Cancer drug resistance as a temporal process: cell plasticity, adaptive buffering, signaling pathways and therapeutic interception

Cancer drug resistance remains a principal barrier to durable cancer control, yet it is still often interpreted through endpoint-centred models that emphasise pre-existing resistant clones, incomplete drug exposure or late-acquired mechanisms. These explanations are essential but incomplete. The question is not only why resistant cells exist, but also how residual cells remain alive long enough to become resistant. Across tumour types and therapeutic modalities, treatment reshapes residual populations while it is acting, driving transitions through non-genetic heterogeneity, stochastic cell-fate variability, reversible tolerance, stress adaptation, microenvironmental support and eventually stable resistance. This Review develops a temporal framework in which resistance is resolved into a phase-structured trajectory organised around two mechanistic transitions: entry into adaptive buffering and later stabilisation into resistance hardening. Adaptive buffering is defined operationally as a transient, support-dependent survival state in which residual tumour cells preserve viability by recruiting physiological support systems while molecular and phenotypic reorganisation proceeds. It is not a single pathway or a synonym for drug tolerance, but a functional demand-capacity state whose dependencies change over time. We distinguish adaptive buffering from drug tolerance, persister states, dormancy, minimal residual disease and established resistance and discuss how buffering can be evaluated experimentally using perturbational, single-cell, spatial, computational and functional readouts. The framework suggests that informative biomarkers should assign residual populations to adaptive phases, so that therapy can intercept transient dependencies before adaptive states become hardened and clinically entrenched.