Is Hydroxyurea Deleterious in Patients with ELN Favorable AML and WBC Inferior to 50.10⁹/L?

Oral hydroxyurea (HU) is often prescribed in hyperleukocytic acute myeloid leukemia (HL-AML). A retrospective study showed that a pre-treatment with HU prior to intensive chemotherapy could improve early survival of patients with HL-AML (Mamez AC et al. Leuk Lymph 2016). We have also shown that a HU prephase before chemotherapy was independently associated with improved overall survival (OS) (aHR=0.44 [0.26-0.76], p=.003), event-free survival (EFS) (aHR=0.51 [0.31-0.85], p=.009) and disease-free survival (DFS) (aHR=0.38 [0.22-0.68], p=.001) in AML with WBC > 50.10⁹/L (Récher C et al. Blood 2016 128:1636). HU is also given in patients with a moderate leukocytosis in the interval before consultation or referral and prior to the induction. In vitro data suggested that a pretreatment of AML cells with HU could enhance the cytotoxic activity of intercalating agents in part through synchronization of AML cells in S phase (Ritch PS et al. Cancer Res 1981 et Minford J et al. Cancer Res 1984). In this retrospective study, we sought to determine whether a HU prephase, could impact response to chemotherapy and outcome in non HL-AML.

We included 646 AML patients with initial WBC <50.10⁹/L treated with induction chemotherapy in our center between 2004 and 2015. Median follow-up was 54.1 months.

91 patients were treated with HU before induction chemotherapy. Delay from AML diagnosis to HU initiation was 1 day (interquartile range [IQR], 0-5.5). Median duration of HU treatment was 5 days (IQR, 3-14.5). Compared to the 555 patients that did not receive HU, HU treated-patients were older (median: 63.3 vs 59.1 years, p=.001), had a poorer ECOG performance status (ECOG 2-4: 36.5 vs 15.7%, p<.0001), more secondary diseases (30.8 vs 22.7% p=.0941), higher WBC (median: 24.7 vs 4.2.10⁹/L, p<.0001), more frequently ELN 2010 intermediate-1 (35.2 vs 20.4%, p=.0046), more FLT3 -ITD mutations (30.8 vs 10.6%, p<.0001), more leukostasis symptoms (5.5 vs 0.4%, p=.0008), higher creatinine level (91 vs 80 µM, p<.0001), higher lactate dehydrogenase level (818 vs 558 UI/L, p<.0001). They were also more frequently admitted in intensive care unit (ICU) (24.2 vs 9.1%, p<.0001) and suffered from more grade 3-4 hemorrhagic complications (8.0 vs 2.9%, p=.0288) during induction phase.

HU treatment was associated with decreased complete remission (CR) (71.4 vs 80.2%, p=.0576) and increased early death rate (ED) (14.3 vs 7.0%, p=.0183). It was also associated with decreased EFS (9.0 vs 14.7, p=.001), DFS (10.3 vs 23 months, p=.006) and OS (11.7 vs 26.6, p<0.001), in univariate analyses. In multivariate analyses including age, secondary status, ECOG, ELN classification, WBC, ICU admission, leukostasis and allogeneic stem-cell transplantation in first CR, HU treatment had no significant impact on ED (HR=1.07, p=.868) and DFS (HR=1.03, p=.875). However, there was a significant interaction between HU treatment and ELN 2010 classification for CR, OS and EFS. Indeed, HU treatment had a significant impact only in ELN 2010 favorable subgroup (CR: HR=0.12, p=.012; OS: HR=3.35, p=.004; EFS: HR=2.19, p=.055). Contrarily, in other ELN 2010 subgroups (intermediate-1 and -2, adverse and unknown), HU treatment had no significant impact on OS and EFS.

In this retrospective study, we uncovered a potential deleterious effect of HU used before intensive chemotherapy in AML patients with WBC < 50.10⁹/L and ELN 2010 favorable risk. Molecular reasons underlying this unexpected result are unknown. The impact of HU prior to induction chemotherapy should be carefully assessed in a multicenter setting. In the absence of leukostasis symptoms or rapid increase in WBC, HU treatment before induction should be avoided in this subgroup of patients.