Fig. 6

Heterogeneity within LSCs mediated therapy resistance. (Upper, CML) (1) Quiescent BCR-ABL⁺ LSC subpopulation, with inflammatory-associated gene upregulated at diagnosis, persisted during TKI treatment. RUNX1 mutation gain was linked to the blast crisis in BCR-ABL⁺ LSC [102]. (2) CD93 was found to be selectively and persistently expressed in an LSC subpopulation, with quiescent gene upregulated at diagnosis. This LSC subpopulation persisted during TKI [99]. (Lower, AML). (3) AML LSC profile at relapse was presumed to be the consequence of recovery from senescence-associated secretory phenotype (SASP), a phenotype induced by Ara-c treatment [107]. (4) Proliferative (Cd36^high) and self-renewal (Cd69^high) subtypes were found to be distinct in LSCs. Only targeting proliferation or self-renewal pathways caused resistance, and simultaneous targeting improved therapeutic outcomes [105]. (5) Monocytic LSC (m-LSC) was distinguished from previously well-defined primitive LSC (p-LSC) and drives resistance to venetoclax(VEN)-based treatment [37]. Co-targeting m-LSCs with cladribine may be clinically important. (6) Mac-1 was found to be differentially expressed in LSCs. Mac-1⁺ subtype is DA resistant with higher RAS pathway activation [106]