Abstract
The day of clearance of peripheral blood blast cells (PBBC) is a prognostic marker in patients with acute myeloid leukemia (AML) treated with induction chemotherapy with cytarabine (ara-C) plus idarubicin (AI regimen: ara-C 1.5g/m2 x 4 days, idarubicin 12mg/m2 x 3 days). Earlier PBBC clearance correlates with improved overall survival (OS). We extended this observation to subsets of patients receiving the AI regimen in combination with a targeted agent such as the FLT3 inhibitor sorafenib (AI regimen plus sorafenib 400mg orally twice daily x 7 days) and vorinostat (vorinostat 500mg three times daily x 3 days followed by AI regimen), and examined the interaction of FLT3 mutational status and PBBC clearance.
PBBC clearance (defined as PBBC=0% by CBC differential) was examined for patients with non-APL AML undergoing AI alone (n=168), AI+sorafenib (n=75), and AI+vorinostat (n=102). Patient characteristics for all patients (n=345) were as follows: median age 54 years (range, 18-72), FLT3-mutated (FLT3+, includes FLT3-ITD and FLT3-D385) (95, 28%), wild-type FLT3 (FLT3-neg) (207, 60%), median WBC 5.0x109/dL (range, 0-229), median PBBC 17% (range, 0-99), median BM blasts 46% (range, 1-98). Patients for each cohort were divided based on day of PBBC clearance and survival evaluated by Kaplan-Meier curves. Comparisons of curves were carried out using a log-rank test.
The overall response rates (ORR=CR+CRp) for the AI, AI+sorafenib, and AI+vorinostat induction cohorts were 63%, 79%, and 76% respectively. We evaluated OS for each of the three cohorts individually. In our first analysis, we divided patients into 5 groups according to PBBC clearance day: group A (0-1 days), B (2-3 days), C (4-5 days), D (6-8 days), and E (>8 days). We found that OS for patients in groups A-E was significantly different only in the AI therapy cohort (p-value<0.01), and not in the AI+sorafenib (p-value=0.15) or AI+vorinostat (p-value=0.1) cohorts. Noting that earlier blast clearance generally correlated with improved OS, we simplified our analysis by dividing patients into only two groups based on blast clearance. We performed two separate analyses using a 3-day cutoff (0-3 days vs. >3 days) or a 5-day cutoff (0-5 days vs. >5 days). For a blast clearance cutoff of 3 days, OS was significantly different in the AI+vorinostat cohort (p-value=0.02) and not in the AI alone (p-value=0.27) or AI+sorafenib (p-value=0.1) cohorts. For a 5-day cutoff, OS was significantly different in the AI (p-value<0.001) and AI+vorinostat (p-value=0.04) cohorts, but not the AI+sorafenib (p-value=0.13) cohort. We next evaluated FLT3+ and FLT3-neg patients individually for all three cohorts. Using a 3-day cutoff for blast clearance, the prognosis of FLT3+ patients could be significantly differentiated for patients treated with the FLT3 targeted regimen AI+sorafenib (p-value=0.04), but not AI alone (p-value=0.64) or AI+vorinostat (p-value=0.48). In distinction, FLT3-neg patients receiving AI+vorinostat could be well differentiated (p-value=0.06), but not FLT3-neg patients receiving AI alone (p-value=0.21) or AI+sorafenib (p-value=0.6).
The prognostic significance of the day of PBBC clearance during induction chemotherapy in AML is dependent on the specific treatment used and the mutation status of the patients being treated. For patients treated with AI alone, disappearance of blasts within 5 days of induction strongly predicts superior survival compared to patients who clear blasts after 5 days. For patients treated with AI+sorafenib, there is not a strong correlation between day of blast clearance and prognosis, unless FLT3+ patients are investigated alone, where there is a significant correlation between clearing blasts within 3 days of induction and prognosis. For patients treated with AI+vorinostat, disappearance of blasts within 3 days suggests a better survival in FLT3-neg but not FLT3+ patients. Additional approaches are needed to evaluate the prognostic value of clearance of PBBC in AML in the context of targeted therapy and mutational status of disease.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
