Abstract
Acute myeloid leukemia (AML) is paradigmatic for therapeutic resistance driven by genetic heterogeneity, epigenetic plasticity and microenvironmental protection. Over the past decade, six targeted or pathway-directed small molecules-midostaurin, gilteritinib, quizartinib, ivosidenib, enasidenib, venetoclax and glasdegib-have changed frontline and relapsed/refractory (R/R) practice in genomically defined subgroups or in patients unfit for intensive chemotherapy. Yet primary refractoriness and early relapse remain common, frequently via adaptive rewiring of apoptotic dependencies, clonal evolution and differentiation resistance. Here we integrate mechanistic insights with clinical evidence to: (i) map resistance biology onto targetable nodes (apoptosis control; signalling kinases; chromatin/lineage programmes; RNA splicing; DNA-damage response; nuclear export; niche adhesion and innate immune evasion); (ii) summarise the clinical trajectory and current limits of approved and emerging small molecules (including menin and LSD1 inhibitors); (iii) propose rules for rational doublets and triplets that are biologically orthogonal yet clinically tolerable; (iv) outline a regulatory timeline for key AML small molecules; and (v) prioritise where drug development should go next, including next-generation BH3 toolkits, clonal-pressure-aware designs, minimal residual disease (MRD)-adapted trials and therapy guided by dynamic functional profiling. The review closes with cross-platform challenges-myelosuppression, infectious risk, resistance monitoring and trial design-and a pragmatic framework for moving beyond incrementalism toward durable control and cure.