Fetal Hemoglobin (HbF) Induction during Initial Decitabine (DAC) Treatment of Elderly High-Risk MDS and AML Patients: A Potential Dynamic Biomarker for Outcome

Introduction:

The full effects of DNA-hypomethylating agent (HMA) treatment, the present standard for many MDS and AML patients, often become apparent only after 4-6 months, and only about half the patients eventually respond. Early indicators for subsequent hematologic response to HMAs are as yet very scarce, and urgently needed for better treatment tailoring. HbF is among the potential novel outcome predictors for epigenetic treatments, being induced in vitro and in vivo by HMAs and HDAC inhibitors (Press et al., Leuk. Lymph. 2017); recently, we could demonstrate that HbF elevated prior to HMA treatment was associated with superior outcome of MDS/AML patients (Lübbert et al., Br. J. Haematol. 2017). Therefore, we now asked whether early induction of HbF during treatment with DAC also had predictive value for subsequent outcome of these patients.

Patients and Methods:

16 MDS and 37 AML patients enrolled on two clinical trials, the 06011 EORTC-GMDSSG phase III trial of higher-risk MDS, and the 00331 phase II trial of older, non-fit AML patients, were treated with DAC. MDS patients received nine 4-hour infusions of 15 mg/m² given over 72 hours, repeated every 6 weeks for a minimum of four courses. AML patients were treated with nine 3-hour infusions of 15 mg/m² given over 72 hours, repeated every 6 weeks for four courses followed by maintenance treatment with DAC at 20 mg/m² when responding (defined by CR, PR or an antileukemic effect). HbF levels were measured in peripheral blood erythrocytes by HPLC before treatment and sequentially, i. e. after the end of each treatment course (every 6 weeks). HbF levels of 0 - 1.0% of total hemoglobin are the normal range in adults, thus HbF > 1.0% was considered induced.

Results:

HbF measurements at the end of the 2nd DAC treatment course (the first on-treatment time point with consistent HbF induction) were available for 32 patients (15/16 with MDS, 17/37 with AML). Compared to the median pre-treatment HbF of 0.8% , we observed overall induction to 1.1% after course 2, and, in patients achieving CR or PR as best overall response, a median HbF of 1.9% (1.9% MDS, 2.0% AML) at this time point compared to 0.8% (1.0% and 0.6%) in patients not attaining CR/PR (p=0.015 via Wilcoxon 2 sample test). Correlation analyses were performed to determine the association of % HbF after DAC course 2 and responses of the different hematopoietic lineages after course 4. For platelet counts, Spearman's correlation coefficient r_s was 0.49 (p=0.01); for neutrophils, r_s=0.35 (p=0.08); for Hb, r_s=0.36 (p=0.08); for % bone marrow blasts, r_s=-0.48 (p=0.03). In patients achieving platelet doubling already after 1 course of DAC, median HbF after course 2 was 1.9%, vs. 0.8% in patients without this platelet response (p=0.006). For MDS patients with induced vs. normal HbF after DAC course 2, median overall survival (OS) from end of treatment course 2 was 22.9 vs. 7.3 months (hazard ratio [HR] 0.21 [CI 0.05 - 0.87], p=0.03), progression-free survival was 7.7 vs. 2.4 months (HR 0.32 [CI 0.10 - 1.10], p=0.07), AML-free survival was 13.1 vs. 7.6 months (HR 0.42 [CI 0.13 - 1.38], p=0.15). For AML patients, median OS from end of course 2 was 20.0 vs. 10.4 months (HR 0.61 [CI 0.21 - 1.75], p=0.35), PFS was not determined. Asking whether clonal cells (derived from the underlying disease) or normal, emerging erythroid precursors synthesize the elevated HbF in response to the HMA, linear regression analyses were performed, which (in conjunction with FISH analyses) strongly suggested that HbF induction observed after 4 treatment courses occurred preferentially in the emerging non-clonal erythroid cells.

Conclusions:

HbF induction at an early time point during DAC treatment was associated with an early platelet doubling, was predictive for subsequent objective responses and, in MDS patients, for longer survival. Thus determination of HbF induction by this approach may provide a dynamic biomarker during hypomethylating therapy. Particularly in AML, HbF time point optimization appears necessary, and is conducted prospectively within the EORTC trial AML21 (inDACtion vs. induction, NCT02172872) where patients in the experimental arm receive a 10-day course of DAC as initial treatment.