Dynamics and Prognostic Impact of Peripheral Blood Blast Clearance in Patients with Acute Myeloid Leukemia (AML) Receiving FLT3 Inhibitor Therapy in Combination with Induction Chemotherapy

We have previously shown that for patients with acute myeloid leukemia (AML) treated with induction chemotherapy with cytarabine (ara-C) plus an anthracycline, the day of blast clearance from peripheral blood (PB) is a powerful prognostic marker. Earlier PB blast (but not white blood cell [WBC]) clearance portends improved overall survival (OS). We expanded our investigation to include patients with AML undergoing induction chemotherapy with ara-C (1.5g/m² ×3 days) plus idarubicin (12mg/m²×3 days) that is, the AI regimen, plus sorafenib (400mg orally twice daily).

We reviewed the clearance of PB blast and WBC (PB blasts = 0%, WBC≤0.1×10⁹/dL), for patients with AML (except APL) undergoing AI alone (n=168) or AI+sorafenib (n=75). Patient characteristics for the AI alone group (n=168) were as follows: median age 55 years (range, 19–72), diploid cytogenetics (n=57, 34%), poor cytogenetics (n=61, 36%), FLT3-ITD positive (n=15, 9%), FLT3-negative (n=122, 73%), median WBC 5.0×10⁹/dL (range, 0.3–132.3), median platelets 37×10⁹/dL (range, 1–581), median hemoglobin 9.1g/dL (range, 4.0–13.2), median PB blasts 15% (range, 1–96), median BM blasts 42% (range, 1–96). Patient characteristics for the AI+sorafenib group (n=75) were as follows: median age 52 years (range, 18–66), diploid cytogenetics (n=38, 51%), poor cytogenetics (n=10, 13%), FLT3-ITD (n=25, 33%), FLT3-negative (n=48, 64%) median WBC 5.9×10⁹/dL (range, 0.6–228.5), median platelets 51×10⁹/dL (range 7–306), median hemoglobin 9.2g/dL (range, 7.4–12.6), median PB blasts 21% (range 0–98), median BM blasts 56% (range 6–98).

The overall response rate (ORR=CR+CRp) in the AI group was 63% (58%+5%) and 50 patients (30%) were resistant to therapy. The ORR in the AI+sorafenib group was 79% (72%+7%) and 11 patients (15%) were resistant. We analyzed OS by dividing patients based on the day of PB blast clearance: group 1 (0–1 days), 2 (2–3 days), 3 (4–5 days), 4 (6–8 days), and 5 (>8 days). For the patients receiving AI induction therapy, a total of 32, 59, 38, 18, and 2 patients were included in groups 1–5, and median OS for groups 1–4 were 167, 48, 67, and 25 weeks respectively. A logrank test comparison revealed these four OS curves were significantly different (p-value=0.0015). For the AI treatment group, earlier blast disappearance corresponded with better OS. In contrast, for patients in the AI+sorafenib group, the correlation between day of PB blast clearance and OS was less clear. A total of 10, 23, 25, 6, and 3 patients were included in groups 1–5. The median OS for groups 1, 3, and 4 were 101, 74, and 66 weeks respectively, and the median survival for group 2 could not be estimated because the curve did not fall below 0.5. The difference between these four survival curves was not significantly different (p-value=0.35). Survival differences for the AI group were more clearly demarcated when patients were divided based on blast disappearance within 0–5 days or >5 days (p-value=0.0004). A similar analysis in the AI+sorafenib group revealed no significantly different OS (p-value=0.13). Among patients in the AI+sorafenib treatment group, the mean day of blast disappearance for FLT3-ITD versus FLT3-negative patients was 5.2 vs. 4.5 days, and this difference was not statistically significant (p-value=0.33). While FLT3-ITD patients who cleared their blasts within 0–3 days (n=10) tended to have better OS survival than FLT3-ITD patients who cleared their blasts after 3 days (n=13), the difference in OS curves did not reach significance (p-value=0.18).

The day of PB blast clearance is prognostic among patients receiving classic induction chemotherapy, where early clearance predicts better long-term outcomes. The addition of a targeted agent to a standard induction regimen limits the prognostic power of early PB blast clearance. We are currently investigating a novel mathematical approach to evaluate the prognostic value of clearance of peripheral blasts compared to WBC after initiation of therapy in patients receiving targeted agents during induction therapy; the approach aims to integrate the prognostic impact of FLT3 and NPM1 in combination with cytoreduction kinetics to better identify prognostically distinct groups.

Off Label Use: Plerixafor plus G-CSF as a part of conditioning in allogenic transplant for AML/MDS.