Phase Transition Window in Metabolic Oncology

1. The Problem

Seyfried observed that once patients enter nutritional ketosis, many old antiparasitic drugs (fenbendazole, mebendazole, ivermectin, chloroquine) can exert anticancer effects. His explanation: cancer and parasites share fermentative metabolic pathways.

Within the SAE framework, this observation admits a heuristic structural reading: cancer cells and parasites can be viewed as emergent phenomena sharing metabolic vulnerabilities — both are highly dependent on fermentative metabolism, operating in the low-Ω region of metabolic modularity. This does not mean cancer and parasites "are the same thing," but that they are anchored on similar foundational-layer components (glycolysis/fermentative metabolism), and may therefore exhibit structural co-vulnerability to the same class of foundational-layer intervention (switching the metabolic environment).

Ketosis does not strike at the emergence layer one target at a time (the logic of conventional chemotherapy). It replaces the bridge at the foundational layer: dismantling the low-Ω-friendly bridge (glucose), laying the high-Ω-friendly bridge (ketone bodies).

The question of this paper: can the ZFCρ [2] phase-transition window Ω ∈ [2.75, 4.01] be mapped to metabolic biology to yield testable quantitative predictions?

2. Definitions

Definition 1: Metabolic Modularity Ω

Ω = the number of functionally coupled independent metabolic pathways per cell.

Initial module inventory: glycolysis, TCA cycle, ETC Complex I–IV (each capable of independent dysfunction, counted separately), β-oxidation, glutamine metabolism.

Correction (after two rounds of literature confrontation): Ω cannot count only classical ATP-producing modules; it must be expanded to "functionally independent metabolic pathways" including non-canonical routes. For example, the AACS bypass (β-OHB → cytosolic acetyl-CoA → lipid synthesis, bypassing TCA) also counts as an independent module. The Ω count here does not refer to gene expression or protein presence/absence, but to structurally coupled nodes capable of maintaining a baseline effective flux threshold — specifically, a module is counted as "independent" if its perturbation can independently cause functional loss and alter the cell's fitness landscape, rather than merely changing flux magnitude. Ω-collapse is not only the physical loss of modules but the structural degradation of flux-carrying capacity.

Definition 2: Ω-Collapse

The Warburg effect in cancer cells is essentially Ω-collapse. Mitochondrial dysfunction decouples multiple TCA and ETC modules, regressing the cell to near-exclusive glycolysis (Ω ≈ 1–2). The metabolic simplification of parasites is an isomorphic Ω-collapse.

Correction: Ω-collapse is not uniform. High-Ω and low-Ω subpopulations coexist within the same tumor (e.g., GBM bulk tumor vs stem cell subpopulations). Ω is a distribution, not a scalar.

Definition 3: Bridge Replacement

Ketosis is not simple bridge removal (demolishing foundational-layer support to kill all emergence indiscriminately), but bridge replacement: dismantling the low-Ω-friendly bridge (glucose-rich environment where Ω ≈ 1–2 suffices for survival) while laying the high-Ω-friendly bridge (ketone environment requiring intact β-oxidation → TCA → ETC chain, Ω ≥ 4 for efficient utilization).

Consequence: low-Ω emergence (cancer cell bulk, parasites) is penalized; high-Ω emergence (normal cells, immune cells) is enhanced.

Boundary statement. This paper discusses the structural logic of metabolic pressure redistribution, not a universal anticancer conclusion for ketogenic diets. Ketosis redistributes selection pressure, but its direction depends on intra-tumor Ω distribution, AACS/MCT dependence, immune ecology, and tissue context. Some cancer cells can use the AACS bypass to directly utilize ketone bodies for proliferation [7]; in some models, ketosis even promotes metastasis. Ketosis is not unidirectionally beneficial — it changes the topology of the game, not the outcome of the game.

3. Core Theorems

Theorem 1: Metabolic Phase-Transition Window Mapping

The ZFCρ phase-transition window Ω ∈ [2.75, 4.01] (data source: [1]; sixth indicator z/√j from [3]) maps to metabolic biology, defining four-phase therapeutic dynamics:

Sprouting (Ω ≈ 2.75). Cancer cell proliferative advantage ceases to dominate on more than half of metabolic channels, but net effect remains negative. Candidate clinical proxy: mild ketosis (GKI ≈ 6–9); metabolic parameters shift but no measurable tumor response. Patients are likely to abandon therapy at this stage.

Spectral flip (Ω ≈ 3.14). Tumor metabolic fluctuations reach their peak. Ketosis is already exerting pressure on cancer cell metabolism ("the heat sink has just turned on"); glycolytic inertia accumulation reaches its maximum, but shielding mechanisms are not yet strong enough. Candidate clinical proxy: peak in tumor metabolic instability — not a change in the mean, but maximum volatility (peak in FDG-PET scan-to-scan variability or day-to-day serum lactate fluctuation amplitude). After this, volatility begins to be suppressed. This is consistent with physical phase transitions: the susceptibility peak precedes the order-parameter flip. Tumor metabolic instability precedes the net decline in tumor metabolism.

Flip (Ω ≈ 3.79). The net energy gain cancer cells derive from fermentation is offset by ketosis-induced metabolic pressure. Metabolic buffering mechanisms (glutamine compensation, autophagy) can no longer fully counteract reduced glucose supply. Candidate clinical proxy (composite): GKI ≤ 2.0 + FDG-PET signal begins to decline + serum lactate no longer rising. This is the point where old antiparasitic drugs suddenly become effective — not because the drug became stronger, but because the bridge's load-bearing capacity dropped below the lethal threshold of the drug's interference.

Establishment (Ω ≈ 4.01). Glycolytic pathways lose competitiveness at every local point to oxidative phosphorylation. Tumor emergence conditions are structurally negated.

Key asymmetry: sprouting to flip = 1.04 (of which sprouting → spectral flip = 0.39, spectral flip → flip = 0.65); flip to establishment = 0.22; ratio approximately 5:1. Entering the window is easy (long period of no visible effect); once past the flip point, only a short distance is needed for structural negation. Abandoning before the flip wastes everything.

The internal structure of the sprouting phase shows two segments: the first (fluctuation accumulation, 0.39) is shorter; the second (shielding begins to intervene but is not yet sufficient, 0.65) is longer. Clinical implication: after entering ketosis, metabolic volatility first increases (first segment), then volatility begins to be suppressed but the tumor has not yet responded (second segment, the most difficult to endure), and only then does the net effect flip.

All numerical coordinates (2.75, 3.14, 3.79, 4.01) are direct mappings from the ZFCρ phase-transition window and constitute candidate coordinates that have not yet been empirically validated in metabolic oncology. Clinical proxies (GKI values, etc.) are likewise candidate correspondences, not calibrated biological constants. The core argument depends only on r >> 1 (asymmetry exists), not on specific values.

Theorem 2: Metabolic Conservation Law

Metabolic reconfiguration is not free. Every "ketone-friendly" cancer, in acquiring the ability to utilize ketone bodies, exposes new vulnerabilities at new coupling points. The cost is not "losing modules" but "reloading coupling points."

Two rounds of literature confrontation identified four cost types:

(1) Signal-coupling dependence. BRAF V600E melanoma: acetoacetate binds to the BRAF V600E–MEK1 signaling axis, making tumor growth dependent on acetoacetate availability. The same molecule serves simultaneously as fuel and signal — bridge reuse. HMGCL/HMGCS1 become synthetic-lethal targets [4, 5].

(2) Redox saturation. Pancreatic cancer (PDAC): ketone body oxidation → elevated NADH/NAD ratio → redox saturation. Yang et al. directly measured that KD + chemotherapy substantially raises tumor NADH, synergistically suppressing growth [6]. Prediction: ketosis + any treatment that increases redox burden should show supra-linear synergy.

(3) Substrate allocation bottleneck. AACS bypass: β-OHB → AACS → cytosolic acetyl-CoA, competing with lipid synthesis, cholesterol synthesis, and histone acetylation (including the newly discovered Kacac modification). AACS itself becomes a target [7].

(4) Ecological dependence deepening. Two-compartment model: stromal cells produce ketones → feed adjacent cancer cells. Cancer cell dependence on MCT1 and ketolysis enzymes deepens. The remainder need not manifest as functional loss; it can also manifest as coupling deepening. MCT1/MCT2 inhibition becomes a target [8].

General pattern: fuel switching repeatedly creates bottlenecks at transport gates (MCT family) and assimilation enzymes (BDH1/OXCT1/HMGCL/AACS/ACSS2). Bridge replacement = exposing new dismantling points at the new bridge's interfaces.

4. Subject Conditions

Condition 1: Emergence Has No Purpose

Emergence does not distinguish good from bad, high from low; it is a result, not a purpose. The literature says "mitochondria are not collapsed but reprogrammed," implying cancer cells "cleverly" adjusted their strategy. In SAE there is no such narrative. Mitochondrial functional reconfiguration and cancer cell proliferative advantage are two emergent results of the same set of foundational-layer conditions, not one causing the other. "Metastasis requires intact Complex I" does not mean "cancer cells preserved Complex I for metastasis," but "in cells where Complex I happens to be intact, the conditions for the emergence called metastasis happen to be satisfied." SAE describes the necessary unfolding of possibilities (constructs) in state space; Darwinian selection is the post-hoc pruning of these emergent results by environmental pressure. Metabolic reprogramming is a topological necessity under foundational-layer constraints, not an intelligent calculation for survival.

Condition 2: The Remainder Cannot Be Eliminated

Ketosis as bridge replacement entails positive-emergence loss (gut microbiome changes, possible transient impairment of acute immune function) — a direct medical manifestation of remainder ineliminability. One cannot dismantle only the construct one wishes to dismantle; the remainder always follows.

Correction: immune cells as high-Ω entities have their sustained/memory function enhanced, not weakened, in the high-Ω environment (ketosis). Ferrere et al.'s data (β-OHB suppresses myeloid PD-L1 via GPR109A, promotes CXCR3+ T cell expansion [9]) directly supports this. The directionality of the remainder is not predetermined — high-Ω emergence benefits from bridge replacement.

Condition 3: Ω Is a Distribution, Not a Scalar

High-Ω subpopulations (stem cells, OXPHOS-dependent) and low-Ω subpopulations (bulk tumor, glycolysis-dependent) coexist within the same tumor. The phase-transition window acts on different parts of the distribution. Pure ketosis may shrink the low-Ω bulk while relatively enriching the high-Ω stem cell subpopulation — "apparent improvement, actual resistance."

5. Rays

Ray 1: Clinical Translation

Sprouting (Ω ≈ 2.75) → Candidate clinical proxy: GKI ≈ 6–9 (mild ketosis).

Spectral flip (Ω ≈ 3.14) → Candidate indicator: peak in FDG-PET scan-to-scan variability or serum lactate day-to-day fluctuation amplitude. Not a mean change, but a volatility peak.

Flip (Ω ≈ 3.79) → Candidate composite proxy: GKI ≤ 2.0 + FDG-PET begins to decline + serum lactate no longer rising. More precise tissue-level candidate: tumor OCR/ECAR ratio drops below a critical value.

Establishment (Ω ≈ 4.01) → Candidate proxy: sustained FDG-PET decline + tumor imaging shrinkage.

Stabilization (Ω ≈ 6.94) → Far outside the window. Long-term maintenance still required after tumor regression; relapse risk persists in the [4.01, 6.94] interval.

All numerical coordinates (2.75, 3.14, 3.79, 4.01) are direct mappings from the ZFCρ phase-transition window and constitute candidate coordinates not yet empirically validated in metabolic oncology. Clinical proxies (GKI values, etc.) are likewise candidate correspondences, not calibrated biological constants. The core argument depends only on r >> 1 (asymmetry exists), not on specific values.

Retrospective test protocol: in case series with dense metabolic monitoring during ketosis therapy, look for two inflection points — (a) metabolic volatility peaks (spectral flip), (b) FDG mean begins to decline (flip). If (a) precedes (b), this is consistent with the susceptibility-before-order-parameter structure of physical phase transitions.

Ray 2: Flip-Point Explanation for Antiparasitic Drugs

Mukherjee's team found in a mouse high-grade glioma model that the maximum efficacy of mebendazole + devimistat occurred only when combined with a ketogenic diet [14]. This cannot directly verify the broad claim that "cancer and parasites share a fermentative bridge," but it is consistent with the flip-point framework's prediction: the drug interference alone is not lethal, but when the metabolic environment is pushed near the flip point by ketosis, the same drug interference can become lethal. Ketosis changes the killing threshold, not the drug itself.

Ray 3: Immune Ω-Stratification

High-Ω cells (immune cells, intact mitochondrial coupling) thrive in the high-Ω environment (ketosis):

• CD8 T cells: ketolysis is required for optimal effector function [10]
• β-OHB suppresses myeloid PD-L1 via GPR109A, promotes CXCR3+ T cell expansion [9]
• Human data: very-low-carbohydrate diets enhance T cell memory formation

Temporal trade-off awaiting verification: acute immune activation depends on glycolytic bursts (M1 macrophage polarization, initial T cell activation); these may be transiently impaired under ketosis. Prediction: ketosis enhances sustained/memory immunity but may transiently impair acute inflammatory responses.

Microbiome–immune axis: ketosis → bifidobacteria depletion → butyrate reduction → Treg differentiation may be affected. But directionality is undetermined: Treg reduction may enhance anti-tumor immunity (since Tregs themselves suppress anti-tumor responses).

Ray 4: Two-Step Therapeutic Architecture

Step one: ketosis as screen — clear low-Ω bulk tumor.

Step two: surviving subpopulations (high-Ω or reconfigured ketone-friendly cells) expose new bridges → design second-round strikes using the new vulnerability faces predicted by the conservation law.

Specific strategies vary by cancer type:

• BRAF V600E melanoma: ketosis + HMGCL/HMGCS1 inhibition (or acetoacetate reduction)
• PDAC: ketosis + treatments that increase redox burden (chemotherapy/radiation, exploiting redox inflexibility)
• Cancers utilizing the AACS bypass: ketosis + AACS inhibition or downstream lipogenesis blockade
• Two-compartment-dependent: ketosis + MCT1/MCT2 inhibition (severing the stromal feeding channel)

6. Non-Trivial Predictions

Prediction 1: Conservation Law — Universal

Every ketone-friendly cancer type has a corresponding new vulnerability face, concentrated at transport gates (MCT family) and assimilation enzymes (BDH1/OXCT1/HMGCL/AACS/ACSS2).

Partial verification: Wu 2025 [11] (MCT1/FASN vulnerability), Kang [5] (HMGCL synthetic lethality), Yang [6] (NADH/NAD redox saturation).

Falsification: A ketone-friendly cancer is found that, upon acquiring ketone utilization capacity, exposes no new targetable vulnerability.

Prediction 2: Selective Enrichment

Pure ketosis shrinks the low-Ω bulk tumor but relatively enriches the high-Ω stem cell subpopulation → "apparent improvement, actual resistance."

Indirect support: GBM stem cell OXPHOS dependence literature [12].

Falsification: In animal models, the Ω distribution of residual tumor after ketosis is identical to pre-ketosis (no selective enrichment).

Prediction 3: Clinical Window

Most patients in existing negative trials did not reach GKI ≤ 2.0 or did not maintain it long enough.

Partial support: ERGO2 internal best-responder analysis (deeper day-6 metabolic state associated with longer PFS/OS).

Falsification: A trial with verified time-in-zone ≥ 4 weeks at GKI ≤ 2.0 is still overall negative.

Prediction 4: Supra-Linear Synergy of Combination Strategies

Ketosis + treatments that increase redox burden show supra-linear synergy in PDAC (effect greater than the sum of both).

Direct verification: Yang et al. KD + chemotherapy synergy data in PDAC models [6].

Falsification: In multiple PDAC models, the effect of ketosis + chemotherapy is strictly equal to the sum of both (purely additive, no synergy).

Prediction 5: Immune Enhancement by Ω-Stratification

Ketosis enhances high-Ω immune function (sustained/memory CD8 T cells, checkpoint blockade response) but may transiently impair low-Ω-dependent acute immune function (glycolytic-burst-dependent M1 polarization, initial T cell activation).

Strong support for first half: Ferrere [9] (checkpoint blockade enhancement), Luda [10] (CD8 ketolysis dependence).

Falsification: CD8 T cell anti-tumor function is not enhanced or is significantly impaired under ketosis.

Prediction 6: Four-Phase Dynamics and Susceptibility Peak

In case series with dense metabolic + imaging monitoring during ketosis therapy, a four-phase sequence should be observed: (a) metabolic volatility increases after mild ketosis (sprouting → spectral flip), (b) volatility peaks then begins to be suppressed while tumor mean indicators remain unchanged (spectral flip → flip, the most difficult phase to endure), (c) tumor response appears within a short time window (flip → establishment). Volatility peak precedes mean change — susceptibility peak precedes order-parameter flip. The duration ratio of (a)+(b) to (c) should be approximately 5:1.

Falsification: Dense longitudinal data shows tumor metabolic volatility and mean change simultaneously (no sequential ordering), or response is linearly gradual with no inflection point and no asymmetry.

7. Conclusion

The mapping of ZFCρ phase-transition window Ω ∈ [2.75, 4.01] to metabolic biology yields a structural framework: the Warburg effect in cancer is Ω-collapse; ketosis is bridge replacement (not bridge removal); metabolic reconfiguration obeys a conservation law (four cost types); the therapeutic window has four-phase dynamics (sprouting → spectral flip → flip → establishment) with a sprouting-to-flip vs flip-to-establishment distance ratio of approximately 5:1. The existence of the spectral flip (metabolic volatility peak preceding mean change) aligns this framework with the susceptibility–order-parameter structure of physical phase transitions.

The power of this framework lies not in explaining existing results (any sufficiently flexible theory can do that) but in yielding six testable non-trivial predictions, each with explicit falsification conditions. The most practically useful — the clinical window prediction — has been independently developed with ERGO2 as a worked example in SAE Methodology Paper VI [1].

Open questions: (1) Precise operationalization of Ω (which modules count, how to count, how to set flux thresholds). (2) Ω distribution atlases for different cancer types. (3) Whether a fifth or sixth cost type remains unidentified in the conservation law. (4) Dynamic evolution of intra-tumor Ω distribution — how the distribution changes over time under ketosis pressure. (5) Long-term running cost of the high-Ω bridge: whether sustained ketosis accelerates immune cell exhaustion through redox-stress overload (e.g., telomere shortening or mitophagy limits), thereby eroding immune benefit on long timescales.

---

Relationship to the SAE framework: this paper is the first application of SAE (Self-as-an-End; foundational paper [13]) to biology. Core concepts — construct, emergence, remainder, Le Chatelier shielding, phase-transition window — are all from the SAE/ZFCρ system. Metabolic modularity Ω is the biological mapping of Ω (number of prime factors of an integer) in ZFCρ. The mathematical foundation of the phase-transition window is in Methodology Paper VI [1].

---

References

[1] Qin, H. "Phase-Transition Windows and Experimental Design." SAE Methodology Paper VI. DOI: 10.5281/zenodo.19464506.

[2] Qin, H. "Why DP min Produces Local Smoothness." ZFCρ Paper 44. DOI: 10.5281/zenodo.19247859.

[3] Qin, H. "The Screened Spectral Bound and the Block Mean Target." ZFCρ Paper 57. DOI: 10.5281/zenodo.19469013.

[4] Xia, S. et al. "Prevention of dietary-fat-fueled ketogenesis attenuates BRAF V600E tumor growth." Cell Metabolism, 2017.

[5] Kang, H.B. et al. "Metabolic rewiring by oncogenic BRAF V600E links ketogenesis pathway to BRAF-MEK1 signaling." Molecular Cell, 2015.

[6] Yang, L. et al. "Ketogenic diet and chemotherapy combine to disrupt pancreatic cancer metabolism and growth." Med, 2022.

[7] Kaluba, F.C. et al. "An alternative route for β-hydroxybutyrate metabolism supports cytosolic acetyl-CoA synthesis in cancer cells." Nature Metabolism, 2025.

[8] Martinez-Outschoorn, U.E. et al. "Ketones and lactate increase cancer cell 'stemness,' driving recurrence, metastasis and poor clinical outcome in breast cancer." Cell Cycle, 2011.

[9] Ferrere, G. et al. "Ketogenic diet and ketone bodies enhance the anticancer effects of PD-1 blockade." JCI Insight, 2021.

[10] Luda, K.M. et al. "Ketolysis drives CD8+ T cell effector function through effects on histone acetylation." Immunity, 2023.

[11] Wu, Z. et al. "Ketone metabolism reprograms lung cancer tumor-initiating cells." Cell Metabolism, 2025.

[12] Janiszewska, M. et al. "Imp2 controls oxidative phosphorylation and is crucial for preserving glioblastoma cancer stem cells." Genes & Development, 2012.

[13] Qin, H. "Systems, Emergence, and the Conditions of Personhood." SAE Paper 1. DOI: 10.5281/zenodo.18528813.

[14] Mukherjee, P. et al. "Therapeutic benefit of combining calorie-restricted ketogenic diet and glutamine targeting in late-stage experimental glioblastoma." Communications Biology, 2019.