Molecular Analysis in the Pacritinib Dose-Finding PAC203 Study in Patients with Myelofibrosis Refractory or Intolerant to Ruxolitinib

PAC203 is a global multicenter dose-finding study of pacritinib (PAC), an oral JAK2/IRAK1 inhibitor in patients with primary or secondary myelofibrosis refractory or intolerant to ruxolitinib, including patients with severe thrombocytopenia. Patients were randomized 1:1:1 (PAC 100mg QD, 100mg BID or 200mg BID) and stratified by baseline platelet count. The mutational landscape of this patient group is not well characterised, and the impact of mutation status on disease response and hematologic parameters is unknown.

We carried out baseline mutational analysis on 105 (of total 164 recruited; 161 treated) patients using an ISO accredited Illumina TruSeq Custom Amplicon Panel, including 32 gene mutation hotspots and exons (~36,000 bp, 287 amplicons). CALR mutation screening was carried out independently. Accepted coverage was achievement of a depth of ≥100 reads per base in ≥95% of targeted bases. Median follow-up time for this cohort was 163 (28-476) days. 57% of patients had a diagnosis of primary myelofibrosis (PMF), whereas 27% and 15% had post-polycythemia vera MF (PPV-MF) and post-essential thrombocythemia MF (PET-MF), respectively. The median baseline platelet count was 64.5 x10⁹/L, with <50 x10⁹/L in 38.3% of patients and baseline Hb <10g/dL in 67.8% of patients. The median age was 67.5 (37-87) years.

The majority of patients were JAK2 V617F-mutated (77%), followed by MPL-mutated (8.6%), with a notably low incidence of CALR-mutated (8.6%; type 1 CALR n=6, type 2 CALR n=3) and "triple-negative" (4.8%) cases. Non-myeloproliferative neoplasm driver mutations (NDM) were present in 77.1% (n=81) with ≥3 NDMs in 18.1% of patients. Similar to previous reports, the most prevalent NDMs were TET2 and ASXL1 (n=27 and n=25 respectively). Splicing factor (SF) mutations were mutually exclusive and detected in 32.3% of patients: SF3B1 [13], U2AF1 [14], SRSF2 [6], ZRSR2 [1]. RAS pathway mutations (KRAS/NRAS) were found at a higher frequency than previously in MF cohorts; 16.2% (n=17).

SF mutations were present in more PMF (45.6%) than PPV-MF (3.7%, n=1/27) or PET-MF patients (33.3%, n=5/15), P=.002. SF3B1-mutated patients had higher trial entry platelet counts (66.7% platelets ≥100x10⁹/L, P=.014). Baseline red cell transfusion dependency was more often associated with U2AF1 mutations 24.1% (n=7/29) as compared with red cell transfusion independence, 4.5% (n=2/44), P=.047 and U2AF1-mutated patients had lower baseline hemoglobin <10g/dL (n=11/12), P=.015. Overall, 41 (39%) of patients had a high molecular risk mutation (HMR; IDH1/2, SRSF2, ASXL1, SRSF2, U2AF1Q157), and 8 patients had TP53 mutations.

In those with molecular analysis available, the highest rates of ≥35% SVR were observed in the 200mg BID arm: 11.1% (n=4/36) followed by 3.2% (n=1/32) in 100mg BID arm and 0% (n=0/30) in 100mg QD arm). Rates of TSS reduction ≥50% were similar across arms. There were no significant correlations between mutations and SVR or TSS response. More grade 3/4 anemia occurred in TET2-mutated patients (OR 5.7, 95% CI 1.6-20.4, P=.007). RAS pathway mutations were associated with grade 3/4 thrombocytopenia (OR 4.4, 95% CI 1.3-14.8, P=0.016). Treatment discontinuation was not influenced by mutation status.

In summary, the PAC203 cohort is molecularly high risk, with a high incidence of HMR and low incidence of CALR mutations. We identified novel associations between mutation profiles and hematologic events in this population with advanced MF.