The Warburg effect

In brief

What the Warburg effect is

A cell can get usable energy from sugar in two ways. It can split glucose in the cytosol to pyruvate and then send that pyruvate into the mitochondrion, where the tricarboxylic acid cycle — the TCA cycle, also called the Krebs cycle — strips the carbons off and feeds electrons to the electron transport chain, the chain of protein complexes that passes them to oxygen and in doing so makes most of the cell's ATP, the molecule cells spend as energy. That is oxidative phosphorylation. Or the cell can stop halfway: reduce pyruvate to lactate and push the lactate out. Fermentation of that kind is what working muscle does when it breaks down glucose faster than its mitochondria can take up pyruvate, as in a sprint, and it is what any cell does when oxygen runs out ^1,2,39.

The Warburg effect is the observation that tumors do a great deal of the second thing while the oxygen is still on. The modern and more exact name for the behavior is aerobic glycolysis, meaning glucose used beyond what feeds oxidative phosphorylation despite oxygen being available ⁴⁰. It is not a theory. It is an arithmetic result from the blood going into and out of a tumor, and it reproduces ^1,3.

What has changed is everything around that result. Warburg believed fermentation in tumors was evidence that their respiration was broken, and that broken respiration was the origin of cancer ⁴. That claim was contested in print within six months ⁵, and the field now states flatly that the increase in aerobic glycolysis "is often erroneously thought to occur instead of mitochondrial respiration and has been misinterpreted as evidence for damage to respiration instead of damage to the regulation of glycolysis," and that many cancers show the effect while keeping mitochondrial respiration ¹. A review from inside the field put the position with unusual candor: the phenomenon is seen "even in the presence of completely functioning mitochondria," has been documented for more than ninety years, has attracted thousands of papers claiming to establish its cause or function, and its function "remains unclear" ⁷.

So the honest sentence is this. The measurement is right, the mechanism was misread, the reason the behavior exists is still open, and the word "disproven" does not fit. In 2023, a hundred years after Warburg's first tumor-slice measurements, the field marked the centenary with a multi-author assessment in Nature Metabolism rather than an obituary ⁴¹, and in 2026 a review of gastric cancer still uses the term, treating it there as a multi-module metabolic state rather than a single glycolytic phenotype ²⁸.

The whole history, including the reversals

By 1926 Warburg's question had become blunt and practical: can tumor cells in a living animal be killed by taking their energy away? The 1927 paper in the Journal of General Physiology, with Franz Wind and Erwin Negelein, answers it carefully and mostly in the negative. Pieces of tumor survived a day without oxygen if glucose was present, and survived without glucose if oxygen was present; only removing both for about four hours killed most cells. In living tumor-bearing rats, insulin convulsions that drove blood sugar very low left tumor respiration and fermentation nearly normal afterwards — his own conclusion was that "even if it were possible to remove the blood-sugar entirely in living animals, the life of the tumor would not be threatened." Breathing 5% oxygen for 40 hours did kill most of the tumor, more than he expected, and he explained the loss of the better-supplied part by oxygen want killing the capillary cells as well as the tumor cells, so that the capillaries stopped carrying blood ³.

That paper also contains the in-body version of the measurement the effect is named for, and a detail almost always left out: the tumor was supplied with glucose three times worse than the worst normal tissue he measured, and the blood leaving it had given up most of its oxygen ³. The popular picture of a greedy tumor swimming in sugar is the opposite of what he recorded.

In 1931 Warburg received the Nobel Prize for the discovery of the nature and mode of action of the respiratory enzyme; he was a controversial figure, intolerant of opposing scientific views yet, in his biographer's words, tolerant toward Nazi abuses, and although accused of collaboration under the Nazi regime he was readmitted to the global scientific community after the Second World War ⁵³. Twenty-five years later, in Science, he turned the measurement into a theory of cancer's origin: respiration irreversibly damaged, fermentation taking over, the cell dedifferentiating ⁴. Sidney Weinhouse answered in the same volume ⁵; Warburg replied in the same issue ⁶. What the record does establish is that the explanation was contested from the start, by a leading biochemist, in the same journal.

The correction then came in pieces across sixty years, and some of the pieces came from the laboratories that had proposed the ideas being corrected ^49,54. The lesson the season draws is not that Warburg was foolish. It is that a solid measurement can carry a wrong explanation for a very long time if the explanation is memorable and the measurement is easy to repeat.

How it works, and who gives the order

The modern account reverses the direction of causation. Fermentation in tumors is not the residue of a damaged engine; it is installed by the same signals that drive growth. MYC, a transcription factor overexpressed in many cancers, directly raises lactate dehydrogenase A, and reducing that enzyme cuts the ability of MYC-transformed cells and Burkitt lymphoma cells to form colonies ⁴². MYC also runs a transcriptional program for glutamine catabolism that makes cells dependent on glutamine to stay alive, an addiction that does not require PI3K or AKT activation and that reduces how much glucose carbon enters the TCA cycle ⁴³. The glutamine story began in 2007, when glioblastoma cells were shown to consume glutamine far faster than building proteins and nucleotides required, using it to refill the TCA cycle as intermediates were drawn off to make fat, and to generate NADPH ⁶⁰. It has since been cut down twice: lung tumors growing in mice used little glutamine at all ¹⁴, and the glutaminase inhibitor telaglenastat added nothing in a 444-patient kidney cancer trial ²¹.

The PI3K-AKT arm acts on the supply side. Constitutively active AKT raised glucose consumption in transformed cells without changing the rate of oxidative phosphorylation, produced high aerobic glycolysis in human glioblastoma cells that carried active AKT, and left those cells more likely to die when glucose was withdrawn ⁴⁴. HIF-1 (hypoxia-inducible factor 1), the transcription factor stabilized when oxygen is low, closes the mitochondrial door from the other side: it directly activates pyruvate dehydrogenase kinase 1, which inactivates pyruvate dehydrogenase, the enzyme that converts pyruvate to acetyl-CoA, so glucose carbon is shunted away from the TCA cycle; in hypoxic cells lacking HIF-1α, forcing that kinase back in raised ATP, cut reactive oxygen species and rescued the cells from hypoxia-induced death ⁴⁵.

Mutant KRAS, the signature lesion of pancreatic ductal adenocarcinoma, maintains the program in established tumors: switching the mutant gene off in mice showed that advanced tumors remained strictly dependent on it, and that it controlled glucose uptake and channeled glucose intermediates into the hexosamine and non-oxidative pentose phosphate pathways, decoupling ribose synthesis from redox control ⁴⁶. Which pathway a tumor leans on, however, is not fixed by the mutation alone. Infusing labeled nutrients into mice with Kras-driven lung tumors showed increased lactate production from glucose as expected, but minimal glutamine use by either tumor or normal lung, and more glucose carbon entering the TCA cycle in tumor than in normal lung — the opposite of the culture result from the same tumor type, and a demonstration that the in vivo environment, not the genotype alone, sets the phenotype ¹⁴.

The best general statement available is deliberately modest: the metabolic adaptations of cancer cells are determined jointly by the physiology of the cell of origin, the identity of the transforming lesions, and the tissue the cells sit in, and some metabolites act as signals as well as substrates ¹³. In the hallmarks framework, reprogramming of energy metabolism entered in 2011 as an 'emerging' hallmark ⁶¹; it stayed in the framework through Hanahan's 2022 update, which added further dimensions to it ⁶², and the framework was restated in 2026 ⁶³. Being a hallmark means tumors reliably show it, not that blocking it is a treatment.

Lactate: waste, then fuel, then signal

The single largest change since Warburg is what became of the lactate. George Brooks argued from the mid-1980s that lactate is produced continuously under fully aerobic conditions and shuttled between producing and consuming cells, with 25% to 50% of the carbohydrate burned in a resting, postabsorptive person passing through the lactate pool ⁴⁸. Thirty years later he set out the mature version: lactate is a major energy source, the major precursor for making new glucose, and a signaling molecule, and in medicine a high blood lactate is better read as a marker of strain than of oxygen debt ³⁹.

Inside tumors, the shuttle has a geography. Hypoxic cells ferment glucose and release lactic acid, creating a lactate gradient that mirrors the oxygen gradient, and oxygenated cells import that lactate through monocarboxylate transporter 1 and burn it. Blocking that transporter in a human cervical carcinoma line switched the cells from lactate-fueled respiration to glycolysis; doing the same in a mouse lung carcinoma and in xenografted human colorectal cells retarded tumor growth, because the hypoxic cells then starved of glucose, and the survivors became more sensitive to radiation ⁴⁷. The first MCT1 inhibitor tested in people, AZD3965, reached target concentrations in a 40-patient phase I trial; its main dose-limiting toxicity was reversible change in retinal function, because the eye also depends on the transporter, and no efficacy has been reported ⁶⁵.

Then the quantitative work landed. In mice, systematic infusions of labeled nutrients showed that lactate has the highest circulatory turnover flux of any metabolite measured, exceeding glucose by 1.1-fold in the fed state and 2.5-fold while fasting, and that in the fasted state glucose reaches tissue TCA cycles mainly indirectly, through circulating lactate, in every tissue except the brain. In genetically engineered lung and pancreatic tumors in fasted mice, circulating lactate contributed more TCA carbon than glucose did, with glutamine larger than lactate in pancreatic cancer ¹². In people, carbon-13 lactate infusion in non-small cell lung cancer labeled tumor TCA metabolites extensively; in mouse xenografts of the same cancer, lactate's contribution predominated over glucose's, and deleting the transporter from mouse tumor cells abolished the labeling, confirming the uptake was by the tumor cells themselves ¹¹. The most Warburg-looking tumors on a PET scan were the ones most clearly eating lactate ¹¹.

Lactate also talks. Lactate-derived lactylation of lysine residues on histones — the proteins DNA is wound around — is a chemical tag that directly stimulates transcription, and 28 such sites were identified on core histones in human and mouse cells; hypoxia and bacterial challenge generate the lactate that drives it, and in macrophages exposed to bacteria the tag appears late and switches on wound-healing genes ³⁷. The mechanism has already been corrected once from inside: because the proposed donor molecule, lactyl-coenzyme A, is about a thousand times less concentrated than acetyl-CoA, the acetyltransferase p300 is a doubtful lactyltransferase, and a 2024 paper identified the alanyl-tRNA synthetase AARS1 as a genuine one that uses lactate and ATP directly, lactylating and activating the YAP-TEAD complex in gastric cancer ⁶⁶.

And lactate defends the tumor. Regulatory T cells — immune cells whose job is to suppress other immune cells — turned out to do badly on glucose and well on lactic acid: high glucose impaired their function and stability, while lactate prevented that destabilization and supplied intermediates for proliferation. Deleting the lactate transporter from these cells made no difference to their function elsewhere in the body but was required inside tumors, and doing so slowed tumor growth and improved the response to immunotherapy in mice ³⁸. A 2026 review of gastric cancer gathers the same theme: lactate acidifies and remodels the microenvironment, impairs CD8-positive T cells and natural killer cells, promotes regulatory T cell accumulation and M2 macrophage polarization, and contributes to resistance to chemotherapy, targeted therapy and immunotherapy ²⁸. The most direct evidence is a 2024 study in which lactate-driven lactylation of the DNA-repair protein NBS1 promoted repair of chemotherapy damage. Lowering lactate, genetically or with stiripentol, an epilepsy drug that inhibits lactate dehydrogenase A, restored chemotherapy sensitivity in laboratory models, and high levels of the mark predicted poor response to chemotherapy given before surgery ⁶⁷. None of this has been tested as a treatment in a trial.

The reverse Warburg effect, and why one tumor is not one metabolism

In 2009 a group proposed the mirror image of Warburg's picture: that epithelial cancer cells induce aerobic glycolysis in neighboring stromal fibroblasts, which then secrete lactate and pyruvate for the cancer cells to oxidize — the authors themselves called it "a type of host-parasite relationship" ⁶⁸. The supporting data were a caveolin-1-deficient fibroblast model showing coordinated upregulation of myofibroblast markers and glycolytic enzymes in normal oxygen, plus staining of human breast cancers lacking stromal caveolin-1, whose loss is associated with recurrence, metastasis and poor outcome ⁶⁸. That is a hypothesis with a biomarker association attached, not a measured transfer of carbon in patients.

Seventeen years later it is still an active framework rather than a discarded one. A 2026 review describes cancer-associated fibroblasts in this state as markedly glycolytic, generating large quantities of lactate that adjacent tumor cells internalize and use, with monocarboxylate transporters as the central nodes, and extends the idea to immune suppression and epigenetic change ⁶⁹. It remains a review of models. No human study has yet measured stromal-origin lactate carbon arriving in cancer-cell TCA intermediates, and the closest functional demonstration — intratumoral lactate shuttling with MCT1 as the importer — was between tumor cell compartments in mice, not from stroma in patients ⁴⁷.

What has become unarguable is heterogeneity. Within single human lung tumors, metabolically different regions sit side by side, and non-glucose nutrients appear to contribute in the better-perfused areas ⁹. Within one organ, two kidney cancers run opposite ways: clear cell renal cell carcinoma showed the clearest glycolytic shift of any human tumor traced this way ¹⁰, while translocation renal cell carcinoma, driven by TFE3 gene fusions, is transcriptionally rewired toward oxidative phosphorylation, "contrasting with the highly glycolytic nature of most other renal cancers," and is vulnerable to NADH reductive stress and to inhibition of EGLN1 ⁷⁰. A single generalization cannot survive that contrast.

Warburg was also not wholly wrong about mitochondrial damage, only wrong about its direction. Repurposed whole-exome sequencing found pathogenic mitochondrial DNA mutations arising in tumors at a rate comparable to the most common cancer driver genes, with loss-of-function mutations accumulating at an elevated rate specifically in complex I, often at homopolymeric hotspots. But complex V, the ATP synthase, is depleted of all non-synonymous mutations, which the authors read as selection against losing ATP synthesis and membrane potential. And pathogenic mitochondrial DNA mutations were associated with substantially better overall survival in colorectal cancer — an association, not a causal claim ⁷¹. Meanwhile tumor cells deprived of mitochondrial DNA entirely grew slowly until they acquired mitochondrial DNA from host cells, recovering respiration in step with recovering aggressiveness ⁵⁷. The best human case is clear cell kidney cancer: mitochondria isolated from these tumors had lower electron transport chain activity than those from adjacent kidney, one cancer in which respiration really is turned down. Yet the metastases regained TCA cycle activity, and in mice stimulating respiration promoted metastasis while blocking complex I reduced it ¹⁶. Respiration can be turned down for growth at the original site and turned back up to spread.

Where metabolism really did cause cancer: the oncometabolites

There is a genuine causal chain from a metabolic enzyme to a tumor, and it is not the one Warburg drew. In 2000, germline mutations in SDHD, which encodes a small subunit of mitochondrial complex II, were found in families with hereditary paraganglioma, findings the authors read as showing that mitochondria play an important role in the pathogenesis of certain tumors ⁷². Two years later, germline mutations in FH, the TCA cycle enzyme fumarate hydratase, were shown to predispose to uterine fibroids, skin leiomyomas and papillary renal cell cancer, with enzyme activity very low or absent in the tumors ⁷³. In 2005 the mechanism was traced: succinate accumulating behind an inhibited complex II leaves the mitochondrion and inhibits the prolyl hydroxylases that mark HIF-1α for destruction, so HIF-1α is stabilized and the hypoxia program runs in normal oxygen ⁷⁴.

The clearest case came from sequencing. Genome-wide sequencing of 22 glioblastomas, followed by screening of a larger series, found recurrent mutations in the active site of isocitrate dehydrogenase 1 in 12% of patients, concentrated in younger patients and in secondary glioblastomas, and associated with longer overall survival ⁷⁵. In 2009 the consequence was identified: the mutant enzyme gains a new activity, reducing alpha-ketoglutarate to R(-)-2-hydroxyglutarate, which accumulates markedly in mutant human gliomas ⁷⁶. In 2011 the damage was explained — 2-hydroxyglutarate is a competitive inhibitor of multiple alpha-ketoglutarate-dependent dioxygenases, including histone demethylases and the TET family of enzymes that hydroxylate methylated cytosine, so the whole epigenetic landscape shifts ⁷⁷.

That chain produced drugs that work. In the INDIGO trial, 331 patients with residual or recurrent grade 2 IDH-mutant glioma who had had only surgery were randomized to vorasidenib or placebo: median progression-free survival 27.7 months against 11.1 months, hazard ratio 0.39 (95% CI 0.27 to 0.56, P < 0.001), and time to the next intervention also improved, hazard ratio 0.26 (95% CI 0.15 to 0.43). Grade 3 or higher adverse events occurred in 22.8% against 13.5%, with grade 3 or higher alanine aminotransferase elevation in 9.6% against none. The trial was funded by Servier ²³. In ClarIDHy, 185 patients with previously treated IDH1-mutant cholangiocarcinoma — cancer of the bile ducts — were randomized 2:1 to ivosidenib or placebo: median progression-free survival 2.7 months (95% CI 1.6 to 4.2) against 1.4 months (1.4 to 1.6), hazard ratio 0.37 (95% CI 0.25 to 0.54). The study was funded by Agios Pharmaceuticals, the company that had identified the oncometabolite ^76,78. Overall survival, the endpoint that matters most to patients, was 10.3 against 7.5 months, a difference that was not statistically significant (hazard ratio 0.79, 95% CI 0.56 to 1.12) because 43 of the 61 placebo patients crossed over to ivosidenib when their cancer grew; adjusted statistically for that crossover, the hazard ratio was 0.49 ⁷⁹. In acute myeloid leukemia, the AGILE trial randomized 146 newly diagnosed patients with IDH1 mutations who could not have intensive chemotherapy to ivosidenib or placebo, each with azacitidine: median overall survival was 24.0 against 7.9 months (hazard ratio 0.44, 95% CI 0.27 to 0.73), in a trial funded by Agios and Servier ²⁵. Vorasidenib was approved by the U.S. Food and Drug Administration in 2024 for grade 2 IDH-mutant glioma after surgery ²⁴.

The bile duct is also where this page meets the season's subject. An international analysis of 489 cholangiocarcinomas from ten countries found that the molecular landscape differs radically by cause: fluke-positive tumors clustered separately from fluke-negative ones, and IDH1 and IDH2 mutations with epigenetic rewiring fell on the fluke-negative side ⁸⁰. The same organ, the same histology, two different routes in — which is the whole argument of the season in one dataset, and a caution against assuming that a metabolic marker found in one population will be present in another.

How it is found and measured, including FDG-PET

The Warburg effect is the only piece of cancer metabolism that most patients meet directly, because it is the basis of a routine scan. Fluorodeoxyglucose is glucose with a fluorine-18 atom in place of one hydroxyl group; cells take it up and hexokinase phosphorylates it, after which it cannot continue down glycolysis and is stuck inside, a design principle described by Gallagher and colleagues in 1978 as metabolic trapping ²⁹. Animal tumor imaging followed in 1980: uptake peaked by 30 minutes and held for an hour, tumor-to-normal-tissue ratios ran from 2.1 to 9.2 depending on tumor type, and a scan of a dog's seminoma showed high uptake in the viable part and none in the necrotic mass, with no acute or chronic toxicity at a thousand times the human tracer dose in mice ⁸². The scan itself, its history and its false positives in parasitic disease have their own Atlas page, FDG-PET.

What the signal actually reports is important and often forgotten. In rat tumor models, FDG uptake correlated with expression of the glucose transporter GLUT1 (r = 0.83) and hexokinase (r = 0.77) but not with proliferation ⁸³; the correlation was with the transcription of glycolysis-associated genes, not with how fast cells were dividing ⁸³. Most consequentially, microautoradiography of mouse tumors showed that newly formed granulation tissue and macrophages infiltrating the margins took up more tracer than the viable tumor cells, and that up to 29% of the tumor's glucose use came from non-tumor tissue ⁸⁴. A hot spot on a PET scan is a glucose-avid region, not a diagnosis.

That is why the test's specificity depends on where the patient lives. Pooling 40 studies and 1,474 pulmonary lesions, the maximum joint sensitivity and specificity of FDG-PET was 91.2% (95% CI 89.1 to 92.9), with practice operating near 96.8% sensitivity and 77.8% specificity ⁸⁵. A later meta-analysis of 70 studies and 8,511 nodules, 5,105 of them malignant, gave pooled sensitivity 89% (95% CI 86 to 91) and specificity 75% (95% CI 71 to 79) — but average specificity was 16 points lower in regions where infectious lung disease is endemic, 61% (95% CI 49 to 72) against 77% (95% CI 73 to 80), with substantial heterogeneity (I² = 87% and 82%). The authors concluded the data do not support using FDG-PET to diagnose lung cancer in endemic regions unless a center can match non-endemic performance ¹⁷. For a season about parasites, that is the sentence to keep: the instrument that counts cancers counts inflammation too, and it does so worst where infectious lung disease is endemic ¹⁷; the meta-analysis does not break that down by parasitic disease.

Two research methods go deeper. Infusing carbon-13-labeled nutrients during surgery and tracing the heavy carbon shows what a tumor is burning in a living person; this is the method behind the human lung, brain and kidney results above, and its limitation is that the series are small, the largest being more than 80 kidney cancer patients ^9,10,11,16. Hyperpolarized carbon-13 pyruvate magnetic resonance imaging raises the signal of injected pyruvate more than ten-thousand-fold and images its conversion to lactate within seconds; the first-in-human study in 31 men with biopsy-proven prostate cancer found no dose-limiting toxicities and elevated lactate-to-pyruvate ratios in regions where biopsy had found cancer ⁸⁶. Neither has an outcome trial behind it.

• FDG-PET (fluorodeoxyglucose positron emission tomography)A glucose molecule with a fluorine-18 atom in place of one hydroxyl group is injected; cells take it up and phosphorylate it, after which it cannot continue down glycolysis and accumulates, a principle Gallagher and colleagues described as metabolic trapping in 1978 ²⁹; rapid tumor uptake in animals was reported in 1980 ⁸². Signal tracks the glucose transporter GLUT1 and hexokinase rather than how fast cells are dividing ⁸³.The test actually ordered, across staging, restaging and treatment response in most solid cancersIn 70 studies and 8,511 pulmonary nodules, pooled sensitivity 89% (95% confidence interval 86-91) and specificity 75% (95% CI 71-79) ¹⁷; an earlier meta-analysis of 40 studies and 1,474 lesions put maximum joint sensitivity and specificity at 91.2% (95% CI 89.1-92.9) ⁸⁵.It is not a tumor test but a glucose test. In regions where infectious lung disease is endemic, average specificity fell to 61% (95% CI 49-72) against 77% elsewhere, a 16-point gap ¹⁷. In mouse tumors, granulation tissue and infiltrating macrophages took up more tracer than the viable tumor cells, and up to 29% of a tumor's glucose use came from non-tumor cells ⁸⁴, the same mechanism by which infectious and inflammatory lesions can light up.
• Intraoperative carbon-13 nutrient infusionCarbon-13-labeled glucose, lactate or glutamine is infused into a patient during surgery and the labeled carbon is then traced through the tumor's metabolites, which shows what the tumor was actually burning inside the living body ^9,11.A research method, in small surgical series at a few centers, not a clinical testDirect and specific: it measures flux rather than uptake. In nine patients with non-small cell lung cancer, enhanced glycolysis and glucose oxidation were common, and every tumor showed evidence of multiple nutrients being oxidized ⁹.Most series are small and the tumor types few; the largest, more than 80 kidney cancer patients in 2024, found labeling differed by subtype and that metastases burned more through the TCA cycle than the primary tumors ¹⁶, so results do not generalize across cancers or even across stages of one cancer ^10,16.
• Hyperpolarized carbon-13 pyruvate magnetic resonance imagingPyruvate is magnetically hyperpolarized to raise its signal more than ten-thousand-fold, injected, and its conversion to lactate is imaged within seconds; in 31 men with biopsy-proven prostate cancer the ratio of labeled lactate to pyruvate was elevated in cancerous regions ⁸⁶.Investigational imaging, in specialist centersFirst-in-human study: no dose-limiting toxicities, and elevated lactate-to-pyruvate where biopsy had found cancer ⁸⁶.A single early study in one tumor type; it measures the conversion step, not survival benefit, and no outcome trial supports acting on it.
• Serum lactate dehydrogenase and lactateBlood markers downstream of fermentation. A 2026 review of gastric cancer grades circulating lactate dehydrogenase and lactate-related markers as candidate prognostic and response measures, not established tests ²⁸.Lactate dehydrogenase is used as a prognostic marker in some cancers; blood lactate is a general illness severity markerNon-specific. Raised blood lactate tracks illness or injury severity in general, and lactate shuttle work notes that some clinicians now read lactatemia as a "strain" rather than a "stress" biomarker ³⁹.It cannot distinguish tumor fermentation from any other source, and raised blood lactate, alongside neurotoxicity, was among the dose-limiting toxicities of a drug that blocks oxidative phosphorylation in trial ¹⁸.
• IDH1 and IDH2 mutation testingSequencing for the single-residue mutations that make isocitrate dehydrogenase produce the oncometabolite 2-hydroxyglutarate, found in 12% of glioblastomas in the study that discovered them ^75,76.Routine in glioma diagnosis, in advanced cholangiocarcinoma, and in acute myeloid leukemia; in each it selects patients for IDH inhibitors shown to work in randomized trials ^23,25,78A genetic test, so it is as reliable as the sequencing; it is the one metabolic marker in this family with randomized trials behind the treatment decision ^23,78.It says nothing about how glycolytic the tumor is, and an IDH-wild-type tumor is not thereby a Warburg tumor.

What is done about it, and what is sold

Start with the drugs, because they were tried properly and they failed properly. IACS-010759, a potent and selective inhibitor of mitochondrial complex I, went into two dose-escalation phase I trials, in relapsed or refractory acute myeloid leukemia (17 patients) and advanced solid tumors (23 patients). It had a narrow therapeutic index, with dose-limiting toxicities including raised blood lactate and neurotoxicity that made target exposure impossible to hold; no recommended phase 2 dose was established, target inhibition was modest, antitumor activity limited, and both trials were discontinued, with the authors concluding that caution is warranted in developing complex I inhibitors as antitumor agents. The compound was discovered and trialed by units of the same cancer center ¹⁸.

The other two were larger and cleaner. Metformin, the first-line diabetes drug, which among other actions inhibits complex I, was tested in MA.32, a phase 3 randomized placebo-controlled trial of 3,649 patients with high-risk non-metastatic breast cancer and no diabetes: in the hormone-receptor-positive primary analysis population of 2,533, invasive disease-free survival events ran at 2.78 against 2.74 per 100 patient-years, hazard ratio 1.01 (95% CI 0.84 to 1.21, P = .93), and the hormone-receptor-negative group was declared futile at a second interim analysis ¹⁹. A second phase 3 test, in 1,874 men starting hormone therapy for metastatic prostate cancer within the STAMPEDE platform, found no significant overall survival benefit either (hazard ratio 0.91, 95% CI 0.80 to 1.03, P = .15), although metformin did reduce the metabolic side effects of hormone therapy ²⁰. Telaglenastat, a glutaminase inhibitor aimed at the glutamine addiction described above, was tested in CANTATA, 444 patients with metastatic clear cell renal cell carcinoma, the first tumor traced by intraoperative infusion to show a convincing glycolytic shift ¹⁰, added to cabozantinib: median progression-free survival 9.2 against 9.3 months, hazard ratio 0.94 (95% CI 0.74 to 1.21, P = .65) ²¹. Two of the authors were employees of the sponsor ²¹. The largest trial of a drug aimed squarely at tumor mitochondrial fuel use is also negative. In AVENGER 500, 528 patients with untreated metastatic pancreatic cancer were randomized to devimistat plus modified FOLFIRINOX or to standard FOLFIRINOX: median overall survival 11.1 against 11.7 months (hazard ratio 0.95, 95% CI 0.77 to 1.18) ²².

The glycolysis inhibitors sold outside oncology have thinner records still. 2-Deoxyglucose, a glucose analog that jams the pathway, reached phase I twice. With docetaxel in 34 patients with advanced solid tumors, the tolerable dose was 63 mg/kg per day; adverse effects at 63 to 88 mg/kg included reversible hyperglycemia in all patients, gastrointestinal bleeding in 6% and reversible grade 3 QTc prolongation in 22%; best responses were stable disease in 11 patients, one partial response and progression in 22 ³³. Alone, in castration-resistant prostate cancer and other advanced malignancies, 45 mg/kg was the recommended phase 2 dose, limited by asymptomatic grade 3 QTc prolongation at 60 mg/kg; five of eight patients scanned showed reduced FDG uptake by day 2, which is competition with the tracer rather than a treatment effect ⁸⁷. Neither trial went further.

Dichloroacetate was the subject of the most widely shared claim of all. The 2010 report described mitochondrial hyperpolarization in freshly isolated glioblastomas from 49 patients, reversed by the drug, plus five patients treated with oral dichloroacetate for up to 15 months with dose-dependent reversible peripheral neuropathy as the dose-limiting toxicity and "indications of clinical efficacy" ⁸⁸. Five patients is not a trial. The randomized test came later and in a different disease: 45 patients with locally advanced head and neck squamous cell carcinoma were randomized, double-blind, to dichloroacetate or placebo added to cisplatin chemoradiotherapy. The drug was safe, serum pyruvate and lactate fell as predicted, complete response at three months was higher (71.4% against 37.5%, P = .0362), and five-year progression-free and overall survival showed no significant difference between groups ³⁴. Fevers and low platelet counts were more common on the drug. Dichloroacetate's genuine medical niche is not cancer but inherited disorders of pyruvate metabolism. Even there, a 2026 phase III crossover trial in 34 children with pyruvate dehydrogenase complex deficiency, dosed according to each child's genotype for GSTZ1, the enzyme that clears the drug, missed its primary motor endpoint, although blood lactate fell by about 20% ⁸⁹.

3-Bromopyruvate is where the gap between claim and evidence is widest. It is a reactive small molecule that inhibits hexokinase and depletes ATP, and in animals it was reported to eradicate advanced tumors in all 19 of 19 treated rodents, a claim made in an editorial by the laboratory that developed it ^90,91. The human record in the indexed literature amounts to a case report: one young adult with fibrolamellar hepatocellular carcinoma treated in Germany on a translational basis, published with a fatal outcome recorded in the indexing ⁹². The 2019 review (online December 2018) restating the animal results was written by authors including the holder of a company founded on the compound, and it describes "a limited number of cancer patients" treated, with no controlled data ⁹¹. Anyone selling this as a therapy is selling animal data and one case.

Which leaves the diets. The honest statement for October 2026 is that ketogenic diets in glioma are safe and feasible and have not been shown to extend survival. A 2025 systematic review of 18 studies found that the two randomized trials, rated high-quality evidence, showed no significant survival benefit, with observational studies inconsistent ⁹³. A 2026 systematic review of 43 reports concluded that clinical evidence primarily supports feasibility and metabolic activity rather than independent efficacy, because the studies are small, mostly uncontrolled, and confounded by surgery, radiotherapy, chemotherapy, bevacizumab and corticosteroids ³⁵.

The first randomized test in brain tumors is worth describing exactly. ERGO2 randomized 50 patients with recurrent brain tumors to re-irradiation with or without a calorie-restricted ketogenic diet and intermittent fasting, and did not meet its primary endpoint: progression-free survival at six months was 20% with the diet and 16% without ⁹⁴. An exploratory finding that diet patients whose glucose fell below the median lived longer ⁹⁴ is an after-the-fact subgroup and cannot show that the diet caused anything. The 20 patients who completed the diet hit their calorie and carbohydrate targets, leptin and insulin fell substantially, uric acid rose, and neither quality of life nor cognition was affected; the control group unexpectedly ate less than planned, at 21 rather than 30 kcal/kg per day, which complicates the comparison ⁹⁵. Against that, a 2026 meta-analysis reports a median overall survival of 29.4 months in adherent cohorts against 14.6 months in historical controls ⁹⁶. Historical controls are not a comparison group, and a figure obtained that way cannot carry the weight the headline gives it.

Outside the brain, the largest randomized test so far is small. In a 2026 randomized phase II screening trial in untreated metastatic pancreatic cancer, 32 evaluable patients received gemcitabine, nab-paclitaxel and cisplatin with or without a medically supervised ketogenic diet. Fifteen of the 16 on the diet reached ketosis, but they were in ketosis on a median of only 39.4% of days. Median progression-free survival was 8.5 against 6.2 months (hazard ratio 0.53, 95% CI 0.21 to 1.37) and overall survival 13.7 against 10.2 months (hazard ratio 0.58, 95% CI 0.25 to 1.37). Both intervals include no effect, and the design used a lenient one-sided significance threshold of 0.20, meant only to decide whether a larger trial is worth running ⁹⁷. A 2026 feasibility trial randomized 19 women with endometrial cancer and excess weight to a very low-carbohydrate diet or a standard diet for three to four weeks before surgery; the diet was tolerated, fasting glucose fell 22% and insulin 60%, and LDL cholesterol rose 17.8% ⁹⁸.

The strongest legitimate version of the dietary argument is narrower and more interesting than the one sold. Because PI3K is the enzyme insulin signals through, inhibiting it raises blood glucose, and the insulin released in response switches the pathway back on inside the tumor; in mice, blocking that feedback with a ketogenic diet or with drugs greatly improved the efficacy-to-toxicity ratio of PI3K inhibitors ⁶⁴. That is diet as an adjunct to make a targeted drug work, in several mouse tumor models, not diet as cancer treatment ⁶⁴. Against it stands the whole-organism claim that cancer is fundamentally a metabolic disease of impaired mitochondria ³⁶, a position published in 2010 and still circulated, which the human tracing studies, the ubiquinol requirement and the mitochondrial DNA acquisition experiments all cut against ^9,57,58.

And there is a plain physiological reason "starving cancer of sugar" was never going to work as described, which Warburg himself set out in 1927, with a nuance the slogan drops. He accepted reports that insulin slowed tumor growth and glucose sped it up, and calculated that because tumors are so poorly supplied, a modest fall in blood sugar would cut tumor fermentation to about a quarter. But he separated slowing a tumor from killing it: tumor cells get their oxygen from the circulation and can live on respiration alone. His words: "Even if it were possible to remove the blood-sugar entirely in living animals, the life of the tumor would not be threatened." He then held tumor-bearing rats at very low blood sugar with insulin, to the point of convulsions, for hours, and found tumor respiration and fermentation nearly normal ³. In a person, diet cannot take blood glucose anywhere near zero, because the liver makes new glucose from lactate and other precursors ³⁹; in a 2026 trial a very low-carbohydrate diet lowered fasting glucose by 22% ⁹⁸. And in mice carrying cancers that produce the inflammatory signal IL-6, a ketogenic diet slowed tumor growth but brought on cachexia, the wasting of muscle and fat, sooner and shortened survival ⁹⁹.

What was believed, and is not

Eight beliefs in this area were held confidently and are now either overturned or demoted. They are set out plainly because every one of them is still in circulation, several are still taught, and several are still being sold. None of them is a straw man: each was the mainstream reading of real data at the time, and in at least two cases, the 2009 biomass explanation and the 2008 pyruvate kinase switch, the correction came from the same laboratories that had proposed them ^49,54.

It is worth naming what the pattern has in common. In every case the error was not in the measurement but in the inference drawn from it — from fermentation to broken respiration ^1,4, from high glucose uptake to high total energy production ¹⁵, from lactate in the vein to waste in the vein ^11,12. That is the reason this page spends so much of its length on what each study actually measured, in which species, in how many subjects. A measurement that is right can carry an explanation that is wrong for a hundred years, and the explanation is the part people buy.

What is still unknown

The central question is open. Two laboratories have published competing explanations about five years apart, in Molecular Cell and Nature Metabolism, both in cell culture: that cells ferment when demand for NAD⁺ exceeds demand for ATP, shown by activating pyruvate dehydrogenase and watching proliferation fail as the NAD⁺/NADH ratio fell, with rescue by uncoupling respiration or increasing ATP turnover ⁵⁰; and that the driver is demand for ATP made in a particular place, the cytosol, with hexokinase detached from the outer mitochondrial membrane doing the work: existing CRISPR screens across hundreds of cell lines show both hexokinase isoforms dispensable in conventional media, and the authors showed HK2 becomes required in human plasma-like medium, which they trace to where HK1 sits ⁵¹. No paper adjudicates between them, and the 2016 verdict that the function remains unclear still holds ⁷. An earlier, ecological proposal holds that the acid produced by fermentation gives tumors an invasive advantage ⁵², and a 2025 review concludes that no explanation yet accounts for every condition in which aerobic glycolysis is seen, while holding that it matters for the growth of many cancers and that blocking glucose uptake or fermentation can slow tumors in experiments ²⁷. Two leading investigators have also argued in print for more careful use of the term itself ¹⁰⁰.

Several quantities the field speaks about are not measured. "Most tumors retain functional oxidative phosphorylation" is the field's wording, not a measured proportion: the human intraoperative tracing literature covers a handful of tumor types. Most series are small, nine patients in the first lung glucose study ⁹ and 35 in the follow-up, only five of them infused with labeled lactate ¹¹, but a 2024 study infused carbon-13 nutrients into more than 80 patients with kidney cancer during surgery: labeling varied by subtype, clear cell tumors had suppressed TCA cycle labeling and lower electron transport chain activity than adjacent kidney, and their metastases, unexpectedly, had more TCA cycle labeling than the primary tumors ¹⁶. The fraction of a tumor cell's ATP that comes from glycolysis versus oxidative phosphorylation has no reliable published figure to quote. And the most consequential recent reversal, that primary solid tumors make ATP more slowly than normal tissue, rests on five mouse models, did not hold for lung metastases of breast cancer ¹⁵, and has no direct human replication, though the human kidney data point the same way ¹⁶. The settling study would be absolute flux and computed ATP synthesis measured in human primary tumors against matched normal tissue.

The reverse Warburg effect remains a model. No human study has measured carbon moving from stromal fibroblasts into cancer-cell TCA intermediates; the origin paper is explicitly a hypothesis with a caveolin-1 biomarker association ⁶⁸, the 2026 status piece is a review ⁶⁹, and the functional demonstration of lactate shuttling was between tumor cell compartments in mice rather than from stroma in patients ⁴⁷. The study that would settle it is cell-type-resolved isotope tracing in human tumors, showing stromal-origin lactate carbon arriving in cancer-cell TCA intermediates. Until that exists, the host-parasite metaphor the original authors used for the stroma should be read as a metaphor ⁶⁸.

Finally, the literature itself needs reading with care. A PubMed search for "Warburg effect" on 6 October 2026 returned 5,352 records, 21 of them flagged as retracted publications (about 0.4 percent), and they are not all obscure: the 1981 Science paper titled "Warburg effect revisited" is indexed as a Retracted Publication ⁵⁵, and a 2022 paper claiming that propofol restrains ovarian cancer glycolysis through a circular-RNA axis is also retracted ¹⁰¹. A field with thousands of mechanism papers and a steady trickle of retractions is a field in which claims should be anchored to flux measurements and randomized trials, not to the most recent cell-line story.

The parasite comparison, at its true strength

The analogy is tempting and it is partly real. Many helminths and protozoa are heavily glycolytic, and some run their mitochondria backwards. Adult Schistosoma mansoni cultured for twelve days used glucose and produced lactic acid at the same rate with 5% oxygen as with none — no Pasteur effect at all, meaning oxygen did not slow fermentation; what oxygen changed was egg production, which reached 118 viable eggs per worm pair between days four and six aerobically and virtually none anaerobically ³⁰. Parasitic helminths in general have a large capacity for anaerobic function in some life-cycle stages and an aerobic metabolism in others, shifting as the availability of oxygen and food changes ¹⁰².

The machinery they use for that was long thought to be theirs alone. In anaerobic conditions these animals reduce fumarate to succinate, the reverse of what complex II normally does, using the electron carrier rhodoquinone, which has a lower redox potential than ubiquinone and so can hand electrons to fumarate, which ubiquinone cannot do ¹⁰³. Rhodoquinone was found in every examined eukaryote that reduces fumarate during anoxia — parasitic helminths, but also freshwater snails, mussels, lugworms and oysters — and the rhodoquinone-to-ubiquinone ratio tracked how much the animal relied on fumarate reduction; across the development of the liver fluke Fasciola hepatica, ubiquinone tracked use of the aerobic chain and rhodoquinone tracked fumarate reduction ¹⁰⁴. Two corrections arrived in 2025. A re-examination of the helminth data argued that most worms do this with their ordinary complex II run in reverse rather than a dedicated parasite version, so the model built on the pig roundworm Ascaris should not be generalized ¹⁰⁵. And rhodoquinone itself was found in mitochondria purified from certain mouse and human tissues, carrying electrons to fumarate whatever the oxygen level; it is undetectable in cultured cells, which is why it was missed for so long ⁵⁹.

The pathway in which malate is split two ways to balance the books, part oxidized and part reduced to succinate, is called malate dismutation, and it is a standard feature of anaerobic mitochondria rather than a parasite invention: no enzyme of core anaerobic energy metabolism is specific to any one of the six eukaryotic supergroups, each being found in at least two, which the authors read as presence in the eukaryotic common ancestor followed by differential loss as lineages specialized to oxygen-rich or oxygen-poor niches ³¹. A 2026 review of helminth metabolism treats glycolysis and malate dismutation as the parasite's core adaptation to a hypoxic, nutrient-limited host environment, and as a drug target ³²; the same pathway is nominated as an understudied target in Echinococcus multilocularis, the fox tapeworm, precisely because mammals do not run it ¹⁰⁶.

Protozoa diverge further. Trypanosomatids confine the first enzymes of glycolysis inside glycosomes, peroxisome-like organelles whose ATP and NAD pools cannot freely equilibrate with the cytosol, so they must balance redox inside the organelle — by a glycerol-3-phosphate shuttle to the respiratory chain aerobically, by reducing part of their phosphoenolpyruvate to succinate when oxygen is short, and by a succinate-fumarate-malate cycle exchanging carbon with the mitochondrion; some survive true anaerobiosis by pyruvate dismutation or propionic acid fermentation ¹⁰⁷. The malaria parasite was long thought to be nearly pure glycolysis, and that was corrected: labeling Plasmodium falciparum blood stages with carbon-13 glucose and glutamine showed a conventional oxidative TCA cycle in both asexual and sexual stages, with low glucose flux into the cycle in asexual parasites and a marked increase in gametocytes, where inhibiting the cycle caused arrested development and death while asexual stages barely noticed ¹⁰⁸.

Now the limits of the analogy, stated plainly. A parasite ferments because that is its evolved metabolism, built over its whole lineage for low-oxygen niches ³¹; a tumor cell ferments while keeping a human electron transport chain that it demonstrably needs ⁵⁸. The 2025 finding that some human tissues also carry rhodoquinone ⁵⁹ narrows the biochemical gap but does not change the logic: shared chemistry is not shared cause. The resemblance is convergence on an environment, not a shared mechanism, and it is not evidence that parasites cause cancer through their metabolism, or that cancer is a parasite. Some parasites do cause cancer: the liver flukes Opisthorchis viverrini and Clonorchis sinensis and the blood fluke Schistosoma haematobium are classified by the International Agency for Research on Cancer as Group 1 carcinogens ¹⁰⁹, but the route is chronic inflammation and injury of the bile duct and bladder, not a shared fuel pathway, and fluke-associated bile duct cancers even lack the IDH mutations common in other bile duct cancers ⁸⁰. What the comparison does teach is narrower. The closest true parallel is in parasites that live in oxygen-rich blood: adult schistosomes make lactic acid at the same rate with or without oxygen ³⁰, and asexual malaria parasites inside red cells send little glucose into their TCA cycle ¹⁰⁸. Like tumor cells, they ferment with oxygen present, and like tumor cells, they are not damaged. Aerobic fermentation is a strategy, not a symptom. The one genuinely shared piece of biochemistry is suggestive and small: human cancer cells under the hypoglycemic, hypoxic conditions of a tumor microenvironment can run the same NADH-fumarate reductase reaction, which is why the anthelmintic pyrvinium pamoate, which inhibits it, became an anticancer candidate ^110,111. Mammalian mitochondria can indeed use fumarate as a terminal electron acceptor when oxygen reduction is blocked, reversing the succinate dehydrogenase complex, and mouse tissues differ in how much they can do it ¹¹². Pyrvinium's human record, however, is a phase I dose-escalation protocol in patients awaiting pancreatic surgery, published as a protocol in 2023, with no results in PubMed as of October 2026 ¹¹³. That is a hypothesis with an interesting pedigree, not a treatment.

There is one practical place where parasite metabolism and the Warburg effect really do collide, and it is diagnostic rather than therapeutic. Live adult Schistosoma mansoni took up FDG in proportion to worm number in vitro, and in athymic mice portal FDG signal tracked worm burden (R² = 0.85, P < 0.001, in the 17 animals carrying more than 50 worms) ¹¹⁴. In a patient, host inflammation around a parasitic lesion would be expected to add signal by the same mechanism seen with macrophages and granulation tissue in mouse tumors ⁸⁴; whether that explains the 16-point specificity drop in regions with endemic infectious lung disease has not been measured, and the meta-analysis does not separate parasitic from other infections ¹⁷. If the season's question is whether the cancers attributed to parasites are the real number, part of the answer is that the instrument used to find and stage those cancers is least specific in the places where those parasites live.

Where it connects

Sources

114 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Koppenol WH, Bounds PL, Dang CV. Otto Warburg's contributions to current concepts of cancer metabolism.Nat Rev Cancer · 2011 · 11(5):325-37doi:10.1038/nrc3038 · PMID 21508971

   The citable statement of the correction: approximately tenfold more glucose to lactate, often erroneously thought to occur instead of respiration, misinterpreted as damage to respiration rather than to the regulation of glycolysis. A historical review, not new data.

2. 2
   Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg effect: the metabolic requirements of cell proliferation.Science · 2009 · 324(5930):1029-33doi:10.1126/science.1160809 · PMID 19460998

   The reframing of aerobic glycolysis as a biosynthetic growth program. A review synthesizing others' data; its own generalization that most cancer cells rely on aerobic glycolysis is what the human tracing studies later qualified.

3. 3
   Warburg O, Wind F, Negelein E. THE METABOLISM OF TUMORS IN THE BODY.J Gen Physiol · 1927 · 8(6):519-30doi:10.1085/jgp.8.6.519 · PMID 19872213

   The primary source, read in full: the 70 mg glucose and 46 mg lactic acid figures, the poor glucose and oxygen supply of tumors, and Warburg's own conclusion that removing blood sugar would not threaten the tumor. It does not contain the respiration-damage theory of cancer, which came later.

4. 4
   Warburg O. On the origin of cancer cells.Science · 1956 · 123(3191):309-14doi:10.1126/science.123.3191.309 · PMID 13298683

   The paper in which Warburg argued that cancer originates in irreversibly damaged respiration. The PubMed record carries no abstract and the article is not in PubMed Central, so nothing about the content of his argument can be verified from the database.

5. 5
   Weinhouse S. On respiratory impairment in cancer cells.Science · 1956 · 124(3215):267-9doi:10.1126/science.124.3215.267 · PMID 13351638

   Establishes that the explanation was challenged in print within six months, in the same journal. No abstract, so what Weinhouse argued must be read in the paper.

6. 6
   Warburg O. On respiratory impairment in cancer cells.Science · 1956 · 124(3215):269-70PMID 13351639

   Warburg's reply, on the following pages of the same issue. No abstract and no DOI in the record; whether he conceded anything cannot be established from PubMed.

7. 7
   Liberti MV, Locasale JW. The Warburg Effect: How Does it Benefit Cancer Cells?Trends Biochem Sci · 2016 · 41(3):211-218doi:10.1016/j.tibs.2015.12.001 · PMID 26778478

   The admission from inside the field that after ninety years and thousands of papers the function remains unclear, and that the effect occurs with fully functioning mitochondria. A review.

8. 8
   Zu XL, Guppy M. Cancer metabolism: facts, fantasy, and fiction.Biochem Biophys Res Commun · 2004 · 313(3):459-65doi:10.1016/j.bbrc.2003.11.136 · PMID 14697210

   A narrative review of forty years of data concluding there is no evidence cancer cells are inherently glycolytic, though some tumors may be so in vivo because of hypoxia. No new measurements.

9. 9
   Hensley CT, Faubert B, Yuan Q, et al. Metabolic Heterogeneity in Human Lung Tumors.Cell · 2016 · 164(4):681-94doi:10.1016/j.cell.2015.12.034 · PMID 26853473

   Intraoperative carbon-13 glucose infusion in nine patients with non-small cell lung cancer: enhanced glycolysis and glucose oxidation both common, multiple nutrients oxidized in every tumor, heterogeneity within and between tumors. Human, observational, n = 9.

10. 10
    Courtney KD, Bezwada D, Mashimo T, et al. Isotope Tracing of Human Clear Cell Renal Cell Carcinomas Demonstrates Suppressed Glucose Oxidation In Vivo.Cell Metab · 2018 · 28(5):793-800.e2doi:10.1016/j.cmet.2018.07.020 · PMID 30146487

    The counter-example from the same group: clear cell renal cell carcinoma showed enhanced glycolytic labeling, suppressed pyruvate dehydrogenase flow and reduced TCA labeling, the first human tumor in this series with a convincing glycolytic shift. Small surgical series.

11. 11
    Faubert B, Li KY, Cai L, et al. Lactate Metabolism in Human Lung Tumors.Cell · 2017 · 171(2):358-371.e9doi:10.1016/j.cell.2017.09.019 · PMID 28985563

    Carbon-13 lactate infusion in patients: extensive labeling of tumor TCA metabolites in 5 lactate-infused patients, clearest in high-FDG aggressive tumors; lactate's contribution predominated over glucose's in mouse xenografts; MCT1 deletion in mouse tumors abolished the labeling. Does not show that blocking lactate uptake helps patients.

12. 12
    Hui S, Ghergurovich JM, Morscher RJ, et al. Glucose feeds the TCA cycle via circulating lactate.Nature · 2017 · 551(7678):115-118doi:10.1038/nature24057 · PMID 29045397

    Mouse infusions quantifying circulating nutrient fluxes: lactate turnover highest of all metabolites, exceeding glucose 1.1-fold fed and 2.5-fold fasted, with lactate outcontributing glucose to TCA intermediates in engineered lung and pancreatic tumors. Animal data.

13. 13
    Pavlova NN, Zhu J, Thompson CB. The hallmarks of cancer metabolism: Still emerging.Cell Metab · 2022 · 34(3):355-377doi:10.1016/j.cmet.2022.01.007 · PMID 35123658

    The framework statement that metabolic adaptations are set jointly by cell of origin, transforming lesion and tissue, and that some metabolites act as signals. A review; no primary data.

14. 14
    Davidson SM, Papagiannakopoulos T, Olenchock BA, et al. Environment Impacts the Metabolic Dependencies of Ras-Driven Non-Small Cell Lung Cancer.Cell Metab · 2016 · 23(3):517-28doi:10.1016/j.cmet.2016.01.007 · PMID 26853747

    Infusions in mice showed minimal glutamine use by lung tumors and more glucose carbon entering the TCA cycle than in normal lung, the opposite of the same tumor type in culture, with glucose oxidation required for tumor formation. Direct evidence that environment, not genotype alone, sets the phenotype.

15. 15
    Bartman CR, Weilandt DR, Shen Y, et al. Slow TCA flux and ATP production in primary solid tumours but not metastases.Nature · 2023 · 614(7947):349-357doi:10.1038/s41586-022-05661-6 · PMID 36725930

    Absolute flux measurements in mice: TCA flux suppressed in all five primary solid tumor models, glycolysis up but insufficient, so total ATP synthesis slower than normal tissue; TCA flux higher in breast cancer lung metastases. Mouse models only; no human replication exists.

16. 16
    Bezwada D, Perelli L, Lesner NP, et al. Mitochondrial complex I promotes kidney cancer metastasis.Nature · 2024 · 633(8031):923-931doi:10.1038/s41586-024-07812-3 · PMID 39143213

    Carbon-13 nutrient infusions in more than 80 kidney cancer patients during surgery: labeling varied by subtype, clear cell tumors had suppressed TCA labeling and lower electron transport chain activity than adjacent kidney, yet metastases had more TCA labeling than the primary tumors; in mice, stimulating respiration promoted metastasis and blocking complex I reduced it. Human plus mouse.

17. 17
    Deppen SA, Blume JD, Kensinger CD, et al. Accuracy of FDG-PET to diagnose lung cancer in areas with infectious lung disease: a meta-analysis.JAMA · 2014 · 312(12):1227-36doi:10.1001/jama.2014.11488 · PMID 25247519

    Seventy studies and 8,511 nodules: pooled sensitivity 89% (86-91) and specificity 75% (71-79), with specificity 16 points lower in endemic regions, 61% (49-72) vs 77% (73-80), and high heterogeneity. The authors advise against using FDG-PET to diagnose lung cancer in endemic regions.

18. 18
    Yap TA, Daver N, Mahendra M, et al. Complex I inhibitor of oxidative phosphorylation in advanced solid tumors and acute myeloid leukemia: phase I trials.Nat Med · 2023 · 29(1):115-126doi:10.1038/s41591-022-02103-8 · PMID 36658425

    The clinical reality check: 17 patients with acute myeloid leukemia and 23 with solid tumors, dose-limiting raised blood lactate and neurotoxicity, no recommended phase 2 dose, both trials discontinued, and a stated caution about developing complex I inhibitors. Institutional conflict: the compound was discovered and trialed by units of the same center.

19. 19
    Goodwin PJ, Chen BE, Gelmon KA, et al. Effect of Metformin vs Placebo on Invasive Disease-Free Survival in Patients With Breast Cancer: The MA.32 Randomized Clinical Trial.JAMA · 2022 · 327(20):1963-1973doi:10.1001/jama.2022.6147 · PMID 35608580

    The definitive null result for the best-known complex I drug in oncology: 3,649 patients, hazard ratio 1.01 (95% CI 0.84-1.21) for invasive disease-free survival in the receptor-positive population, futility declared in the receptor-negative group, more grade 3 non-hematological toxicity on metformin.

20. 20
    Gillessen S, Murphy L, James ND, et al. Metformin for patients with metastatic prostate cancer starting androgen deprivation therapy: a randomised phase 3 trial of the STAMPEDE platform protocol.Lancet Oncol · 2025 · 26(8):1018-1030doi:10.1016/S1470-2045(25)00231-1 · PMID 40639383

    Phase 3, 1,874 men with metastatic hormone-sensitive prostate cancer: no significant overall survival benefit (hazard ratio 0.91, 95% CI 0.80-1.03, P = .15), though metformin reduced the metabolic side effects of androgen-deprivation therapy. A second large negative phase 3 for metformin.

21. 21
    Tannir NM, Agarwal N, Porta C, et al. Efficacy and Safety of Telaglenastat Plus Cabozantinib vs Placebo Plus Cabozantinib in Patients With Advanced Renal Cell Carcinoma: The CANTATA Randomized Clinical Trial.JAMA Oncol · 2022 · 8(10):1411-1418doi:10.1001/jamaoncol.2022.3511 · PMID 36048457

    The glutamine-addiction hypothesis tested properly and failed: 444 patients, median progression-free survival 9.2 vs 9.3 months, hazard ratio 0.94 (95% CI 0.74-1.21, P = .65). Two authors were employees of Calithera Biosciences, the sponsor.

22. 22
    Philip PA, Sahai V, Bahary N, et al. Devimistat (CPI-613) With Modified Fluorouarcil, Oxaliplatin, Irinotecan, and Leucovorin (FFX) Versus FFX for Patients With Metastatic Adenocarcinoma of the Pancreas: The Phase III AVENGER 500 Study.J Clin Oncol · 2024 · 42(31):3692-3701doi:10.1200/JCO.23.02659 · PMID 39088774

    Phase 3, 528 patients with untreated metastatic pancreatic cancer: median overall survival 11.1 vs 11.7 months, hazard ratio 0.95 (95% CI 0.77-1.18); progression-free survival 7.8 vs 8.0 months. The largest trial of a drug aimed at tumor mitochondrial fuel entry, and negative. The title typo 'Fluorouarcil' is as indexed.

23. 23
    Mellinghoff IK, van den Bent MJ, Blumenthal DT, et al. Vorasidenib in IDH1- or IDH2-Mutant Low-Grade Glioma.N Engl J Med · 2023 · 389(7):589-601doi:10.1056/NEJMoa2304194 · PMID 37272516

    INDIGO: 331 patients randomized, median progression-free survival 27.7 vs 11.1 months, hazard ratio 0.39 (95% CI 0.27-0.56), grade 3 or higher adverse events 22.8% vs 13.5%. Funded by Servier. Progression-free survival, not overall survival.

24. 24
    Nakhate V, Youssef G, Wen PY. Current and emerging therapies in IDH-mutant glioma.Neurotherapeutics · 2026 · 23(6):e00913doi:10.1016/j.neurot.2026.e00913 · PMID 42085879

    Review; the source for vorasidenib's 2024 US Food and Drug Administration approval for grade 2 IDH-mutant glioma after surgery, on the INDIGO trial [47].

25. 25
    Montesinos P, Recher C, Vives S, et al. Ivosidenib and Azacitidine in IDH1-Mutated Acute Myeloid Leukemia.N Engl J Med · 2022 · 386(16):1519-1531doi:10.1056/NEJMoa2117344 · PMID 35443108

    AGILE, phase 3, 146 newly diagnosed IDH1-mutant AML patients ineligible for intensive chemotherapy: median overall survival 24.0 vs 7.9 months (hazard ratio 0.44, 95% CI 0.27-0.73) with ivosidenib plus azacitidine. Funded by Agios and Servier. A randomized overall-survival win for an oncometabolite-enzyme inhibitor.

26. 26
    Szlosarek PW, Creelan BC, Sarkodie T, et al. Pegargiminase Plus First-Line Chemotherapy in Patients With Nonepithelioid Pleural Mesothelioma: The ATOMIC-Meso Randomized Clinical Trial.JAMA Oncol · 2024 · 10(4):475-483doi:10.1001/jamaoncol.2023.6789 · PMID 38358753

    Randomized, 249 patients: arginine depletion with pegargiminase added to chemotherapy extended median overall survival from 7.7 to 9.3 months (hazard ratio 0.71, 95% CI 0.55-0.93). A randomized success from starving a tumor of a nutrient it cannot make, not from blocking glucose.

27. 27
    Li Z, Munim MB, Sharygin DA, Bevis BJ, Vander Heiden MG. Understanding the Warburg Effect in Cancer.Cold Spring Harb Perspect Med · 2025 · 15(12)doi:10.1101/cshperspect.a041532 · PMID 39284669

    A 2025 review from the laboratory behind the growth-program idea: no proposed explanation fits all the conditions in which aerobic glycolysis is seen, yet the effect matters for the proliferation of many cancers and blocking glucose uptake or fermentation can impair tumor growth. A review.

28. 28
    Liu G, Li P, Li Z, et al. The evolving landscape of the Warburg effect in gastric cancer: From molecular mechanisms to targeted therapy.Clin Transl Med · 2026 · 16(8):e70772doi:10.1002/ctm2.70772 · PMID 42557756

    The clearest dated statement of current usage: a multi-module metabolic state rather than a single glycolytic phenotype, with biomarker-guided combinations rather than isolated glycolytic inhibition named as the translational path. A review; it grades biomarkers by maturity and reports no new data.

29. 29
    Gallagher BM, Fowler JS, Gutterson NI, et al. Metabolic trapping as a principle of oradiopharmaceutical design: some factors resposible for the biodistribution of [18F] 2-deoxy-2-fluoro-D-glucose.J Nucl Med · 1978 · 19(10):1154-61PMID 214528

    The design principle behind every FDG scan: the tracer is phosphorylated and then cannot proceed, so it accumulates. The record has no abstract and no DOI; the title's typographical errors are as indexed.

30. 30
    Schiller EL, Bueding E, Turner VM, Fisher J. Aerobic and anaerobic carbohydrate metabolism and egg production of Schistosoma mansoni in vitro.J Parasitol · 1975 · 61(3):385-89PMID 1169294

    Adult schistosomes used glucose and made lactic acid at the same rate with or without oxygen — no Pasteur effect — while egg output depended on oxygen: 118 viable eggs per worm pair aerobically against virtually none anaerobically. In vitro; no DOI in the record.

31. 31
    Müller M, Mentel M, van Hellemond JJ, et al. Biochemistry and evolution of anaerobic energy metabolism in eukaryotes.Microbiol Mol Biol Rev · 2012 · 76(2):444-95doi:10.1128/MMBR.05024-11 · PMID 22688819

    The comprehensive map: anaerobic energy metabolism, including malate dismutation, is ancestral and widespread across eukaryotes rather than a parasite invention, with compartmentalized metabolic maps for 16 species. A review.

32. 32
    Roy S, Lyndem LM. Decoding the metabolic checkpoints in helminths: Opportunities for novel anthelmintic drug development.Acta Trop · 2026 · 279:108151doi:10.1016/j.actatropica.2026.108151 · PMID 42173387

    The current review of helminth energy metabolism: glycolysis and anaerobic pathways including malate dismutation sustaining ATP across life stages, with host immune pressure shaping it and glycolytic inhibitors proposed as drugs. A review; no primary data. Rhodoquinone itself has since been found in some mouse and human tissues [90], so the selectivity of drugs aimed at this route needs re-testing.

33. 33
    Raez LE, Papadopoulos K, Ricart AD, et al. A phase I dose-escalation trial of 2-deoxy-D-glucose alone or combined with docetaxel in patients with advanced solid tumors.Cancer Chemother Pharmacol · 2013 · 71(2):523-30doi:10.1007/s00280-012-2045-1 · PMID 23228990

    34 patients: 63 mg/kg per day judged clinically tolerable with docetaxel, with reversible hyperglycemia in all patients at 63-88 mg/kg, gastrointestinal bleeding in 6% and grade 3 QTc prolongation in 22%; one partial response. A dose-finding study, not an efficacy trial.

34. 34
    Powell SF, Mazurczak M, Dib EG, et al. Phase II study of dichloroacetate, an inhibitor of pyruvate dehydrogenase, in combination with chemoradiotherapy for unresected, locally advanced head and neck squamous cell carcinoma.Invest New Drugs · 2022 · 40(3):622-633doi:10.1007/s10637-022-01235-5 · PMID 35312941

    The randomized test: 45 patients, safe, with falls in pyruvate and lactate and a higher 3-month complete response rate (71.4% vs 37.5%, P = .0362) but no significant difference in 5-year progression-free or overall survival. Drug-related fevers and low platelets were more common.

35. 35
    Ghoche MT, Miki K, Yuan F, et al. Metabolic modulation of glioblastoma by dietary intervention: a systematic review.J Neurooncol · 2026 · 179(3)doi:10.1007/s11060-026-05779-x · PMID 42758370

    Forty-three reports to August 2026: clinical evidence supports feasibility and metabolic activity rather than independent efficacy, with effects inseparable from surgery, radiotherapy, chemotherapy, bevacizumab and corticosteroids. The current best statement of where ketogenic diets stand.

36. 36
    Seyfried TN, Shelton LM. Cancer as a metabolic disease.Nutr Metab (Lond) · 2010 · 7:7doi:10.1186/1743-7075-7-7 · PMID 20181022

    The position that impaired mitochondrial function and energy metabolism underlie essentially all cancer hallmarks, which much of the metabolic-therapy trade rests on. A hypothesis review with no primary data; the human tracing, ubiquinol and mitochondrial DNA transfer experiments cut against it.

37. 37
    Zhang D, Tang Z, Huang H, et al. Metabolic regulation of gene expression by histone lactylation.Nature · 2019 · 574(7779):575-580doi:10.1038/s41586-019-1678-1 · PMID 31645732

    Identifies 28 lactylation sites on core histones in human and mouse cells and shows the modification stimulates transcription, with hypoxia and bacterial challenge supplying the lactate. In vitro and mouse; it does not show that lactylation drives any human cancer.

38. 38
    Watson MJ, Vignali PDA, Mullett SJ, et al. Metabolic support of tumour-infiltrating regulatory T cells by lactic acid.Nature · 2021 · 591(7851):645-651doi:10.1038/s41586-020-03045-2 · PMID 33589820

    Regulatory T cells do badly on high glucose and well on lactate; deleting MCT1 was dispensable peripherally but required intratumorally, slowing tumor growth and improving immunotherapy response. Mouse plus in vitro.

39. 39
    Brooks GA. The Science and Translation of Lactate Shuttle Theory.Cell Metab · 2018 · 27(4):757-785doi:10.1016/j.cmet.2018.03.008 · PMID 29617642

    The mature statement: lactate as major fuel, major gluconeogenic precursor and signaling molecule, with raised blood lactate better read as a strain than a stress marker. A review.

40. 40
    Vaishnavi SN, Vlassenko AG, Rundle MM, et al. Regional aerobic glycolysis in the human brain.Proc Natl Acad Sci U S A · 2010 · 107(41):17757-62doi:10.1073/pnas.1010459107 · PMID 20837536

    Aerobic glycolysis is a normal feature of the healthy resting human brain, regionally distributed and not strictly tied to total energy metabolism, in 33 neurologically normal young adults. Shows the phenotype is not cancer-specific.

41. 41
    Thompson CB, Vousden KH, Johnson RS, et al. A century of the Warburg effect.Nat Metab · 2023 · 5(11):1840-1843doi:10.1038/s42255-023-00927-3 · PMID 37990075

    Eleven contributors, including Koppenol and Bartman, assessing the concept at its centenary. No abstract in the record, so individual positions cannot be quoted from PubMed; what it shows is that the field convened rather than retired the idea.

42. 42
    Shim H, Dolde C, Lewis BC, et al. c-Myc transactivation of LDH-A: implications for tumor metabolism and growth.Proc Natl Acad Sci U S A · 1997 · 94(13):6658-63doi:10.1073/pnas.94.13.6658 · PMID 9192621

    Links an oncogene directly to lactate production: MYC induces lactate dehydrogenase A, overexpression alone makes cells overproduce lactic acid, and antisense suppression reduces colony formation by transformed and Burkitt lymphoma cells. Cell culture and rodent lines.

43. 43
    Wise DR, DeBerardinis RJ, Mancuso A, et al. Myc regulates a transcriptional program that stimulates mitochondrial glutaminolysis and leads to glutamine addiction.Proc Natl Acad Sci U S A · 2008 · 105(48):18782-7doi:10.1073/pnas.0810199105 · PMID 19033189

    The origin of glutamine addiction as an oncogene-driven phenotype, independent of PI3K and AKT, with reduced glucose carbon entering the TCA cycle as a consequence. Cell culture.

44. 44
    Elstrom RL, Bauer DE, Buzzai M, et al. Akt stimulates aerobic glycolysis in cancer cells.Cancer Res · 2004 · 64(11):3892-9doi:10.1158/0008-5472.CAN-03-2904 · PMID 15172999

    Activated AKT raised glucose consumption without changing oxidative phosphorylation and made cells dependent on glucose for survival, with human glioblastoma lines sorting by AKT activity. Cell and mouse work.

45. 45
    Kim JW, Tchernyshyov I, Semenza GL, Dang CV. HIF-1-mediated expression of pyruvate dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia.Cell Metab · 2006 · 3(3):177-85doi:10.1016/j.cmet.2006.02.002 · PMID 16517405

    Shows HIF-1 actively suppresses the TCA cycle by installing PDK1, and that forcing PDK1 into hypoxic HIF-1α-null cells raises ATP, cuts reactive oxygen species and prevents apoptosis. Cell culture with mouse fibroblasts.

46. 46
    Ying H, Kimmelman AC, Lyssiotis CA, et al. Oncogenic Kras maintains pancreatic tumors through regulation of anabolic glucose metabolism.Cell · 2012 · 149(3):656-70doi:10.1016/j.cell.2012.01.058 · PMID 22541435

    An inducible mouse model showing advanced pancreatic tumors remain strictly dependent on mutant Kras, which drives glucose uptake and routes intermediates into hexosamine and non-oxidative pentose phosphate pathways. Mouse.

47. 47
    Sonveaux P, Végran F, Schroeder T, et al. Targeting lactate-fueled respiration selectively kills hypoxic tumor cells in mice.J Clin Invest · 2008 · 118(12):3930-42doi:10.1172/JCI36843 · PMID 19033663

    The functional demonstration of metabolic symbiosis: MCT1 as the lactate importer, inhibition switching oxygenated cells back to glycolysis, hypoxic cells then dying of glucose starvation, tumor growth retarded and survivors radiosensitized. Mice and cell lines; MCT1 inhibition has not delivered a practice-changing human result.

48. 48
    Brooks GA. Lactate production under fully aerobic conditions: the lactate shuttle during rest and exercise.Fed Proc · 1986 · 45(13):2924-9PMID 3536591

    The original lactate shuttle argument, including the estimate that 25% to 50% of carbohydrate burned at postabsorptive rest passes through the lactate pool. A review of physiology, not cancer data; no DOI in the record.

49. 49
    Hosios AM, Hecht VC, Danai LV, et al. Amino Acids Rather than Glucose Account for the Majority of Cell Mass in Proliferating Mammalian Cells.Dev Cell · 2016 · 36(5):540-9doi:10.1016/j.devcel.2016.02.012 · PMID 26954548

    From Vander Heiden's own laboratory, and a partial refutation of his 2009 explanation: most cell carbon mass comes from amino acids other than glucose and glutamine, despite those being consumed fastest. Cultured cells.

50. 50
    Luengo A, Li Z, Gui DY, et al. Increased demand for NAD⁺ relative to ATP drives aerobic glycolysis.Mol Cell · 2021 · 81(4):691-707.e6doi:10.1016/j.molcel.2020.12.012 · PMID 33382985

    Cell-culture test of one current explanation: activating pyruvate dehydrogenase impaired proliferation by lowering the NAD+/NADH ratio, with rescue by uncoupling respiration or raising ATP turnover. The PubMed title drops the superscript plus sign. In vitro, A549 and HeLa cells.

51. 51
    Huggler KS, Flickinger KM, Forsberg MH, et al. Hexokinase detachment from mitochondria drives the Warburg effect to support compartmentalized ATP production.Nat Metab · 2026 · 8(1):215-236doi:10.1038/s42255-025-01428-1 · PMID 41545565

    The most current primary explanation, and the dated proof the question is open: CRISPR screens across hundreds of lines, both hexokinase isoforms dispensable in conventional media, HK2 required in human plasma-like medium, cytosolic rather than membrane-docked hexokinase supporting growth. In vitro.

52. 52
    Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis?Nat Rev Cancer · 2004 · 4(11):891-9doi:10.1038/nrc1478 · PMID 15516961

    The acid-mediated invasion hypothesis: fermentation is selected because the acid it produces damages neighboring normal tissue and aids invasion, after pre-malignant lesions adapt to intermittent hypoxia. A review and hypothesis.

53. 53
    Weisz GM. Dr. Otto heinrich warburg-survivor of ethical storms.Rambam Maimonides Med J · 2015 · 6(1):e0008doi:10.5041/RMMJ.10183 · PMID 25717390

    The source for the 1931 Nobel Prize being awarded for the nature and mode of action of the cellular respiratory enzyme, and for Warburg's contested personal history. A biographical article, not evidence about tumor metabolism.

54. 54
    Israelsen WJ, Dayton TL, Davidson SM, et al. PKM2 isoform-specific deletion reveals a differential requirement for pyruvate kinase in tumor cells.Cell · 2013 · 155(2):397-409doi:10.1016/j.cell.2013.09.025 · PMID 24120138

    From the same laboratories: deleting PKM2 accelerated Brca1-driven breast tumors in mice, and proliferating tumor cells had no detectable pyruvate kinase, so M2 is not necessary for tumor cell proliferation. The reversal of the 2008 claim [102]. Mouse genetics.

55. 55
    Racker E, Spector M. Warburg effect revisited: merger of biochemistry and molecular biology.Science · 1981 · 213(4505):303-7doi:10.1126/science.6264596 · PMID 6264596

    RETRACTED PUBLICATION per the PubMed record's article types. Cited here only as a documented failure in this literature; none of its claims are used.

56. 56
    Christofk HR, Vander Heiden MG, Harris MH, et al. The M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth.Nature · 2008 · 452(7184):230-3doi:10.1038/nature06734 · PMID 18337823

    The 2008 claim: swapping the fetal M2 form of pyruvate kinase for the adult M1 form reduced lactate, raised oxygen consumption and reduced tumor formation in mice, concluding M2 is necessary for aerobic glycolysis. Human cell lines and mouse xenografts; reversed in 2013 [103].

57. 57
    Tan AS, Baty JW, Dong LF, et al. Mitochondrial genome acquisition restores respiratory function and tumorigenic potential of cancer cells without mitochondrial DNA.Cell Metab · 2015 · 21(1):81-94doi:10.1016/j.cmet.2014.12.003 · PMID 25565207

    Cells without mitochondrial DNA grow slowly until they take mitochondrial DNA from host cells, with respiration and growth recovering in step from primary tumor to circulating cells to lung metastases. Mouse models; the clearest refutation of damaged respiration as the driver.

58. 58
    Martínez-Reyes I, Cardona LR, Kong H, et al. Mitochondrial ubiquinol oxidation is necessary for tumour growth.Nature · 2020 · 585(7824):288-292doi:10.1038/s41586-020-2475-6 · PMID 32641834

    Why tumors need the electron transport chain: complex III loss impairs growth, alternative oxidase rescues it, further loss of complex I, II or DHODH diminishes the rescue, and NAD+ regeneration alone does not rescue. Also states that human brain and lung tumors show robust mitochondrial glucose oxidation. Mouse xenografts plus cell biology.

59. 59
    Valeros J, Jerome M, Tseyang T, et al. Rhodoquinone carries electrons in the mammalian electron transport chain.Cell · 2025 · 188(4):1084-1099.e27doi:10.1016/j.cell.2024.12.007 · PMID 39909039

    Rhodoquinone, long treated as a signature of oxygen-starved worms, was detected in mitochondria purified from certain mouse and human tissues, delivering electrons to fumarate by reversing succinate dehydrogenase whatever the oxygen level; it is undetectable in cultured cells. In vivo mouse and human tissue, with genetic and pharmacologic tools.

60. 60
    DeBerardinis RJ, Mancuso A, Daikhin E, et al. Beyond aerobic glycolysis: transformed cells can engage in glutamine metabolism that exceeds the requirement for protein and nucleotide synthesis.Proc Natl Acad Sci U S A · 2007 · 104(49):19345-50doi:10.1073/pnas.0709747104 · PMID 18032601

    The founding glutamine-addiction study: glioblastoma cells consumed glutamine far faster than protein and nucleotide synthesis required, using it to refill the TCA cycle as carbon left for fatty-acid synthesis and to generate NADPH. Cultured cells.

61. 61
    Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation.Cell · 2011 · 144(5):646-74doi:10.1016/j.cell.2011.02.013 · PMID 21376230

    The source for reprogramming of energy metabolism entering the hallmarks framework as an emerging rather than core hallmark. A review.

62. 62
    Hanahan D. Hallmarks of Cancer: New Dimensions.Cancer Discov · 2022 · 12(1):31-46doi:10.1158/2159-8290.CD-21-1059 · PMID 35022204

    The 2022 update to the hallmarks framework, adding phenotypic plasticity, nonmutational epigenetic reprogramming and polymorphic microbiomes as new dimensions; deregulated cellular metabolism remains in the framework. A perspective.

63. 63
    Hanahan D. Hallmarks of cancer-Then and now, and beyond.Cell · 2026 · 189(8):2254-2277doi:10.1016/j.cell.2025.12.049 · PMID 41616779

    A 2026 restatement of the hallmarks framework, 25 years on. A review.

64. 64
    Hopkins BD, Pauli C, Du X, et al. Suppression of insulin feedback enhances the efficacy of PI3K inhibitors.Nature · 2018 · 560(7719):499-503doi:10.1038/s41586-018-0343-4 · PMID 30051890

    The legitimate version of the dietary argument: inhibiting PI3K raises glucose, the insulin released reactivates the pathway in tumors, and preventing that with a ketogenic diet or drugs improved the efficacy-to-toxicity ratio. Mouse models; not evidence that diet treats cancer.

65. 65
    Halford S, Veal GJ, Wedge SR, et al. A Phase I Dose-escalation Study of AZD3965, an Oral Monocarboxylate Transporter 1 Inhibitor, in Patients with Advanced Cancer.Clin Cancer Res · 2023 · 29(8):1429-1439doi:10.1158/1078-0432.CCR-22-2263 · PMID 36652553

    The first MCT1 inhibitor tested in people: 40 patients, target concentrations reached, dose-limiting toxicities mainly reversible retinal changes, plus one cardiac troponin rise and one acidosis; no efficacy reported. Phase I dose escalation.

66. 66
    Ju J, Zhang H, Lin M, et al. The alanyl-tRNA synthetase AARS1 moonlights as a lactyltransferase to promote YAP signaling in gastric cancer.J Clin Invest · 2024 · 134(10):e174587doi:10.1172/JCI174587 · PMID 38512451

    A correction inside the lactylation story: lactyl-CoA is about a thousandfold less concentrated than acetyl-CoA, making p300 a doubtful lactyltransferase, and AARS1 is identified as a genuine one acting on YAP-TEAD. In vitro and mouse, with a human expression and prognosis association.

67. 67
    Chen H, Li Y, Li H, et al. NBS1 lactylation is required for efficient DNA repair and chemotherapy resistance.Nature · 2024 · 631(8021):663-669doi:10.1038/s41586-024-07620-9 · PMID 38961290

    Lactate-driven lactylation of the DNA-repair protein NBS1 promoted repair of chemotherapy damage; lowering lactate genetically or with stiripentol restored chemotherapy sensitivity in laboratory models, and high levels predicted poor neoadjuvant response. Cell lines and human tissue; not tested as a treatment in a trial.

68. 68
    Pavlides S, Whitaker-Menezes D, Castello-Cros R, et al. The reverse Warburg effect: aerobic glycolysis in cancer associated fibroblasts and the tumor stroma.Cell Cycle · 2009 · 8(23):3984-4001doi:10.4161/cc.8.23.10238 · PMID 19923890

    The origin of the idea, and explicitly a proposal: caveolin-1-deficient stromal cells upregulate myofibroblast markers and glycolytic enzymes in normoxia, validated by staining human breast cancers lacking stromal caveolin-1. It does not measure lactate transfer in patients [94].

69. 69
    Xiao L, Na J, Dang F, et al. Current research status of the reverse Warburg effect in cancer-associated fibroblasts of solid tumors.Int Immunopharmacol · 2026 · 179:116586doi:10.1016/j.intimp.2026.116586 · PMID 41931960

    The 2026 standing of the idea: an active, reviewed framework with monocarboxylate transporters as the central nodes and lactate also shaping immunosuppression and epigenetics. A review with no primary data; it reports what the field holds.

70. 70
    Li J, Huang K, Thakur M, et al. Oncogenic TFE3 fusions drive OXPHOS and confer metabolic vulnerabilities in translocation renal cell carcinoma.Nat Metab · 2025 · 7(3):478-492doi:10.1038/s42255-025-01218-9 · PMID 39915638

    A human cancer wired the other way: TFE3 fusions rewire translocation renal cell carcinoma toward oxidative phosphorylation, contrasting with the glycolytic nature of other renal cancers, and create vulnerability to NADH reductive stress and EGLN1 inhibition. Cell lines, CRISPR screens and mouse work; not a clinical study.

71. 71
    Gorelick AN, Kim M, Chatila WK, et al. Respiratory complex and tissue lineage drive recurrent mutations in tumour mtDNA.Nat Metab · 2021 · 3(4):558-570doi:10.1038/s42255-021-00378-8 · PMID 33833465

    Pathogenic mitochondrial DNA mutations arise at a rate comparable to common driver genes, enriched for loss of function in complex I but depleted of non-synonymous change in complex V, and were associated with better overall survival in colorectal cancer. Human genomics; association, not causation.

72. 72
    Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma.Science · 2000 · 287(5454):848-51doi:10.1126/science.287.5454.848 · PMID 10657297

    Germline SDHD (complex II) mutations in hereditary paraganglioma families, read by the authors as evidence that mitochondria play an important role in certain tumors. Human genetics.

73. 73
    Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer.Nat Genet · 2002 · 30(4):406-10doi:10.1038/ng849 · PMID 11865300

    Identifies fumarate hydratase, a TCA cycle enzyme, as a tumor suppressor, with activity reduced in carriers' cells and very low or absent in their tumors. Human genetics.

74. 74
    Selak MA, Armour SM, MacKenzie ED, et al. Succinate links TCA cycle dysfunction to oncogenesis by inhibiting HIF-alpha prolyl hydroxylase.Cancer Cell · 2005 · 7(1):77-85doi:10.1016/j.ccr.2004.11.022 · PMID 15652751

    The mechanism joining a broken TCA enzyme to a cancer program: succinate leaves the mitochondrion and inhibits the prolyl hydroxylases that degrade HIF-1α, stabilizing it. Cell work.

75. 75
    Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis of human glioblastoma multiforme.Science · 2008 · 321(5897):1807-12doi:10.1126/science.1164382 · PMID 18772396

    The unbiased sequencing study that found recurrent active-site IDH1 mutations in 12% of glioblastoma patients, concentrated in younger patients and secondary tumors, and associated with longer overall survival. Discovery sequencing of 22 tumors with prevalence from a larger screen; an association for the survival finding.

76. 76
    Dang L, White DW, Gross S, et al. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate.Nature · 2009 · 462(7274):739-44doi:10.1038/nature08617 · PMID 19935646

    Identifies the gain of function: mutant IDH1 reduces alpha-ketoglutarate to R(-)-2-hydroxyglutarate, which is markedly elevated in mutant human gliomas. Note the first author's affiliation is Agios Pharmaceuticals, the company that later developed ivosidenib.

77. 77
    Xu W, Yang H, Liu Y, et al. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases.Cancer Cell · 2011 · 19(1):17-30doi:10.1016/j.ccr.2010.12.014 · PMID 21251613

    Explains the damage: 2-hydroxyglutarate competitively inhibits histone demethylases and TET 5-methylcytosine hydroxylases, and IDH1-mutant gliomas show increased histone methylation and decreased 5-hydroxymethylcytosine. Cell and tissue work.

78. 78
    Abou-Alfa GK, Macarulla T, Javle MM, et al. Ivosidenib in IDH1-mutant, chemotherapy-refractory cholangiocarcinoma (ClarIDHy): a multicentre, randomised, double-blind, placebo-controlled, phase 3 study.Lancet Oncol · 2020 · 21(6):796-807doi:10.1016/S1470-2045(20)30157-1 · PMID 32416072

    185 patients randomized 2:1; median progression-free survival 2.7 vs 1.4 months, hazard ratio 0.37 (95% CI 0.25-0.54). Funded by Agios Pharmaceuticals. A real but small benefit on a surrogate endpoint.

79. 79
    Zhu AX, Macarulla T, Javle MM, et al. Final Overall Survival Efficacy Results of Ivosidenib for Patients With Advanced Cholangiocarcinoma With IDH1 Mutation: The Phase 3 Randomized Clinical ClarIDHy Trial.JAMA Oncol · 2021 · 7(11):1669-1677doi:10.1001/jamaoncol.2021.3836 · PMID 34554208

    Final overall survival for ClarIDHy: 10.3 vs 7.5 months, not significant (hazard ratio 0.79, 95% CI 0.56-1.12) because 43 of 61 placebo patients crossed over; crossover-adjusted hazard ratio 0.49. Funded by Agios.

80. 80
    Jusakul A, Cutcutache I, Yong CH, et al. Whole-Genome and Epigenomic Landscapes of Etiologically Distinct Subtypes of Cholangiocarcinoma.Cancer Discov · 2017 · 7(10):1116-1135doi:10.1158/2159-8290.CD-17-0368 · PMID 28667006

    489 cholangiocarcinomas from ten countries: fluke-positive and fluke-negative tumors fall into different molecular clusters, with IDH mutations and epigenetic rewiring on the fluke-negative side. Shows cause shapes molecular landscape; it does not quantify fluke-attributable incidence.

81. 81
    Isaacs JS, Jung YJ, Mole DR, et al. HIF overexpression correlates with biallelic loss of fumarate hydratase in renal cancer: novel role of fumarate in regulation of HIF stability.Cancer Cell · 2005 · 8(2):143-53doi:10.1016/j.ccr.2005.06.017 · PMID 16098467

    The mechanism behind the fumarate-hydratase row: loss of fumarate hydratase raises intracellular fumarate, which competitively inhibits HIF prolyl hydroxylase and so stabilizes HIF. Human renal cancer and cell biology.

82. 82
    Som P, Atkins HL, Bandoypadhyay D, et al. A fluorinated glucose analog, 2-fluoro-2-deoxy-D-glucose (F-18): nontoxic tracer for rapid tumor detection.J Nucl Med · 1980 · 21(7):670-5PMID 7391842

    Early animal tumor imaging with FDG: uptake peaking by 30 minutes, tumor-to-normal ratios of 2.10 to 9.15 by tumor type, no uptake in a necrotic mass, and no toxicity at a thousand times the human tracer dose in mice. Animal work.

83. 83
    Haberkorn U, Ziegler SI, Oberdorfer F, et al. FDG uptake, tumor proliferation and expression of glycolysis associated genes in animal tumor models.Nucl Med Biol · 1994 · 21(6):827-34doi:10.1016/0969-8051(94)90162-7 · PMID 9234332

    What the signal reports: FDG uptake correlated with GLUT1 (r = 0.83) and hexokinase (r = 0.77) expression and not with proliferation. Rat tumor models.

84. 84
    Kubota R, Yamada S, Kubota K, et al. Intratumoral distribution of fluorine-18-fluorodeoxyglucose in vivo: high accumulation in macrophages and granulation tissues studied by microautoradiography.J Nucl Med · 1992 · 33(11):1972-80PMID 1432158

    The engine of every false positive: granulation tissue and infiltrating macrophages took up more tracer than viable tumor cells, with up to 29% of tumor glucose use from non-tumor tissue. Mouse tumors.

85. 85
    Gould MK, Maclean CC, Kuschner WG, et al. Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: a meta-analysis.JAMA · 2001 · 285(7):914-24doi:10.1001/jama.285.7.914 · PMID 11180735

    Forty studies, 1,474 lesions: maximum joint sensitivity and specificity 91.2% (95% CI 89.1-92.9), with practice near 96.8% sensitivity and 77.8% specificity. Few data for nodules under 1 cm.

86. 86
    Nelson SJ, Kurhanewicz J, Vigneron DB, et al. Metabolic imaging of patients with prostate cancer using hyperpolarized [1-13C]pyruvate.Sci Transl Med · 2013 · 5(198):198ra108doi:10.1126/scitranslmed.3006070 · PMID 23946197

    First-in-human hyperpolarized pyruvate imaging in 31 patients: no dose-limiting toxicities and elevated lactate-to-pyruvate ratios in biopsy-proven cancer. A feasibility and safety study, not an outcome trial.

87. 87
    Stein M, Lin H, Jeyamohan C, et al. Targeting tumor metabolism with 2-deoxyglucose in patients with castrate-resistant prostate cancer and advanced malignancies.Prostate · 2010 · 70(13):1388-94doi:10.1002/pros.21172 · PMID 20687211

    Phase I: recommended dose 45 mg/kg, limited by asymptomatic grade 3 QTc prolongation at 60 mg/kg; five of eight patients showed reduced FDG uptake by day 2, which demonstrates competition with the tracer rather than antitumor effect.

88. 88
    Michelakis ED, Sutendra G, Dromparis P, et al. Metabolic modulation of glioblastoma with dichloroacetate.Sci Transl Med · 2010 · 2(31):31ra34doi:10.1126/scitranslmed.3000677 · PMID 20463368

    The paper behind the dichloroacetate phenomenon: mitochondrial hyperpolarization in tumors from 49 patients reversed by the drug, plus five patients treated for up to 15 months with reversible peripheral neuropathy as the dose-limiting toxicity. Five patients is not a trial, and the clinical claims are described as indications.

89. 89
    Stacpoole PW, Abdenur JE, Bedoyan JK, et al. Phase III trial of sodium dichloroacetate for pyruvate dehydrogenase complex deficiency in children.JCI Insight · 2026 · 11(18)doi:10.1172/jci.insight.200149 · PMID 42770298

    Phase III crossover trial, 34 children with pyruvate dehydrogenase complex deficiency, dosed by GSTZ1 genotype: the primary motor endpoint was not met, though blood lactate fell about 20%. Dichloroacetate's genuine medical niche is this inherited disorder, not cancer.

90. 90
    Pedersen PL. 3-Bromopyruvate (3BP) a fast acting, promising, powerful, specific, and effective "small molecule" anti-cancer agent taken from labside to bedside: introduction to a special issue.J Bioenerg Biomembr · 2012 · 44(1):1-6doi:10.1007/s10863-012-9425-4 · PMID 22382780

    The source of the figure most often quoted for 3-bromopyruvate: 19 rodents with advanced cancer, all eradicated. It is an editorial introducing a special issue, written by the laboratory head whose own program produced the compound, with no new data and no controls.

91. 91
    Ko YH, Niedźwiecka K, Casal M, et al. 3-Bromopyruvate as a potent anticancer therapy in honor and memory of the late Professor André Goffeau.Yeast · 2019 · 36(4):211-221doi:10.1002/yea.3367 · PMID 30462852

    The review restating the animal results and referring to a limited number of treated patients. Commercial conflict visible in the record: the first author's affiliation is KoDiscovery, LLC. No controlled human data.

92. 92
    Ko YH, Verhoeven HA, Lee MJ, et al. A translational study "case report" on the small molecule "energy blocker" 3-bromopyruvate (3BP) as a potent anticancer agent: from bench side to bedside.J Bioenerg Biomembr · 2012 · 44(1):163-70doi:10.1007/s10863-012-9417-4 · PMID 22328020

    The entire published human record for 3-bromopyruvate at the time: one young adult with fibrolamellar hepatocellular carcinoma, indexed with a fatal outcome. A case report, with authors holding commercial interests in the compound.

93. 93
    Gritsch D, Baselga-Garriga C, Gonzalez Castro LN. Nutritional strategies in the management of diffuse gliomas: A systematic review.Neurooncol Pract · 2025 · 12(5):773-786doi:10.1093/nop/npaf049 · PMID 41080197

    Eighteen studies, including two randomized trials rated high-quality, which showed no significant survival benefit from dietary intervention; observational studies inconsistent. Feasibility and safety supported, efficacy not.

94. 94
    Voss M, Wagner M, von Mettenheim N, et al. ERGO2: A Prospective, Randomized Trial of Calorie-Restricted Ketogenic Diet and Fasting in Addition to Reirradiation for Malignant Glioma.Int J Radiat Oncol Biol Phys · 2020 · 108(4):987-995doi:10.1016/j.ijrobp.2020.06.021 · PMID 32619561

    The primary ERGO2 report: 50 patients randomized, progression-free survival at 6 months 20% with the diet vs 16% without, not significant; an exploratory subgroup with below-median glucose lived longer. The randomized primary endpoint, as distinct from the later diet-diary analysis [69].

95. 95
    Voss M, Wenger KJ, von Mettenheim N, et al. Short-term fasting in glioma patients: analysis of diet diaries and metabolic parameters of the ERGO2 trial.Eur J Nutr · 2022 · 61(1):477-487doi:10.1007/s00394-021-02666-1 · PMID 34487222

    ERGO2 randomized 50 patients with recurrent brain tumors to re-irradiation with or without a calorie-restricted ketogenic diet and intermittent fasting and did not meet its primary endpoint; the diet was tolerated, leptin and insulin fell, cognition and quality of life were unaffected, and the control group ate less than planned.

96. 96
    Firdous J, Asif AE, Haris HM, et al. Efficacy and safety of ketogenic diet in glioblastoma: an updated systematic review and meta-analysis.Neurol Sci · 2026 · 47(5)doi:10.1007/s10072-026-09035-y · PMID 42032215

    The favorable synthesis, included so the claim can be examined: 41 heterogeneous studies including case series and abstracts, reporting median overall survival of 29.4 months in adherent cohorts against 14.6 months in historical controls. Historical controls are not a randomized comparison and the authors themselves call for phase III trials.

97. 97
    Jameson GS, Roe DJ, Borazanci E, et al. A randomized phase II trial of gemcitabine, nab-paclitaxel, cisplatin with or without a medically supervised ketogenic diet for patients with metastatic pancreatic cancer.Cancer · 2026 · 132(6):e70343doi:10.1002/cncr.70343 · PMID 41817106

    Randomized phase II screening trial, 32 evaluable patients: progression-free survival 8.5 vs 6.2 months (hazard ratio 0.53, 95% CI 0.21-1.37), overall survival 13.7 vs 10.2 months (HR 0.58, 95% CI 0.25-1.37); both intervals include no effect, and the design used a lenient one-sided threshold of 0.20. Not powered for a definitive result.

98. 98
    Dantas E, Hootman KC, Moyer J, et al. Very low-carbohydrate ketogenic diet in treatment-naïve women with endometrial cancer and overweight: a randomized feasibility study.Nat Commun · 2026 · 17(1)doi:10.1038/s41467-026-73519-w · PMID 42192131

    Randomized feasibility trial, 19 women: a very low-carbohydrate diet for three to four weeks before surgery was tolerated; fasting glucose fell about 22% and insulin about 60%, while LDL cholesterol rose about 17.8%. A feasibility endpoint, not an efficacy trial.

99. 99
    Ferrer M, Mourikis N, Davidson EE, et al. Ketogenic diet promotes tumor ferroptosis but induces relative corticosterone deficiency that accelerates cachexia.Cell Metab · 2023 · 35(7):1147-1162.e7doi:10.1016/j.cmet.2023.05.008 · PMID 37311455

    In mice carrying IL-6-producing cancers, a ketogenic diet slowed tumor growth but accelerated cachexia and shortened survival; dexamethasone reversed the harm. A caution that diet effects on the host matter as much as on the tumor. Mouse models.

100. 100
     DeBerardinis RJ, Chandel NS. We need to talk about the Warburg effect.Nat Metab · 2020 · 2(2):127-129doi:10.1038/s42255-020-0172-2 · PMID 32694689

     A commentary by two leading investigators calling for terminological discipline. The record carries no abstract, so only the existence and subject of the piece are verifiable from PubMed; its recommendations must be read in the full text.

101. 101
     Qu D, Zou X, Liu Z. Propofol modulates glycolysis reprogramming of ovarian tumor via restraining circular RNA-zinc finger RNA-binding protein/microRNA-212-5p/superoxide dismutase 2 axis.Bioengineered · 2022 · 13(5):11881-11892doi:10.1080/21655979.2022.2063649 · PMID 35543376

     RETRACTED PUBLICATION per the PubMed record. Cited only as an example of the quality problem in this literature; none of its claims are used.

102. 102
     Tielens AG. Energy generation in parasitic helminths.Parasitol Today · 1994 · 10(9):346-52doi:10.1016/0169-4758(94)90245-3 · PMID 15275412

     The introduction to helminth bioenergetics: large anaerobic capacity in some life-cycle stages, aerobic metabolism in others, with external conditions driving the switch. A narrative review with no pooled figures.

103. 103
     Tielens AG, Van Hellemond JJ. The electron transport chain in anaerobically functioning eukaryotes.Biochim Biophys Acta · 1998 · 1365(1-2):71-8doi:10.1016/s0005-2728(98)00045-0 · PMID 9693724

     Explains why ubiquinone cannot substitute for rhodoquinone in fumarate reduction, and that eukaryotic fumarate reductase is structurally close to succinate dehydrogenase but functionally reversed. A review.

104. 104
     Van Hellemond JJ, Klockiewicz M, Gaasenbeek CP, et al. Rhodoquinone and complex II of the electron transport chain in anaerobically functioning eukaryotes.J Biol Chem · 1995 · 270(52):31065-70doi:10.1074/jbc.270.52.31065 · PMID 8537365

     Rhodoquinone was present in every examined eukaryote that reduces fumarate during anoxia, including snails, mussels, lugworms and oysters as well as parasitic helminths, and its ratio to ubiquinone tracked reliance on fumarate reduction across Fasciola hepatica development.

105. 105
     Vairoletti F, Saiz C, Salinas G. Pharmacological targeting of the helminth complex II: are there any rhodoquinone-driven adaptations?Trends Parasitol · 2025 · 41(9):746-752doi:10.1016/j.pt.2025.06.016 · PMID 40675863

     A 2025 re-examination arguing that most helminths run their ordinary complex II in reverse as a fumarate reductase rather than using a dedicated parasite complex II, so the Ascaris-based model should not be generalized. A review.

106. 106
     Lundström-Stadelmann B, Rufener R, Ritler D, et al. The importance of being parasiticidal… an update on drug development for the treatment of alveolar echinococcosis.Food Waterborne Parasitol · 2019 · 15:e00040doi:10.1016/j.fawpar.2019.e00040 · PMID 32095613

     Names the malate dismutation pathway as an understudied and essential target in Echinococcus multilocularis that mammals do not have. A drug-development review; the claim is about parasite biology, not about cancer.

107. 107
     Opperdoes FR, Alencar MB, Silber AM, et al. Common pathways and outliers in glucose catabolism across Trypanosomatidae.FEBS J · 2026 · advance online publicationdoi:10.1111/febs.70672 · PMID 42552839

     How protozoan glycolysis differs from ours: the first enzymes sit inside glycosomes whose nucleotide and NAD pools do not equilibrate with the cytosol, forcing three specific redox-balancing routes, with some species surviving true anaerobiosis by pyruvate dismutation or propionate fermentation. A review.

108. 108
     MacRae JI, Dixon MW, Dearnley MK, et al. Mitochondrial metabolism of sexual and asexual blood stages of the malaria parasite Plasmodium falciparum.BMC Biol · 2013 · 11:67doi:10.1186/1741-7007-11-67 · PMID 23763941

     Corrects the picture of malaria as pure glycolysis: a conventional oxidative TCA cycle in both asexual and sexual blood stages, low glucose flux into it in asexual parasites and much higher in gametocytes, where inhibiting it caused arrest and death. Parasite culture work.

109. 109
     Bouvard V, Baan R, Straif K, et al. A review of human carcinogens--Part B: biological agents.Lancet Oncol · 2009 · 10(4):321-2doi:10.1016/s1470-2045(09)70096-8 · PMID 19350698

     The International Agency for Research on Cancer classification placing the liver flukes Opisthorchis viverrini and Clonorchis sinensis and the blood fluke Schistosoma haematobium among Group 1 human carcinogens. A summary of the IARC monograph.

110. 110
     Sakai C, Tomitsuka E, Esumi H, et al. Mitochondrial fumarate reductase as a target of chemotherapy: from parasites to cancer cells.Biochim Biophys Acta · 2012 · 1820(5):643-51doi:10.1016/j.bbagen.2011.12.013 · PMID 22226661

     The explicit bridge: the NADH-fumarate reductase system of helminths was also found in some human cancer cells under tumor-like conditions, with an anthelmintic identified as a specific inhibitor. A review of the authors' own program; not clinical evidence.

111. 111
     Ishii I, Harada Y, Kasahara T. Reprofiling a classical anthelmintic, pyrvinium pamoate, as an anti-cancer drug targeting mitochondrial respiration.Front Oncol · 2012 · 2:137doi:10.3389/fonc.2012.00137 · PMID 23061049

     The rationale for repurposing an anthelmintic: pyrvinium suppressed the NADH-fumarate reductase system under tumor-like hypoglycemic and hypoxic conditions and was cytotoxic to cancer lines during glucose starvation. A review of preclinical work.

112. 112
     Spinelli JB, Rosen PC, Sprenger HG, et al. Fumarate is a terminal electron acceptor in the mammalian electron transport chain.Science · 2021 · 374(6572):1227-1237doi:10.1126/science.abi7495 · PMID 34855504

     Mammalian cells deprived of oxygen reduction accumulate ubiquinol and drive the succinate dehydrogenase complex in reverse onto fumarate, with mouse tissues differing in capacity. Shows the parasite-like reaction is available to our own mitochondria; cell and mouse work.

113. 113
     Ponzini FM, Schultz CW, Leiby BE, et al. Repurposing the FDA-approved anthelmintic pyrvinium pamoate for pancreatic cancer treatment: study protocol for a phase I clinical trial in early-stage pancreatic ductal adenocarcinoma.BMJ Open · 2023 · 13(10):e073839doi:10.1136/bmjopen-2023-073839 · PMID 37848297

     A published protocol, not results: a 3+3 dose escalation from 5 to 20 mg/kg given for three days before surgery, with grade 3 or higher dose-limiting toxicity as the primary endpoint. Cited to show exactly how early the human work is.

114. 114
     Salem N, Balkman JD, Wang J, et al. In vivo imaging of schistosomes to assess disease burden using positron emission tomography (PET).PLoS Negl Trop Dis · 2010 · 4(9):e827doi:10.1371/journal.pntd.0000827 · PMID 20877718

     Live adult worms took up FDG in proportion to worm number in vitro, and portal FDG signal tracked worm burden in athymic mice, R2 = 0.85 (P < 0.001) in the 17 animals with more than 50 worms. Mouse work in a T-cell-deficient host, so it isolates the parasite signal in a way an intact patient does not.