Results from a phase II trial of decitabine and ruxolitinib for patients with MPN-AP/BP

Myeloproliferative neoplasms (MPN) are chronic malignancies that can transform to the accelerated phase (AP) with 10–19% blast cells or the blast phase (BP), with ≥ 20% of blast cells. Overall survival (OS) in patients with MPN-BP with cytotoxic chemotherapy is only ~3−5 months which, therefore, highlights the urgent need for treatment options to improve outcomes. Ruxolitinib is a Janus kinase (JAK) inhibitor approved for the treatment of MPN, and decitabine is an established therapy for patients with myelodysplastic syndrome and acute myeloid leukemia, which has shown to extend survival in patients with MPN-AP.

Following the results of a phase I study, which showed that the combination of ruxolitinib and decitabine had an acceptable safety profile and was well tolerated, with an OS of 7.9 months (4.1–not reached months) and an overall response rate (ORR) of 53% in patients with MPN-AP/BP, a phase II trial was initiated. In a paper published in Blood Advances by the MPN Hub Steering Committee member John O. Mascarenhas and colleagues, the results of this trial were described.¹

Study design

Dosing schedule:

• Induction (Cycle 1, Days 1–28): Ruxolitinib, 25 mg twice daily along with decitabine on Days 8–12)
• Subsequent cycles: Ruxolitinib, 10 mg twice daily + decitabine 20 mg/m² for 5 consecutive days (Days 8–12) in a 28-day cycle

Eligibility criteria:

• ≥ 18 years
• MPN-AP or MPN-BP diagnosis coupled with prior essential thrombocythemia, polycythemia vera, or myelofibrosis diagnosis
• Eastern Cooperative Oncology Group performance status (ECOG PS) score: 0−3

Exclusion criteria included treatment with chemotherapy (except hydroxyurea) or investigational therapy within 4 weeks of the study.

The primary endpoint for this study was the best response at 6 months, while the secondary endpoints were safety and tolerability of ruxolitinib and decitabine.

A total of 25 patients were included in the phase II trial and baseline characteristics are shown in Table 1. The median age overall was 71 years and 40% of patients were diagnosed with MPN-AP while 60% presented with MPN-BP. In the MPN-AP group, there was a high proportion of female patients (70%) whereas the split was more even in the BP group. In addition, a greater proportion of patients with MPN-AP had an ECOG PS of 2 or more, compared with the BP group. Previous treatment with ruxolitinib was more common than decitabine, but still low overall.

Table 1. Baseline patient characteristics¹

Sequencing information was also available for 24 of the 25 patients included in the study, by next-generation sequencing. The most common mutations were:

• JAK2 V617F, 58%
• TP53, 38%
• TET2, 29%
• SRSF2, 29%
• EZH2, 25%
• ASXL1, 21%

To examine the effect of treatment with ruxolitinib and decitabine on the cytokine profile of patients, 42 baseline samples obtained from a previous phase I and II study (MPN-RC 109 trial), and further 34 samples, were collected at Cycle 1 Day 8, and Cycle 2 Day 1. Results were compared with control plasma cells derived from six health donors (three male, three female, all aged 30–50 years) and eight patients with chronic phase myelofibrosis.

Key findings

Notably, no complete response (CR) was achieved within the entire cohort, although two patients reached CR with incomplete count recovery (CRi). Both patients had a bone marrow with increased cellularity and increased blast rate (25% and 10%, respectively) without signs of myelofibrosis at baseline, and developed characteristics of chronic phase myelofibrosis at the end of treatment.

A partial response (PR) was recorded in 36% of patients. A total of 56% of patients did not respond, including 28% for whom a response could not be determined. The ORR (CR + CRi + PR) was 44%. Best response at 6 months is depicted in Table 2.

Table 2. Best response by 6 months¹

Best response was achieved after a median number of four cycles (range, 2–11 cycles) in 11 responders, which lasted a median duration of 3.4 months (95% CI, 2.1 to not estimated). Only 8/25 patients received six or more cycles of therapy.

Initially, the median peripheral blood blasts was 34% in the MPN-BP group compared with 12% in the MPN-AP group. Following treatment, a median reduction in peripheral blasts of 54.8% (−100% to 71.4%) was seen. A total of 89% of patients who achieved PR demonstrated a blast count ≤ 5% at time of assessment. A median reduction in splenic size of −70.5% (−100% to 0%) was achieved in patients who had a palpably enlarged spleen at baseline. Regarding mutations, it was noted that none of the three patients with the PHF6 mutation responded to treatment. No other associations were noted between mutation profile and response. Three patients achieved a molecular response which led to longer survival in two patients (10.2 and 20.0 months).

Safety

The adverse events experienced by patients are listed in Table 3. Fatigue was the most commonly experienced adverse event, closely followed by diarrhea, febrile neutropenia, and pneumonia. Grade ≥ 3 events were similar among both groups.

Table 3. AEs and treatment discontinuation¹

The median OS for the entire cohort was 9.5 months (95% CI, 4.3–12.0 months). The median OS in responders was estimated to be 9.5 months (5.8–not estimated) and 9.7 months (3.6–not estimated) in those who did not respond to treatment. The presence of the TP53 mutation was associated with a reduced survival time of 7.6 months compared with 9.6 in patients with wild-type TP53 (p = 0.78). Also, there was no impact on survival when comparing patients with less than four, and four and more, gene mutations.

To the best of the author’s knowledge, this is the first time the cytokine profile for MPN-AP/BP has been described. Certain cytokines were found to be elevated in MPN-AP/BP compared with normal controls including interleukin (IL)-8 (p < 0.01), tumor necrosis factor-α (p < 0.01), and interferon-gamma-induced protein 10 (p < 0.05). Endothelial growth factor (p < 0.01), platelet-derived growth factor-AA (p < 0.01), and platelet-derived growth factor-AB/BB (p < 0.05) were demonstrated to be reduced compared with controls. Eotaxin (p < 0.05) and GRO-α (p < 0.01) were shown to be elevated when baseline MPN-AP/BP patient samples were compared with patients with chronic-phase myelofibrosis. Patients who responded to treatment were shown to have significantly higher baseline levels of IL-8 (p = 0.05) and vascular endothelial growth factor (p = 0.05), compared with those who did not respond. The alterations in cytokine levels over the course of treatment are shown in Table 4, and it is evident that the cytokine profile was not constant during treatment and may be correlated with alterations in symptom burden.

Table 4. Changes in cytokine levels over the course of treatment, compared with baseline values¹

Conclusion

The prognosis for patients with MPN-AP/BP is generally poor and there is a great need for effective therapies. This study could confirm that the combination of decitabine and ruxolitinib is a viable treatment option for patients with MPN-AP/BP. Although numerically lower, survival of patients with the TP53 mutation did not differ significantly to those with wild-type TP53, but warrants further evaluation in larger studies given the low number of patients in the presented study. High levels of IL-8 and TNF-α compared with controls were also noted. This is in alignment with previous data showing that IL-8 is associated with poorer outcomes in myelofibrosis, while overexpression of TNF-α in patients with MPN correlates with JAK2 V617F allele burden. The description of the cytokines involved in MPN is an important step for assessing the role they play, although more work is required to ascertain the exact connection. The authors acknowledged the heterogeneity of the study population as a limitation, as both patients with MPN-AP and BP were involved. In addition, the lack of standard response criteria and challenges obtaining bone marrow assessments in these patients compromises the interpretation of results.