Outcome of Newly Diagnosed Acute Myeloid Leukemia with Hyperleukocytosis: Retrospective Study of JALSG AML201

Introduction:

Acute myeloid leukemia (AML) with hyperleukocytosis is considered to have a poor prognosis due to a high early death rate secondary to complications such as disseminated intravascular coagulation (DIC), tumor lysis syndrome (TLS), and leukostasis. The usefulness of pre-treatment reduction of leukemic cells is controversial. We reviewed the clinical characteristics and outcome of adult patients with newly diagnosed AML with or without hyperleukocytosis registered to the Japan Adult Leukemia Study Group (JALSG) AML201study and examined whether early death in induction therapy for this AML subtype influenced the outcome.

Methods:

We retrospectively reviewed the records of 1,022 patients enrolled in the JALSG AML201 study, which was a phase 3 clinical trial for AML newly diagnosed from December 2001 to December 2005. The JALSG AML201 study administered cytarabine from day 1 to 7 as the induction therapy combined with idarubicin or daunorubicin (Ohtake S, et al. Blood. 2011, 24;117:2358-65.). None of the patients received pre-treatment with oral hydroxyurea or leukapheresis. Patients with performance status (PS) of 4 or severe comorbidities such as septic shock were excluded. Patients' characteristics and outcomes such as rate of early death (defined as death occurring within the first thirty days of treatment), achievement of complete remission (CR), overall survival (OS) rate and cumulative incidence of recurrence (CIR) were compared between patients with hyperleukocytosis and those without hyperleukocytosis.

Result:

We analyzed retrospectively the records of 1,057 patients with AML registered with the JALSG AML201 study and excluded 35 patients whose leukocyte count or outcome was not recorded. We found 113 AML patients with an initial leukocyte count greater than 100 × 10⁹L⁻¹ (HiLC group). There were 66 males and 47 females, and the median age was 45 years (range: 16-64). Median of leukocyte count before treatment was 160 × 10⁹L⁻¹ (range: 102-382 × 10⁹L⁻¹). According to the French-American-British (FAB) classification, there were four as M0, thirty-four as M1, fifty-six as M2, thirty as M4, and nineteen as M5. Cytogenetic analysis was performed in 107 patients. Eight patients had favorable karyotype, 90 had intermediate karyotype, and nine had unfavorable karyotype. Fifteen percent of patients with hyperleukocytosis had PS of two or greater. We compared the outcome of this AML subset with 909 patients with a leukocyte count lower than 100 × 10⁹L⁻¹ (median count: 10,900, range: 0.77-99.5× 10⁹L⁻¹) who received the same induction chemotherapy as the JALSG AML201 protocol (LoLC group). There was no significant difference between the two groups for age and sex; median age of the patients in LoLC group was 47 years (range: 15-64) with 542 males and 367 females. The frequency of the patients with a favorable karyotype or PS of 0 or 1 was lower in HiLC group than in LoLC group, whereas the rate of FAB M4 or M5 was higher in HiLC as compared to LoLC group. Eighty-two patients (72.5%) in HiLC group and 714 patients (78.5%) in LoLC group achieved CR. Sixty (53%) and twenty-two (19.5%) achieved CR after one and two cycles of treatment, respectively. In contrast, 577 (63%) and 137 (15%) in LoLC group achieved CR after one and two cycles of chemotherapy, respectively. The rate of CR was not significantly different between two groups (p=0.118). However, 5 year OS was significantly lower in HiLC as compared to LoLC group (26.9% vs. 49.4%, p < 0.001). Interestingly, the rate of early death was not different between the two groups (1.65% vs. 1.77%, p = 1.0), although the rate of severe complications, such as DIC, was higher in HiLC compared to LoLC group (32% vs. 10%, p < 0.001). The frequency of brain hemorrhage (1% vs. 1%, p = 1.0) and gastrointestinal hemorrhage (4% VS. 3%, p = 0.294) was not different between two groups. One patient died of brain hemorrhage and one of DIC in HiLC group, whereas five patients died of brain hemorrhage, four of sepsis, and three of pneumonia in LoLC within the first thirty days of treatment. CIR was higher in HiLC (64.7%) as compared to LoLC group (51.6%) (p = 0.0048).

Conclusion:

We conclude that hyperleukocytosis on presentation did not significantly affect the rate of early mortality during the induction phase of the treatment. It is considered that intensive chemotherapy without pre-treatment would be possible if appropriate supportive cares were given to manage fatal complications such as DIC.