Therapy in Patients with Chronic Myeloid Leukemia (CML).

There is an increasing number of reports of Philadelphia-negative (Ph-) clonal hematopoiesis developing in patients (pts) with CML treated with imatinib. The significance of these abnormalities remains unclear. To examine the incidence, natural history and risk factors for the development of Ph- clones, we undertook a retrospective review of 142 consecutive pts with CML treated with imatinib and followed at VGH or BCCA between June 1999 and July 2004. Four pts were excluded due to incomplete records. Out of 138 pts, there were 76 (55%) males. The median age at diagnosis was 48.6 yrs (range 19.3–82.8). At the time of initiation of imatinib, 93 (67%) pts were in stable phase, 33 (24%) in accelerated phase and 12 (9%) in blast phase. 126 pts had received prior therapy with hydroxyurea (81%), interferon (53%), hematopoietic stem cell transplantation (17%), cytarabine (11%), or other therapy (17%). The median number of prior therapies was 2 (range 0–4). The median time from diagnosis to treatment with imatinib was 12.7 mos (range 0–240.1) and the median duration of imatinib therapy was 15.2 mos (range 0.6–45.1). Best responses to imatinib were complete hematologic response (96%), major cytogenetic response (53%) and complete cytogenetic response (43%). The event-free survival (EFS) at 3 yrs from the start of imatinib was 59% (95%CI 47–70%). Of the 138 pts, 7 (5%) pts developed Ph- clonal abnormalities: −7 and +8 (2 pts), +8 (2 pts), −7 (1 pt), −X and −22 (1 pt) and t(12;16) (1 pt). One pt had −7 and 1 pt had +8 in both Ph- and Ph+ cells. For the pts who developed Ph- clonal abnormalities, the median time from diagnosis to the start of imatinib was 28.4 mos (range 0.7–143.9). Six of these 7 pts had received prior therapy. The median time from the start of imatinib therapy to the development of the Ph- clone was 11.9 mos (range 2.7–23.5). All but two of the Ph- clones were transient and none was associated with myelodysplasia. Two of the 7 pts had progressed on imatinib. There was no difference in EFS or overall survival between pts with or without Ph- clones. In univariate analysis comparing characteristics of pts with and without Ph- clones, no predictive factors could be identified. In conclusion, the development of Ph- clonal hematopoiesis in pts with CML treated with imatinib occurred in ~5% of this series and did not appear to confer a worse prognosis. The majority of the Ph- clones included −7 and/or +8 and were transient without associated myelodysplasia. Finally, the observation of −7 (to our knowledge the first such case reported) or +8 in both Ph- and Ph+ cells in the same patient could be explained by the existence of a Ph- monoclonal state prior to the BCR/ABL mutation.