Phase 2 study of azacytidine plus sorafenib in patients with acute myeloid leukemia and FLT-3 internal tandem duplication mutation

Abstract

Patients received 5-azacytidine (AZA) 75 mg/m(2) intravenously daily for 7 days and sorafenib 400 mg orally twice daily continuously; cycles were repeated at ~1-month intervals. Forty-three acute myeloid leukemia (AML) patients with a median age of 64 years (range, 24-87 years) were enrolled; 37 were evaluable for response. FMS-like tyrosine kinase-3 (FLT3)-internal tandem duplication (ITD) mutation was detected in 40 (93%) patients, with a median allelic ratio of 0.32 (range, 0.009-0.93). They had received a median of 2 prior treatment regimens (range, 0-7); 9 had failed prior therapy with a FLT3 kinase inhibitor. The response rate was 46%, including 10 (27%) complete response with incomplete count recovery (CRi), 6 (16%) complete responses (CR), and 1 (3%) partial response. The median time to achieve CR/CRi was 2 cycles (range, 1-4), and the median duration of CR/CRi was 2.3 months (range, 1-14.3 months). Sixty-four percent of patients achieved adequate (defined as >85%) FLT3 inhibition during their first cycle of therapy. The degree of FLT3 inhibition correlated with plasma sorafenib concentrations. FLT3 ligand levels did not rise to levels seen in prior studies of patients receiving cytotoxic chemotherapy. The combination of AZA and sorafenib is effective for patients with relapsed AML and FLT-3-ITD. This trial was registered at clinicaltrials.gov as #NCT01254890.

Figure 1

Previous FLT3 kinase inhibitor exposure.

Figure 2

Patient outcomes. (A) Remission duration in responders. (B) OS. (C) Event-free survival.

Figure 3

FL levels and in vivo FLT3 inhibition. (A) Plasma FL levels determined by enzyme-linked immunosorbent assay using samples collected from the current trial of 5-azacitidine/sorafenib (solid columns). For comparison (hatched columns) is shown FL levels from plasma samples collected from relapsed/refractory FLT3-ITD AML patients on the Cephalon 204 trial 2 or 4 weeks after starting salvage chemotherapy (mitoxantrone/etoposide/cytarabine). (B) FLT3 inhibition grouped according to treatment time point. PIA assays (see Materials and methods) were performed on individual plasma samples collected during therapy. Densitometry measurements from individual FLT3 PIA assays are plotted against time point. The solid lines indicate the mean level of P-FLT3 for the patient group at each time point. C, cycle; D, day.

Figure 4

Plasma levels of sorafenib versus in vivo FLT3 inhibition. Plasma levels of sorafenib at individual time points from 2 different patients are compared with the results of PIA assays from those same time points. The PIA results are expressed inversely, as percent inhibition. (A) and (B) represent results from 2 different patients.

Figure 5

Survival grouped by PIA assay results. The Inhibited group consisted of 14 patients whose P-FLT3 levels were <15% of baseline at ≥1 point during the first cycle of therapy. The Not inhibited group consisted of 8 patients whose P-FLT3 levels were never less than 15% of baseline during the first cycle of therapy. Although there was a trend toward improved survival in the Inhibited group, this finding did not reach statistical significance.

Figure 6

Outcomes by FLT3-ITD allele burden. (A) Baseline FLT3-ITD mutation burden of responders vs nonresponders. (B) FLT3-ITD mutation burden of responders versus nonresponders at 1 to 3 months after treatment initiation.