Therapy resistance and disease relapse in patients with MPN-blast phase are associated with clonal evolution and transcriptomic changes in MPN- blast phase stem cells Free

Background:

Myeloproliferative neoplasms in accelerated or blast phase (MPN-AP/BP) are associated with a median overall survival (OS) of 3-5 months. The Myeloproliferative Neoplasm Research Consortium (MPN-RC) phase 1 and 2 trials (MPN-RC 109) demonstrated that combination therapy with ruxolitinib and decitabine for these patients was tolerable and improved median OS to 9.5 months, with an overall response rate of 46% (Mascarenhas Blood Adv 2018, 2020). Despite initial efficacy, 54% of patients failed to respond and most responders (81%) eventually relapsed.

Methods:

To investigate the molecular and cellular basis of response, resistance, and relapse, we performed longitudinal analyses of peripheral blood samples from 24 patients with MPN-AP/BP treated with the ruxolitinib and decitabine combination. Approaches included targeted next-generation sequencing (NGS) of 576 hematologic malignancy-associated genes, single-cell mutational and epitope profiling, cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq), and xenograft repopulation assays by transplanting CD3⁺ cell-depleted mononuclear cells (MNCs) into NSG mice.

Results:

Targeted NGS of baseline and on-treatment (C4D1/C10D1) MNCs from 24 patients revealed significant changes in clonal dynamics. Reduction of some clones was observed in 100% of patients with complete remission (CR) + incomplete platelet recovery (CRi), 61.5% of partial remission (PR), and 33.3% of no response (NR) patients. However, expansion of other clones and/or emergence of new clones were noted in 100% of CRi, 84.6% of PR, and 83.3% of NR patients, suggesting that clonal evolution occurred despite clinical response.

Transplantation of drug-treated MNCs from responders (2 CRi, 2 PR) in xenograft models resulted in a 49-94% reduction in human (h) myeloid blast burden in NSG mouse bone marrow versus baseline (BL) samples (p<0.05), indicating partial depletion of functional MPN-BP stem cells (SCs). By contrast, MNCs from NRs generated similar or increased leukemic burdens. NGS of CD34⁺ cells selected from xenografts showed that mutational profiles largely resembled those of the corresponding patient's primary samples, suggesting disease-initiating SC clones persisted or evolved despite treatment.

To elucidate the mechanisms that are responsible for disease relapse, we characterized changes in the function, clonal architecture and transcriptomic landscape of MPN-BP SCs from 9 relapsed samples. Seven samples showed an increased hCD34⁺/hCD45^dimCD33⁺blast burden compared to BL in xenografts, reflecting enhanced disease-initiating capacity of relapsed SCs.

Single-cell mutational and epitope profiling of one pair of primary samples identified 8 subclones containing 2-3 oncogenic mutations in WT1, ETV6 and KRAS with differing variant allele frequencies within the CD34⁺ cells. Although the prevalence of 1 major subclone harboring heterogenous ETV6, WT1 and KRAS mutations decreased at relapse (BL: 83.54%, relapse: 70.46%), all other 6 minor subclones' prevalences increased (BL: 1.09-5.8%, relapse: 1.31-13.6%) and one new subclone emerged at relapse. Moreover, 4 minor subclones each containing homozygous mutations in ETV6, WT1 or KRAS respectively and heterogenous mutations in the other two genes expanded at disease relapse. Clonal expansion and/or emergence of new clones were detected using NGS in primary CD34⁺ cells from 4 additional cases, but one case showed stable clonal architecture.

CITE-seq of matched BL and relapse primary samples identified shifts in hematopoietic stem and progenitor cell populations, including increased HSC (1.38-fold) and CMP (0.49-fold) and decreased MkP (0.5-fold). Relapsed CD34⁺ HSCs contained upregulated genes including those associated with poor survival, worse prognosis, and chemotherapy resistance (e.g., SLC38A2, RACK1), cancer development and progression (EEF1A, FTL, BTF3), and cell growth and survival (UBC, DNAJB1, SYF2), highlighting potential therapeutic targets.

Conclusions:These studies demonstrate that therapy resistance and relapse in MPN-BP are driven by persistent and evolving MPN-BP SCs. Clonal evolution and transcriptomic reprogramming confer enhanced leukemogenic potential of MPN-BP SCs and resistance to combination therapy. These findings suggest that more effective therapeutic strategies for MPN-BP should be directed towards eliminating MPN-BP SCs and targeting their molecular drivers.