In plain terms

The claim being assessed: Chromatin dysregulation can autonomously initiate malignant transformation in mammals and is a primary driver of human cancer rather than a consequence of genetic oncogenic lesions.

Where the evidence stands: Γ — Plausible mechanism. There is a believable biological story and partial support, but the causal step has not been demonstrated.

What this mechanism does: This makes the process worse or keeps it going, but something else starts it.

Strongest evidence for it: Scope-matched SC4 yields Γ for the registered mammalian-autonomy and general human-primacy claim. Transient fly initiation is Ε and its circuit perturbations provide Δ causal contribution, while mammalian epimutational and reprogramming causality is retained in defined engineered/predisposition settings. These positive settings do not establish representative autonomous chromatin-first human origin.

What decisive evidence would look like

In plain terms, researchers would change the proposed cause, confirm it changed as intended, and measure a meaningful health outcome while checking other explanations. The result should be able to support or count against the claim.

Relative explanatory fit: Alternative ranked ahead

What this review means

Chromatin changes can affect cancer initiation, persistence and treatment response. This review separates epigenetic first hits and state-dependent tumor behavior from the claim that autonomous chromatin disruption is the primary cause of human cancer.

Implications for health and treatment: Scope-matched SC4 yields Γ for the registered mammalian-autonomy and general human-primacy claim. Transient fly initiation is Ε and its circuit perturbations provide Δ causal contribution, while mammalian epimutational and reprogramming causality is retained in defined engineered/predisposition settings. These positive settings do not establish representative autonomous chromatin-first human origin.

Limits for individual care: 60 primary reports with 8 selected partial-body examinations; density, heterogeneous controls and incomplete access set the confidence, not correctness. The grade applies to the exact registered claim; stronger findings for narrower subclaims remain separate. No complete-paper audit or blinded calibration is claimed. This educational review does not establish an individual treatment recommendation.

Medical information notice: This review provides general health information, not medical advice. It cannot diagnose a condition or replace an individual discussion with a qualified clinician.

Source checking used complete abstracts/status records and selected original partial-body passages. No complete-paper audit or independent scientific peer review is claimed. Detailed access counts and limitations appear in the review.

A low grade is not an accusation. It means the causal work has not been done yet, which is the ordinary state of most actively studied mechanisms in biology.

At a glance

Sub-claims assessed separately

Figures

Explanatory figures summarizing this review. They carry no evidential weight of their own; where a figure and the review differ, the review governs.

Full assessment

From here the review is written for a technical reader: caps and thresholds, competing models, the independence basis, and the citation ledger.

Sources: 60 primary reports; claims: 65; verified: 65; unverified: 0. Verification is bounded attribution. Access: 8 partial-body and 52 complete abstract/status examinations. Literature cutoff: 2026-10-08; source examination UTC date: 2026-10-09. No complete-paper audit or blinded calibration claimed.

Adversarial review: Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-14. 6 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: Approved as revised.

AT A GLANCE

Epigenetic changes can cause cancer predisposition and sustain malignant states; autonomous chromatin-first initiation as the primary explanation of human cancer remains unestablished.

Confidence basis: Moderate reflects evidence density and reproducibility, not reviewer access depth. Human tumour epigenome mapping and the genetic chromatin-regulator literature are large and well replicated, but the evidence that actually bears on the registered autonomy claim is thin and hard to reproduce: a single fly programme behind the transient-initiation result, one engineered p16-methylation cohort of 6/22 versus 0/18 on a small mixed background, reprogramming tumours that share one OSKM programme, an already-immortalised bronchial line for the RYBP/TET result, bulk-assay detection bounds standing in for proof that no driver variant exists, and one source carrying an Editorial Expression of Concern.

Deciding experiment: Independent native mammalian transient chromatin-state induction and restoration with longitudinal malignant lineage and genome ascertainment

Belief versus evidence: Scope-matched SC4 yields Γ for the registered mammalian-autonomy and general human-primacy claim. Transient fly initiation is Ε and its circuit perturbations provide Δ causal contribution, while mammalian epimutational and reprogramming causality is retained in defined engineered/predisposition settings. These positive settings do not establish representative autonomous chromatin-first human origin.

Best-supported rival: Epigenetic co-driver and maintenance dependency. Co-driver and genotype-first models overlap, but natural epigenetic first hits and targeted methylation incidence demonstrate that chromatin is not always only a downstream consequence; neither model explains all tumors through one order.

COMPETING MECHANISTIC MODELS

Rank 1: Epigenetic co-driver and maintenance dependency (Ζ). Chromatin state and regulators contribute to initiation, persistence and therapy response alongside genotype. Endogenous regulator perturbations, p16 incidence, epimutation predisposition and reversible tumor dependencies converge (25061879,20473912,20951942,27941798,28823558). Specific EZH2 suppression fails to halt progression; interactions and lineage boundaries prevent universal necessity (41223330,25625625).

Rank 2: Genetic oncogenesis with epigenetic adaptation (Ε). Genetic lesions can initiate disease and remodel chromatin; adapted states can be causal thereafter. KRAS/ARID1A/IDH/H3 models and human second hits demonstrate genotype-first or cooperating routes (35134344,25625625,21130701,20473912). Transient fly state and mouse reprogramming/targeted methylation prevent a consequence-only interpretation (38658752,24529372,25061879).

Rank 3: Lineage plasticity and reversible state (Ε). Malignant cells can change differentiation and drug tolerance through reversible regulatory programmes. KDM5A tolerance, RB/p53 prostate lineage and WNT-independent CRC state support functional reversibility (20371346,28059768,28059767,32546576). State reversal need not remove DNA lesions or every malignant cell and is not an origin test.

Rank 4: Chromatin-protein pleiotropy with DNA-damage initiation (Ε). Some chromatin perturbations cause repair failure, insertional events or chromosome changes that cooperate in malignancy. Sustained PH, Dnmt1 and RYBP/TET bodies support genomic alternatives (38888809,12702876,17621273,37272752). Transient PH tumors have retained repair and no recurrent detected driver; methylation-specific controls support a direct state effect (38658752,25061879).

Rank 5: Autonomous epigenetic initiation (Γ). Chromatin state alone is the autonomous initiating cause and primary explanation of human cancer. Fly transient initiation and mammalian reprogramming/methylation are substantial positive cases (38658752,24529372,25061879). Engineering, baseline immortalization, later genomic hits, limited ascertainment and nonrepresentative scope block broad primacy (37272752,20473912,32025007).

The tournament does not alter the nominal level. Compatible pathways and flexible alternatives are not automatically proved by explaining observations.

FULL REVIEW

Registered claim and scope

Chromatin dysregulation can autonomously initiate malignant transformation in mammals and is a primary driver of human cancer rather than a consequence of genetic oncogenic lesions.

The ranking specifies Cancer Biology; preregister only the cancer arm of the index cancer-and-aging entry. Prior aging reviews remain separate. Epigenetic initiation without a DNA driver is distinct from mutations in chromatin regulators, co-driver activity and therapy-induced plasticity.

Selection: public-interest rank 45 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

Evidence and causal interpretation

Human tumor chromatin abnormalities are extensive and measurable: early DNA hypomethylation and pan-cancer accessibility maps establish the phenomenon (6185846,30361341). Cross-sectional signatures do not settle whether the state initiated disease, followed lineage/genotype, or reflects mixture. The primary claim is autonomous mammalian chromatin-first origin and human primacy, not the weaker existence of abnormal epigenomes.

Parreno is a strong native fly experiment, not merely an association (38658752). Brief PH depletion followed by restored protein creates persistent neoplasia; serial allografts at conditions excluding ongoing depletion continue over ten passages and metastasize. Another PRC1 component and altered backgrounds support the effect. Later serial progression may acquire additional genetic or epigenetic changes; that is distinct from initiating driver dependence.

Original genome sections compare four controls with twelve primary fly tumors spanning transient and constant depletion and find no recurrent deleterious variant/CNV pattern across all tumors (38658752). Repair kinetics and karyotype controls further narrow detectable genetic-driver alternatives. These are bulk-assay detection bounds, not a proof that every cell has no variant. Consensus secondary FISH-only wording omits the original WGS and was rejected. The defined transient fly result earns Ε; transfer to mammals requires separate evidence.

zfh1 and Stat perturbations are introduced together with PH knockdown (38658752). They significantly reduce growth and partially restore polarity/differentiation and gene expression; circuit knockdown alone has no obvious eye defect. This supplies Δ causal contribution. It does not show complete elimination of the malignant state or isolated late maintenance after restoration, so the exact necessity assertion is not graded as established.

Rawal uses the same fly platform but prolonged PH depletion (38888809). It yields more baseline damage, delayed break repair and sixfold more karyotypic rearrangements than controls. Transient repair/genome stability is companion evidence, not an independent replication. Its progression diagram is a model of the sustained intervention, not proof every transient tumor inevitably becomes genetically driven.

Mammalian causal positives must remain visible. Ohnishi reports endogenous partial-reprogramming kidney tumors and reprogramming reversibility; a later germ-cell-like model implicates DMRT1 and imprint erasure (24529372,34413299). Both are currently complete-abstract evidence and share an OSKM programme. Reversal demonstrates state plasticity; genome-ascertainment and lineage limits constrain autonomous generalization, not the existence of a causal intervention.

Yu offers an independent mouse incidence experiment (25061879). A promoter-proximal DNA insert recruits p16 methylation; a control insert and cellular demethylation support methylation-dependent silencing. Spontaneous tumors occur in 6/22 homozygous engineered mice versus 0/18 wild-type and survival is shortened. This is strong causal epimutational predisposition. The inserted sequence, small mixed-background cohorts and combined knockout arm are explicit limits; cellular expression restoration is not an in-vivo tumor-reversal test.

Human MLH1 constitutional epimutations provide an epigenetic first-hit equivalent rather than mere tumor correlation (20473912,30283143). Original cases at 18/20 years show allele silencing across normal tissues and loss of the opposite active allele in tumors; other tumors acquire predicted damaging MLH1 substitutions. No critical promoter sequence change was identified, but unknown trans causes remain. Predisposition is supported, while autonomous mutation-free malignancy and population-wide primary cause are different claims.

Targeted methylation in healthy primary breast cells prevents senescence, with catalytic/no-guide controls and multiple donors (29133799). Original body tests report no anchorage-independent colonies and later crisis; hTERT alleviates that later arrest. Off-target methylation exists even though alternative-guide controls narrow it. Proliferation and senescence escape do not equal malignant transformation.

Cui transforms RYBP/TET-deficient bronchial cells with strong xenograft results, but the original starting line is already CDK4/telomerase-immortalized and four genes are genetically disrupted (37272752). QKO tumors arise at 29/31 sites; cells also gain chromosome 12p including KRAS without a KRAS point mutation. This supports chromatin-regulator cooperation and a genomic alternative, not native non-genetic autonomous initiation.

Dnmt1 hypomorphism produces mouse lymphomas with trisomy 15; related tumors include Notch1 retroelement insertions (12702876,17621273). Other methylation-loss experiments produce checkpoint arrest, context-specific tumor promotion/suppression or chromosomal instability; one CRISPRi study finds CIN without increased base mutation/SV burden (22305267,16174748,16204030,37106015). Genomic consequences cannot be asserted uniformly in either direction.

Native genetic chromatin-regulator experiments reject a consequence-only model. Snf5/Ezh2 epistasis, ARID1A liver/colon lesions, Tet2 restoration and KRAS-dependent chromatin rewiring establish functional roles (20951942,26569409,27941798,28823558,35134344). The initiating changes are genetic in these models. Ovarian ARID1A alone fails without PIK3CA, and H3K27M depends on lineage and cooperating lesions (25625625,33545065,35524725).

IDH mutations alter TET activity and differentiation, SMARCB1-mutant rhabdoid tumors have simple but not mutation-free genomes, and BRD9/DNMT3A structural work explains specific regulatory dependencies (21130701,22797305,27941797,31015438,39196936). A separate EZH2 dependency report 26552009 has Editorial Expression of Concern 42760338. It remains in the examined count with Weak weight and no threshold credit; the notice body and other corrections are not presumed harmless.

Reversible drug tolerance and lineage switching are independently causal state effects in existing tumors (20371346,28059768,28059767,32546576). The 2026 EZH2 prostate study does not halt progression/resistance and expands lineages, a useful negative boundary (41223330). Other KMT2D results affect AR signalling or nonhistone AURKA stability, showing that chromatin-protein perturbation can be pleiotropic (41379538,41507359).

Gong jointly perturbs MYB, HDAC2 and FOXA2 in three established colorectal cancer cell lines (39661721). Original bodies verify normal-like expression and reduced proliferation/xenograft growth, not a healthy-cell origin test, elimination of all malignant cells or genetic normalization. Its correction and the colorectal plasticity paper corrections remain metadata-level limitations (39478232).

PCAWG body section 6 explains why some drivers were initially missed: normal contamination, TERT coverage and filtering of clonal blood variants, with candidate aneuploid contributions (32025007). Even after adjustments 5.3% lack identified drivers. This neither proves all cancers have known genomic drivers nor proves those cases originated through autonomous epigenetic changes. Related evolutionary and LINE1 analyses share the cohort (32025013,32024998).

The convergent result is a strong context-dependent epigenetic co-driver/maintenance model, with specific initiating and predisposition cases. The strict registered mammalian-autonomy and primary-human-origin conjunction remains Γ with Moderate confidence. Its experimental gaps require functional malignant outcomes, longitudinal lineage/genome measurements and independent native restoration contrasts, without demanding an impossible clinical initiation RCT.

HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

Subclaim matrix and aggregation

Preregistered aggregation: scope-matched. Scope-matched SC4 yields Γ for the registered mammalian-autonomy and general human-primacy claim. Transient fly initiation is Ε and its circuit perturbations provide Δ causal contribution, while mammalian epimutational and reprogramming causality is retained in defined engineered/predisposition settings. These positive settings do not establish representative autonomous chromatin-first human origin. Sensitivities: weakest-link Β; primary-subclaim Γ; scope-matched Γ.

Decisive-perturbation inventory and causal tally

[1] Transient PRC1 loss followed by PH restoration creates persistent fly neoplasia with serial transplantation and metastasis; bulk genome and repair assays find no recurrent canonical DNA driver. Four controls and twelve tumors spanning transient/constant depletion constrain detectable shared drivers, not all rare variants; zfh1/Stat concomitant knockdown partially rescues induction and phenotype, not proved complete late-maintenance necessity.

[2] Sustained five-day PH depletion causes increased DNA damage, defective break repair and sixfold higher chromosomal rearrangement frequency than controls. Transient tumor genome stability is cited from Parreno rather than independently replicated here; shared fly platform. Sustained damage does not show all transient epigenetic tumors inevitably acquire drivers.

[3] Transient reprogramming induces tumors and tumor-derived iPSCs yield nonneoplastic kidney cells. Engineered expression includes oncogenic/reprogramming factors; normal descendants do not prove no context-dependent sequence changes.

[4] Reprogramming and imprint erasure create tumors with reversible differentiation competence. Same reprogramming programme; nonphysiological factor induction and lineage scope limit primary human inference.

[5] Transient dedifferentiation enables mutant-Kras signaling and tumor development; acinar factors inhibit lesions. Chromatin and genetic lesions cooperate; this supports a causal state contribution rather than autonomous mutation-free initiation.

[6] RYBP/TET loss induces broad CGI methylation and squamous carcinomas after transplantation. HBEC3-KT is already telomerase/CDK4-immortalized, four regulator genes are genetically deleted, and QKO cells gain chromosome 12p including KRAS; not autonomous non-genetic native initiation.

[7] Hit-and-run methylation represses p16 and bypasses senescence. No anchorage-independent colonies; crisis after 60–80 days and hTERT only alleviates later arrest. Off-target methylation exists, despite catalytic/dead-editor and different-guide controls.

[9] Ezh2 loss blocks tumors caused by Snf5 loss. A genetically initiated chromatin dependency, not non-genetic autonomy; linked erratum PMID absent and correction body unaudited.

[11] Epigenetic Hic1 loss cooperates with p53 genetic lesions and alters tumor phenotypes. Predisposed genetic context and cooperating lesions prevent autonomous broad primacy inference.

[12] Hypomethylation increases chromosomal instability without increased assessed sequence-mutation burden. Reduced proliferation and culture scope; genomic instability differs from base-mutation burden and malignant transformation was not established.

[14] Hypomethylation promotes early lesions/liver tumors with Apc LOH but suppresses large intestinal tumors. Genetic predisposition and stage/lineage-dependent opposing effects defeat universal direction.

[15] Hypomethylated tumors carry Notch1 retroelement insertion in 7 of 16 plus chromosomal instability. Shared original Dnmt1 model; followup clarifies genomic contribution rather than independent initial tumor induction.

[18] Genome hypomethylation increases aggressive T-cell lymphomas with chromosome 15 trisomy. Genetic Dnmt1 allele and chromosomal instability coexist; does not establish persistent chromatin-only malignant state.

[19] Primary cells arrest through p14ARF/p53; checkpoint bypass permits aneuploidy. Additional suppressor depletion is required for bypass and malignant initiation is not demonstrated.

[22] Mutant IDH drives methylation, differentiation block and implanted sarcomas. Neomorphic genetic lesion is the initiating intervention; transplantation scope and pleiotropic metabolism remain.

[23] Rhabdoid cancers can have very few coding mutations with recurrent biallelic SMARCB1 loss. Low mutation burden is not mutation-free; exomes do not assess all structural/noncoding events.

[26] Arid1a loss produces inflammation, steatohepatitis and HCC. A genetic chromatin-regulator deletion with inflammatory intermediates, not isolated non-genetic autonomous initiation.

[27] ARID1A regulates ATR-associated repair and checkpoint function and creates PARP sensitivity. Chromatin proteins also govern genome integrity; tumor response is not initiation attribution.

[28] Loss causes invasive colon tumors without canonical APC pathway deregulation and alters enhancers. Arid1a deletion is itself genetic; colon/small-intestine difference limits broad tissue primacy.

[29] Arid1a alone fails tumor induction; concurrent Pik3ca activation enables rapid OCCC. Context-specific epistasis and inflammation undermine a universal autonomous chromatin-only account.

[31] H3K27M effects depend on cell type and tumors require NSC context with TP53 loss. Genetic histone and TP53 lesions; engineered transplantation does not establish native epigenetic-only initiation.

[32] H3K27M changes latency in Nestin progenitors but has reduced effects in Olig2 progenitors. Growth-factor induction and p53 deletion cooperate; cell-of-origin specificity defeats uniform primacy.

[34] Genetic lesions plus transient TGF-beta create persistent YAP/TAZ lineage reversion and WNT-independent resistance. A state transition in genetically altered tumors, not primary non-genetic initiation.

[35] SOX2-dependent lineage switching creates antiandrogen resistance and genetic/SOX2 restoration reverses it. Maintenance and therapy resistance in genetically altered models; not cancer origin.

[36] Rb1/Trp53 loss in Pten-initiated prostate tumors promotes lineage plasticity/metastasis and resistance; EZH2 inhibition restores AR and antiandrogen sensitivity. Genetic lesions initiate the model; regulatory restoration affects progression/resistance, not autonomous epigenetic origin. Related 28059768 programme and author summary 28401192 are not independent replications.

[37] Endogenous Tet2 restoration reverses aberrant self-renewal and suppresses leukemia progression. Predisposed/genetically altered disease and vitamin-C pleiotropy; maintenance response does not identify original initiator.

[39] EZH2 suppression expands lineage diversity without suppressing progression/resistance. Selected advanced disease context; null efficacy is not absence of all EZH2 functions.

[41] KMT2D loss promotes squamous switching and an AURKA dependency. Genetic/state cooperation and therapy-selected lineage transition; also changes kinase stability apart from histone chemistry.

[42] Extended lifespan induces methylation changes while oncogenes confer malignancy traits. Culture-adaptation model; epigenetic changes can reflect proliferation history rather than primary malignant origin.

[44] AP1/SWI-SNF remodeling drives lineage-specific transformation downstream of Kras. A real chromatin co-driver within an upstream genetic programme, not mutually exclusive genetic-versus-epigenetic origin.

[49] Combined regulator inhibition induces enterocyte-like differentiation and reduces malignancy. Three established cancer cell lines and short xenograft growth tests show normal-like differentiation/reduced proliferation, not origin, eradication or normal genome restoration; correction 42138980 body unaudited.

[52] Systemic allelic methylation predisposes to very early cancers, usually with a genetic second hit. Two new early-onset constitutional epimutation cases have somatic loss of the opposite active allele; earlier tumors include predicted deleterious MLH1 missense changes. Epigenetic first-hit predisposition is real, while later genetic hits and unknown trans cues remain.

[53] Most analyzed carriers lack cis nucleotide/structural variants; methylation can erase between generations. Restricted locus ascertainment and rare predisposition, not general human-cancer primacy or absence of tumor second hits.

[55] Retroelement events can generate oncogenic deletions, translocations and amplifications. Shared consortium cohort and genomic reconstruction; cannot prove every tumor requires these events and correction body is unaudited.

[56] Driver sequences often precede diagnosis and later evolution gains instability. Model-based timing and shared cohort; cannot exclude earlier unobserved epigenetic predisposition and correction body unaudited.

[57] Reversible chromatin states mediate drug tolerance in a small cancer-cell fraction. Preexisting malignancy and treatment challenge; cell-state reversibility does not show a non-genetic initial cause.

[58] Most cancers have identified drivers, but around 5 percent have none recognized. Five-point-three percent remain without identified drivers after technical corrections; TERT coverage, normal-panel filtering and aneuploidy expose limits. No identified driver is not proof of no genomic contribution. Shared PCAWG cohort.

[60] Targeted p16 methylation and silencing increase spontaneous tumor incidence and shorten survival in mice; homozygous engineered allele yields 6/22 tumors versus 0/18 wild-type. An inserted cis DNA element recruits methylation; small mixed-background cohorts and combined knockout experiments do not establish mutation-free spontaneous human cancer primacy. Methylation reversal rescues expression in cells, not demonstrated tumor reversal in vivo.

Tally: 38 discriminating contrasts. Null and negative contrasts are discriminators, not positive threshold crossings. Reports and replications are not independent causal lines. Fly companion results share a platform; OSKM follow-ups share induction; PCAWG papers share genomes; report count is not independent causal replication. The EZH2 Expression-of-Concern paper is explicitly excluded from discriminating threshold crossings.

Cap audit, precedence and Β/Γ/Δ boundary

Association cap Β governs SC1 direction-neutral tumor signatures. Fly transient restoration crosses endogenous Δ/Ε in its own scope; partial circuit rescue crosses Δ contribution but not full necessity. Genetically engineered regulator dependencies, human predisposition and targeted p16 methylation collectively support stronger co-driver causality. Their success is not capped simply because one source is a transplant; the broad SC4 claim remains Γ because the combined evidence still lacks representative autonomous human initiation.

Gamma hard-rule audit: Tumor chromatin signatures do not establish initiating direction. Genetic loss of a chromatin regulator is not a non-genetic initiating event. Xenograft growth of immortalized engineered cells does not establish endogenous autonomous mammalian initiation. Normal-like differentiation of existing tumor cells does not establish their non-genetic origin or eliminate every malignant cell.

SC1 cap: “Association-only cancer claims cap at Β.” Human tumor epigenomic abnormalities are well established observationally; existence does not determine initiating direction, composition or causal primacy.

SC4 cap: “Mechanistically plausible cell culture or transplantation-linked models without endogenous in vivo dependency cap at Γ.” Endogenous mouse reprogramming tumors, cis-directed p16 methylation and human MLH1 epimutations are meaningful causal evidence. Genomic engineering, permissive genotype, later genetic hits, transplantation and unresolved driver ascertainment leave the broader autonomy-and-primacy conjunction below Δ.

SC1 field-valid definition: “ - Endogenous genetic manipulation in autochthonous models - Patient-derived system perturbation with matched rescue AND in vivo confirmation - Clonal tracking demonstrating persistence of dependency through selection pressure” Positive full-scope result: no. Human tumor epigenomic abnormalities are well established observationally; existence does not determine initiating direction, composition or causal primacy.

SC2 field-valid definition: “ - Endogenous genetic manipulation in autochthonous models - Patient-derived system perturbation with matched rescue AND in vivo confirmation - Clonal tracking demonstrating persistence of dependency through selection pressure” Positive full-scope result: yes. Transient native PH depletion, restoration, persistent neoplasia and serial propagation establish context-bounded fly initiation. Bulk sequencing, repair and karyotype controls narrow detectable shared DNA-driver alternatives without proving absolute absence of every variant or late evolution.

SC3 field-valid definition: “ - Endogenous genetic manipulation in autochthonous models - Patient-derived system perturbation with matched rescue AND in vivo confirmation - Clonal tracking demonstrating persistence of dependency through selection pressure” Positive full-scope result: yes. Concomitant zfh1 or Stat co-knockdown significantly reduces growth and partly restores polarity/differentiation; a meaningful causal contribution crosses Δ. Incomplete rescue and intervention during induction do not yet establish complete necessity or isolated late maintenance, so Ε for the exact necessity assertion is not claimed.

SC4 field-valid definition: “ - Endogenous genetic manipulation in autochthonous models - Patient-derived system perturbation with matched rescue AND in vivo confirmation - Clonal tracking demonstrating persistence of dependency through selection pressure” Positive full-scope result: no. Endogenous mouse reprogramming tumors, cis-directed p16 methylation and human MLH1 epimutations are meaningful causal evidence. Genomic engineering, permissive genotype, later genetic hits, transplantation and unresolved driver ascertainment leave the broader autonomy-and-primacy conjunction below Δ.

Upward threshold ladder

SC1 Δ: “Requires at least one meaningful causal perturbation showing the proposed mechanism contributes to initiation, maintenance, or relapse.” Satisfied: no. Human longitudinal lineage-resolved epigenomic evidence distinguishing state, composition and initiating direction.

SC2 Δ: “Requires at least one meaningful causal perturbation showing the proposed mechanism contributes to initiation, maintenance, or relapse.” Satisfied: yes.

SC2 Ε: “Requires context-specific causal validity with explicit limits. Cancer stem cells often belong here when a stem-like hierarchy exists in some tumor contexts but fails as a universal model.” Satisfied: yes.

SC2 Ζ: “Requires: - Independent causal confirmation across models - Reduced viability of major alternatives - Relevance beyond one assay class” Satisfied: no. Independent transient perturbation/restoration in another fly lineage, serial genomic lineage tracking and orthogonal local chromatin rescue.

SC3 Δ: “Requires at least one meaningful causal perturbation showing the proposed mechanism contributes to initiation, maintenance, or relapse.” Satisfied: yes.

SC3 Ε: “Requires context-specific causal validity with explicit limits. Cancer stem cells often belong here when a stem-like hierarchy exists in some tumor contexts but fails as a universal model.” Satisfied: no. Separate late circuit interruption after PH restoration, matched controls and circuit-specific rescue that abolish and restore the defined malignant phenotype rather than only reduce induction.

SC4 Δ: “Requires at least one meaningful causal perturbation showing the proposed mechanism contributes to initiation, maintenance, or relapse.” Satisfied: no. Independent endogenous mammalian transient state-specific initiation/restoration with clone-resolved genomic ascertainment, malignant lineage tracking and representative human natural equivalents that distinguish epigenetic first hit from genetic cooperation.

Upper ladder: Human constitutional MLH1 epimutations and endogenous mammalian partial reprogramming/engineered methylation are live causal equivalents, not dismissed for lack of a clinical RCT. Multiple genotypes and orthogonal mouse perturbations establish epigenetic contribution. They do not resolve mutation-free autonomy or the representative primacy of chromatin-first human initiation.

Independence: Convergent reprogramming, locus methylation, human epimutation and endogenous regulator interventions jointly establish a context-dependent epigenetic causal contribution; no single weak leg is re-capped over that set. Shared fly, OSKM, Dnmt1, TCGA and PCAWG components are disclosed. The common unresolved boundary for the registered claim is representative chromatin-first autonomy, not the mere existence of functional epigenetic effects.

Transient fly restoration and functional propagation [1]: Separates continued initiator loss from persistent neoplasia; genome assays narrow common detectable DNA drivers, but one fly programme remains.

Mouse epigenetic initiation and human first-hit predisposition [3, 60, 52]: Independent reprogramming, targeted methylation and natural epimutation address transplantation-only/correlation explanations; their engineering/genetic cooperation prevents universal autonomy inference.

Genetic endogenous chromatin-regulator causality [9, 26, 28, 29, 44]: Independent native lineages and gene interactions reject a consequence-only model, while genetic lesions remain initiating events.

Reversible malignant-state dependency [57, 37, 35, 36, 34]: Different pathways and systems support maintenance/plasticity rather than inferring cause from bulk signatures.

Falsification and pitfalls

Senescence escape without malignant transformation. Primary healthy breast methylation targeting produces proliferation but no anchorage-independent growth and later crisis (29133799). Additional downgrade: False.

Transformation requires genetic cooperation or genomic changes. Immortalized bronchial QKO cells gain 12p; ARID1A ovary needs PIK3CA; H3K27M models require origin/genotype cooperation (37272752,25625625,33545065,35524725). Additional downgrade: False.

Sustained chromatin perturbation creates genetic instability. Constant PH depletion disrupts break repair and karyotype, unlike transient companion assays; Dnmt1 lymphomas acquire trisomy/insertional events (38888809,12702876,17621273). Additional downgrade: False.

Chromatin intervention does not generally improve therapy response. EZH2 suppression expands prostate lineages without blocking progression or antiandrogen resistance in defined models (41223330). Additional downgrade: False.

Mutation of chromatin regulators relabeled as mutation-free initiation. ARID1A, SMARCB1, H3K27M, IDH and TET/DNMT genetic models.

Absence of identified drivers treated as absence of all genomic contribution. Bulk fly assays and residual PCAWG cases; coverage/purity and unknown biology remain.

Existing-tumor reversion treated as origin or cure. Drug-tolerant states, lineage plasticity and three-cell-line normal-like CRC reversion.

Co-knockdown during induction treated as late-maintenance necessity. Parreno zfh1/Stat partial-rescue assays.

Editorial warning ignored. 26552009 has Expression of Concern 42760338; not a threshold-crossing source.

Adversarial steelman: above, then below

A Δ steelman cites native mouse partial reprogramming tumors, p16 methylation-linked incidence and human MLH1 predisposition, plus the fly restoration experiment. These justify causal credit for narrower epigenetic initiation/predisposition and defeat a consequence-only dismissal.

A Β steelman treats human chromatin signatures as correlations and genetic regulator models as ordinary mutation-driven disease. It misses intervention-linked mouse incidence, reversibility, human epigenetic first hits and causal native state dependencies.

Γ is retained for SC4 rather than erasing mammalian causality or upgrading general primacy. Δ is not crossed for the exact autonomy-and-primary-human-driver conjunction; Β is too low for the mechanistic and intervention base.

Final score, causal direction and M/S/D

Scope-matched SC4 yields Γ for the registered mammalian-autonomy and general human-primacy claim. Transient fly initiation is Ε and its circuit perturbations provide Δ causal contribution, while mammalian epimutational and reprogramming causality is retained in defined engineered/predisposition settings. These positive settings do not establish representative autonomous chromatin-first human origin.

Step 1 consequence-only fails because defined state interventions and epimutations can precede and alter malignancy. Step 2 mere association is exceeded. Step 3 general primary/rate-limiting status is not established; step 4 uniform permissive necessity is not established. Step 5 captures context-dependent co-driver and maintenance/plasticity roles before the tournament. Defined fly initiation and p16/MLH1 first-hit settings can have earlier causal roles without assigning them to all human cancers.

Momentum: Mechanistic only. The named artifacts are completed primary perturbation studies (the 2024 transient-PRC1 fly experiment and its companion sustained-depletion report, the 2026 prostate EZH2 study) that refine mechanistic and functional boundaries; no work is named that is presently being run and could cross the next threshold for the registered mammalian-autonomy claim.

Stability: Stable. No named pending result is asserted to be likely to move the registered score imminently.

Divergence: Contested; no single implied-level arithmetic. These dimensions do not change level or confidence.

PROPOSED DECISIVE EXPERIMENTS

Independent native mammalian transient chromatin-state induction and restoration with longitudinal malignant lineage and genome ascertainment

Readout: Persistent endogenous malignant behavior after state-initiator restoration, clone-linked genome/epigenome trajectories and context-matched normal controls. Controls: Match genotype, baseline lineage, regulator abundance and nonchromatin functions; use orthogonal local-state rescue and genomic-driver positive/negative controls. Supporting: Multiple independent native lineages acquire persistent malignancy after a reversible state-only initiating change, with convergent controls excluding measured genetic and pleiotropic alternatives; human natural equivalents show the same order. Falsifying: Malignancy requires permissive genomic lesions, fails after state restoration, tracks acquired genetic events or shows only reversible proliferation/differentiation without malignant behavior.

RESEARCH TIMELINE

KNOWLEDGE GAPS

Independent mammalian transient chromatin-state initiation/restoration beyond one reprogramming or inserted-sequence platform.

Clone-linked genome ascertainment for rare subclones, structural/noncoding events and post-initiation evolution rather than absolute mutation absence assertions.

Human constitutional epimutation incidence, penetrance and pre-tumor trajectories with identified versus unknown cis/trans causes separated.

Late interruption and matched rescue of fly zfh1/JAK-STAT after PH restoration, distinct from co-knockdown during induction.

Local methylation/histone interventions preserving regulator abundance and repair/nonchromatin functions.

Malignant-function tests distinguished from senescence bypass, anchorage growth, short xenograft growth and normal-like markers.

Prespecified tissue/genotype boundaries for chromatin regulator single-hit versus genetic-cooperation models.

Durability, residual malignant cells and genome state after regulatory reversion or therapy-associated lineage switching.

Independent cohort/platform replication with explicit shared TCGA/PCAWG/OSKM/fly components.

Full original supplements, unresolved correction bodies and the EZH2 editorial Expression-of-Concern investigation before publication-grade reliance.

METHODS AND CONSENSUS CONTRIBUTION

Consensus was used through the signed-in browser for discovery, counterevidence, research-gap and timeline suggestions, and attempted checks of specific indexed full-text passages. Provider statements about Methods/Results and any returned passages were compared with original-paper passages where available; failed retrievals, source substitutions and unverified aspects are recorded in the case reconciliation. Successful original-passage comparisons are distinguished from provider summaries and assertions of access. Provider screening counts, citation counts and meters were not used as primary counts or HSEF scores. This source log records access depth; it does not claim complete-paper or supplement review. The monthly API allowance was conserved: zero Consensus API calls in this batch.

REVIEWER NOTES

60 original primary abstract/status records examined, including one Expression-of-Concern original explicitly excluded from causal threshold credit. Eight selected original bodies partially examined, no complete-paper/supplement audit. Cancer-only registered scope; exact archived v5.3, not 5.4. Consensus 2 Pro / 0 API / 0 Deep supplies one correctly attributed original-body comparison; source substitutions, unavailable texts and omitted original WGS are corrected. Literature cutoff 2026-10-08, examinations 2026-10-09 UTC. Coordinator unblinded adjudication, not independent scientific peer review.

Search leads, reviews, protocols and correction notices do not pad the primary count. Downloaded XML does not earn examination credit. Linked notices remain attached; unexamined correction details limit inference. This is a coordinator source/claim consistency assessment, not independent blinded calibration.

ADVERSARIAL REVIEW AND APPROVAL

Authoring model: Sol 6.1. Adversarial reviewer: Claude Opus 5.5 (Anthropic). Date: 2026-10-14.

Scope of checks. PMIDs verified 60/60 against PubMed get_article_metadata: first author, title, journal and DOI matched for every reference; 0 unresolvable, 0 wrong-record substitutions. Correction and concern screen run on all 60 records. The Editorial Expression of Concern recorded against [30] was confirmed — PMID 42760338, Nature Medicine, 2026-09-18, naming "SWI/SNF-mutant cancers depend on catalytic and non-catalytic activity of EZH2" (PMID 26552009) — and the review's handling of it (retained in the examined count, Weak weight, no threshold credit) is correct and is upheld. The other recorded corrections were likewise confirmed and attach to the right parents: PMID 31558726 (Author Correction, BRD9/rhabdoid), PMID 36944736 (LINE-1 retrotransposition), PMID 36697833 (evolutionary history of 2,658 cancers), PMID 36697834 (Pan-cancer analysis of whole genomes) and PMID 42138980 (Correction to "Control of Cellular Differentiation Trajectories for Cancer Reversion", [49]). One further claimed correction — asserted for Wilson 2010 [9] with no identifier — has no erratum record indexed in PubMed and the claim was withdrawn. 11 load-bearing claims spot-checked (Parreno transient PH depletion with restoration, serial allograft and no recurrent driver across 4 controls and 12 tumours; Rawal sustained depletion with sixfold rearrangement increase; Yu 6/22 versus 0/18 p16-methylation incidence; MLH1 constitutional epimutation cases; Gracia-Aznarez/Stolzenberg targeted methylation without anchorage-independent growth; Cui QKO 29/31 sites with 12p/KRAS gain; Dnmt1 hypomorph lymphomas with trisomy 15 and Notch1 insertions; PCAWG 5.3% residual driverless after correction; ARID1A ovary requiring PIK3CA; the 2026 prostate EZH2 negative boundary; Gong three-cell-line reversion) — all matched. Scoring logic re-derived against hsef-hypothesis-review-v5/SKILL.md and references/hsef-framework.md. Internal consistency checked across review.md, scorecard.json and validation-report.json (60 ledger rows, 65 claims, 8 partial-body + 52 abstract/status = 60).

Scoring logic — checked and upheld. Caps are applied per subclaim before thresholds and the cap quotations match the rubric text. The declared scope-matched rule is applied correctly, and this is the case where the framework's warned-about misread would have been easiest to make: SC2 sits at Ε on a Drosophila scope and SC3 at Δ on a fly-PRC1-loss scope, while the central claim is mammalian autonomy plus human primacy. The review takes the overall letter from SC4 — the only subclaim whose scope matches — and reports the narrower fly results separately rather than letting an explicitly scope-restricted Ε set a broadly scoped headline. That is the rule as written. The alternatives are arithmetically correct (weakest-link Β from SC1; primary-subclaim Γ from SC4). The cap-precedence rule is applied in both directions: the convergent mammalian set is not re-capped because one leg is a transplant, and the transplantation/engineering caps do bind the inference from the QKO xenografts and the engineered p16-methylation allele to native human initiation. Role of mechanism resolves at derivation step 5 (reinforcing / stabilizing factor), and the derivation walks the steps in order rather than reading the role off the tournament — correctly, since the nominal model ranks last in this tournament. Divergence Contested is permitted without implied-level arithmetic where no single field-implied level is defensible.

Rejected findings (checked, no change). 1. "Best-supported rival" in AT A GLANCE naming the epigenetic co-driver model. Score card confirms rank_in_tournament: 1 with is_nominal: false; the review's own nominal (M1, autonomous epigenetic initiation) ranks 5. The renderer defect seen elsewhere did not occur here. 2. Six references whose print year differs by one from the PubMed e-pub year (Howard 2008/2007; 25398940 2015/2014; 27941797 and 27941798 2017/2016; 24371281 2014/2013; Gong 2025/2024; Moorman 2025/2024). Standing convention. 3. Two apparent title mismatches against PubMed — "Kras^G12D induces" rendered by PubMed as "Krasinduces", and "DNMT3A^R882" rendered as "DNMT3A". The review's text is correct and the PubMed field is lossy. Standing convention. 4. Bare-numeral citation form in the prose (e.g. "(38658752,24529372)") rather than the "PMID 38658752" form used in REFERENCES. A systematic house convention in this document, applied consistently; not a dropped space inside a word and not rewritten. 5. v5.claim_form "phenomenon" warning. Standing convention: the registered claim states autonomous initiation plus primary-driver status. 6. REVIEWER NOTES vs the spliced Consensus section. No contradiction: the notes record the Consensus usage (two Pro queries, zero API, zero Deep) and one correctly attributed original-body comparison, consistent with the Consensus section and the Methods statement. 7. Collective authors. No reference in this list has a consortium first author; no leading-comma defect present. The PCAWG consortium entries are cited under personal first authors, which matches their PubMed records. 8. Stability "Stable". Correct by definition: no trigger is named, and the framework states that where no trigger can be named the answer is Stable.

Residual limitations. Access depth is unchanged by this pass: eight partial-body and 52 abstract/status examinations, no complete-paper or supplement audit. The Expression of Concern body for PMID 26552009 and the six confirmed corrections remain unexamined in substance, so the review's Weak weighting of [30] rests on the existence of the notice rather than on its content. The partial-body figures for Parreno, Cui, Yu, Gong and PCAWG section 6 could not be re-verified beyond their abstracts and are not contradicted by any record examined.

Approval: Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-14.

REFERENCES

1. Parreno V, Loubiere V, Schuettengruber B et al. Transient loss of Polycomb components induces an epigenetic cancer fate. Nature. 2024. PMID 38658752. DOI: 10.1038/s41586-024-07328-w. Access: full_text_partial. Study: fly-PRC1-thermosensitive-platform.

2. Rawal CC, Loubiere V, Butova NL et al. Sustained inactivation of the Polycomb PRC1 complex induces DNA repair defects and genomic instability in epigenetic tumors. Histochemistry and cell biology. 2024. PMID 38888809. DOI: 10.1007/s00418-024-02302-z. Access: full_text_partial. Study: fly-PRC1-thermosensitive-platform-sustained.

3. Ohnishi K, Semi K, Yamamoto T et al. Premature termination of reprogramming in vivo leads to cancer development through altered epigenetic regulation. Cell. 2014. PMID 24529372. DOI: 10.1016/j.cell.2014.01.005. Access: abstract_metadata. Study: OSKM-partial-reprogramming-mouse-programme.

4. Taguchi J, Shibata H, Kabata M et al. DMRT1-mediated reprogramming drives development of cancer resembling human germ cell tumors with features of totipotency. Nature communications. 2021. PMID 34413299. DOI: 10.1038/s41467-021-25249-4. Access: abstract_metadata. Study: OSKM-partial-reprogramming-mouse-programme-germline-followup.

5. Shibata H, Komura S, Yamada Y et al. In vivo reprogramming drives Kras-induced cancer development. Nature communications. 2018. PMID 29802314. DOI: 10.1038/s41467-018-04449-5. Access: abstract_metadata. Study: PMID29802314.

6. Cui W, Huang Z, Jin SG et al. Deficiency of the Polycomb Protein RYBP and TET Methylcytosine Oxidases Promotes Extensive CpG Island Hypermethylation and Malignant Transformation. Cancer research. 2023. PMID 37272752. DOI: 10.1158/0008-5472.CAN-23-0269. Access: full_text_partial. Study: PMID37272752.

7. Saunderson EA, Stepper P, Gomm JJ et al. Hit-and-run epigenetic editing prevents senescence entry in primary breast cells from healthy donors. Nature communications. 2017. PMID 29133799. DOI: 10.1038/s41467-017-01078-2. Access: full_text_partial. Study: PMID29133799.

8. Pantoja C, de Los Ríos L, Matheu A et al. Inactivation of imprinted genes induced by cellular stress and tumorigenesis. Cancer research. 2005. PMID 15665276. DOI: none assigned (no DOI in the PubMed record). Access: abstract_metadata. Study: PMID15665276.

9. Wilson BG, Wang X, Shen X et al. Epigenetic antagonism between polycomb and SWI/SNF complexes during oncogenic transformation. Cancer cell. 2010. PMID 20951942. DOI: 10.1016/j.ccr.2010.09.006. Access: abstract_metadata. Study: PMID20951942.

10. Farwell DG, Shera KA, Koop JI et al. Genetic and epigenetic changes in human epithelial cells immortalized by telomerase. The American journal of pathology. 2000. PMID 10793065. DOI: 10.1016/S0002-9440(10)65025-0. Access: abstract_metadata. Study: PMID10793065.

11. Chen W, Cooper TK, Zahnow CA et al. Epigenetic and genetic loss of Hic1 function accentuates the role of p53 in tumorigenesis. Cancer cell. 2004. PMID 15488761. DOI: 10.1016/j.ccr.2004.08.030. Access: abstract_metadata. Study: PMID15488761.

12. Besselink N, Keijer J, Vermeulen C et al. The genome-wide mutational consequences of DNA hypomethylation. Scientific reports. 2023. PMID 37106015. DOI: 10.1038/s41598-023-33932-3. Access: abstract_metadata. Study: PMID37106015.

13. Hervouet E, Lalier L, Debien E et al. Disruption of Dnmt1/PCNA/UHRF1 interactions promotes tumorigenesis from human and mice glial cells. PloS one. 2010. PMID 20613874. DOI: 10.1371/journal.pone.0011333. Access: abstract_metadata. Study: PMID20613874.

14. Yamada Y, Jackson-Grusby L, Linhart H et al. Opposing effects of DNA hypomethylation on intestinal and liver carcinogenesis. Proceedings of the National Academy of Sciences of the United States of America. 2005. PMID 16174748. DOI: 10.1073/pnas.0506612102. Access: abstract_metadata. Study: PMID16174748.

15. Howard G, Eiges R, Gaudet F et al. Activation and transposition of endogenous retroviral elements in hypomethylation induced tumors in mice. Oncogene. 2008. PMID 17621273. DOI: 10.1038/sj.onc.1210631. Access: abstract_metadata. Study: Dnmt1-hypomorph-mouse-programme-followup.

16. Karpf AR, Matsui S. Genetic disruption of cytosine DNA methyltransferase enzymes induces chromosomal instability in human cancer cells. Cancer research. 2005. PMID 16204030. DOI: 10.1158/0008-5472.CAN-05-1961. Access: abstract_metadata. Study: PMID16204030.

17. Cheray M, Pacaud R, Nadaradjane A et al. Specific inhibition of one DNMT1-including complex influences tumor initiation and progression. Clinical epigenetics. 2013. PMID 23809695. DOI: 10.1186/1868-7083-5-9. Access: abstract_metadata. Study: PMID23809695.

18. Gaudet F, Hodgson JG, Eden A et al. Induction of tumors in mice by genomic hypomethylation. Science (New York, N.Y.). 2003. PMID 12702876. DOI: 10.1126/science.1083558. Access: abstract_metadata. Study: Dnmt1-hypomorph-mouse-programme.

19. Barra V, Schillaci T, Lentini L et al. Bypass of cell cycle arrest induced by transient DNMT1 post-transcriptional silencing triggers aneuploidy in human cells. Cell division. 2012. PMID 22305267. DOI: 10.1186/1747-1028-7-2. Access: abstract_metadata. Study: PMID22305267.

20. Figueroa ME, Abdel-Wahab O, Lu C et al. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation. Cancer cell. 2010. PMID 21130701. DOI: 10.1016/j.ccr.2010.11.015. Access: abstract_metadata. Study: PMID21130701.

21. Kernytsky A, Wang F, Hansen E et al. IDH2 mutation-induced histone and DNA hypermethylation is progressively reversed by small-molecule inhibition. Blood. 2015. PMID 25398940. DOI: 10.1182/blood-2013-10-533604. Access: abstract_metadata. Study: PMID25398940.

22. Lu C, Venneti S, Akalin A et al. Induction of sarcomas by mutant IDH2. Genes & development. 2013. PMID 24065766. DOI: 10.1101/gad.226753.113. Access: abstract_metadata. Study: PMID24065766.

23. Lee RS, Stewart C, Carter SL et al. A remarkably simple genome underlies highly malignant pediatric rhabdoid cancers. The Journal of clinical investigation. 2012. PMID 22797305. DOI: 10.1172/JCI64400. Access: abstract_metadata. Study: PMID22797305.

24. Wang X, Lee RS, Alver BH et al. SMARCB1-mediated SWI/SNF complex function is essential for enhancer regulation. Nature genetics. 2017. PMID 27941797. DOI: 10.1038/ng.3746. Access: abstract_metadata. Study: PMID27941797.

25. Wang X, Wang S, Troisi EC et al. BRD9 defines a SWI/SNF sub-complex and constitutes a specific vulnerability in malignant rhabdoid tumors. Nature communications. 2019. PMID 31015438. DOI: 10.1038/s41467-019-09891-7. Access: abstract_metadata. Study: PMID31015438.

26. Fang JZ, Li C, Liu XY et al. Hepatocyte-Specific Arid1a Deficiency Initiates Mouse Steatohepatitis and Hepatocellular Carcinoma. PloS one. 2015. PMID 26569409. DOI: 10.1371/journal.pone.0143042. Access: abstract_metadata. Study: PMID26569409.

27. Shen J, Peng Y, Wei L et al. ARID1A Deficiency Impairs the DNA Damage Checkpoint and Sensitizes Cells to PARP Inhibitors. Cancer discovery. 2015. PMID 26069190. DOI: 10.1158/2159-8290.CD-14-0849. Access: abstract_metadata. Study: PMID26069190.

28. Mathur R, Alver BH, San Roman AK et al. ARID1A loss impairs enhancer-mediated gene regulation and drives colon cancer in mice. Nature genetics. 2017. PMID 27941798. DOI: 10.1038/ng.3744. Access: abstract_metadata. Study: PMID27941798.

29. Chandler RL, Damrauer JS, Raab JR et al. Coexistent ARID1A-PIK3CA mutations promote ovarian clear-cell tumorigenesis through pro-tumorigenic inflammatory cytokine signalling. Nature communications. 2015. PMID 25625625. DOI: 10.1038/ncomms7118. Access: abstract_metadata. Study: PMID25625625.

30. Kim KH, Kim W, Howard TP et al. SWI/SNF-mutant cancers depend on catalytic and non-catalytic activity of EZH2. Nature medicine. 2015. PMID 26552009. DOI: 10.1038/nm.3968. Access: abstract_metadata. Study: PMID26552009.

31. Haag D, Mack N, Benites Goncalves da Silva P et al. H3.3-K27M drives neural stem cell-specific gliomagenesis in a human iPSC-derived model. Cancer cell. 2021. PMID 33545065. DOI: 10.1016/j.ccell.2021.01.005. Access: abstract_metadata. Study: PMID33545065.

32. Tomita Y, Shimazu Y, Somasundaram A et al. A novel mouse model of diffuse midline glioma initiated in neonatal oligodendrocyte progenitor cells highlights cell-of-origin dependent effects of H3K27M. Glia. 2022. PMID 35524725. DOI: 10.1002/glia.24189. Access: abstract_metadata. Study: PMID35524725.

33. Brien GL, Bressan RB, Monger C et al. Simultaneous disruption of PRC2 and enhancer function underlies histone H3.3-K27M oncogenic activity in human hindbrain neural stem cells. Nature genetics. 2021. PMID 34294917. DOI: 10.1038/s41588-021-00897-w. Access: abstract_metadata. Study: PMID34294917.

34. Han T, Goswami S, Hu Y et al. Lineage Reversion Drives WNT Independence in Intestinal Cancer. Cancer discovery. 2020. PMID 32546576. DOI: 10.1158/2159-8290.CD-19-1536. Access: abstract_metadata. Study: PMID32546576.

35. Mu P, Zhang Z, Benelli M et al. SOX2 promotes lineage plasticity and antiandrogen resistance in TP53- and RB1-deficient prostate cancer. Science (New York, N.Y.). 2017. PMID 28059768. DOI: 10.1126/science.aah4307. Access: abstract_metadata. Study: PMID28059768.

36. Ku SY, Rosario S, Wang Y et al. Rb1 and Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, and antiandrogen resistance. Science (New York, N.Y.). 2017. PMID 28059767. DOI: 10.1126/science.aah4199. Access: abstract_metadata. Study: PMID28059767.

37. Cimmino L, Dolgalev I, Wang Y et al. Restoration of TET2 Function Blocks Aberrant Self-Renewal and Leukemia Progression. Cell. 2017. PMID 28823558. DOI: 10.1016/j.cell.2017.07.032. Access: abstract_metadata. Study: PMID28823558.

38. Guan Y, Greenberg EF, Hasipek M et al. Context dependent effects of ascorbic acid treatment in TET2 mutant myeloid neoplasia. Communications biology. 2020. PMID 32895473. DOI: 10.1038/s42003-020-01220-9. Access: abstract_metadata. Study: PMID32895473.

39. Jacobi JJ, Wadosky KM, Jaiswal N et al. EZH2 Suppression Diversifies Prostate Cancer Lineage Variant Evolution and Lacks Efficacy in Inhibiting Disease Progression. Cancer research. 2026. PMID 41223330. DOI: 10.1158/0008-5472.CAN-25-2747. Access: abstract_metadata. Study: PMID41223330.

40. Kittane S, Ladewig E, Li T et al. The Histone Methyltransferase KMT2D Is a Critical Mediator of Lineage Plasticity and Therapeutic Response in Castration-Resistant Prostate Cancer. Cancer research. 2026. PMID 41379538. DOI: 10.1158/0008-5472.CAN-25-2053. Access: abstract_metadata. Study: PMID41379538.

41. Chen N, Fang M, Zhong L et al. KMT2D loss drives adeno-to-squamous transition and sensitizes TKI-resistant lung cancer to AURKA inhibition. Cell death and differentiation. 2026. PMID 41507359. DOI: 10.1038/s41418-025-01657-7. Access: abstract_metadata. Study: PMID41507359.

42. Gordon K, Clouaire T, Bao XX et al. Immortality, but not oncogenic transformation, of primary human cells leads to epigenetic reprogramming of DNA methylation and gene expression. Nucleic acids research. 2014. PMID 24371281. DOI: 10.1093/nar/gkt1351. Access: abstract_metadata. Study: PMID24371281.

43. Mathison AJ, Kerketta R, de Assuncao TM et al. KrasG12D induces changes in chromatin territories that differentially impact early nuclear reprogramming in pancreatic cells. Genome biology. 2021. PMID 34649604. DOI: 10.1186/s13059-021-02498-6. Access: abstract_metadata. Study: PMID34649604.

44. Kadur Lakshminarasimha Murthy P, Xi R, Arguijo D et al. Epigenetic basis of oncogenic-Kras-mediated epithelial-cellular proliferation and plasticity. Developmental cell. 2022. PMID 35134344. DOI: 10.1016/j.devcel.2022.01.006. Access: abstract_metadata. Study: PMID35134344.

45. Field MG, Kuznetsov JN, Bussies PL et al. BAP1 Loss Is Associated with DNA Methylomic Repatterning in Highly Aggressive Class 2 Uveal Melanomas. Clinical cancer research : an official journal of the American Association for Cancer Research. 2019. PMID 31285370. DOI: 10.1158/1078-0432.CCR-19-0366. Access: abstract_metadata. Study: PMID31285370.

46. López-Moyado IF, Tsagaratou A, Yuita H et al. Paradoxical association of TET loss of function with genome-wide DNA hypomethylation. Proceedings of the National Academy of Sciences of the United States of America. 2019. PMID 31371502. DOI: 10.1073/pnas.1903059116. Access: abstract_metadata. Study: PMID31371502.

47. Lu R, Wang J, Ren Z et al. A Model System for Studying the DNMT3A Hotspot Mutation (DNMT3AR882) Demonstrates a Causal Relationship between Its Dominant-Negative Effect and Leukemogenesis. Cancer research. 2019. PMID 31164355. DOI: 10.1158/0008-5472.CAN-18-3275. Access: abstract_metadata. Study: PMID31164355.

48. Corces MR, Granja JM, Shams S et al. The chromatin accessibility landscape of primary human cancers. Science (New York, N.Y.). 2018. PMID 30361341. DOI: 10.1126/science.aav1898. Access: abstract_metadata. Study: TCGA-accessible-chromatin-cohort.

49. Gong JR, Lee CK, Kim HM et al. Control of Cellular Differentiation Trajectories for Cancer Reversion. Advanced science (Weinheim, Baden-Wurttemberg, Germany). 2025. PMID 39661721. DOI: 10.1002/advs.202402132. Access: full_text_partial. Study: PMID39661721.

50. Moorman A, Benitez EK, Cambuli F et al. Progressive plasticity during colorectal cancer metastasis. Nature. 2025. PMID 39478232. DOI: 10.1038/s41586-024-08150-0. Access: abstract_metadata. Study: PMID39478232.

51. Feinberg AP, Vogelstein B. Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature. 1983. PMID 6185846. DOI: 10.1038/301089a0. Access: abstract_metadata. Study: PMID6185846.

52. Goel A, Nguyen TP, Leung HC et al. De novo constitutional MLH1 epimutations confer early-onset colorectal cancer in two new sporadic Lynch syndrome cases, with derivation of the epimutation on the paternal allele in one. International journal of cancer. 2011. PMID 20473912. DOI: 10.1002/ijc.25422. Access: full_text_partial. Study: constitutional-MLH1-carrier-programme.

53. Dámaso E, Castillejo A, Arias MDM et al. Primary constitutional MLH1 epimutations: a focal epigenetic event. British journal of cancer. 2018. PMID 30283143. DOI: 10.1038/s41416-018-0019-8. Access: abstract_metadata. Study: constitutional-MLH1-carrier-mechanistic-followup.

54. Timp W, Bravo HC, McDonald OG et al. Large hypomethylated blocks as a universal defining epigenetic alteration in human solid tumors. Genome medicine. 2014. PMID 25191524. DOI: 10.1186/s13073-014-0061-y. Access: abstract_metadata. Study: PMID25191524.

55. Rodriguez-Martin B, Alvarez EG, Baez-Ortega A et al. Pan-cancer analysis of whole genomes identifies driver rearrangements promoted by LINE-1 retrotransposition. Nature genetics. 2020. PMID 32024998. DOI: 10.1038/s41588-019-0562-0. Access: abstract_metadata. Study: PCAWG-2658-cohort.

56. Gerstung M, Jolly C, Leshchiner I et al. The evolutionary history of 2,658 cancers. Nature. 2020. PMID 32025013. DOI: 10.1038/s41586-019-1907-7. Access: abstract_metadata. Study: PCAWG-2658-cohort.

57. Sharma SV, Lee DY, Li B et al. A chromatin-mediated reversible drug-tolerant state in cancer cell subpopulations. Cell. 2010. PMID 20371346. DOI: 10.1016/j.cell.2010.02.027. Access: abstract_metadata. Study: PMID20371346.

58. ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium. Pan-cancer analysis of whole genomes. Nature. 2020. PMID 32025007. DOI: 10.1038/s41586-020-1969-6. Access: full_text_partial. Study: PCAWG-2658-cohort.

59. Gretarsson KH, Abini-Agbomson S, Gloor SL et al. Cancer-associated DNA hypermethylation of Polycomb targets requires DNMT3A dual recognition of histone H2AK119 ubiquitination and the nucleosome acidic patch. Science advances. 2024. PMID 39196936. DOI: 10.1126/sciadv.adp0975. Access: abstract_metadata. Study: PMID39196936.

60. Yu DH, Waterland RA, Zhang P et al. Targeted p16(Ink4a) epimutation causes tumorigenesis and reduces survival in mice. The Journal of clinical investigation. 2014. PMID 25061879. DOI: 10.1172/JCI76507. Access: full_text_partial. Study: PMID25061879.

Source-specific evidence cards

[1] Transient loss of Polycomb components induces an epigenetic cancer fate.

Shows: Transient PRC1 loss followed by PH restoration creates persistent fly neoplasia with serial transplantation and metastasis; bulk genome and repair assays find no recurrent canonical DNA driver.

Does not show: Four controls and twelve tumors spanning transient/constant depletion constrain detectable shared drivers, not all rare variants; zfh1/Stat concomitant knockdown partially rescues induction and phenotype, not proved complete late-maintenance necessity.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Epigenetic perturbations initiate tumours:paragraph0, 1:Epigenetic perturbations initiate tumours:paragraph2, 1:Epigenetic perturbations initiate tumours:paragraph3, 1:Epigenetic perturbations initiate tumours:paragraph4, 1:Epigenetic perturbations initiate tumours:paragraph5, 5:Tumorigenesis requires STAT92E and ZFH1:paragraph0, 5:Tumorigenesis requires STAT92E and ZFH1:paragraph2, 5:Tumorigenesis requires STAT92E and ZFH1:paragraph3, 6:EICs are autonomous immortal tumours:paragraph0, 6:EICs are autonomous immortal tumours:paragraph1, 6:EICs are autonomous immortal tumours:paragraph2. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3, SC4.

[2] Sustained inactivation of the Polycomb PRC1 complex induces DNA repair defects and genomic instability in epigenetic tumors.

Shows: Sustained five-day PH depletion causes increased DNA damage, defective break repair and sixfold higher chromosomal rearrangement frequency than controls.

Does not show: Transient tumor genome stability is cited from Parreno rather than independently replicated here; shared fly platform. Sustained damage does not show all transient epigenetic tumors inevitably acquire drivers.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 2:Results:paragraph0, 2:Results:paragraph25, 2:Results:paragraph27, 2:Results:paragraph28, 2:Results:paragraph29. No complete-paper or supplement audit claimed. UpdateOf: PMID38746379. Subclaims: SC2, SC3, SC4.

[3] Premature termination of reprogramming in vivo leads to cancer development through altered epigenetic regulation.

Shows: Transient reprogramming induces tumors and tumor-derived iPSCs yield nonneoplastic kidney cells.

Does not show: Engineered expression includes oncogenic/reprogramming factors; normal descendants do not prove no context-dependent sequence changes.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[4] DMRT1-mediated reprogramming drives development of cancer resembling human germ cell tumors with features of totipotency.

Shows: Reprogramming and imprint erasure create tumors with reversible differentiation competence.

Does not show: Same reprogramming programme; nonphysiological factor induction and lineage scope limit primary human inference.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[5] In vivo reprogramming drives Kras-induced cancer development.

Shows: Transient dedifferentiation enables mutant-Kras signaling and tumor development; acinar factors inhibit lesions.

Does not show: Chromatin and genetic lesions cooperate; this supports a causal state contribution rather than autonomous mutation-free initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[6] Deficiency of the Polycomb Protein RYBP and TET Methylcytosine Oxidases Promotes Extensive CpG Island Hypermethylation and Malignant Transformation.

Shows: RYBP/TET loss induces broad CGI methylation and squamous carcinomas after transplantation.

Does not show: HBEC3-KT is already telomerase/CDK4-immortalized, four regulator genes are genetically deleted, and QKO cells gain chromosome 12p including KRAS; not autonomous non-genetic native initiation.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 2:Results:paragraph0, 2:Results:paragraph17, 2:Results:paragraph19, 2:Results:paragraph23. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[7] Hit-and-run epigenetic editing prevents senescence entry in primary breast cells from healthy donors.

Shows: Hit-and-run methylation represses p16 and bypasses senescence.

Does not show: No anchorage-independent colonies; crisis after 60–80 days and hTERT only alleviates later arrest. Off-target methylation exists, despite catalytic/dead-editor and different-guide controls.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results:paragraph6, 1:Results:paragraph10, 1:Results:paragraph12, 1:Results:paragraph13, 1:Results:paragraph14, 1:Results:paragraph17, 1:Results:paragraph24, 1:Results:paragraph25. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[8] Inactivation of imprinted genes induced by cellular stress and tumorigenesis.

Shows: Stress triggers imprinted-gene silencing and tumor-associated methylation.

Does not show: Stress response and tumor signatures do not establish autonomous malignant initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[9] Epigenetic antagonism between polycomb and SWI/SNF complexes during oncogenic transformation.

Shows: Ezh2 loss blocks tumors caused by Snf5 loss.

Does not show: A genetically initiated chromatin dependency, not non-genetic autonomy; linked erratum PMID absent and correction body unaudited.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. Correction screen: no erratum, corrigendum or retraction record is indexed in PubMed for this report; status_checked_on 2026-10-14. Subclaims: SC1, SC4.

[10] Genetic and epigenetic changes in human epithelial cells immortalized by telomerase.

Shows: p16 methylation or chromosome9p deletion accompanies immortalization and Rb changes.

Does not show: Clonal chromosomal changes and expression engineering confound purely epigenetic initiation; immortalization is not cancer.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[11] Epigenetic and genetic loss of Hic1 function accentuates the role of p53 in tumorigenesis.

Shows: Epigenetic Hic1 loss cooperates with p53 genetic lesions and alters tumor phenotypes.

Does not show: Predisposed genetic context and cooperating lesions prevent autonomous broad primacy inference.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[12] The genome-wide mutational consequences of DNA hypomethylation.

Shows: Hypomethylation increases chromosomal instability without increased assessed sequence-mutation burden.

Does not show: Reduced proliferation and culture scope; genomic instability differs from base-mutation burden and malignant transformation was not established.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[13] Disruption of Dnmt1/PCNA/UHRF1 interactions promotes tumorigenesis from human and mice glial cells.

Shows: DNMT complex disruption links hypomethylation and tumor-like traits to instability.

Does not show: Genetic and epigenetic alterations jointly occur; phosphorylation perturbs additional functions.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[14] Opposing effects of DNA hypomethylation on intestinal and liver carcinogenesis.

Shows: Hypomethylation promotes early lesions/liver tumors with Apc LOH but suppresses large intestinal tumors.

Does not show: Genetic predisposition and stage/lineage-dependent opposing effects defeat universal direction.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[15] Activation and transposition of endogenous retroviral elements in hypomethylation induced tumors in mice.

Shows: Hypomethylated tumors carry Notch1 retroelement insertion in 7 of 16 plus chromosomal instability.

Does not show: Shared original Dnmt1 model; followup clarifies genomic contribution rather than independent initial tumor induction.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[16] Genetic disruption of cytosine DNA methyltransferase enzymes induces chromosomal instability in human cancer cells.

Shows: DNMT deficiency creates spontaneous persistent chromosome translocations and aneuploidy.

Does not show: Already malignant cell line; genomic consequences support pleiotropic progression, not autonomous native initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[17] Specific inhibition of one DNMT1-including complex influences tumor initiation and progression.

Shows: Different DNMT interaction disruptions promote tumors or sensitize existing cells to therapy.

Does not show: Peptide specificity and preexisting tumor context limit autonomy; not all complex disruption has the same direction.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[18] Induction of tumors in mice by genomic hypomethylation.

Shows: Genome hypomethylation increases aggressive T-cell lymphomas with chromosome 15 trisomy.

Does not show: Genetic Dnmt1 allele and chromosomal instability coexist; does not establish persistent chromatin-only malignant state.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[19] Bypass of cell cycle arrest induced by transient DNMT1 post-transcriptional silencing triggers aneuploidy in human cells.

Shows: Primary cells arrest through p14ARF/p53; checkpoint bypass permits aneuploidy.

Does not show: Additional suppressor depletion is required for bypass and malignant initiation is not demonstrated.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[20] Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation.

Shows: Mutant IDH produces2HG, inhibits TET and causes methylation/differentiation defects.

Does not show: Genetic enzyme alteration precedes chromatin changes; marker/differentiation changes are not sole malignant initiation evidence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[21] IDH2 mutation-induced histone and DNA hypermethylation is progressively reversed by small-molecule inhibition.

Shows: IDH inhibition reverses histone changes rapidly and DNA changes more slowly.

Does not show: Engineered preexisting leukemia background and upstream genetic driver; methylation reversal is not proof of epigenetic origin.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[22] Induction of sarcomas by mutant IDH2.

Shows: Mutant IDH drives methylation, differentiation block and implanted sarcomas.

Does not show: Neomorphic genetic lesion is the initiating intervention; transplantation scope and pleiotropic metabolism remain.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[23] A remarkably simple genome underlies highly malignant pediatric rhabdoid cancers.

Shows: Rhabdoid cancers can have very few coding mutations with recurrent biallelic SMARCB1 loss.

Does not show: Low mutation burden is not mutation-free; exomes do not assess all structural/noncoding events.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[24] SMARCB1-mediated SWI/SNF complex function is essential for enhancer regulation.

Shows: SMARCB1 loss remodels enhancer targeting and creates lineage-specific survival dependencies.

Does not show: Genetic loss underlies epigenetic effects and survival is distinct from autonomous initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[25] BRD9 defines a SWI/SNF sub-complex and constitutes a specific vulnerability in malignant rhabdoid tumors.

Shows: BRD9 complex integrity rather than its bromodomain mediates a selective tumor survival dependency.

Does not show: Genetic SMARCB1 context and protein scaffolding; linked author-correction body unaudited.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID31558726. Subclaims: SC1, SC4.

[26] Hepatocyte-Specific Arid1a Deficiency Initiates Mouse Steatohepatitis and Hepatocellular Carcinoma.

Shows: Arid1a loss produces inflammation, steatohepatitis and HCC.

Does not show: A genetic chromatin-regulator deletion with inflammatory intermediates, not isolated non-genetic autonomous initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[27] ARID1A Deficiency Impairs the DNA Damage Checkpoint and Sensitizes Cells to PARP Inhibitors.

Shows: ARID1A regulates ATR-associated repair and checkpoint function and creates PARP sensitivity.

Does not show: Chromatin proteins also govern genome integrity; tumor response is not initiation attribution.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[28] ARID1A loss impairs enhancer-mediated gene regulation and drives colon cancer in mice.

Shows: Loss causes invasive colon tumors without canonical APC pathway deregulation and alters enhancers.

Does not show: Arid1a deletion is itself genetic; colon/small-intestine difference limits broad tissue primacy.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[29] Coexistent ARID1A-PIK3CA mutations promote ovarian clear-cell tumorigenesis through pro-tumorigenic inflammatory cytokine signalling.

Shows: Arid1a alone fails tumor induction; concurrent Pik3ca activation enables rapid OCCC.

Does not show: Context-specific epistasis and inflammation undermine a universal autonomous chromatin-only account.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[30] SWI/SNF-mutant cancers depend on catalytic and non-catalytic activity of EZH2.

Shows: The original report proposed catalytic and noncatalytic EZH2 dependencies.

Does not show: Editorial Expression of Concern 42760338; excluded from threshold-crossing evidence pending notice investigation, although the original abstract/status record was examined.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ExpressionOfConcernIn: PMID42760338. Subclaims: SC1, SC4.

[31] H3.3-K27M drives neural stem cell-specific gliomagenesis in a human iPSC-derived model.

Shows: H3K27M effects depend on cell type and tumors require NSC context with TP53 loss.

Does not show: Genetic histone and TP53 lesions; engineered transplantation does not establish native epigenetic-only initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[32] A novel mouse model of diffuse midline glioma initiated in neonatal oligodendrocyte progenitor cells highlights cell-of-origin dependent effects of H3K27M.

Shows: H3K27M changes latency in Nestin progenitors but has reduced effects in Olig2 progenitors.

Does not show: Growth-factor induction and p53 deletion cooperate; cell-of-origin specificity defeats uniform primacy.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[33] Simultaneous disruption of PRC2 and enhancer function underlies histone H3.3-K27M oncogenic activity in human hindbrain neural stem cells.

Shows: Histone mutation locks immature lineage through enhancer and PRC2 disruption.

Does not show: Mutation-induced chromatin response; state maintenance and mechanistic culture assays do not establish mutation-free native initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[34] Lineage Reversion Drives WNT Independence in Intestinal Cancer.

Shows: Genetic lesions plus transient TGF-beta create persistent YAP/TAZ lineage reversion and WNT-independent resistance.

Does not show: A state transition in genetically altered tumors, not primary non-genetic initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[35] SOX2 promotes lineage plasticity and antiandrogen resistance in TP53- and RB1-deficient prostate cancer.

Shows: SOX2-dependent lineage switching creates antiandrogen resistance and genetic/SOX2 restoration reverses it.

Does not show: Maintenance and therapy resistance in genetically altered models; not cancer origin.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[36] Rb1 and Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, and antiandrogen resistance.

Shows: Rb1/Trp53 loss in Pten-initiated prostate tumors promotes lineage plasticity/metastasis and resistance; EZH2 inhibition restores AR and antiandrogen sensitivity.

Does not show: Genetic lesions initiate the model; regulatory restoration affects progression/resistance, not autonomous epigenetic origin. Related 28059768 programme and author summary 28401192 are not independent replications.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[37] Restoration of TET2 Function Blocks Aberrant Self-Renewal and Leukemia Progression.

Shows: Endogenous Tet2 restoration reverses aberrant self-renewal and suppresses leukemia progression.

Does not show: Predisposed/genetically altered disease and vitamin-C pleiotropy; maintenance response does not identify original initiator.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[38] Context dependent effects of ascorbic acid treatment in TET2 mutant myeloid neoplasia.

Shows: Ascorbate effects vary with mutation and acetylation state.

Does not show: Biochemical context limits general treatment or causal restoration claims; no primary human cancer origin demonstrated.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[39] EZH2 Suppression Diversifies Prostate Cancer Lineage Variant Evolution and Lacks Efficacy in Inhibiting Disease Progression.

Shows: EZH2 suppression expands lineage diversity without suppressing progression/resistance.

Does not show: Selected advanced disease context; null efficacy is not absence of all EZH2 functions.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[40] The Histone Methyltransferase KMT2D Is a Critical Mediator of Lineage Plasticity and Therapeutic Response in Castration-Resistant Prostate Cancer.

Shows: KMT2D sustains both AR-dependent and stem-like chromatin states and therapeutic response.

Does not show: Protein loss and cofactor recruitment pleiotropy; advanced-state maintenance, not initial origin.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[41] KMT2D loss drives adeno-to-squamous transition and sensitizes TKI-resistant lung cancer to AURKA inhibition.

Shows: KMT2D loss promotes squamous switching and an AURKA dependency.

Does not show: Genetic/state cooperation and therapy-selected lineage transition; also changes kinase stability apart from histone chemistry.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[42] Immortality, but not oncogenic transformation, of primary human cells leads to epigenetic reprogramming of DNA methylation and gene expression.

Shows: Extended lifespan induces methylation changes while oncogenes confer malignancy traits.

Does not show: Culture-adaptation model; epigenetic changes can reflect proliferation history rather than primary malignant origin.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[43] KrasG12D induces changes in chromatin territories that differentially impact early nuclear reprogramming in pancreatic cells.

Shows: Oncogenic Kras precedes a coordinated time-dependent chromatin response.

Does not show: Enhancer-mark modeling and cell systems do not isolate primary chromatin initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[44] Epigenetic basis of oncogenic-Kras-mediated epithelial-cellular proliferation and plasticity.

Shows: AP1/SWI-SNF remodeling drives lineage-specific transformation downstream of Kras.

Does not show: A real chromatin co-driver within an upstream genetic programme, not mutually exclusive genetic-versus-epigenetic origin.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[45] BAP1 Loss Is Associated with DNA Methylomic Repatterning in Highly Aggressive Class 2 Uveal Melanomas.

Shows: BAP1 loss reproduces tumor methylomic repatterning.

Does not show: Chromosome3 loss and BAP1 mutation precede the inferred epigenomic transition; depletion does not reconstruct full initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[46] Paradoxical association of TET loss of function with genome-wide DNA hypomethylation.

Shows: TET loss creates compartment-specific hyper/hypomethylation, repeat activation and genomic alterations.

Does not show: Genetic deficiency and acquired alterations coexist; undermines a generic no-DNA-change assumption.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[47] A Model System for Studying the DNMT3A Hotspot Mutation (DNMT3AR882) Demonstrates a Causal Relationship between Its Dominant-Negative Effect and Leukemogenesis.

Shows: DNMT3A dominant-negative activity and enhancer hypomethylation support leukemogenic function.

Does not show: Genetic mutation and preexisting leukemia-cell background; synergy with IDH mutation constrains autonomy.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[48] The chromatin accessibility landscape of primary human cancers.

Shows: Tumors show lineage-specific accessibility and candidate regulatory differences.

Does not show: Cross-sectional TCGA data and bulk composition do not determine initiation order or chromatin-only primacy.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[49] Control of Cellular Differentiation Trajectories for Cancer Reversion.

Shows: Combined regulator inhibition induces enterocyte-like differentiation and reduces malignancy.

Does not show: Three established cancer cell lines and short xenograft growth tests show normal-like differentiation/reduced proliferation, not origin, eradication or normal genome restoration; correction 42138980 body unaudited.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Result:paragraph64, 1:Result:paragraph65, 1:Result:paragraph66, 1:Result:paragraph68, 1:Result:paragraph70, 1:Result:paragraph72. No complete-paper or supplement audit claimed. ErratumIn: PMID42138980. Subclaims: SC1, SC4.

[50] Progressive plasticity during colorectal cancer metastasis.

Shows: Conserved mutations coexist with increasing fetal-like and noncanonical cell-state plasticity.

Does not show: Metastatic progression is distinct from primary initiation; two author corrections and preprint replacement logged.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. UpdateOf: PMID37662289; ErratumIn: PMID39789339; ErratumIn: PMID42557335. Subclaims: SC1, SC4.

[51] Hypomethylation distinguishes genes of some human cancers from their normal counterparts.

Shows: Specific-gene hypomethylation is observed in four of five cases.

Does not show: Small observational series, tissue composition and temporal direction unresolved.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1.

[52] De novo constitutional MLH1 epimutations confer early-onset colorectal cancer in two new sporadic Lynch syndrome cases, with derivation of the epimutation on the paternal allele in one.

Shows: Systemic allelic methylation predisposes to very early cancers, usually with a genetic second hit.

Does not show: Two new early-onset constitutional epimutation cases have somatic loss of the opposite active allele; earlier tumors include predicted deleterious MLH1 missense changes. Epigenetic first-hit predisposition is real, while later genetic hits and unknown trans cues remain.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[53] Primary constitutional MLH1 epimutations: a focal epigenetic event.

Shows: Most analyzed carriers lack cis nucleotide/structural variants; methylation can erase between generations.

Does not show: Restricted locus ascertainment and rare predisposition, not general human-cancer primacy or absence of tumor second hits.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[54] Large hypomethylated blocks as a universal defining epigenetic alteration in human solid tumors.

Shows: Large hypomethylated blocks occur early across sampled common solid tumors.

Does not show: Early association does not establish causal dominance; array coverage and normal-cell composition limit inference.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[55] Pan-cancer analysis of whole genomes identifies driver rearrangements promoted by LINE-1 retrotransposition.

Shows: Retroelement events can generate oncogenic deletions, translocations and amplifications.

Does not show: Shared consortium cohort and genomic reconstruction; cannot prove every tumor requires these events and correction body is unaudited.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID36944736. Subclaims: SC1, SC4.

[56] The evolutionary history of 2,658 cancers.

Shows: Driver sequences often precede diagnosis and later evolution gains instability.

Does not show: Model-based timing and shared cohort; cannot exclude earlier unobserved epigenetic predisposition and correction body unaudited.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID36697833. Subclaims: SC1, SC4.

[57] A chromatin-mediated reversible drug-tolerant state in cancer cell subpopulations.

Shows: Reversible chromatin states mediate drug tolerance in a small cancer-cell fraction.

Does not show: Preexisting malignancy and treatment challenge; cell-state reversibility does not show a non-genetic initial cause.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[58] Pan-cancer analysis of whole genomes.

Shows: Most cancers have identified drivers, but around 5 percent have none recognized.

Does not show: Five-point-three percent remain without identified drivers after technical corrections; TERT coverage, normal-panel filtering and aneuploidy expose limits. No identified driver is not proof of no genomic contribution. Shared PCAWG cohort.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 6:PCAWG tumours with no apparent drivers. No complete-paper or supplement audit claimed. ErratumIn: PMID36697834. Subclaims: SC1, SC4.

[59] Cancer-associated DNA hypermethylation of Polycomb targets requires DNMT3A dual recognition of histone H2AK119 ubiquitination and the nucleosome acidic patch.

Shows: Dual histone-ubiquitin/acidic-patch recognition recruits DNMT3A and creates promoter methylation.

Does not show: Mechanistic methylation and differentiation blockade, not autochthonous malignant initiation; final replaces preprint.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. UpdateOf: PMID38562823. Subclaims: SC1, SC4.

[60] Targeted p16(Ink4a) epimutation causes tumorigenesis and reduces survival in mice.

Shows: Targeted p16 methylation and silencing increase spontaneous tumor incidence and shorten survival in mice; homozygous engineered allele yields 6/22 tumors versus 0/18 wild-type.

Does not show: An inserted cis DNA element recruits methylation; small mixed-background cohorts and combined knockout experiments do not establish mutation-free spontaneous human cancer primacy. Methylation reversal rescues expression in cells, not demonstrated tumor reversal in vivo.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 0:Publisher HTML paragraphs:paragraph38, 0:Publisher HTML paragraphs:paragraph39, 0:Publisher HTML paragraphs:paragraph40, 0:Publisher HTML paragraphs:paragraph42, 0:Publisher HTML paragraphs:paragraph44, 0:Publisher HTML paragraphs:paragraph45, 0:Publisher HTML paragraphs:paragraph46. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

Documentary record

Every grade on this site is reproducible from its record. The files below are the posted edition exactly as assessed.

• ADVERSARIAL_REVIEW.md
• SHA256.txt
• adversarial-findings.json
• chromatin-dysregulation-cancer.review.md
• chromatin-dysregulation-cancer.scorecard.json
• codex-correction-provenance.json
• consensus-addendum.md
• consensus-validation.json
• edition-approval-audit.json
• reader-summary.json
• validation-report.json
• validation-round2-2026-10-14.json
• validation.txt

Sources: 60 · load-bearing claims: 65 · verified: 65 · unverified: 0 · contradicted: 0 · schema 5.3