Trial update of the MANIFEST study of CPI-0610 in patients with myelofibrosis from EHA 2020

MANIFEST (NCT02158858) is a phase II clinical trial of CPI-0610, a small molecule inhibitor of bromodomain and extraterminal (BET) proteins, with and without ruxolitinib, a selective inhibitor of the Janus-associated tyrosine kinases (JAK) 1 and 2, in patients with myelofibrosis (MF). Coverage of the previous report from the American Society of Hematology (ASH) Annual Meeting and Exhibition 2019 can be found here. This report is from a larger data set presented at the European Hematology Association (EHA) annual 2020 virtual meeting and may supersede the data in the published abstracts.

Study design^1,2,3,4

• Phase II, open-label, multicenter, trial with three different arms, as follows:
  • Arm 1: CPI-0610 monotherapy in patients who are resistant to, intolerant of, or ineligible for treatment with a JAK inhibitor. Stratified into two cohorts based on transfusion dependent (TD) status
  • Arm 2: CPI-0610 as add-on therapy in combination with ruxolitinib in patients with a suboptimal response to ruxolitinib or MF progression. Stratified into two cohorts based on TD status
  • Arm 3: CPI-0610 in combination with ruxolitinib in JAK-inhibitor-naïve MF patients who are anemic (hemoglobin < 10 g/dL)
• Primary endpoints: proportion of patients with a ≥ 35% spleen volume reduction from baseline (SVR35) after 24 weeks of treatment, and conversion of transfusion dependence status from TD to transfusion independent (TI) for 12 consecutive weeks
• Secondary endpoints: Duration of spleen response after 24 weeks, rate of response after 24 weeks, 50% improvement in total symptom score, patients' global impression of change after 24 weeks, and pharmacokinetic parameters

Key findings^1,2,3

• Since ASH 2019, an additional seven patients were included in the analysis for Arm 1 and an additional 16 patients were included in the analysis for Arm 2
• 64 patients were enrolled in Arm 3, of which 30 patients were evaluable for efficacy analysis and all were evaluable for safety analysis
• The baseline characteristics remained similar to those presented at ASH 2019, in that most patients had Dynamic International Prognostic Scoring System (DIPSS) ≥ intermediate, anemia, high-risk mutations and high symptom burden
• The 24-week best response analysis for Arms 1 and 2 are summarized in Table 1
  • In line with the ASH 2019 data, CPI-0610 in combination with ruxolitinib led to the conversion of more patients from TD to TI than CPI-0610 alone (34.4% vs 21.4%, respectively)
  • In contrast to the ASH 2019 data, TI patients seemed to benefit more than TD patients from CPI-0610 with or without ruxolitinib

Table 1. Best response data after 24-weeks^1,2,

• Arm 3 data at 12 weeks and 24 weeks are summarized in Table 2 
• • SVR35 response at 24 weeks was better than the standard for ruxolitinib per label in a JAKi-naïve patient population (63.3% compared with the standard of 28.5–41.9%)
  • There was no evidence of a correlation between SVR35 response and baseline risk status, platelet counts, or spleen volume

Table 2. Best responses of Arm 3 at 12 and 24 weeks³

•  Safety data can be seen in Table 3
• • The results followed the same trend as the ASH2019 data in that thrombocytopenia was the most common ≥ Grade 3 treatment-emergent adverse event followed by anemia

Table 3. Safety analysis of TEAEs per study arm^1,2,3

 Conclusion

• CPI-0610 monotherapy and combination therapy with ruxolitinib was generally well tolerated
• The spleen response rates at 12 weeks and 24 weeks for first-line treatment were similar with those previously reported and differentiated from standard of care treatment
• Conversion to transfusion independence and improvements in hemoglobin levels suggest possible disease-modifying effects of CPI-0610
• Based on these results, the TD cohorts of Arm 1 and 2 are to be expanded to 60 patients, and a phase III randomized, double-blind trial of CPI-0610 plus ruxolitinib vs placebo plus ruxolitinib, in a JAKi naïve patient population is planned to start later this year based on the Arm 3 data