An Open-Label, Multicenter, 2-Arm, Dose-Finding, Phase 1b Study of the Combination of Ruxolitinib and Buparlisib (BKM120) in Patients with Myelofibrosis: Results from HARMONY Study

BACKGROUND: Ruxolitinib (RUX), a JAK1/JAK2 inhibitor, has demonstrated rapid and durable reductions in splenomegaly, improvements in symptoms and quality-of-life measures, and prolonged survival in patients (pts) with intermediate-2 (int-2) or high risk myelofibrosis (MF) in COMFORT trials. Phosphatidylinositol-3-kinase (PI3K) is a downstream regulator of the JAK-STAT pathway and constitutive PI3K activation has been implicated in MF. Preliminary data suggest that inhibition of PI3K/mTOR pathway has beneficial effects in MF. The aim of this study (HARMONY) is to evaluate the safety and efficacy of combining RUX and buparlisib, a potent oral pan-class I PI3K inhibitor, in pts with int or high-risk MF.

METHODS: HARMONY is a dose finding, phase 1b, two-arm open-label study investigating the combination of RUX and buparlisib in int or high risk MF pts with a palpable splenomegaly ≥ 5 cm. The study consisted of 2 treatment arms; JAK inhibitor–naive pts (arm A) and pts pre-treated with JAK inhibitors (arm B). The study completed enrollment and is currently ongoing. Dose escalation was guided by a Bayesian logistic regression model with overdose control, dependent on dose-limiting toxicities (DLTs) in cycle 1 and other safety findings. Four dose levels were investigated with each dose level consisting of ≥ 3 evaluable pts (RUX [mg bid]/buparlisib [mg qd]) 10/60, 15/60, 15/80, 20/80). Nine evaluable pts were required to determine maximum tolerated dose (MTD) and proceed to the expansion phase, where additional pts were enrolled to confirm the MTD (n=11, each arm). Efficacy was evaluated based upon reductions in spleen length (escalation phase) and in spleen volume (expansion phase). Additional end points included change in bone marrow fibrosis.

RESULTS: At data cutoff of June 01, 2015, 22 pts (11 in both escalation and expansion phase) and 20 pts (9 in escalation and 11 in expansion phase) were treated at MTD in arm A and arm B, respectively. Baseline (BL) characteristics of pts at MTD are summarized in Table. MTD was established at RUX15 mg bid/buparlisib 60 mg qd. DLTs in the dose escalation phase included grade 4 thrombocytopenia [n=3 (2 in arm A; 1 in arm B)], and grade 3 depression (n=1, arm B).  

16 pts (73%) in arm A and 11 pts (55%) in arm B remained on treatment at MTD. Five pts in arm A and 3 pts in arm B discontinued due to adverse events (AE). There were 3 deaths reported at MTD in arm B; two deaths attributed to myeloid leukemia and one to multi organ failure secondary to progressive MF (n=1).

At MTD, Grade 3/4 non-hematological AEs reported in more than 1 pt included anxiety (Arm A, n=2/22) and multiorgan failure (Arm B, n=2/20). The events of multiorgan failure were attributed to myeloid leukemia & progressive MF. Grade 3/4 hematologic AEs included anemia [Arm A, n=2/22; Arm B n= 7/20] and thrombocytopenia [Arm A, n=5/22; Arm B, n=6/20].

Overall, 82% (18/22) and 55% (11/20) of MTD pts in arms A and B achieved ≥ 50% reduction from BL in palpable spleen length at any point of time, including 15 (36%) pts who had a complete resolution of palpable splenomegaly (12 in arm A and 3 in arm B). Corresponding proportion of patients achieving ≥50% reduction in palpable spleen length at week 24 was 55% (12/22) and 20% (4/20) in arm A and arm B, respectively (Figure). In the expansion phase at week 24, 45%(5/11)  of pts vs 18%(2/11)  of pts in arm A and arm B, respectively, achieved a ≥ 35% reduction from BL in spleen volume.

At week 24, a median (range) change of -3.35 (-26.9 to 2.7) was observed in JAK2V617F allele burden from BL in arm A and 0.60 (-12.6 to 24.7) in arm B at MTD. After 24 weeks of treatment at MTD, 4 pts (n= 3 in arm A; 1 in arm B) and 19 pts (n=9 in arm A; 10 in arm B) showed an improvement and stabilization, respectively. Two pts in arm A (0 in arm B) experienced worsening of bone marrow fibrosis

CONCLUSION:

The MTD for the combination was determined to be RUX15 mg bid/buparlisib 60 mg qd. The combination was generally well tolerated and provided a clinically relevant efficacy, with a significant number of MF pts achieving complete resolution of splenomegaly, including some who previously received JAKi monotherapy.

Table: Baseline Characteristics at MTD

	Arm A (n = 22)	Arm B (n = 20)
Age (y ) median (range)	63.0 (37-83)	63.5 (50-79)
Age ≥ 65 y (%)	40.9	50.0
Male (%)	68.2	75.0
Hemoglobin (g/L) median (range)	113.5 (85-141)	94.0  (71-156)
Platelet (×109/L) median (range)	237.0- (113-627)	169.0 (81-939)
JAK2V617F positive (%)	68.2	65.0