Platelet Response in Pacritinib-Treated Patients with Cytopenic Myelofibrosis: A Retrospective Analysis of PERSIST-2 and PAC203 Studies

Background: Myelofibrosis (MF) is a myeloid malignancy characterized by clonal hematopoisis, bone marrow fibrosis (BMF) and ineffective extramedullary hematopoiesis, resulting in progressive cytopenias. Thrombocytopenia is both prognostic of poor outcomes and predictive of treatment intolerance with the JAK1/2 inhibitor ruxolitinib, which exacerbates cytopenias. Pacritinib is a JAK1-sparing inhibitor of JAK2/IRAK1/ACVR1 that can be administered at full dose to patients regardless of baseline platelet count (PLT). While prior studies have shown that pacritinib is associated with PLT stability in most patients, PLT improvement has not been described outside of a recently published case report (Yacoub A, et al. JCO Precis Oncol. 2023;7:e2200523). Here, we report rates of hematologic improvement in PLTs (HI-P) with pacritinib across two clinical trials.

Methods: Patients with baseline PLT ≤100x10⁹/L on the pacritinib 200 mg BID arms of the phase 3 PERSIST-2 study (randomized ≥12 weeks prior to study termination) or the phase 2 dose-finding PAC203 study were included. HI-P was defined per International Working Group (IWG) criteria (baseline PLT <20x10⁹/L: increase to >20x10⁹/L and by at least 100%; baseline PLT 20-100 x 10⁹/L: absolute increase of ≥30x10⁹/L) in the absence of PLT transfusions, sustained over any 8 weeks while on treatment. Efficacy outcomes (reduction in spleen volume, symptom score, and BMF) were compared between HI-P responders vs non-responders (those who did not meet IWG criteria). Rates of PLT improvement were also analyzed on the best available therapy (BAT) arm of the PERSIST-2 study over the treatment period (end of study treatment).

Results: Of 117 patients randomized to pacritinib (75 from PERSIST-2, 42 from PAC203), 16% (n=19) experienced HI-P on study (as defined in methods). Additionally, 14 of the 19 HI-P patients had sustained platelet improvement over ≥12 weeks). By contrast, only 5% (4/77) of patients on BAT achieved HI-P response prior to end of study treatment.

HI-P responders on pacritinib compared to non-responders, had numerically higher median baseline PLT count (63 vs 45x10⁹/L) and hemoglobin (Hb; 10.3 vs 8.8 g/dL). A similar percentage had prior (<30 days) JAK2 inhibitor exposure (68.4% vs 66.3%) and JAK2V617F mutation (78.9 vs 74.5%); median JAK2 allele burden was low in both responders and non-responders (26.7% vs 36.9%). Among HI-P responders with prior ruxolitinib exposure, most (85%, 11/13) had recent exposure within the prior 30 days, with median dose of 10 mg BID.

PLT improvement was noted within the first 12 weeks in most HI-P responders, whereas PLT count remained stable, on average, among non-responders (Figure 1a). To evaluate if recent ruxolitinib exposure and washout of drug was responsible for HI-P, we analyzed the subset of patients with ruxolitinib exposure in previous 30 days (11/19). The mean platelet count increased from 56 to 141x10⁹/L at week 12. In the remaining 8/19 patients without recent ruxolitinib exposure (n=6 were treatment naïve), platelets increased from 59 to 102 x 10⁹/L at week 12.

While there was no difference in the magnitude of spleen volume or symptom score reduction in HI-P responders vs non-responders, a correlation with change in BMF was noted. Among the 36 patients with available bone marrow data from baseline and week 24, HI-P responders were numerically more likely to have BMF reduction (67%, n=4/6) compared to non-responders (23%, n=7/30, P=0.057, Figure 1b).

There was no difference in the rate of hemorrhagic events (by standardized MedDRA queries) in PLT responders vs non-responders (47% vs 46%), though grade ≥3 bleeding was observed at lower frequency in responders compared to non-responders (10.5% vs 16%). Other commonly reported adverse events with pacritinib occurred at similar frequency between responders and non-responders.

Discussion: PLT improvement meeting IWG criteria occurs in a subset of MF patients treated with pacritinib and does not seem to be explained by changes in spleen volume. Further studies are warranted to assess the correlation between hematologic improvement and BMF reduction with pacritinib. It is possible that pacritinib’s unique mechanism of action as an IRAK1 inhibitor could result in modulation of the bone marrow microenvironment and thrombopoiesis.

Study funded by CTI BioPharma Corp., a Sobi company.