AUTHORS
Amirthagowri Ambalavanan, PhD, Karishma Tailor2, Oyeronke Ayansola, PhD3, Azin Keshavarz, MHSc.4, Danielle Pyne, BSc5, Maria Perusini, MD, Msc6, Hassan Sibai, MD, Msc, BSc7, Dennis Dong Hwan Kim, MD, PhD8
AFFILIATIONS
1 Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
2 Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
3 Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
4 Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
5 Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
6 Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada, Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
7 Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada, Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
8 Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada, Hans Messner Allogenic Transplant Program, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada
ABSTRACT
Background
Chronic myeloid leukemia (CML) is a clonal disorder initiated from hematopoietic stem cells that can progress from chronic phase (CP) to blast phase (BP) with either myeloid (MBP) or lymphoid blast (LBP) phenotypes, yet the divergent transformation process remain incompletely characterized. We investigated paired changes in hematopoietic cell states from CP to BP using single cell RNA sequencing (scRNAseq) with reference-derived pseudotime analysis using hematopoietic cell atlas generated from normal bone marrows as a reference atlas.
Patients and Methods
14 CML patient pairs transformed from CP to BP were included. Unsorted mononuclear cells from BM or PB were processed using whole-transcriptome scRNAseq (Parse Biosciences Evercode WT). BoneMarrowMap using Symphony-based reference mapping was applied for cell-type composition (Zeng et al, Blood Cancer Discovery 2025), while hematopoietic reference states and reference-derived pseudotime analysis were performed using Seurat. Median pseudotime, complete pseudotime distribution changes and mean cell-type proportions were compared within lineage groups (paired Wilcoxon signed-rank test) or between MBP and LBP (Wilcox rank-sum test). To evaluate lineage-associated progression, cell-state delta values were calculated per patient and cell types as the change in cell-type proportion from CP to BP (BP minus CP) and compared between MBP and LBP (Wilcox rank-sum test). P-values across 24 broad cell type comparisons were adjusted for multiple testing using the Benjamini Hochberg (BH) method, with adjusted P< 0.05 considered significant.
Results
Of 13,091 cells were analyzed, 12,764 met QC. 14 matched pairs (28 samples), comprising 8 MBP and 6 LBP pairs, were analyzed. Reference-derived pseudotime analysis demonstrated heterogeneous, patient-specific remodeling during CP-to-BP transformation. MBP showed a larger median decrease in pseudotime (median CP 14.7 vs. BP 5.40; median paired shift -8.73), whereas LBP demonstrated relatively stable pseudotime distributions (median CP 7.55 vs. BP 9.83; median paired shift -0.42). However, neither within-lineage paired comparison was statistically significant (MBP P=0.141; LBP P=1.00), nor did the magnitude of CP-to-BP pseudotime change differ significantly between MBP and LBP by median (P=0.272) or mean (P=0.175). Likewise, comparison of the complete pseudotime distributions showed no significant difference in overall trajectory distance (P=0.847) or direction of developmental state shift (P=0.175).
Projection of BP samples onto the BoneMarrowMap identified distinct lineage-associated cellular architectures. LBP was characterized by marked enrichment of Pro-B-like cells compared with MBP (38.5% vs 8.3%; unadjusted P=0.033) although this did not remain significant after multiple-testing correction (BH-adjusted P=0.099). In contrast, MBP showed significantly greater representation of primitive and myeloid progenitor-like states including LMPP (6.5% vs 0%; adjusted P=0.026), HSC/MPP (10.0% vs 0.44%; adjusted P=0.026), MEP (7.7% vs 0.20%; adjusted P=0.026), cycling progenitor (2.9% vs 0%; adjusted P=0.028), eosinophil/basophil/mast precursor (4.0% vs 0.10%; adjusted P=0.049), and late GMP states (1.46% vs 0%; adjusted P=0.049).
Cell-state delta analysis identified significant MBP-vs-LBP differences involving LMPP and Early GMP (adjusted P=0.042 each), as well as HSC/MPP, MEP, cycling Progenitors, and late GMP (adjusted P= 0.047 each), suggesting differential remodeling of progenitor states during BP transformation. Within MEP compartment, differentially expressed gene (DEG) analysis identified higher expression of CCR7, MME (CD10), HLA-DQB2, PEG10, and RGR in LBP. Functional enrichment of lineage associated DEGs showed supporting distinct immune and hematopoietic transcriptional programs in MBP and LBP.
Conclusions
CML progression and transformation of CP-to-BP involves heterogeneity in developmental states based on pseudotime trajectory. Although global pseudotime changes did not significantly distinguish MBP from LBP, exploratory cell-state composition suggested biologically coherent lineage-associated patterns: expansion of primitive and myeloid progenitor-like states in MBP, and B-lymphoid precursor-like states in LBP, with distinct immune and hematopoietic transcriptional programs distinguish MBP and LBP.
DISCLOSURES
Dennis Dong Hwan Kim, MD, PhD: Novartis - Research Funding; Enliven - Research Funding; Abbvie - Research Funding; Incyte - Research Funding; Sanofi - Research Funding; Ascentage - Research Funding; JAZZ - Consultancy (Includes expert testimony); Abbvie - Consultancy (Includes expert testimony); Ascentage - Consultancy (Includes expert testimony); Daiichi-Sankyo - Consultancy (Includes expert testimony); Novartis - Consultancy (Includes expert testimony); Sanofi - Consultancy (Includes expert testimony); J&J - Consultancy (Includes expert testimony); Astellas - Consultancy (Includes expert testimony)