Telomerase activation enables replicative immortality in the majority of cancers including acute myeloid leukemia (AML). Imetelstat competitively inhibits telomerase activity with recently reported clinical efficacy in myelodysplasia and myelofibrosis. Here we show final results on a randomized preclinical trial in AML patient-derived xenografts (PDX; n = 30 individual AML; each treated with imetelstat vs. vehicle control; n = 6 per condition). Using drug sensitivity scores (DSS) to define in vivo PDX responses, we provide integrated phenotypic, cytogenetic, genomic and transcriptomic analysis of samples to understand AML response to imetelstat. We use this mechanistic insight to rationally sequence induction chemotherapy with imetelstat to optimize in vivo responses, even in initially poorly responding samples.

Methods: An AML PDX cohort was established using bone marrow (BM) or peripheral blood (PB) samples collected from 30 AML patients and transplanted into NOD/SCID/IL2gR-/-/hIL3,CSF2,KITLG (NSGS; n = 12 recipients / AML patient sample). AML onset was defined by reconstitution of BM and spleen with CD45+ CD33+ donor cells, circulating blasts, anemia (HCT < 35%) or thrombocytopenia (PLT < 400 x 10^6/ml). After initial engraftment was confirmed by PB chimerism (> 1%), recipients were treated with imetelstat (15 mg/kg I.P.) or control every 48 - 72h.

Results: Across the entire cohort, survival was improved for imetelstat- vs. PBS-treated PDX (Median survival PBS: 88d, Imetelstat: 138d; Hazard ratio PBS: 2.87; 2.20 to 3.74; Cox proportional hazards model: p < 0.0001), however there were clear differences in the quality and duration of response to imetelstat treatment. In order to quantify the sensitivity of each individual AML patient sample to therapy, DSS were derived using the area under the curve of PB donor chimerism in PBS versus imetelstat-treated recipients (Mean DSS entire cohort: 3.21 ± 0.36). DSS positively correlated with survival benefit (R square: 0.87), and values higher than 2.3 gained at least 30 days of survival benefit. Based on DSS, AML patient samples were grouped into "Sustained" (17 AML patient samples; 57%) or "Poor" responders (13 AML patient samples; 43%) with DSS 2.3 as cut-off value. European LeukemiaNet 2017 adverse prognostic samples were trend-wise enriched in poor vs. sustained responders to imetelstat (83.3% poor vs. 16.7% sustained response; p = 0.06, Fisher's exact test). RNA sequencing and mutational profiling was performed on all baseline AML samples. Differential expression analysis revealed a gene signature specifically expressed in sustained compared to poor responders (p < 0.05; log2FC > 1.5; 109 annotated genes) that was functionally annotated in replicative stress including nucleosome (p = 0.01), transcription factor activity (p = 0.04), and sequence-specific DNA binding (p = 0.04).

We therefore examined whether the induction of replicative stress could sensitize response to imetelstat therapy. Seven independent AML patient samples with poor response to imetelstat monotherapy were transplanted into NOD.Rag1-/-Il2Rg-/-/ hIL3,CSF2,KITLG (NRGS). Mice received standard induction chemotherapy intravenously as a cytarabine (Arac; 50 mg/kg) and doxorubicin (Doxo; 1.5 mg/kg) 5+3 dosing regimen followed by continuous imetelstat therapy vs. imetelstat alone, standard induction chemotherapy alone, or vehicle control (n = 6 / group, n = 24 / AML sample). Combination therapy significantly prolonged overall survival (Median survival: Control: 38.5d; Imetelstat: 41d; Arac+Doxo: 61d; Arac+Doxo + Imetelstat: 71d post-start of treatment; Hazard ratio Arac+Doxo alone vs. Arac+Doxo+Imetelstat: 2.34; 1.19 to 4.586; Cox proportional hazards model: p < 0.001).

In summary, imetelstat prolonged overall survival in a large cohort of AML PDX. In 57% (17 out of 30 individual AML patient samples), sustained responses to imetelstat were associated with marked improvements in survival and a baseline replicative stress-annotated transcriptional signature. The rational sequencing of standard induction chemotherapy to induce replicative stress provides proof-of-concept to sensitize imetelstat-resistant AML patient samples and suppress expansion and prevent relapse in AML.


Lane: novartis: Consultancy; janssen: Consultancy, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution