Pacritinib in transplant-eligible myelofibrosis: Final analysis of the phase 2 HOVON-134 trial Free

Introduction: Current EBMT/ELN-guidelines recommend pre-transplant spleen-directed management with JAK-inhibitors in symptomatic myelofibrosis (MF) patients, but emphasize the unmet need for data on the use of newer generation JAK-inhibitors in this setting. Transplant-eligible patients are typically excluded from registration trials in MF. The HOVON-134 trial aimed to answer this question: Is the use of the selective JAK2/IRAK1/ACVR1 inhibitor pacritinib (PAC) feasible in transplant-eligible MF-patients, particularly with regards to spleen and symptom responses, safety and ability to proceed toward intended allogeneic stem cell transplantation (alloHSCT)?

Methods: HOVON-134 was a prospective, single-arm, phase 2 study for MF-patients with DIPSS Plus intermediate-2 and high-risk disease with intent-to-transplant, WHO performance status 0-2 and platelet count ≥25x10⁹/L. Prior treatment with ruxolitinib (RUX) was allowed if tapered and discontinued before inclusion. PAC was given orally at a dose of 200 mg twice daily in 3 to 4 cycles of 28 days (depending on donor availability) and stopped one day before start of conditioning. Per protocol alloHSCT was done with a matched related donor (MRD) or 10/10 matched unrelated donor (MUD) and a standardized reduced-intensity conditioning protocol: intravenous busulfan 3.2 mg/kg adjusted ideal body weight (days -7 to -6) and 1.6 mg/kg adjusted ideal body weight (day -5), fludarabine 30 mg/m2 (days -7 to -2) and anti-T lymphocyte globulin 10 mg/kg (days -3 to -1) for MRD and 20 mg/kg (days -3 to -1) for MUD. Prophylaxis for graft-versus-host disease (GVHD) consisted of mycophenolate mofetil (days 0 to 28) and cyclosporine A (days -3 to 100, followed by tapering). This strategy was deemed feasible if at least 80% of included patients successfully reached their intended alloHSCT. Symptom response on PAC was defined as ≥50% reduction in the MPN symptom assessment form total symptom score; spleen response was defined as ≥50% reduction in palpable spleen size below the lower costal margin. Transplant outcomes are shown for per and off protocol treated patients.

Results: Between 11 July 2018 and 16 February 2022, 61 patients were registered. Among them, 64% were JAK2V617F-mutated, 75% had DIPSS Plus intermediate-2 risk disease and 38% had prior RUX exposure. Median follow-up from registration was 55.8 months (IQR, 50.6-67.4). 36% and 43% of evaluable patients showed symptom and spleen responses respectively. In the group of patients with prior RUX exposure, 22% achieved symptom response and 18% achieved spleen response on PAC. 44% of patients had at least one treatment-emergent CTCAE grade 3 or 4 adverse event during PAC, mainly gastro-intestinal (8% CTCAE grade 3, no grade 4-5, PAC dose-modification in 3%). No patients were disqualified for alloHSCT for cardiac reasons.

58 patients (95%) proceeded to alloHSCT: 38 per protocol treatment, 20 off protocol (including use of alternative donors). Reasons for going off protocol were mainly related to study design (strict donor criteria and timeline for alloHSCT). Three patients did not proceed to alloHSCT because of disease progression in one and two deaths (triventricular hydrocephalus during PAC, possibly due to extramedullary hematopoiesis, and undetermined neurological event during conditioning). From registration, overall survival at 1 and 5 years was 80% (95% CI, 68-88) and 57% (95% CI, 43-70). Cumulative incidence of non-relapse mortality at 5 years post-transplant (on and off protocol) was 25%. For patients transplanted on protocol, one-year progression-free survival was 79% (95% CI, 62-89) and GVHD-free relapse-free survival was 42% (95% CI, 26-57). At 12 months post-transplant on protocol, grade 3-4 acute GVHD was observed in 3% of patients and severe chronic GVHD was observed in 26% of patients.In this group, no primary graft failure occurred, and a cumulative incidence of secondary graft failure of 5% was observed. Six patients received a CD34-selected stem cell boost for poor graft function (n=5) or graft failure (n=1) and seven patients received donor lymphocyte infusion for loss of chimerism (n=5) or hematological relapse (n=2).Conclusion: PAC is a feasible and effective symptomatic treatment in transplant-eligible MF-patients and its pre-transplant use does not limit higher-risk patients in proceeding to their intended alloHSCT. Ours is one of few studies in MF showing outcome of alloHSCT by intent-to-transplant.