Phase 2 Study of Ruxolitinib in Patients with Chronic Neutrophilic Leukemia or Atypical Chronic Myeloid Leukemia

Background

Chronic neutrophilic leukemia (CNL) and atypical chronic myeloid leukemia (aCML) are rare myeloid malignancies with an estimated median survival of 2 years. No effective therapy has been established. The majority of CNL and <30% of aCML harbor an oncogenic mutation in colony stimulating factor-3 receptor (CSF3R-T618I) leading to ligand-independent activation of CSF3R and granulocytic proliferation. Preclinical studies reveal an important role for constitutive JAK/STAT signaling in disease manifestations. These data prompted our group to conduct the first clinical trial evaluating the safety and efficacy of ruxolitinib (a JAK1/2 inhibitor) in patients diagnosed with CNL and aCML.

Methods

The clinical trial is an open-label, phase 2 study of single agent ruxolitinib in CNL/aCML patients, regardless of CSF3R mutation status. A minimum of a 7-day washout period was required for patients on therapy aimed at reducing white blood cell count. The primary endpoint is overall response rate (partial response [PR] and complete response [CR]), which was evaluated in patients who completed 6 continuous 28-day cycles. Patients who did not reach the end of cycle 6 were considered non-responders. CR was defined as normalization of the white blood cell count, spleen volume, and marrow hyperplasia and/or dysplasia. Secondary endpoints include evaluation of ruxolitinib safety and tolerability, and clinical and laboratory correlates of response. The starting dose of ruxolitinib was guided in part by baseline platelet count. Dose modifications were allowed based on investigator discretion. Patients who remained on treatment for ongoing clinical benefit could enter an extension phase after 24 cycles of ruxolitinib treatment.

Results

At the time of data cutoff (October 2017), we enrolled 40 patients (median age 73.2 years; 17 females and 23 males) with a diagnosis of CNL (N=20) or aCML (N=20). The median daily starting dose was 20mg (range 10-40mg) and daily dose during the study was 30mg (range 10-50mg), administered twice daily. Twenty (50%) patients were positive for CSF3R-T618I mutation (15 CNL and 5 aCML). The median number of cycles completed was 7.0 cycles (range <1 cycle to 33 cycles). Among the 40 patients, 23 (58%) completed cycle 6 or greater, 5 (13%) have not completed cycle 6 but are in active treatment, and 12 (30%) discontinued treatment before the end of cycle 6. Disease progression and lack of clinical benefit were the main reasons for premature discontinuation of therapy. The overall response rate is 37% (13 of 35 patients: 11 PR [8 CNL and 3 aCML] and 2 CR [2 CNL]). Response rate by mutation status is 50% (10 of 20) in the CSF3R-T618I group and 15% (3 of 20) in the CSF3R-wildtype group (odds ratio 5.67, Fisher's exact p = 0.041, 95% CI: 1.05, 37.99). In those patients not achieving a CR (i.e., those with PR or stable disease [SD]), clinical benefit was observed in 33% of patients in one or more categories: hemoglobin improvement, platelet count improvement, ≥50% symptom burden reduction, and ≥50% spleen volume reduction as defined by an MDS/MPN International Working Group (Savona, et al. Blood 2015). At the end of cycle 6, 52% (11 of 21 evaluable cases) exhibited a ≥35% reduction in spleen volume. By the end of cycle 2 and 6, a mean reduction of 21% and 42%, respectively, was observed in the MPN-SAF total symptom score. Grade ≥3 anemia and thrombocytopenia were observed in 38% and 15% of patients, respectively. There was one grade ≥3 non-hematologic adverse event that was considered probably related or related to ruxolitinib (nutritional weight gain). The remaining grade ≥3 non-hematologic adverse events, including bleeding, infection, and death, were not considered probably related or related to ruxolitinib. There were 11 deaths (3 PR and 8 SD or no response). Disease progression was the most common cause of death.

Conclusions

Ruxolitinib provides clinical benefit in at least a third of CNL and aCML patients and is well-tolerated in this patient population. The greatest clinical benefit was among those with CSF3R-T618I mutation. However, most hematologic responses are partial in nature, reflecting a composite evaluation of the white blood cell count, spleen volume, and marrow hyperplasia and/or dysplasia. Further characterization of the clinical and genetic correlates of response, including a subset of patients with complete responses, may help optimize selection of patients for ruxolitinib therapy.