Imatinib mesylate in the treatment of c-kit–positive acute myeloid leukemia: is this the real target?

Kindler et al¹  reported that 5 of 21 acute myeloid leukemia (AML) patients responded to treatment with imatinib mesylate, which is a c-kit inhibitor. From February 2003 to November 2003, we treated with imatinib 36 c-kit–positive AML patients who were not amenable to conventional chemotherapy. Of the patients, 51% were refractory/relapsed and 49% were previously untreated. The median age was 66 years (range, 26-79 years). The median white blood cell count was 4.8 × 10⁹/L (range, 0.7-88 × 10⁹/L). In addition to a diagnostic panel of monoclonal antibodies, the leukemic blast cells of all the patients were assessed by flow cytometry for c-kit and platelet-derived growth factor receptor β (PDGF-Rβ; phycoerythrin-conjugated anti-CD117 and anti-CD140 monoclonal antibodies; Becton Dickinson, Mountain View, CA) and by reverse-transcription–polymerase chain reaction (RT-PCR) for bcr/abl rearrangement. All of the patients had c-kit–positive blasts (median, 55%; range, 16%-96%), with a high mean fluorescence index (median, 30; range, 17-100; evaluated on 17 cases). Only one patient overexpressed PDGF-Rβ on blast cells (46%). No patient was positive for bcr/abl rearrangements. The median administered dose was 600 mg/d (range, 200-700 mg/d), for a median of 31 days (range, 2-311+ days). Nonhematologic toxicity was mild and not different from the one found in chronic myeloid leukemia imatinib trials. While on therapy with imatinib, 6 patients died (2 with multiorgan failure, 2 of disease progression, and 2 of cerebral stroke). During the follow-up, 15 patients died mainly due to disease progression. No patient achieved a complete or a partial remission. In 2 patients, the disease remained stable as defined by no peripheral blood or bone marrow modification. They are now continuing imatinib after 270 and 311 days. The patient with PDGF-Rβ overexpression had a platelet increase that lasted for one year with transfusion independence. In the meanwhile, blast cells became PDGF-Rβ negative.

Our purpose was to evaluate the in vivo effect of imatinib on c-kit–positive leukemic blasts. In accordance with the findings of Cortes et al²  and Heinrich et al,³  we did not find a significant in vivo activity of imatinib, even though the drug is a proven c-kit in vitro inhibitor. On the other hand, Kindler et al¹  obtained 5 hematologic responses, complete and partial. We agree with the hypothesis that multiple gene aberrations sustain leukemic proliferation and that c-kit itself may be useful but not necessary for leukemic cell survival and proliferation. One positive effect was observed in the PDGF-Rβ–positive patient. In vivo PDGF-Rβ–positive blast cells may be more sensitive to the drug, as previously reported.⁴  This is in line with the reports on the efficacy of imatinib in the hypereosinophilic syndrome, where PDGF-Rβ expression is secondary to the rearrangement of the PDGF-R gene with the FIP1L1 gene.^5,6  Since the percentage of AML patients with PDGF-Rβ rearrangement or overexpression is unknown, it could be interesting to screen all patients at diagnosis, for a prospective and proper use of imatinib in the patients who are positive or rearranged.