A Phase 3 Pilot Study of Continuous Imatinib Versus Pulsed Imatinib with or without G-CSF in Patients with Chronic Phase CML Who Have Achieved a Complete Cytogenetic Response to Imatinib.

Imatinib mesylate (IM) induces complete cytogenetic responses (CCR) in the majority of patients with CML in chronic phase (CP). Despite this reduction in disease burden many patients continue to have detectable Bcr-Abl transcripts in peripheral blood. We believe that IM-resistant Ph+ haemopoietic stem cells (HSC) contribute to this residual disease. IM has been shown to selectively target proliferating CML HSC, while simultaneously exerting an antiproliferative effect on the quiescent population-causing accumulation. We have previously characterised these quiescent cells and shown that intermittent in-vitro exposure of Ph+HSC to exogenous G-CSF leads to a reduction in number of quiescent and total HSC in culture (Clin Can Res 2006, 12: 626). This randomised, multi-centre, pilot phase 3 study was established to determine the safety and efficacy of combining G-CSF with IM in patients with CP CML who had achieved CCR on IM. 45 patients were equally randomised between 3 arms: continuous IM (cIM); pulsed IM (pIM 3 weeks IM and 1 week no drug); and pIM-G-CSF (pIM-G 3 weeks IM and 1 week G-CSF). The dose of IM administered was the dose on which the patient had achieved CCR. Patients had a median age of 59y (25–76y) and had been diagnosed with CML for a median of 36m (8–171m). 21 patients had prior IFN therapy with 4 autografted. Hasford scores were 36% low, 46% Intermediate, and 18% high. At baseline 25 patients had a major molecular response (MMR cIM 7, pIM 8, pIM-G 10). Patients were assessed monthly for 1 year with clinical examination, routine blood tests and Bcr-Abl Q-RT-PCR. Significant cardiovascular events were seen in 2 patients - 1 patient died of myocardial infarction (pIM-G) and 1 survived a stroke (cIM). There was a trend to less IM-associated side effects (diarrhoea, muscle cramps) in the experimental arms, though bony pain was reported in pIM-G (5 patients). Statistical analysis (ANCOVA) was performed on Bcr-Abl readings taken during the trial and no significant difference was detectable between the 3 arms. Further endpoint analyses showed that 4 achieved MMR (4 cIM) 21 maintained MMR (6 cIM, 6 pIM, 9 pIM-G) and 7 showed disease progression with loss of MMR (1 cIM, 2 pIM, 1 pIM-G) and/or loss of CCR (2 pIM, 1 pIM-G). 4 patients were withdrawn as a result of disease progression (3 pIM, 1 pIM-G). In conclusion both experimental arms appeared safe and well tolerated. In this pilot study there was no clear difference in efficacy when either pIM or pIM-G was compared to cIM, despite the 25% effective dose reduction of IM in the experimental arms. The absence of a clear benefit means that this approach cannot be recommended as an alternative to standard IM dosing, however may be applicable to a selected group of patients who cannot tolerate standard daily dosing.