Phronesis← papers
Phronesis · working note

Autodisplasia

Rincón, D., with Claude · phronesis · 2026 · a coinage, examined and cut down

The phrase began as “cancer = effigual displacia” and was cut down to one word. Auto-: the body’s own cells, which is what makes this hard to target. Dis-, not the medical dys-: displaced rather than merely disordered. -plasia: growth. Self-displaced growth. The surprising part is how much real ground it lands on: there is a published, actively defended tradition holding that cancer is a disease of tissue organization rather than only of genes, and one much-cited experiment in which malignant cells resumed behaving normally without a single change to their genome. The equals sign is still wrong, the coinage explains nothing on its own, and none of this is a cause of anything.
Stated once, plainly. Nothing here is medical information, and nothing here should influence any decision about screening, diagnosis or treatment. This note is about the shape of a concept and where a framework's vocabulary does and does not reach. Cancer care is not a place for a nice metaphor.

The kernel, granted

The dominant account is the somatic mutation theory: cancer begins in a single cell that accumulates mutations in the genes governing proliferation. On that account the lesion is inside the cell, and position is a consequence rather than a cause.

It is not the only account with a literature. The tissue organization field theory, argued at length by Ana Soto and Carlos Sonnenschein, holds that carcinogenesis begins with a disruption of tissue architecture and the signalling between cells — that the primary event is organizational, at the level of the tissue, and that genomic damage may follow rather than lead. The two differ on something as basic as what a cell does when left alone: quiescence, or proliferation. That is a genuine live disagreement in the field, published in BioEssays and argued over since.

And there is an experiment that makes the organizational case concrete rather than philosophical. Working in three-dimensional culture, Bissell’s group treated malignant human breast cells with an antibody blocking β1-integrin — a receptor by which a cell reads its physical surroundings. The cells reverted. They reassembled a basement membrane, re-established E-cadherin–catenin junctions, reorganized their cytoskeletons, downregulated cyclin D1, upregulated p21, and stopped growing. Injected into mice, the treated cells produced fewer and smaller tumours. A stimulating antibody did nothing; the effect came specifically from blocking the signal.

The genome was not touched. Whatever mutations those cells carried, they carried them still. What changed was the cell’s reading of where it was — and the malignant behaviour went with it. However that result is finally interpreted, it is not a metaphor and it is not disputed.

The mutations stayed. The behaviour left with the context.

Conditioned, and conditional

Those two words are not the same word twice, and the difference is the whole finding. Conditioned is history: what has already happened to a cell and cannot be undone in it. Conditional is dependency: whether the consequence of that history shows up at all, given present circumstances. Bissell’s cells were conditioned permanently and malignant conditionally. The mutations were not reversed. The expression of them was.

This is not a way of speaking. Genetics has carried conditional in exactly this technical sense for decades: a conditional mutant retains normal function under permissive conditions and loses it under restrictive ones. Same genome, two phenotypes, decided by the surround. The teaching example is the Siamese cat, whose coat is pale on the warm trunk and dark at the cool ears, paws and tail — one animal, one genotype, pigment switched by local condition.

Read through that vocabulary, the reversion experiment stops being surprising and becomes a name for something. Malignancy behaved as a conditional phenotype: expressed under a restrictive condition — the wrong matrix contact — and not expressed once the β1-integrin signal reporting that contact was blocked. Whether malignancy is generally conditional in this sense is exactly the open question between the two theories in the first section, and this note does not settle it.

One thing to keep straight, because it would be easy to take wrongly. The Siamese cat is a temperature-conditional mutant, and that is incidental. The relevant condition in the tissue case is matrix contact, not heat. Temperature is doing the work in the cat and no work at all here.

The correction

Three things the phrase claims that the evidence does not.

The equals sign. Cancer is not one disease and does not admit an identity. A leukaemia and a glioma share a name and not much else, and any sentence of the form cancer = X is answering a question that does not have one answer.

The standing of the theory. Tissue organization field theory is a minority position, vigorously argued and vigorously contested. Others hold that the two theories are not even incompatible — that organizational disruption and somatic mutation describe different levels of one process. Granting that displacement-language lands on real ground is not the same as granting that the organizational account has won, and it has not.

The word itself. Effigual, in the first version, was doing aesthetic work, not analytic work. It suggests a likeness reproduced out of place, which is an evocative image for a tumour and is not a measurement, a mechanism, or a prediction. Naming a thing more beautifully does not increase what is known about it. This site has an entire note on the difference, and the same discipline applies to its own coinages — more so, since these are the ones nobody else will check.

The same distinction, without the contested theory

Added 2026-07-20. The argument above leans on a reversion experiment and on tissue organization field theory, and the theory is a minority position — which the section above says plainly. It turns out the distinction does not need it. The same conclusion is reachable from the mutation-centred programme's own data.

Sequencing of histologically normal tissue has found cancer driver mutations to be ordinary rather than exceptional. Roughly one per cent of normal colorectal crypts in middle-aged people carry at least one. Progression is vanishingly rarer than that: fewer than one crypt in 375,000 becomes an adenoma, and fewer than one in three million becomes a carcinoma. The driver mutation is on the order of ten thousand times more common than the disease it is named for.

The mutation is ordinary. The cancer is not. Something between them is doing the deciding.

What the reviews of this literature nominate for that role is clonal competition, the microenvironment, and conditions like chronic inflammation, carcinogen exposure and tissue injury that shift which clones expand. In the vocabulary of the previous section, a driver mutation is a conditioning that most cells carrying it never express, and the surround is what decides. That is the same shape as Bissell's reverted cells, arrived at from the opposite direction and without anyone having to grant the organizational theory.

This does not vindicate that theory, and it is not evidence for the coinage. Nothing here says the primary lesion is organizational rather than genetic — the finding is compatible with a mutation-first account in which most mutant clones simply never acquire the rest of what they need. What it does establish is narrower and worth having: carrying the mutation and having the disease are separated by something, that separation is enormous, and context is in it. The conditioned/conditional distinction can stand on that alone.

The temptation this section creates should be named, because it is strong and it is the reason the section nearly did not get written. A number this lopsided invites being read as a missing link — as though the gap between the mutation and the disease were a space this note's vocabulary had been waiting to occupy. It is not. The mainstream account explains the same gap without difficulty, by multiple hits and clonal competition, and a result both accounts predict cannot choose between them. Nothing above is evidence for displacement-language over any other language; it is evidence that the gap exists, which nobody disputes. The gap is old, well documented, and not a vacancy.

What would change that is one prediction this vocabulary makes and the mutation-first account does not, specified tightly enough to come out against it. Here is where such a prediction would have to live, and the awkward fact that it has already been tested.

The one place the accounts come apart, and it was settled without us

Added 2026-07-20. Reversion is not all-or-nothing across cell lines, and the ordering is the interesting part. The HMT-3522 T4–2 line reverts on inhibition of any one of EGFR, β1-integrin, PI3K or MAPK. The aggressive lines do not: MDA-MB-231 and Hs578T resist single-agent reversion entirely.

Both accounts explain that ordering, which is why it settles nothing on its own. The lines that resist are also the more genomically deranged, so mutation-first reads the ordering as burden and the organizational account reads it as lost machinery for sensing the surround. In these lines the two are confounded: the cells with more mutations are the same cells that have shed their adhesion apparatus. The discriminating experiment is therefore any that pulls those two apart.

It exists. MDA-MB-231 lacks E-cadherin, is among the most metastatic breast lines available, and is refractory to reversion. Transfected with an E-cadherin cDNA it reverts partially; add an inhibitor of β1-integrin, PI3K or MAPK and reversion is nearly complete — assessed by morphology and polarity, growth in three-dimensional matrix and in soft agar, invasiveness, and tumour formation in mice. Its mutational burden was not reduced by any of this. One restored component of the apparatus a cell reads its surroundings with, plus quieter signalling, moved a highly metastatic genome into a nearly normal phenotype.

Revertibility tracked what the cell could read, not what it carried.

That is the prediction this note was missing, and it had been run before the note was written. It is a real result and it favours reading the malignant phenotype as conditional. Two things it is not. It does not discriminate this framework from any other organizational account — every one of them predicts the same outcome, so the coinage is still not the thing doing the work. And reversion of a culture phenotype is not cure: the cells kept everything they had, and what changed was the expression of it.

Which is the honest shape of this whole note. Each time the evidence has come in, it has supported the general claim that context governs expression, and never the particular vocabulary proposed here. A framework that keeps being right in the company of every rival is not yet earning its name.

It also cuts against a tempting reading of this note. If selection among mutant clones were the whole engine, driver-bearing clones should sweep toward malignancy; overwhelmingly they sit, for decades, and stop. Selection is necessary here and plainly not sufficient.

A sharp prediction, and the literature that breaks it

Added 2026-07-20, from a dialogue that pushed the coinage further than the sections above. The two sections above look like they license a discriminating claim: if malignancy is a failure of the cell to read its surround, then editing should target the sensing apparatus — the second term — rather than the driver mutation, because the driver is not where the disease lives. That is sharp, falsifiable, and it is the prediction this note otherwise admits it lacks. It is also, tested against the literature, wrong.

Oncogene addiction refutes the sharp form. Cancers driven by MYC, RAS, BCR-ABL, MET and BRAF regress when the single driver is switched off — and reactivating it restores the malignancy. So the driver is a real lever in an established tumour; "editing the driver cannot help" is false. The mutation-first account wins this round outright.

But the mechanism of that regression is the framework's own sentence. Withdrawing the oncogene reverts the tumour by restoring normal programs — proliferative arrest, apoptosis, senescence, and above all differentiation: the cell falls back to a settled ground state once the displacing force is removed. And it is symmetric — the force returns, the displacement returns. Conditioned and conditional, exactly. So the driver-removal route and the sensor-restoration route are not rivals; they are the same event reached from two ends — release the hold by removing the force, or by rebuilding the cell's capacity to find ground — and the cell returns either way.

Release the hold by whichever end you can reach. The cell was only ever held.

And that is exactly why it earns nothing. The field already holds this, under its own names. Oncogenic competence is the mainstream synthesis that mutation, cell state and microenvironment together gate transformation. MErT — mesenchymal-to-epithelial reverting transition — is reversion driven by the surround, already named and measured. Light is a third route to the same place: photodynamic therapy reverts the immunosuppressive microenvironment and remodels the very cancer-associated fibroblasts that breach a basement membrane, and photobiomodulation steers cell fate, adhesion and matrix remodelling through light rather than through any gene. Every one of these releases the hold without editing the lesion, and not one of them was predicted by this vocabulary — they were found without it.

A field splits the frame in two, which is worth seeing clearly. Physical fields reach a tumour by two different logics, and only one of them is reversion. Tumour-treating fields — the approved glioblastoma therapy — are alternating electric fields that exploit the geometry of a dividing cell: they interfere with the large dipole moment of the mitotic spindle and tear the cleavage furrow, killing cells in the act of division while sparing still ones. That is cytotoxicity selective for the displaced state, not a return to ground — it destroys the cell while it is held, it does not release the hold. The other logic is the one that matters here. Magnetic nanoparticles bound to integrins, driven by an oscillating field, mechanically fire the sensor: force through the integrin recruits talin and vinculin, gates mechanosensitive channels, and drives differentiation — demonstrated for bone, from mesenchymal stem cells. Integrin is the exact receptor Bissell blocked to revert a tumour; this is the same receptor driven from outside to commit a cell's fate. Same second term, opposite direction, and a physical knob on it that already works.

So the sharp claim is false and the true claim is non-discriminating, which is this note's recurring verdict arriving once more — with two seams left where the vocabulary might yet buy a prediction rather than a paraphrase. First: tumours that regress then relapse after the driver is withdrawn are a return-cost story the oncogene literature treats case by case, and a measure that predicted recurrence risk from how far and how expensively an architecture has been displaced would be saying something new. Second, and more concrete: the magnetomechanical integrin tool is demonstrated for differentiation and has never been pointed at cancer reversion. The transfer — take the field that drives a stem cell to commit, and aim it at a malignant cell's failure to read its surround — is a non-obvious move the bone literature has no reason to make and the framework does. Neither seam is built or tested. They are only the first two places the reach could become a finding.

The displacement, imaged — and a claim of this note walked back

Added 2026-07-20. This note has said, more than once, that in cancer the return cost is a metaphor — that nothing measures the displacement in real numbers the way heart rate variability measures it in the autonomic case. That is too strong, and magnetic resonance is why. Magnetic resonance elastography measures tissue stiffness non-invasively, by reading shear waves through the tissue, and a tumour is stiff: extracellular matrix laid down and cross-linked until the tissue's mechanics have visibly departed from normal. The stiffness correlates with grade, stage and prognosis, and in hepatocellular carcinoma it predicts recurrence after resection — independently of stage and grade, the only independent predictor of overall survival in one study. That is the first of the two seams above, and it is not half-answered but answered: a measure of how far the architecture has been displaced predicts recurrence beyond what staging captures, by an instrument already in clinics.

So the mechanical part of the displacement is not a metaphor. It is imaged, graded, and prognostic. What magnetic resonance measures is the current departure from a mechanical ground — not, strictly, the cost to return from it, which is still inferred rather than read; but the displacement itself, in the one dimension that turns out to matter most, has a number and a picture.

The stiffness is the displacement, and you can see it.

And the stiffness is not a marker — it is a cause. This is where the imaging meets everything else in the note. Matrix stiffness drives the malignant phenotype through the exact receptor the reversion experiment turned on: integrins. Stiff matrix clusters integrins, raises Rho-dependent cytoskeletal tension, disrupts polarity and lumen formation, and pushes cells toward invasion — and reducing that tension, or blocking integrin activity, represses the malignant behaviour. The seminal statement of it is a 2005 paper titled, in the field's own words, tensional homeostasis and the malignant phenotype: malignancy as the loss of a mechanical ground state. That the mechanobiology of cancer independently named its ground state homeostasis and its disease a departure from it is either the strongest outside vindication the displacement vocabulary has found, or the plainest demonstration that the vocabulary was never needed — the field arrived at the same picture without it. Honestly, it is both: the abstraction is right, and it is not this note's.

The one real divergence from gene-first, and it works

Added 2026-07-20, and it walks back this note's headline claim. The note has repeatedly reached the verdict that the displacement vocabulary makes no discriminating oncology prediction — that every sharp claim was refuted and every survivor belonged to the field already. That is too flat. There is one discriminating direction here, it diverges cleanly from the gene-first account, and it is borne out: reduce the displacement — soften the matrix — and the disease recedes.

The enzyme that stiffens tumour matrix is lysyl oxidase, which cross-links collagen; the stiffness it builds drives progression through FAK/Src and integrin signalling. Inhibit it — β-aminopropionitrile is the classic tool — and the numbers move: metastatic frequency down by a quarter to a half, whole-body tumour burden down by roughly half to three-quarters, chemotherapy re-sensitised. The intervention targets neither a driver mutation nor the cell at all; it targets the surround, the mechanical ground the cell reads itself against. That is exactly the move the displacement account would make and the somatic-mutation account would not: do not edit the lesion, lower the field it sits in.

Do not edit the lesion. Lower the field it sits in.

So my own summary of this note was wrong in one specific place, and the correction is worth more than the summary. The displacement framework does carry a discriminating oncology direction — mechanical over genetic, surround over lesion — and the direction is real, prognostic (elastography beyond staging) and therapeutic (matrix-softening reducing burden). What it is not is the framework's to own: this is the mechanobiology of cancer, a large and active field, and it built the whole picture — stiffness drives malignancy, so measure it and lower it — without the vocabulary. And the direction is conditional, like every other bridge this pair of notes has crossed: LOX inhibition is tumour-suppressing in some settings and tumour-promoting in others, so “soften the matrix” is a direction, not a prescription. The honest grade rises from no discriminating direction to one, borne out, shared, and conditional — which is a better and truer thing than the flat verdict it replaces.

One more attempt, and it failed, which is worth recording so it is not re-attempted as if it were open. The framework's single real scoop lives in the autonomic note: that the return dynamics matter, not just the displacement — critical slowing down as an early warning. The obvious move is to transfer that here: apply the same slowing-return statistics to a tumour's mechanics over time, and predict transitions the way heart-rate dynamics predict mood ones. It is not novel. Critical slowing down and dynamic-network early-warning signals are already an active cancer field, applied to the epithelial-to-mesenchymal transition and the normal-to-tumour one; and serial elastography already tracks stiffness across chemotherapy to predict response. Both halves exist. So the transfer buys nothing — the framework is, once again, the place two live fields meet rather than the source of either. The narrowest corner that might still be unoccupied is the join itself — critical-slowing-down statistics computed on serial mechanical readings specifically, rather than on gene-network trajectories — and that is offered as a question, unverified, not as a claim of open ground. Re-checked against the current literature, 2026-07-20: the pass only sharpens it. The molecular early-warning field is more crowded than before — dynamic-network biomarkers, entropy-production and mean-flux variants, work through 2025 on cancer-formation and hepatocellular-metastasis tipping points — and every bit of it runs on gene-network trajectories; the same slowing-return statistics computed on serial mechanical readings are still not among them, though longitudinal elastography already exists to supply exactly that stiffness series. The corner holds: still a question, not a claim of open ground, but a specific and testable one.

The displacement, in time

Added 2026-07-20. Everything above has treated the tumour's displacement as spatial — a stiffness, a matrix, an architecture out of place. There is a second axis, and it is temporal. The healthy cell keeps a clock: its core genes oscillate on a daily cycle, and that oscillation is part of its ground state. The malignant cell loses it. Circadian gene rhythms are found arrhythmic or absent in leukaemia, in colorectal liver metastases, and in prostate and breast cancer lines, and — the part that matters — the loss is graded: increasing breast-tumour grade and aggressiveness track increasingly disrupted cellular rhythms, through PER1/PER2 silencing and BMAL1 inactivation. So the cell is displaced in time as it is displaced in space, and by the same kind of measure: how far the rhythm has flattened, like how far the matrix has stiffened.

Displaced in space, and displaced in time — one departure with two projections.

And the temporal axis carries the same two directions the mechanical one did. Restore the clock and the disease recedes: enhancing circadian clock function in cancer cells inhibits their growth — the temporal twin of softening the matrix. And time the intervention to whatever rhythm remains: chronomodulated chemotherapy, timing the drug to the daily cycle of the enzymes that metabolise it, cut severe toxicity in most of the trials that tested it while holding efficacy, and improved survival in some — never worsening it. It is real, it is decades old, and it is still not standard practice, which is the honest shape of the result: a true effect that has not broken through.

None of this is the framework's — chrono-oncology is a deep field, the clock-as-tumour-suppressor and chronotherapy are theirs, and the autonomic note already traced the same clock through the body's daily return. What the vocabulary adds is only the seeing-together: that a tumour is one displacement with two readable projections, spatial and temporal, each graded, each reversible in the same two ways. Whether the two are coupled — whether a stiffening matrix flattens the clock, or a flattening clock softens nothing and the two are independent departures — is the natural next question and is offered as exactly that, a question, not a finding.

What the vocabulary is actually good for

Read carefully, the displacement framework offers this subject a vocabulary and not a cause. Its terms have obvious counterparts here: a tissue’s ground state as its correct architecture; displacement as the departure from it; the cost of holding a displacement as the continuous energetic expense a tumour imposes; return cost as the question of whether an architecture can be recovered once lost. The reversion experiment is, in exactly this language, a return achieved by restoring the second term a cell reads itself against — which is the principle that nothing reads its own state except against a surround.

That is a mapping, and mappings are cheap. It becomes worth something only if it produces a claim that could be false and that someone could test. This note does not contain one. It records where the language reaches, so that the reach is not mistaken later for a finding.

That paragraph now has a qualification, added later and kept here rather than quietly revised. There is a testable claim in the neighbourhood, it is set out in the two sections just above, and it has already been tested — revertibility tracks a cell's capacity to read its surroundings rather than its mutational burden. But it is a claim the organizational reading owns, shared with every rival account of that kind. The sentence above stays true as written: no claim that could be false belongs to this vocabulary in particular.

Where return cost stops being payable

The framework’s distinctive claim is not that displacement happens. It is that displacement is reversible and the expense lies in holding it — so the question with teeth is always what does return cost, and when does it stop being payable. Asked here, that question has an answer already, and it is a physical object.

A neoplasm confined above the basement membrane — a thin sheet of extracellular matrix under the epithelium — is carcinoma in situ. It cannot reach blood vessels or lymphatics, so it cannot seed anywhere else. Breach that sheet and the same lesion is invasive, with the capacity to metastasise. Pathology treats this as the decisive line, and the literature puts it plainly: breaching the basement membrane is the step that separates a lesion which is easily treatable from one that is highly lethal.

So return cost here is not a smooth gradient. It steps, and it steps at a membrane.

The threshold is not a quantity. It is a sheet of matrix, and it is either intact or it is not.

What makes that worth writing down is the second thing the same structure does. The basement membrane is not only the boundary whose breach defines invasion — it is the surround a cell reads itself against. Bissell’s reversion worked by blocking β1-integrin, the receptor through which a cell senses exactly this matrix. The malignant cells were not repaired; their reading of where they were was changed, and the behaviour followed.

One structure, both roles: the thing whose loss makes return unaffordable is the thing whose signal made return happen. In the language of the principles, it is the second term — nothing reads its own state except against a surround, and this is the surround.

That is an observation, not a discovery, and every part of it belongs to the cited work rather than to this framework. What the framework contributes is the question it makes natural to ask: if return cost steps at a boundary rather than climbing, where else does it do that, and is the boundary always physical? This note does not answer it. It is the first thing here that could be answered.

The autonomic surround, and a merger declined

Added 2026-07-20, from a question worth answering in the open: is cancer autonomic displasia — the disorder the companion note describes in the nervous system? The answer is no. Why it is tempting is worth recording, because the connection is real and, lately, large.

The autonomic nervous system does not merely sit near a tumour; it wires into it. Solid tumours are innervated — sympathetic nerves in most, parasympathetic (vagal) nerves in the thoracic and abdominal ones — and in 2025 the field reported the first direct evidence of functional neuron–tumour synapses, neural circuits signalling to cancer cells directly. The signalling drives disease: sympathetic β-adrenergic input promotes progression in several cancers, prostate among the clearest, and noradrenergic stress signalling suppresses anti-tumour immunity, driving T-cell exhaustion through PD-1. This is cancer neuroscience, an active and growing field, and in the framework's language it says one clean thing: the nervous system is part of the surround a tumour reads itself against. The second term here is not only matrix and basement membrane. It has nerves in it.

And it carries the site's own recurring lesson unbidden. The parasympathetic half does not simply oppose the sympathetic: cholinergic signalling is context-dependent — the acetylcholine that damps inflammation elsewhere promotes colorectal cancer, and the vagal anti-inflammatory arc that helps in one setting harms in another. The same signal helps or harms according to the state it lands in — the framework's deepest principle, found again in someone else's field.

The surround has nerves in it — which makes cancer a thing the autonomic system modulates, not a thing it is.

Which is why the merger is declined. Modulates and is are different claims, and the whole discipline of these notes is to keep them apart. Autonomic displasia names a displacement of the return machinery itself, read directly as heart-rate variability. Cancer is a displacement of a cell from its tissue, which the autonomic system — among many surrounds — happens to modulate. To call cancer “autonomic displasia” would fuse the modulator with the disease, the surround with the thing it surrounds: the exact collapse the framework is always one step from making, and the exact thing the grading note exists to refuse. The two notes touch, genuinely, at the nerve endings growing into the tumour. They are not one note.

On the naming, since it took five tries

The term arrived as effigual displacia and grew before it shrank: effusive autodisplasia, then autoeffugal effusive autodisplasia, then a version ending in invasion. Recorded because the discarding is the useful part.

A later pass offered conditioned conditional displasia. That one earned its doubling — conditioned and conditional mean different things, and the pair names the finding exactly — but it was not folded into the title, because conditional is already a term of art in genetics and the section above can use it properly instead of absorbing it into a coinage.

What survived is the part that was doing work from the start: auto-, and the swap of dys- for dis-. A term should fail the removal test — take a piece out and the meaning should change. This one now does.

The limits, plainly

Sources. Reversion of the malignant phenotype: Weaver VM, Petersen OW, Wang F, Larabell CA, Briand P, Damsky C, Bissell MJ, “Reversion of the malignant phenotype of human breast cells in three-dimensional culture and in vivo by integrin blocking antibodies,” J. Cell Biol. 137(1):231–45, 1997 — full text, PubMed. Tissue organization field theory: Soto AM & Sonnenschein C, “The tissue organization field theory of cancer: a testable replacement for the somatic mutation theory,” BioEssays 33(5):332–40, 2011 — full text, doi:10.1002/bies.201100025. Reversion across cell lines, the E-cadherin transfection of MDA-MB-231, and the combined-pathway result: Wang F, Weaver VM, Petersen OW, et al., and the review by Bissell MJ, Radisky DC and colleagues, “Phenotypic reversion or death of cancer cells by altering signaling pathways in three-dimensional contexts,” J. Natl. Cancer Inst. 94(19):1494–503, 2002 — full text, journal. Driver mutations in histologically normal tissue, and the gap between carrying one and progressing: “Somatic Mutations in Normal Tissues: Calm before the Storm,” Cancer Discovery 14(4):605, 2024 — full text; and “Somatic Mutations in Normal Tissues: New Perspectives on Early Carcinogenesis,” Annual Review of Cancer Biology — record. The crypt figures above are quoted from the first of these; they are population estimates for colorectal tissue in middle age and should not be read across to other organs. The basement membrane as the threshold: Breast Ductal Carcinoma in Situ, StatPearls; Chang & Chaudhuri, “Force-dependent breaching of the basement membrane,” Matrix Biology, 2017; and the original observation that invasive tumours lose basement membrane components where benign ones do not, Barsky et al., 1983. Conditional mutants, permissive and restrictive conditions: Singh, Introduction to Genetics §6.7 and the temperature-sensitive case, of which the Siamese cat is the standard illustration. On whether the two theories conflict at all: Rosenfeld S, “Are the somatic mutation and tissue organization field theories of carcinogenesis incompatible?”, Cancer Informatics 12, 2013, doi:10.4137/CIN.S13013.

On the links. Two DOIs are given as identifiers rather than hyperlinks: those publishers refuse automated requests, so the links could not be checked and this site does not carry a link it cannot stand behind. Every hyperlink above returns 200.

These get worked out in the open, at whatever length the problem takes. I do the same thing on a problem of yours — one thing diagnosed and written up plainly, no build. what that costs