PERSIST-ence in Myelofibrosis: Q&A With Dr. John Mascarenhas

The Late-Breaking Abstract session at the 2016 ASH Annual Meeting featured much-anticipated data from the PERSIST-2 trial, which evaluated the use of the oral JAK inhibitor pacritinib in patients with myelofibrosis (MF) and thrombocytopenia. Here, Dr. John Mascarenhas from the Mount Sinai School of Medicine discusses the study’s design, rationale, and results, and potential future directions with this agent for patients with MF.

How does pacritinib differ from ruxolitinib, the only FDAapproved JAK inhibitor for myelofibrosis?

Pacritinib is a multikinase inhibitor with specificity for JAK2, FLT3, IRAK1, and CSF1R. This profile is different than that of the only currently FDA approved therapy for myelofibrosis (MF), ruxolitinib, a highly selective JAK1/2 inhibitor. Although they share anti-JAK2 activity, the other kinases that are inhibited by pacritinib likely provide an explanation for differences in the clinical responses and may underlie why pacritinib has been shown to be relatively less myelosuppressive than ruxolitinib and has even been associated with an anemia response (25% achieved transfusion independence) and platelet improvement (35% increase in mean platelet count in those with a baseline platelet count lower than 50 × 10⁹/L) in the PERSIST-1 trial.

What was the rationale for PERSIST-2?

PERSIST-2 was designed to evaluate pacritinib in patients with low platelet counts as either a direct function of their disease or as a consequence of ruxolitinib therapy. The data from phase III, randomized PERSIST-1 clinical trial indicated both safety and efficacy in patients with MF irrespective of the baseline platelet count. Thus, even in the severe thrombocytopenia cohort of patients enrolled in PERSIST-1, there was a statistically significant difference in the primary endpoint of spleen volume reduction: 33.3 percent in the pacritinib arm versus 0 percent in the best available therapy (BAT) arm (p=0.037) among evaluable patients. Also, looking at symptom improvement as a secondary endpoint in the cohort with a platelet count less than 50 × 10⁹/L, there appeared to be some benefit with pacritinib over BAT (31.8% vs. 11.1%; p=0.379), though this did not reach statistical significance. These data supported the evaluation of pacritinib in PERSIST-2 with a target population of individuals with thrombocytopenia and prior ruxolitinib therapy.

What was the patient population and study design for the PERSIST-2 trial?

This trial evaluated patients with intermediate-1, intermediate-2, and high-risk MF (either primary or post–essential thrombocythemia/polycythemia vera MF) with a baseline platelet count of less than 100 × 10⁹/L, irrespective of driver mutation status. Participants were allowed to have received previous treatment with ruxolitinib. Patients were randomized 1:1:1 to pacritinib 400 mg daily (PERSIST-1 dose), pacritinib 200 mg BID (chosen based on pharmacokinetic simulations in which 200 mg BID was predicted to have a higher Cmin and lower Cmax than 400 mg daily), and BAT, which allowed ruxolitinib therapy and was administered in 45 percent of the BAT patients. The co-primary endpoint was 35 percent reduction in spleen volume and 50 percent reduction in the total symptom score (TSS) at 24 weeks of therapy. Patients enrolled on BAT could cross over at 24 weeks or prior to 24 weeks for progression of disease.

What were the major efficacy findings regarding spleen volume and total symptom score for the three study arms?

Although the co-primary endpoint of spleen reduction (18.1% vs. 2.8%; p=.001) and symptom improvement (24.8% vs. 13.9%; p=.079) was not met in the combined pacritinib arm compared with the BAT arm, it was met in the 200 mg BID cohort (spleen reduction, 21.6% vs. 2.8%, p=.001; symptom improvement, 32.4% vs. 13.9%, p=.011). It is possible that the endpoint of symptom improvement may have been statistically significant if the full safety population (311 patients enrolled and evaluable for toxicity) was actually evaluable for the 24-week response assessment (221 patients), which was a function of the FDA-imposed clinical hold. When looking at the survival curves censored at the date of clinical hold, there is no difference in outcome.

What were the major hematologic and nonhematologic adverse events between the two doses of pacritinib and best available therapy? Were there any significant differences in the frequency of cardiac events or hemorrhagic events between the three arms?

Treatment-emergent anemia was seen in 28 percent, 24 percent, and 15 percent of treated patients in the pacritinib 400 mg daily, 200 mg BID, and BAT arms, respectively. It is worth noting that patients treated with pacritinib had a lower red blood cell transfusion requirement than those treated with BAT at both week 12 and week 24. Thrombocytopenia was seen in 33 percent, 34 percent and 24 percent of patients, respectively. Importantly, there was no significant change in mean platelet count at 24 weeks in the pacritinib treatment arms. As expected, gastrointestinal toxicity across the three arms (due to FLT3 inhibition) in the form of nausea (38%, 32%, and 11%), vomiting (21%, 19%, and 5%), and diarrhea (67%, 48%, and 15%) were more frequent in the pacritinib arms compared with BAT, and these were mostly low grade, occurring within the first month of therapy, and easily managed. This was not a major reason for study drug discontinuation.

Cardiac events, including QTC prolongation, occurred at similar frequencies — 32%, 32%, and 28% for all grade events, and 13%, 7%, and 9% for grade 3-4 events in the 400 mg daily, 200 mg BID, and BAT arms, respectively; bleeding events of all grades occurred at similar frequencies (36%, 32%, and 41%); and grade 3-4 bleeding events were more common in the BID arm (7%, 14%, and 7% respectively for QD, BID and BAT). The most common grade 3-4 bleeding events in the BID arm were epistaxis and postprocedural bleeding.

Given the findings of the study, what is your opinion about the unmet need(s) that pacritinib fulfills in the current treatment landscape of MF, and what are the next steps?

Thrombocytopenia is a recognized adverse prognostic clinical variable in MF. MF patients with thrombocytopenia (specifically platelet count <50 × 10⁹/L) are at considerable risk for bleeding, disease progression, significant symptom burden, and poor survival. Ruxolitinib is approved for MF patients with a minimum platelet count of 50 × 10⁹/L due to the expected myelosuppression seen with this agent. Therefore, those patients that are most in need of therapy are unable to enjoy the benefit of JAK2 inhibition due to limiting thrombocytopenia. Pacritinib can potentially fill this void and provide a therapeutic option for this subpopulation.

I believe that the PERSIST-2 data show the benefit of pacritinib 200 mg BID with a reasonable toxicity profile and minimal detrimental effect on the platelet count. This could be the initial approved indication, and perhaps further studies could be completed to demonstrate safety and efficacy at doses below 200 mg BID, with a goal to expand the approval for second line after 1) inability to tolerate ruxolitinib due to thrombocytopenia or 2) loss of initial response to ruxolitinib therapy. The full clinical hold placed on all pacritinib studies in February 2016 was due to the concern by the FDA of an increased risk of bleeding and cardiovascular events with pacritinib coupled with survival curves from PERSIST-1 that appeared to indicate worse survival in the pacritinib arm. However, closer analysis at 72-week follow up (presented by Dr. Ruben Mesa at the 2016 ASCO Annual Meeting) demonstrated that the survival curves were the same up to 24 weeks, and then there was a trend toward improved survival with BAT that is likely explained by imbalance, with a greater proportion of pacritinib-treated patients with DIPSS highrisk features such as advanced age, leukocytosis, and circulating blast percentage. Additionally, when looking at the BAT patients who did not cross over to pacritinib (17 patients) there seemed to be a higher frequency of deaths (35%) compared to the cohort that crossed over to pacritinib. Also, attaining a spleen volume reduction of at least 20 percent correlated with improved survival with pacritinib therapy, but not with BAT. There was no significant difference in occurrence of cardiovascular events in these two arms until after 24 weeks, and reasons for death in the BAT arm were more frequently attributed to adverse events. Taken together, pacritinib offers a therapeutic option with a benefit-to-risk ratio that is favorable for patients with low platelets, and attaining a spleen response may predict those patients who are most likely to derive a survival benefit from this novel agent.

Note added in press: On January 5, 2017, CTI BioPharma announced that the FDA’s full clinical hold on trials of pacritinib had been removed.