A Four-Domain Regime-Transition Framework for Carcinogenesis

Cancer is shaped by somatic mutation and selection, but these processes unfold within tissue, metabolic, immune and ecological conditions that influence which biological states are generated, accessible, persistent and executable. Thermodynamic adaptation here denotes context-dependent persistence, state change or differential fitness under specified material, energetic, redox and execution constraints; the qualifier is retained only where that constraint structure adds value beyond matched models using the same measurements without it. A worked same-input discriminator is specified below as an analytic design, but its discriminating performance has not yet been empirically implemented or demonstrated in the present framework. We present a theoretical framework for epithelial cancer systems in which the specified applicability conditions can be operationalized, organized around a directed D1→D2→D3 spine. This ordering is probabilistic and non-obligate; neutral drift, stochastic birth–death dynamics, mutation supply and selection remain admissible. Domain 1 denotes reorganization of organism-integrative interface coupling across host tissue, epithelial barriers, immune states and, where functionally relevant, host-associated biological components. Domain 2 denotes a path-dependent distribution regime and thermodynamic execution constraints affecting the accessibility of metabolic, redox, biosynthetic, organellar, regulated-death and tissue-restoration programmes. Domain 3 distinguishes genomic and epigenomic state generation, selection, stabilization and entrenchment; causal locking is reserved for D3→D2 feedback in which valid state or state-maintenance perturbation changes the defining genomic or epigenomic state and the prespecified downstream trajectory through that state change. Domain 4 is a conditional reinforcement domain requiring demonstrated reciprocity rather than microbial detection alone. Treatment is represented across these domains as a clinically initiated second encounter regime comprising six non-equivalent, potentially causally coupled processes; incomplete elimination leaves a residual state space that those processes may reshape, whereas complete malignant elimination may permit organism-level restoration. The integrated architecture remains unvalidated and is supported only if prespecified longitudinal and perturbational tests distinguish its directed relations from reverse-order, independent-effects, domain-standard and same-input alternatives. A treatment-linked negative GC-direction shift is retained as a separable, currently uninstantiated auxiliary research programme. No treatment-specific negative-direction confirmatory hypothesis is activated in the present manuscript; a future hypothesis becomes confirmatory only when the treatment context, endpoint, negative directional expectation, comparator and falsifier are independently justified and locked before outcome analysis. A negative GC-direction result is not by itself evidence for genomic economy, thermodynamic adaptation, the four-domain architecture or a clinical biomarker. The framework does not replace genetic causation, establish a universal sequence or constitute clinical guidance.