Clival chordoma research
Home › Diet, supplements and claims

Metabolism, immune visibility and chordoma immune escape

Agent report for Kya, 15 September 2026, written after her partner asked two questions in em-research: "do tumours become more identifiable to the immune system when switching metabolism?" and "why isn't the patient's immune system attacking the tumour successfully?" Working frame: probable clival chordoma, chondrosarcoma the main imaging lookalike, dedifferentiated and poorly differentiated (INI1-deficient) subtypes still on the list, no histology yet, 51 kg, on morphine, gabapentin and diclofenac, currently eating a high-fat low-carbohydrate diet. Every PMID here was verified through NCBI E-utilities on 15 September 2026, abstract opened and read. No survival or prognosis figures for the patient, by rule. Whether a metabolic switch makes the tumour more aggressive is a separate thread (25) and is not covered here. Brachyury vaccines, the checkpoint review, and opioids and NK cells are in thread 24 and are cited rather than repeated.

Evidence levels. L1 randomised trial, meta-analysis or guideline. L2 prospective human study. L3 retrospective human series. L4 case report or very small series. L5 animal or cell work (stated as in vivo or in vitro), and expert reasoning.

Tier A. The two answers, in plain English

Question A: do tumours become more identifiable to the immune system when they switch metabolism?

Mostly the opposite. Metabolic switching is one of the main ways tumours hide, not a way they get seen. The literature is large, it is almost entirely mouse and cell work, and it runs in one dominant direction: a tumour that is eating hard changes the chemistry of the space around it, and every one of those changes is bad for T cells and NK cells. It eats the glucose they need (Chang 2015, Ho 2015), it floods the space with lactic acid which shuts off their interferon production (Brand 2016), it creates hypoxia and then adenosine, which is a direct off-switch on T cells (Ohta 2006, Hatfield 2015), and it fills the space with lipids that poison the T cells that do get in while feeding the regulatory T cells that suppress them (Manzo 2020, Xu 2021, Wang 2020).

Worse for the specific hope in her partner's question: starving a tumour cell has been shown to make it less visible, not more. Combined oxygen and glucose deprivation stops cancer cells responding to interferon gamma, which is the signal the immune system uses to order a cell to put its antigens on display, and MHC class I presentation falls even with interferon present (Marijt 2019). The same stresses push up HLA-E, a non-classical molecule that actively tells CD8 T cells to stand down (Sasaki 2016).

There are two honest exceptions, and they are worth knowing. First, one 2025 in vitro study found the opposite: prolonged glucose deprivation in melanoma cells lowered EZH2, which lifted a brake on the MHC class I machinery, raised antigen display and increased T cell killing (Edmondson 2025). So the direction is not universal; it depends on the cell and the duration. Second, and more interesting for her, the fuel a ketogenic diet produces is good for her own T cells: beta-hydroxybutyrate is a preferred fuel for CD8 effector T cells and raises their cytokine production and killing (Luda 2023), and in mice a ketogenic diet or oral 3-hydroxybutyrate restored the effect of PD-1 blockade that had failed on a normal diet (Ferrere 2021). A mouse glioma study found a ketogenic diet acted as an immune adjuvant, with more tumour-reactive CD8 and NK activity and less inhibitory receptor expression (Lussier 2016).

The distinction that matters: those last three papers are about the diet helping her immune cells, not about the tumour becoming more identifiable. Nothing in them shows the tumour displaying more antigen. And there is no human trial of any of this in any cancer, let alone chordoma. So the answer to the question as asked is: no, switching metabolism is not a route to making this tumour visible, and the idea of starving it into the open is not supported. If the diet does anything immunological at all, it is on her side of the fight, not the tumour's, and that is a mouse finding.

Question B: why is her immune system not attacking it successfully?

Six reasons, and they stack. This tumour is close to the hardest case in oncology for the immune system.

  1. There is almost nothing to see. Chordoma carries very few mutations, so it makes very few of the abnormal proteins (neoantigens) the immune system recognises as foreign (Tarpey 2017, Zhang 2026). Its one magnificent marker, brachyury, is a self protein from her own embryonic development, so her T cells were trained not to attack it. That is exactly why the brachyury vaccine idea is so attractive and why the randomised trial of it was negative (thread 24).
  2. The shop window is broken. Chordoma tumours have defects in the HLA class I antigen presentation machinery (Patel 2021, 24 tumours), and in a 126-patient sarcoma survey the transporter TAP2 was specifically altered in chordoma (Renne 2024). Deletions of 9p and 10q also silence the cGAS-STING alarm pathway (Zhang 2026). A cell that cannot display its antigens cannot be killed by a T cell however many T cells arrive.
  3. The immune cells are physically kept out. Chordoma is now described as an "immune-excluded" tumour: dense fibrous septa and cancer-associated fibroblast matrix hold T cells in the stroma, with CXCL12 anchoring them there and TGF-beta suppressing them (Zhang 2026). Single-cell plus spatial work on 35 chordomas found immune cells clustered significantly more in the fibrous septa than in the tumour regions themselves (Arrieta 2026). This is the myxoid matrix her partner has read about, and there is real human data behind the idea.
  4. The ones that get in are switched off. Macrophages dominate the chordoma microenvironment, largely M2-polarised (Dridi 2021, Arrieta 2026), CD8 T cells inside show an exhausted phenotype with restricted clonality (Arrieta 2026), and regulatory T cells and PD-1-positive T cells track with worse outcomes (Zou 2019).
  5. Immunoediting plus slow growth. The standard framework is elimination, equilibrium, escape (Schreiber 2011), and the equilibrium state has been demonstrated experimentally: transformed cells held in check by adaptive immunity for long periods, unedited, until they escape and grow out (Koebel 2007, mouse in vivo). A slow tumour gives the immune system almost nothing to react to and gives itself years to edit out anything visible. This is the most likely explanation for the shape of the patient's own history: 12 to 14 months of pain, an earlier head MRI reported as clear while she was symptomatic, and a 3.8 cm mass now.
  6. Possibly the site and the lineage, but this one is speculation. The nucleus pulposus, the adult remnant of the notochord that chordoma comes from, is a classically immune-privileged tissue and its cells express Fas ligand, which kills invading T cells (Takada 2002, rat and human immunohistochemistry). Nobody has shown that chordoma inherits this, and it should be labelled as a hypothesis. Note also what is not true: the clivus is bone outside the dura, so blood-brain-barrier and brain immune privilege arguments do not apply to it.

And the important qualifier: her immune system is not doing nothing, and chordoma is not unconditionally invisible. In one paediatric poorly differentiated (INI1-deficient) chordoma, nivolumab produced a clinical and radiographic response, and a dominant T cell receptor clone recognising a brachyury peptide on MHC was found in the tumour (Williamson 2021). In another INI1-negative case, an EZH2 inhibitor followed by radiotherapy produced a response at untreated distant sites, with CD8 T cells flooding in (Gounder 2019). Both are single patients, both are the SMARCB1-deficient subtype, and that subtype is on the patient's differential. So the honest answer to her partner is: her immune system is losing for six structural reasons, but this tumour family is not beyond being seen, and which version she has changes that answer.

Tier B. Question A in detail: metabolism and immune visibility, both directions

B1. The case that a metabolic switch makes a tumour LESS visible

Glucose competition. The tumour eats the T cells' food. Chang 2015 (Cell, PMID 26321679, L5, mouse sarcoma in vivo) showed that glucose consumption by tumours metabolically restricts tumour-infiltrating T cells, lowering their mTOR activity, glycolytic capacity and interferon gamma production, and that this alone allows tumour progression. Forcing an otherwise immune-controlled "regressor" tumour to be more glycolytic was enough to override T cell control. The same paper showed checkpoint antibodies (anti-CTLA-4, anti-PD-1, anti-PD-L1) partly work by restoring glucose in the microenvironment.

Ho 2015 (Cell, PMID 26321681, L5, mouse melanoma) is the companion: a glucose-poor tumour microenvironment limits T cell glycolysis, and the glycolytic metabolite phosphoenolpyruvate is the specific checkpoint, sustaining calcium-NFAT signalling. Engineering T cells to make more of it restricted tumour growth in mice.

Lactic acid. The waste product is the weapon. Brand 2016 (Cell Metab, PMID 27641098, L5, mouse melanoma in vivo with a human database correlation) showed that melanomas engineered to make less lactic acid grew significantly more slowly and drew in more interferon-gamma-producing T and NK cells, and that this advantage vanished in mice lacking lymphocytes and NK cells or lacking interferon gamma, proving the effect was immune-mediated. Pathophysiological lactic acid concentrations blocked NFAT upregulation in both T and NK cells. In human melanoma datasets, LDHA expression correlated negatively with T cell activation markers.

Hypoxia and adenosine. The off-switch. Ohta 2006 (PNAS, PMID 16916931, L5, mouse in vivo) showed that deleting the A2A adenosine receptor in the host led to rejection of established immunogenic tumours in roughly 60 per cent of mice versus none of the wild-type controls, and that A2A antagonists (caffeine among them) improved tumour control by antitumour T cells. The authors framed this as the explanation for the old paradox of tumours coexisting with antitumour immune cells. Hatfield 2015 (Sci Transl Med, PMID 25739764, L5, mouse in vivo) showed the reverse manoeuvre works: breathing supplemental oxygen reduced intratumoral hypoxia and adenosine, improved CD8 infiltration, weakened regulatory T cell suppression and improved tumour regression, entirely dependent on T and NK cells. Hypoxia also directly disarms NK cells, which lose the ability to upregulate their activating receptors NKp46, NKp30, NKp44 and NKG2D (Balsamo 2013, Eur J Immunol, PMID 23913266, L5 in vitro, human NK cells). This matters for a hypoxic, matrix-dense tumour and it is relevant to thread 22's hyperbaric oxygen question.

Fat. A lipid-rich microenvironment poisons the killers and feeds the suppressors. Manzo 2020 (J Exp Med, PMID 32491160, L5, mouse and human pancreatic tumours) found CD8 T cells inside lipid-rich tumours progressively accumulate long-chain fatty acids which, rather than fuelling them, impair mitochondrial function and reprogramme them away from fat oxidation; restoring the enzyme VLCAD improved T cell persistence in vivo. Xu 2021 (Immunity, PMID 34102100, L5, mouse in vivo) showed the scavenger receptor CD36 takes oxidised lipids into tumour-infiltrating CD8 T cells, causing lipid peroxidation and dysfunction, and that CD36-deficient T cells kept their effector function. Wang 2020 (Nat Immunol, PMID 32066953, L5, mouse in vivo) showed the mirror image: intratumoral regulatory T cells upregulate CD36 to adapt to a lactic-acid-rich environment, and deleting it selectively starved them and suppressed tumour growth. Herber 2010 (Nat Med, PMID 20622859, L5) established the same principle for dendritic cells: lipid accumulation makes them dysfunctional antigen presenters.

This is the part of the literature that deserves a flag given the current food plan. It is mouse and cell work about lipids inside the tumour, and there is no evidence that eating more fat raises intratumoral lipid content in a person or changes any of this. But it is the reason the fat-immunity story cannot simply be told as good news.

OXPHOS-high tumours and resistance to PD-1. Najjar 2019 (JCI Insight, PMID 30721155, L5 with a prospective human correlative component) is the paper her partner's question implies. Patient-derived melanoma lines varied widely in metabolic derangement; deranged oxidative, not glycolytic, metabolism was associated with generating more hypoxia in vivo. Murine melanoma lines with high oxygen consumption drove more T cell exhaustion and less immune activity, and the lines lacking oxidative metabolism were the only ones that responded to anti-PD-1. In prospectively analysed patient samples, oxidative but not glycolytic metabolism was associated with progression on PD-1 blockade. Read across to the patient with care: this is the metabolic direction a tumour moves in when it starts burning fat and ketones rather than glucose, and in melanoma that direction was associated with worse immune control and worse checkpoint response. It is one paper, in one disease, and there is nothing equivalent in chordoma.

Metabolic stress directly shuts down antigen display. This is the most direct answer to the question and it goes the wrong way. Marijt 2019 (J Immunother Cancer, PMID 31196219, L5 in vitro, mouse TC1 and B16F10 lines) deprived tumour cells of oxygen and glucose to mimic a solid tumour interior. MHC class I antigen presentation fell, even in the presence of interferon gamma, because STAT1 failed to be phosphorylated, so the tumour cell became deaf to the immune system's main "display your antigens" order. Forcing active STAT1 rescued presentation fully, and PI3K inhibitors restored both the signalling and recognition by CD8 T cells. Sasaki 2016 (Eur J Immunol, PMID 26711740, L5 in vitro) found that hypoxia or glucose deprivation quantitatively reduced classical MHC class I peptide production and simultaneously increased HLA-E (Qa-1 in mouse), which engages the inhibitory NKG2A/CD94 receptor on activated CD8 T cells and attenuated their response.

B2. The case that a metabolic switch makes a tumour MORE visible

One in vitro paper, and it is specific. Edmondson 2025 (iScience, PMID 40546964, L5 in vitro, human melanoma) found that prolonged glucose deprivation reduced EZH2 and therefore H3K27me3 at MHC class I regulatory genes, raising MHC class I expression and increasing CD8 T cell killing. The authors propose EZH2 inhibition as an immunotherapy adjuvant. Note the resonance with chordoma: the EZH2 inhibitor tazemetostat plus radiotherapy produced exactly this kind of immune activation in an INI1-negative chordoma patient (Gounder 2019). So the "metabolic stress raises antigen display" mechanism exists, it runs through EZH2, and in this tumour family there is already a drug that hits that node without needing a diet to do it.

[Correction 7 Oct 2026: tazemetostat was withdrawn from all markets in March 2026 (Ipsen, after SYMPHONY-1 showed second blood cancers in 5.7% vs 0); trials and compassionate access stopped. PMID 42599162; research 12 A3. Read anything above or below about it as history, not an option.]

The ketone arm, which is about her T cells and not about the tumour. Luda 2023 (Immunity, PMID 37516105, L5, mouse in vitro and in vivo) identified ketone bodies, beta-hydroxybutyrate and acetoacetate, as essential fuels for CD8 effector T cells: beta-hydroxybutyrate directly increased their cytokine production and cytolytic activity, CD8 effector cells preferentially used ketones over glucose to run the TCA cycle both in vitro and in vivo, and ketone oxidation was required for their response to bacterial infection and tumour challenge. Mechanism: beta-hydroxybutyrate was the major source of acetyl-CoA and acted through histone acetylation. Ferrere 2021 (JCI Insight, PMID 33320838, L5, mouse in vivo) showed a ketogenic diet, or 3-hydroxybutyrate alone, produced T-cell-dependent tumour growth retardation, and in settings where anti-PD-1 with or without anti-CTLA-4 had failed, the diet or the ketone restored the response. Adding sucrose (which stops ketogenesis) or blocking the ketone receptor GPR109A abolished the effect. Mechanistically 3-hydroxybutyrate prevented PD-L1 upregulation on myeloid cells and expanded CXCR3-positive T cells. Lussier 2016 (BMC Cancer, PMID 27178315, L5, GL261 intracranial mouse glioma) found mice on a ketogenic diet had more tumour-reactive CD8 and NK cytokine production and cytolysis, more CD4 infiltration, unchanged regulatory T cell numbers, and reduced inhibitory receptor and ligand expression; the authors describe the diet as a possible immune adjuvant.

How to hold these together honestly. The ketone papers are the best pro-diet immunological evidence that exists and they are all mice and cells. They say her T cells can run on ketones and may work better on them. They do not say the tumour becomes more identifiable. And they sit directly against thread 24's finding that chondrosarcoma cells switch on ketolysis under nutrient deprivation and use ketones for proliferation, and that glioblastoma progressed through a ketogenic diet in mice. Same fuel, both sides of the fight. Nobody has measured which side wins in a person, in this disease. This is not a reason to go stricter on the diet, and not a reason to abandon it. Calories first, as agreed, still stands.

Tier C. Question B in detail: the chordoma immune landscape, verified

C1. How little there is to see

C2. HLA class I and the antigen presentation machinery

That combination, few neoantigens plus a defective display system, is the core of the answer to her partner's question. Even a perfectly primed T cell needs the tumour cell to hold up the antigen.

C3. Checkpoint molecules: the picture is real but the numbers disagree by method

C4. Lymphocytes, macrophages and the physical barrier

C5. Immunoediting, and why a slow tumour can grow for years under surveillance

The framework is elimination, equilibrium and escape: immunity destroys some cancers, holds others in check, and selects the survivors for reduced immunogenicity, so the tumour that finally appears clinically is the one that has already been sculpted to be invisible (Schreiber 2011, Science, PMID 21436444, review). The middle state is not just theory. Koebel 2007 (Nature, PMID 18026089, L5, mouse primary chemical carcinogenesis in vivo) demonstrated equilibrium directly: neoplastic cells held by adaptive immunity, transformed but proliferating poorly in vivo, remaining unedited while restrained, and becoming edited only when they spontaneously escaped and grew into clinically apparent tumours.

Applied to the patient, as reasoning and not as data (L5). A tumour with a handful of drivers, a self-protein lineage marker, defective antigen display and a dense matrix is the archetype of a tumour that spends years in equilibrium or slow escape. It also fits the timeline in the master file: trigeminal-distribution pain for 12 to 14 months, a head MRI reported as clear roughly 10 months before admission while she was already symptomatic, ptosis from July and diplopia from late August. The immune system's failure here is not an acute collapse and there is no reason to look for one. It is a slow structural mismatch, and nothing she or her partner did caused it.

C6. Does checkpoint blockade do anything in chordoma? Sometimes, in the salvage setting

C7. Chondrosarcoma, briefly, because it is still on the differential

Tier D. What in the patient's current situation plausibly affects her immunity

Ranked by how much evidence there is and how much can be done about it. Everything here is about general immune competence, particularly around surgery. None of it is a treatment for the tumour.

  1. Being 51 kg and having lost 12 to 15 per cent of her body weight. The biggest item, and the most fixable. Protein-energy malnutrition is the commonest cause of secondary immunodeficiency worldwide (standard teaching, thread 13). The best recent human data: 43 colorectal cancer patients assessed prospectively before resection, in whom malnutrition was associated with more tissue-infiltrating neutrophils and altered T cell phenotypes, including reduced effector cytokine expression in conventional T-helper and iNKT cells inside the tumour (Perillo 2026, Front Nutr, PMID 42317870, L2 prospective pilot). Different cancer, small, and not proof of causation, but it is the only human study here that links the thing she actually has, undernutrition before an operation, to the thing her partner is asking about, anti-tumour T cell function. Calories and protein are the immune intervention with the best evidence in this whole file.
  2. Opioids. Covered in threads 13 and 24, and not repeated. The short version stands: the NK suppression signal is real, mu-mediated, and most relevant in the perioperative window, and the one large randomised test of the regional-versus-opioid anaesthesia question (Sessler 2019, Lancet, PMID 31645288, L1, 2,108 women) was negative for recurrence. Do not undertreat her pain because of it.
  3. Pain and stress hormones, which cut the other way from the opioid worry. Catecholamines and prostaglandins released around surgery are implicated in pro-metastatic processes, and in a randomised double-blind placebo-controlled biomarker trial of 34 colorectal cancer patients given propranolol plus etodolac for 20 perioperative days, treated tumours showed reduced epithelial-to-mesenchymal transition and increased tumour-infiltrating CD56 natural killer cells (Haldar 2020, Cancer, PMID 32533792, L1, biomarker endpoints, small; three-year recurrence 2 of 16 versus 6 of 18, p = 0.239, not significant). Separately, gabapentin given before hysterectomy in a 60-patient randomised trial lowered pain scores, plasma cortisol (p < 0.001) and catecholamines (p < 0.05) 24 hours after surgery, with modest shifts in lymphocyte subsets (Karbić 2014, Int Immunopharmacol, PMID 25448495, L1, small). The reframe worth giving her partner: uncontrolled pain is itself an immunosuppressive stressor, so controlling her pain properly is not simply a trade against immunity.
  4. Sleep. Sleep loss reduces NK activity and vaccine responses and, when chronic, produces low-grade systemic inflammation (Besedovsky 2019, Physiol Rev, PMID 30920354, review; already in thread 13). Free, and currently one of the worse-controlled variables in the house.
  5. Diclofenac and COX-2. The counter-intuitive one, and it is reassuring. Tumour-derived prostaglandin E2 is immunosuppressive: in mouse melanoma, breast and colorectal models, genetic ablation of cyclooxygenases or prostaglandin E synthases rendered tumours susceptible to immune control and shifted the inflammatory profile toward anti-cancer pathways, and COX inhibition synergised with anti-PD-1 to eradicate tumours; the COX-dependent suppressive signature was conserved in human melanoma biopsies (Zelenay 2015, Cell, PMID 26343581, L5 in vivo). Diclofenac is a non-selective COX inhibitor. So her regular diclofenac sits on the immunologically favourable side of this question, not the unfavourable one. That is mechanism, not evidence of benefit in chordoma, and it is not a reason to change any dose in either direction. It is a reason to stop worrying about that particular drug. (Her diclofenac needs watching for kidneys and bleeding, per the medicine card and thread 13, which is a separate matter.)
  6. Dexamethasone, if it is started. Iorgulescu 2021 (Clin Cancer Res, PMID 33239433, L5 mouse in vivo plus L3 human) is the sharpest data. In immunocompetent mice with GL261 and CT-2A glioblastoma, concurrent dexamethasone reduced survival on anti-PD-1 in a dose-dependent way, abrogating benefit with or without radiotherapy; dexamethasone reduced T lymphocyte numbers by increasing apoptosis, reduced their functional capacity, and generally reduced myeloid and NK populations. In 181 consecutive IDH wild-type glioblastoma patients treated with PD-(L)1 blockade, baseline dexamethasone was the strongest predictor of poor survival on multivariable adjustment, with a hazard ratio of 2.16 even below 2 mg. What this does and does not mean: it is not a reason to refuse dexamethasone if she needs it for oedema or for cranial nerve function, and steroids around skull base surgery are standard. It is a strong reason to ask for the lowest effective dose, the shortest course, and a written taper.
  7. What has no evidence either way. Morphine's effect on the specific mechanisms in this file (antigen presentation, T cell exclusion) is unstudied. Gabapentin has no cancer-immunity literature. Neither is a reason to change anything.

Tier E. What this means for the patient this week

The honest headline for her partner, in one paragraph. There is no way to make this tumour more visible by changing what the patient eats, and the published mechanisms mostly run the other way: metabolic stress on a tumour cell tends to shut its antigen display down, not open it up. Her immune system is not failing through weakness or through anything that was neglected. It is facing a tumour with almost no foreign-looking proteins, a damaged display system, and a wall of matrix that holds T cells in the stroma, and that combination has been measured in human chordoma tissue by several independent groups. The two things in the house that genuinely bear on her immune competence are her weight and her sleep, and the things that bear on it most are what happens around the operation. The one variable that could change the whole immunotherapy conversation is the histology, and specifically the INI1 stain.

No change to the food plan on the strength of this file. Calories and protein first, as thread 24 concluded. The pro-diet immune evidence (ketones fuel CD8 T cells, ketones restored PD-1 responses in mice) is real but it is mouse and cell work about her T cells, not about tumour visibility, and it is offset by the ketolysis findings in thread 24. Nothing here justifies going stricter, and nothing here justifies stopping.

Questions worth putting to the surgeon on Wednesday (17 September). Four of these are additions to the consultation sheet; the rest are already there in some form.

  1. At biopsy, alongside brachyury, INI1/SMARCB1, TP53, p16 and Ki-67, can the pathologist also run PD-L1 and an HLA class I or MHC class I stain, and can enough tissue be banked fresh frozen for sequencing? Reason to give them: the only chordoma patients with convincing checkpoint inhibitor responses in the literature are INI1-deficient, and PD-L1 plus HLA class I status is what any future immunotherapy discussion would be built on. Ask them to note where in the tumour the sample came from, because antigen presentation defects vary between areas of the same tumour (Renne 2024).
  2. If PD-L1 is reported, which assay and which threshold? Worth asking because tumour cell PD-L1 in chordoma has been reported at 0 per cent by one method and 86 per cent by another, so a bare positive or negative is not interpretable on its own.
  3. If the histology is dedifferentiated chondrosarcoma, will the immune markers be assessed on the dedifferentiated component specifically? That is the only chondrosarcoma subtype with PD-L1 expression, lymphocyte infiltration and retained HLA class I, and the expression was found exclusively in the dedifferentiated part.
  4. If dexamethasone is started, what is the lowest effective dose, and can we have the planned duration and taper in writing? The immunological cost of steroids is dose-dependent and measurable, and the ask is for the minimum that controls her symptoms, not for avoidance.
  5. Can we have the dietitian referral and a target weight this week? Frame it as the immune and surgical-recovery item it is, not as a lifestyle question. Ask for her weight to be recorded at the clinic and for the number to be given to her partner.
  6. Is there any research immune-profiling or tissue-banking study her sample could contribute to? Chordoma immune profiling is active (Arrieta 2026 and others) and the national MDT or the Chordoma Foundation route is the natural way to ask. This costs her nothing and is worth one sentence.
  7. Already on the sheet and not to be dropped: the perioperative pain plan and the written opioid taper agreed before the operation, not after (thread 24 and the 14 September log). Nothing in this file changes that priority; if anything, the stress-hormone data strengthens the case for good analgesia rather than weakening it.

Sources

Verified through NCBI E-utilities on 15 September 2026; abstract opened for each. Level in brackets.

Metabolism and immune visibility, against (tumour becomes less visible)

Metabolism and immune visibility, for (more visible, or better T cells)

Immunoediting framework

Chordoma immune landscape

Checkpoint blockade in chordoma and chondrosarcoma

the patient's own situation

Deliberately not quoted here: survival and prognosis figures from the chordoma and sarcoma series above, per the standing rule; the per-component HLA class I percentages from Patel 2021 and the tumour mutational burden figure from Hsia 2025, because neither appears in the abstract and the full texts were not read.

Research thread 26 · source file em-research/26-metabolism-immune-visibility-and-chordoma-immune-escape.md · last edited 07 October 2026