Wilson disease (WD) has traditionally been understood through a straightforward causal sequence: pathogenic ATP7B variants impair the incorporation of copper into ceruloplasmin and, more importantly, reduce biliary copper excretion; copper then accumulates in the liver and subsequently in other organs, producing hepatic, neurological, psychiatric, ophthalmological, and systemic manifestations^[1-3]. This model has enormous explanatory and therapeutic value. It established the molecular basis of WD, justified lifelong copper-lowering therapy, and remains the foundation of current clinical guidelines^[1,2].

Its success, however, may have encouraged an overly static interpretation of the disease. In the classical model, ATP7B loss is the initiating event, copper is treated largely as a cumulative burden, and tissue injury is positioned as a downstream consequence. That sequence is necessary, but it is no longer sufficient. We argue that ATP7B deficiency creates an unstable copper-handling state in which copper entry, intracellular distribution, adaptive buffering, organelle vulnerability, intercellular signaling, and systemic modifiers continuously reshape disease expression. WD is therefore better conceptualized as a dynamic copper systems disorder than as a passive copper storage disease.

Several long-recognized observations expose the limits of a burden-centered model. First, WD is monogenic but phenotypically diverse. Patients with similar or identical ATP7B variants may differ markedly in age at onset, dominant organ involvement, rate of progression, and treatment response. In a large international cohort, age and sex influenced clinical presentation more effectively than ATP7B genotype^[4]. This does not diminish the causal role of ATP7B; rather, it indicates that genotype establishes susceptibility while additional biological processes shape phenotype.

Second, total copper content does not consistently predict the timing or severity of injury. In ATP7B-deficient mice, liver pathology was related more closely to the duration of copper exposure than to the absolute hepatic copper concentration^[5]. At the single-cell level, temporal and spatial copper distribution was more informative than total copper alone, with compartment-specific copper relocation accompanied by distinct metabolic responses^[6]. Thus, where copper is located, in what chemical form, and for how long it remains there may matter as much as how much copper is present.

Third, ATP7B deficiency does not produce a uniform cellular response. Comparisons of different Atp7b-deficient mouse models have shown that the magnitude and character of hepatic injury can vary despite copper overload, and that proteomic changes may correlate more closely with buffering capacity, including metallothionein abundance, than with total liver copper^[7]. Copper burden is therefore not synonymous with copper toxicity. Toxicity emerges when copper flux exceeds the capacity of sequestration, trafficking, metabolic adaptation, immune control, and tissue repair.

We propose a flux-compartment-response framework for WD [Figure 1]. The framework does not replace the classical ATP7B model; it expands it. ATP7B deficiency remains the initiating lesion, but the disease state at any point in time reflects the interaction of three dimensions: copper flux, intracellular compartmentalization, and cellular/tissue response. These dimensions are modified by age, sex, diet, extrahepatic ATP7B deficiency, other metals, the intestine and microbiome, immune-cell interactions, treatment, and time [Table 1].

The conceptual contribution is not the individual components, many of which have been described separately. The added value is to treat flux, compartmentalization, and response as coupled, time-dependent state variables rather than as a linear list of downstream events. This yields three testable propositions. First, patients with similar total copper burdens may have different risks because they occupy different biological copper states. Second, the dominant mechanism may shift during disease progression or after treatment. Third, biomarkers and therapeutic combinations should be selected to identify and modify the mechanism active at a particular time, rather than relying on a single cumulative copper value. In this sense, the framework converts mechanistic heterogeneity into a falsifiable strategy for longitudinal phenotyping and treatment selection.

A dynamic model has immediate implications for biomarkers. Current diagnosis and monitoring rely on combinations of ceruloplasmin, serum copper, 24-h urinary copper, hepatic copper, clinical findings, and genetic testing^[1,2]. These tools are indispensable, but most are indirect or static. Calculated non-ceruloplasmin-bound copper is vulnerable to analytical error, and total copper measurements do not distinguish buffered from biologically active pools. Exchangeable copper and relative exchangeable copper were developed to estimate a labile circulating fraction^[24], and accurate measurement of non-ceruloplasmin-bound copper by high-performance liquid chromatography (HPLC) coupled to inductively coupled plasma tandem mass spectrometry (ICP-MS/MS) may improve assessment and monitoring^[25].

Metabolic biomarkers may add information about the state of response. Serum sphingolipid profiling has identified candidate markers that distinguish patients with WD from healthy controls^[26], while untargeted metabolomics has demonstrated a distinct metabolic signature involving lipid, amino-acid, and energy pathways^[27]. Such signals could potentially identify metabolic stress before irreversible organ injury, but most available studies are cross-sectional and include patients with established disease. Demonstration of an early-warning role will require longitudinal studies in presymptomatic genetically diagnosed individuals, newly diagnosed untreated patients, and patients undergoing decoppering therapy.

The next step is therefore to define a biological copper state rather than a single copper value. This state would integrate burden with bioavailability, organ distribution, mitochondrial and metabolic injury, inflammatory activity, and adaptive capacity. Research could combine exchangeable copper, organ-specific imaging, liver and brain injury markers, lipidomic or metabolomic profiles, inflammatory signatures, and spatial or single-cell methods. The objective is not to create an impractically large panel, but to identify the smallest set of measures that predicts disease onset, progression, or therapeutic response better than conventional copper indices alone.

Hypothetical retention-dominant, uptake-amplified, mitochondrial-injury, immunometabolic, or mixed states are testable rather than definitive categories. Their value will depend on reproducibility, longitudinal stability, and clinical utility. They may ultimately prove more therapeutically informative than a purely hepatic-versus-neurological classification, but this remains a research hypothesis.

Current therapies reduce intestinal copper absorption, increase copper excretion, or replace the failing organ in advanced disease^[1-3]. Their effectiveness confirms the central importance of copper lowering. The proposed framework does not weaken this rationale. Instead, it explains why clinical response may vary despite apparently acceptable conventional copper indices and why adjunctive therapies should be evaluated against a defined biological state rather than added empirically.

Restoration of ATP7B function is the most direct way to correct the initiating defect. In preclinical models, adeno-associated virus (AAV)-mediated delivery of a shortened but functional ATP7B construct produced long-term hepatic expression, normalized copper handling, and prevented liver injury^[28]. These results establish proof of concept for gene replacement, but AAV therapy for WD remains investigational. Vector dose, pre-existing immunity, durability in a growing liver, transgene expression, redosing, and long-term safety must be resolved before claims of complete or permanent correction can be made in patients.

Other mechanism-informed strategies could reduce pathological copper entry, redirect intracellular copper away from vulnerable organelles, strengthen adaptive buffering, or limit mitochondrial, immunometabolic, and fibrotic injury. PrP-directed intervention is one experimental example, but human pathway activation, tissue-selective delivery, physiological functions of PrP, and interactions with chelators or zinc would require careful evaluation^[9]. Likewise, microbiome, immune, or metabolic interventions should be regarded as adjunctive research directions rather than alternatives to established decoppering therapy.

Translation will also require better clinical infrastructure. Comprehensive phenotyping should be collected with standardized definitions, sampling schedules, copper assays, treatment-exposure data, neurological and hepatic outcomes, and biobanking protocols. International, well-curated repositories would allow data sharing across centers and provide sufficiently large datasets for externally validated machine-learning models. Such models may assist diagnosis, identify clinically meaningful subgroups, forecast treatment response, and detect nonadherence or early deterioration. They should complement, not replace, mechanistic reasoning and must be evaluated for calibration, interpretability, transportability, and bias.

Human studies should determine whether noncanonical uptake pathways, lipid and bile-acid responses, immune-cell states, and microbiome-derived metabolites are activated in specific subgroups and how these responses change with treatment. Spatial methods should relate cellular and organelle copper distribution to mitochondrial function, histology, and phenotype. Longitudinal cohorts should track the transition from presymptomatic disease to hepatic or neurological manifestations and should test whether metabolic or inflammatory signatures predict onset before conventional organ injury markers become abnormal.

Experimental models should be interpreted as representations of different phases or components of WD rather than ranked simply as accurate or misleading. Liver-specific Atp7b knockout models, for example, are valuable because they can reveal early hepatocyte compensation before systemic manifestations emerge. Systemic knockout mice, LEC rats, cell models, and organoids provide complementary information about chronic adaptation, mitochondrial failure, immune amplification, and tissue-specific vulnerability. Therapeutic testing should therefore define which response state a model reproduces and match that state to the proposed mechanism of action.

The framework must remain falsifiable. It will be useful only if measures of flux, compartmentalization, or response predict phenotype, progression, or therapeutic response better than total copper burden alone, or identify clinically actionable subgroups. The aim is not to replace a simple model with untestable complexity, but to identify the minimum additional biology needed to explain meaningful variation and improve care.

The classical copper-retention model of WD remains correct, indispensable, and incomplete. ATP7B deficiency initiates copper accumulation, but it does not by itself explain when, where, and how injury occurs. A dynamic flux-compartment-response framework offers a broader interpretation: disease expression reflects the balance between copper entry and export, the intracellular destination and chemical availability of copper, and the capacity of cells, immune and stromal partners, and extrahepatic tissues to adapt to or amplify stress.

Future research should move beyond total copper burden toward biological copper states that can be measured longitudinally and linked to outcomes. Copper lowering remains the therapeutic foundation; the opportunity is to combine it, when supported by evidence, with interventions targeting pathological flux, organelle vulnerability, immunometabolic injury, or restoration of ATP7B. The value of the framework will ultimately be judged by whether it improves prediction, patient stratification, and treatment response.