Dynamics of Mutations in Patients with ET Treated with Imetelstat

Background: Imetelstat, a first in class specific telomerase inhibitor, induced hematologic responses in all patients (pts) with essential thrombocythemia (ET) in a recent phase-2 study, and molecular responses were seen within 1-12 months in the majority of pts carrying JAK2 V617F (JAK2m) and CALR mutations (CALRm) (Baerlocher et al., ASH 2014). It has been reported that the treatment response to imetelstat in pts with myelofibrosis (Tefferi et al., ASH 2014) was negatively influenced by mutations in ASXL1 and favorably impacted by mutations in SF3B1 and U2AF1. In pts with CALRm ET treated with Interferon-alpha (INFa) the response rate was lower if the patient carried more than one mutation in ASXL1, TET2, IDH2, CSF3R and SH2B3 (Kiladjian et al., ASH 2014). In addition, in JAK2m pts with polycythemia vera, TET2-mutated clones have been demonstrated to be resistant to IFNa therapy (Kiladjian et al., Leukemia 2010).

Aims: Our aim was to assess the dynamics of additional mutations besides JAK2 V617F, CALR and MPL mutations in pts with ET treated with imetelstat, and to investigate their association with hematologic and molecular response.

Methods: The study enrolled 18 pts with ET who had failed or were intolerant to ≥1 prior therapy, or refused standard therapy. Pts were treated with imetelstat 7.5 mg/kg or 9.4 mg/kg IV weekly until attainment of platelet count ~250-300x10⁹/L followed by a maintenance phase with dosing titrated according to platelet count. DNA was extracted from granulocytes or leukocytes. Mutation analysis was performed by high-throughput sequencing of selectively amplified target sequences on a PGM Ion Torrent instrument covering the coding and adjacent intronic sequences of 15 genes known for mutations in MPN (ASXL1, CBL, DNMT3A, EZH2, IDH1, IDH2, JAK2, MPL, SF3B1, SRSF2, SOCS1, TET2, TP53, U2AF1 and ZRSR2). Samples were taken at baseline and up to 8 time points during treatment through cycle 26, with approximately 3 cycles between samples.

Results: As a driver mutation, at baseline, 9 pts had JAK2V617F, 5 pts had CALR and 2 pts had MPL mutations (MPLm; one L and one K). Two pts were triple negative. A partial molecular response (according to Barosi et al., Blood 2009) was seen in 7/8 JAK2m pts and reductions in CALRm allele burden were between 15% and 55%. At baseline, 19 additional somatic mutations (11 missense, 4 frameshift, 3 nonsense, 1 splice site) were detected in 6/9 JAK2m and 2/5 CALRm pts, affecting the genes ASXL1 (n=3), DNMT3A (n=5), EZH2(n=1), SF3B1 (n=1), SOCS1 (n=2), TET2 (n=4) and TP53 (n=3). Two mutations in DNMT3A and ZRSR2 were detected in 1/2 MPLm pts and none were found in the triple negative pts. Of note, all but one mutated pts carried at least 2 mutations in addition to their driver mutation (up to 5 additional mutations). ASXL1 and SOCS1 mutations were only present in JAK2m pts, and these pts reached hematologic and partial molecular response. At time of best molecular response, a reduction of mutant allele burden corresponding to the reduction of the driver mutations was observed for mutations in ASXL1, EZH2, SOCS1 and some DNMT3A, TET2 and TP53 mutations, but not for SF3B1 and ZRSR2 mutations. Sequential analysis in a JAK2m patient with 4 additional mutations showed that all 4 mutated clones were sensitive to imetelstat treatment and followed the dynamics of the JAK2 mutation, and in a patient with 5 mutations in addition to the CALR mutation, 3/5 mutated clones were responsive. Three pts with a weaker molecular response had TP53 mutations which persisted over time, and 2 were accompanied by additional mutations.

Conclusions: 9/18 (50%) pts in this study carried no additional mutations at baseline, and 50% carried 1-5 mutations in addition to the driver mutation, suggesting genetic instability in a subset of pts. Genetic instability might be enhanced in this pretreated patient cohort with a median time since diagnosis of 7.2 years (range 0.3-24.9). Clones with ASXL1 mutations, a known poor prognostic marker in MPN, appear to be sensitive to imetelstat treatment, and pts with 2-5 additional mutations had both hematologic and molecular responses. TP53 mutations were an exception, being associated with weaker molecular responses to imetelstat treatment. Additional analyses of associations between mutations, disease characteristics and response will be presented.

.