evipedia.ai
← Back

High-Dose Vitamin C to Treat Cancer

Evidence Review created on 08/11/2026 using AI4L / Opus 5

Also known as: Intravenous Vitamin C, IVC, Pharmacological Ascorbate, High-Dose Ascorbic Acid, Sodium Ascorbate, Vitamin C Megadosage

Motivation

Vitamin C is a nutrient the body cannot make and normally takes in from food. At everyday amounts it protects cells from damage and helps enzymes build connective tissue. At the far larger amounts reached only by dripping it into a vein, it behaves like a different substance: inside tissue it generates hydrogen peroxide, which harms some tumour cells more than healthy ones.

The idea dates to the 1970s, when a hospital surgeon and a chemist reported that patients with advanced cancer given very large vitamin C doses lived longer than comparison patients. The trials that followed used capsules rather than infusions, found nothing, and the approach was set aside for decades. Later work measuring blood levels showed that swallowed and infused vitamin C are not interchangeable, which reopened the question.

This review examines what is known about high-dose vitamin C given for cancer: how it is thought to work, what human studies show and where they disagree, the doses and schedules used in clinics, the harms and the people for whom it is unsafe, and how response and safety are tracked.

Benefits - Risks - Protocol - Conclusion

High-level overviews of high-dose vitamin C in cancer from expert platforms and non-systematic academic sources.

Content from four of the six priority platforms could not be included: the only relevant material on peterattiamd.com is a segment of AMA #120 on vitamin C in cancer treatment whose audio and show notes sit behind the member paywall, hubermanlab.com and chriskresser.com carry no material discussing high-dose or intravenous vitamin C in a cancer context, and lifespan.io covers vitamin C only in relation to ageing biology, not cancer treatment. Five qualifying items were found, so the list is not padded.

Grokipedia

  • Vitamin C megadosage

    Covers the dose-response ceiling of oral vitamin C, the Pauling-era cancer claims, and the later infusion literature, with unusually detailed treatment of the Mayo Clinic dispute.

Examine

  • Vitamin C

    Notes that studied doses run to 24 g intravenously, and flags oxalate nephropathy (crystal damage in the kidney’s filtering tubes) and kidney failure above 2 g as the precaution that matters most here.

ConsumerLab

  • Vitamin C Supplements Review & Top Picks

    Tests retail oral products for label accuracy and covers the forms question; relevant here only for the oral maintenance dosing used between infusions, not the infusions themselves.

Systematic Reviews

Systematic reviews and meta-analyses of high-dose vitamin C in cancer, covering both survival claims and safety.

Mechanism of Action

At dietary intakes vitamin C donates electrons to protect molecules from oxidation and helps enzymes build collagen and neurotransmitters. Infused at 50–100 g, plasma reaches millimolar concentrations and the chemistry inverts: ascorbate reduces catalytically available iron in the fluid around cells, generating the ascorbate radical and then hydrogen peroxide, which diffuses inward. Many tumour cells carry more loosely bound iron and less catalase (the enzyme that clears hydrogen peroxide), so peroxide becomes hydroxyl radicals inside them, damaging DNA and draining adenosine triphosphate (ATP, the cell’s energy currency).

Two non-oxidative routes also operate. Vitamin C is a required cofactor for ten-eleven translocation enzymes (TET, which strip methyl tags off DNA and reawaken silenced genes) and for prolyl hydroxylases that destabilise hypoxia-inducible factor 1-alpha (HIF-1α, the switch tumours use to tolerate low oxygen). In cells carrying mutations in the RAS gene family (including KRAS) or in BRAF (genes that drive uncontrolled growth), oxidised vitamin C enters through glucose transporters and inactivates glyceraldehyde-3-phosphate dehydrogenase (GAPDH, an enzyme of the sugar-burning pathway).

Pharmacologically, ascorbate is water-soluble, essentially unbound, and distributes into extracellular water while concentrating in adrenal, pituitary, brain and white-cell tissue. It is cleared largely unchanged by the kidney with a plasma half-life near two hours at these doses, is not a cytochrome P450 substrate (the liver’s main drug-processing enzymes), and is metabolised in small part to dehydroascorbic acid and then oxalate.

The competing explanation holds that circulating catalase and glutathione quench peroxide before it reaches tumours, which predicts no anti-tumour effect.

Historical Context & Evolution

Vitamin C entered oncology through Ewan Cameron, a hospital surgeon who reasoned that ascorbate would strengthen the connective-tissue barrier containing tumours, and Linus Pauling, who supplied the biochemical framing. Their 1976 report described terminal patients given 10 g daily, largely starting intravenously, surviving several times longer than controls drawn from hospital records. Mayo Clinic investigators tested the claim twice, in 1979 and 1985, in randomised double-blind trials (studies where participants are assigned by chance and neither side knows the assignment) using 10 g of oral vitamin C, and found no effect on survival or symptoms.

The findings on both sides stand; the dispute is over what each measured. Cameron’s controls were retrospective and not matched for treatment intensity, a genuine weakness. The Mayo trials were rigorously run but used capsules, and pharmacokinetic work published in 2004 showed that oral dosing plateaus near 220 µmol/L in plasma while a 50 g infusion is predicted to reach 13,400 µmol/L — a sixty-fold gap at exactly the concentrations where tumour cell kill is observed. Whether that voids the Mayo result or merely bounds its scope is still argued.

The practical consequence was a forty-year split. Academic oncology treated the question as settled, while integrative and naturopathic clinics kept infusing, producing case series but little controlled data. Randomised infusion trials resumed only after 2010, and the first phase 3 attempt reported in 2022.

Expected Benefits

High 🟩 🟩 🟩

Reliable Attainment of Millimolar Ascorbate Exposure

Infusion is the only route that raises plasma vitamin C from the micromolar into the millimolar range where tumour cell kill occurs, because gut absorption saturates and the kidney excretes the excess. This is the best-established effect of the intervention, resting on controlled pharmacokinetic modelling (Padayatty et al., 2004) and confirmed by direct measurement in modern trials that sampled plasma (Bodeker et al., 2024). Reaching the concentration is a prerequisite for the other benefits, not an outcome in itself.

Magnitude: Maximum tolerated oral dosing (3 g every 4 hours) plateaus at a predicted 220 µmol/L, while 50 g infused is predicted to reach 13,400 µmol/L; 87.5 g measured ≥20,000 µmol/L in patients with glioblastoma (an aggressive brain cancer).

Medium 🟩 🟩

Reduced Symptom Burden and Better Quality of Life

Across early-phase trials and clinic series, infusions were followed by lower fatigue, nausea, insomnia, constipation and pain scores and better global quality-of-life ratings, attributed to reduced inflammation plus correction of deficiency rather than tumour shrinkage. Systematic reviews call this the most consistent clinical signal in the field (Fritz et al., 2014), but almost every supporting study was open-label with self-reported endpoints, and the one adequately randomised trial found no quality-of-life difference either way (Bodeker et al., 2024).

Magnitude: Direction is toward milder fatigue, nausea and pain, largest in patients starting with low vitamin C status and active chemotherapy; the literature reports no pooled outcome figure for symptom scores.

Adding infused ascorbate to platinum and taxane chemotherapy was associated with fewer and milder bone-marrow, kidney, liver and nerve toxicities in an ovarian cancer trial (Ma et al., 2014), and the randomised pancreatic trial recorded no increase in adverse-event frequency or severity despite 75 g three times weekly. The proposed basis is that normal tissue, which retains catalase, is not damaged by the peroxide load. Sample sizes are small and toxicity grading was not blinded in the ovarian study.

Magnitude: Direction is toward fewer grade 2 and higher toxicities when infusions accompany platinum–taxane regimens; the literature reports no pooled outcome figure across trials.

Longer Overall Survival Added to Chemotherapy ⚠️ Conflicted

A meta-analysis of 8 studies in 2,722 adults found longer median overall survival with infused vitamin C (Qu et al., 2025), and a randomised trial in metastatic pancreatic cancer nearly doubled it (Bodeker et al., 2024). Against this, the only phase 3 trial, in metastatic colorectal cancer, missed its primary endpoint with no survival separation (Wang et al., 2022). The pooled effect was larger in cohort studies than in randomised ones, the classic signature of confounding.

Magnitude: Pooled median survival ratio 1.83 (95% CI 1.40–2.40; CI is the confidence interval, the range within which the true value probably lies); pancreatic trial 16.0 versus 8.3 months, HR 0.46 (hazard ratio, the relative rate of death between groups); colorectal 20.7 versus 19.7 months, not significant.

Low 🟩

Progression-Free Gain in RAS-Mutant Colorectal Cancer

In a prespecified subgroup of the phase 3 colorectal trial, patients whose tumours carried RAS mutations gained progression-free time, while unmutated tumours did not (Wang et al., 2022). This matches laboratory work showing mutation-driven sugar metabolism creates the vulnerability (Yun et al., 2015). One subgroup, one trial, not replicated.

Magnitude: Median progression-free survival 9.2 versus 7.8 months, HR 0.67 (95% CI 0.50–0.91).

Radiosensitisation in Glioblastoma and Lung Cancer

Ascorbate increased radiation and chemotherapy kill of glioblastoma and non-small cell lung cancer cells by disrupting their iron handling (Schoenfeld et al., 2017), and the first-in-human glioblastoma trial reported survival above the historical benchmark (Allen et al., 2019). Single-arm design prevents attribution.

Magnitude: Median overall survival 18 months (23 months in the subgroup whose tumours lacked the chemical tag that normally predicts a better response to standard chemotherapy) against roughly 14 months historically, in 11 evaluable patients.

Reduced Systemic Inflammation

Repeated infusions of 7.5–50 g lowered C-reactive protein (CRP, a general marker of inflammation) and several inflammatory signalling proteins in a clinic series (Mikirova et al., 2012). That series came from the Riordan Clinic, which sells the infusions it studied — a direct financial interest in the finding.

Magnitude: CRP fell in 75% of 45 patients and rose in 25%; uncontrolled design with no comparison group.

Speculative 🟨

Restoration of Epigenetic Control in Myeloid Blood Cancers

Vitamin C is required for the enzymes that strip methyl tags off DNA, so restoring their activity could amplify drugs acting the same way. The basis is laboratory work only; trials are still running.

Enhanced Response to Immune Checkpoint Drugs

Mouse work suggests pharmacological ascorbate reshapes the immune cells inside tumours and improves checkpoint-inhibitor activity (drugs that unblock immune attack). No human outcome data exist; the first large pairing trial is only now recruiting.

Benefit-Modifying Factors

  • RAS and BRAF mutation status: Tumours carrying these mutations depend on the sugar-burning enzyme that oxidised vitamin C inactivates, and were the only colorectal subgroup to gain progression-free time; unmutated tumours showed no signal.

  • Baseline plasma ascorbate: People starting deficient have the most to gain on symptom and inflammation endpoints, since part of the observed benefit is simply repletion. Those already replete depend entirely on the pharmacological mechanism.

  • Tumour catalase and iron content: Selectivity rests on tumour cells clearing peroxide poorly and holding more loosely bound iron. Glioblastoma and lung tumour lines are unusually susceptible; catalase-rich tumours are predicted to resist.

  • Sex: At equal milligram-per-kilogram dosing, women reach higher peak plasma concentrations because of smaller extracellular water volume. Weight-based dosing narrows but does not erase the gap; no sex difference in outcome has been reported.

  • Kidney function: Ascorbate is cleared renally, so reduced filtration raises peak exposure but also oxalate retention. Dose ceilings imposed for safety in impaired kidneys can drop exposure below the millimolar threshold.

  • Age: Older adults reach target concentrations at lower absolute doses through reduced clearance, but tolerate the fluid volume less well. Trials enrolled few people beyond their late seventies, so the benefit estimate thins there.

  • Concurrent chemotherapy: The pooled survival signal was larger in non-chemotherapy combinations than alongside chemotherapy, the opposite of the mechanistic prediction — an unexplained finding that weakens confidence in the adjuvant framing.

  • Iron overload states: Hereditary iron loading supplies the catalytic metal the mechanism needs, which could raise both tumour kill and collateral oxidative damage. Direction of the net effect is untested.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Oxalate Nephropathy and Acute Kidney Injury

A fraction of infused ascorbate is metabolised to oxalate, which can crystallise in kidney tubules and cause abrupt kidney failure, sometimes needing dialysis. Cases cluster in people with pre-existing kidney disease, dehydration or very large repeated doses, and have followed both oral and infused high doses. The most recent benefit-and-risk synthesis names this the principal serious hazard and the reason for mandatory pre-screening (Alangari et al., 2026). Onset is typically days after exposure and recovery is sometimes incomplete.

Magnitude: Direction is toward crystal-driven kidney injury, essentially confined to impaired baseline kidney function or dehydration; prospective oncology trials excluding renal impairment reported no cases, and the literature reports no incidence figure for the screened population.

Haemolysis in Glucose-6-Phosphate Dehydrogenase Deficiency

Glucose-6-phosphate dehydrogenase (G6PD, the enzyme that shields red blood cells from oxidative damage) is absent or weak in a sizeable minority of people. Without it, the hydrogen peroxide generated by infusion ruptures red cells, and published cases include severe haemolysis with kidney failure and death (Padayatty et al., 2010). Because the deficiency is silent until challenged, enzyme testing before the first infusion is treated as mandatory rather than advisable.

Magnitude: The deficiency affects roughly 5–10% of people of African, Mediterranean, Middle Eastern and Southeast Asian ancestry and under 1% of northern Europeans; 2 of the deaths in the practitioner survey occurred in patients known to carry a contraindication.

Falsely Elevated Point-of-Care Glucose Readings

Ascorbate is chemically similar enough to glucose to interfere with glucose-oxidase and glucose-dehydrogenase test strips, so finger-prick meters read far above the true value for hours after an infusion. Insulin given on such a reading has caused severe hypoglycaemia (dangerously low blood sugar), including a fatal case. The effect is analytical rather than physiological and is documented in device and regulatory warnings; laboratory assays using hexokinase (an enzyme that tags glucose for breakdown) are unaffected.

Magnitude: Direction is a large upward bias on strip meters lasting several hours after a 50–100 g infusion, largest at the time of peak plasma concentration; the literature reports no single quantified bias figure across meter types.

Medium 🟥 🟥

The common effects are self-limited: lethargy, fatigue, altered mental status, vein irritation, thirst, nausea, tremor and osmotic diuresis (extra urine output pulled out by the sugar-like load). Sodium ascorbate also delivers a substantial sodium dose, which has caused hypernatraemia (blood sodium above the normal range) and matters in heart failure, ascites (fluid collecting in the abdomen) or advanced kidney disease. A practitioner survey covering more than 9,000 patients found these mild events dominated the profile (Padayatty et al., 2010), consistent with the tolerability reported in academic trials.

Magnitude: Side effects in 101 of 9,328 surveyed patients (about 1%), of which 59 were lethargy or fatigue, 21 mental-status change and 6 phlebitis (inflammation of the infused vein); a 60 g dose given entirely as sodium ascorbate carries roughly 7 g of sodium.

Displacement of Effective Conventional Treatment

The largest measurable harm attached to this intervention is not the infusion but what it can substitute for. Patients who take up complementary therapies and then refuse or delay surgery, chemotherapy or radiation die sooner (Johnson et al., 2018). The risk is behavioural rather than pharmacological and concentrates where clinics market infusions as a standalone alternative rather than an addition, so it scales with how the intervention is framed rather than with dose.

Magnitude: Roughly a two-fold higher risk of death among complementary-medicine users with curable cancers, driven entirely by refusal of at least one conventional treatment; no excess mortality when complementary treatment was added rather than substituted.

Antagonism of Bortezomib-Based Myeloma Therapy

Ascorbate chemically inactivates boronic-acid proteasome inhibitors (drugs that block the cell’s protein-recycling machinery), and animal work showed vitamin C abolished the anti-tumour activity of bortezomib (Perrone et al., 2009). Practice in myeloma (a cancer of antibody-producing bone-marrow cells) therefore advises avoiding vitamin C around bortezomib cycles. Human outcome data are limited to observation, but the chemistry is unambiguous and the affected drug is a backbone therapy, so the consequence of getting it wrong is loss of an effective treatment.

Magnitude: Direction is near-complete loss of bortezomib activity in animal models at ascorbate concentrations far below those reached by infusion; the literature reports no quantified human survival effect.

Low 🟥

Calcium Oxalate Kidney Stones

Repeated oxalate loads raise stone risk, most clearly in men and in people with prior stones (Thomas et al., 2013). The evidence comes from high-dose oral cohorts rather than infusion protocols, so the transfer to infused dosing is inferred rather than measured.

Magnitude: Roughly a two-fold higher stone incidence with habitual oral supplement use in a male cohort; infusion-specific incidence is not quantified.

Iron Overload Aggravation

Ascorbate mobilises stored iron and increases its absorption, worsening oxidative tissue damage in hereditary haemochromatosis (an inherited disease of iron accumulation) or transfusion-dependent iron loading. The basis is case reports of cardiac deterioration in iron-loaded patients receiving chelation and the mechanism of catalytic iron; no infusion trial enrolled this group.

Magnitude: Direction is toward increased iron-mediated tissue damage, and only in people with existing overload; the literature reports no outcome figure.

Tumour Lysis Syndrome

Rapid tumour cell death can release potassium, phosphate and uric acid faster than the kidneys clear them, causing rhythm disturbance and kidney failure. Isolated cases have followed infusion in bulky, highly sensitive disease.

Magnitude: Direction is a rising risk with tumour bulk and treatment sensitivity; the literature reports no incidence figure.

Speculative 🟨

Antioxidant Blunting of Radiation or Chemotherapy ⚠️ Conflicted

A long-standing concern that antioxidants shield tumours as well as normal tissue. Laboratory work at pharmacological doses shows added kill, not protection, while lower antioxidant doses may interfere; no human trial resolves it.

Selection for Ascorbate-Resistant, More Metastatic Cells

Pancreatic tumour cells made resistant to pharmacological ascorbate in the laboratory acquired greater metastatic capacity (Pope et al., 2025). Whether repeated clinical exposure selects for the same phenotype is untested.

Risk-Modifying Factors

  • G6PD gene variants: Hundreds of variants exist with widely different residual enzyme activity. Class I and II variants make haemolysis after high-dose infusion close to predictable; mild variants may tolerate lower doses. Testing measures activity, not genotype.

  • Baseline kidney function and hydration: Reduced filtration and low urine flow both concentrate oxalate in tubules. This is the single strongest determinant of whether an infusion is routine or dangerous.

  • Prior kidney stones or hyperoxaluria (persistently high urinary oxalate): A stone history signals existing oxalate handling limits, raising the chance that repeated infusions precipitate a new stone or crystal nephropathy.

  • Sex: Men carry the higher oxalate stone risk in cohort data; women reach higher peak plasma concentrations at equal weight-adjusted doses, shifting exposure-related risks toward them.

  • HFE variants (the gene controlling how much iron the gut absorbs) and transfusion history: Both supply the catalytic iron the mechanism needs, amplifying oxidative damage in normal tissue as well as tumour.

  • Heart failure, ascites or oedema: These conditions limit tolerance of the fluid and sodium volume delivered with buffered ascorbate, independent of any oxidative mechanism.

  • Age: Falling filtration rate, lower total body water and higher rates of unrecognised kidney disease all raise exposure and oxalate retention in adults past their seventies.

  • Insulin-treated diabetes: Glucose meter interference converts a benign analytical artefact into a hypoglycaemia risk only in people who dose insulin from strip readings.

Key Interactions & Contraindications

  • Proteasome inhibitors (bortezomib, ixazomib) — absolute avoidance during dosing: Ascorbate chemically inactivates the boronic acid group, risking loss of anti-myeloma activity. Mitigation is to stop all vitamin C above dietary amounts throughout the treatment cycle.

  • Platinum and taxane chemotherapy (carboplatin, cisplatin, paclitaxel) — monitor: No loss of anti-tumour effect was seen and toxicity was lower, but data are from small trials. Mitigation is to infuse on non-chemotherapy days or immediately after.

  • Gemcitabine and erlotinib — monitor: A systematic review of infusion trials found no interference with anti-tumour effect or added toxicity. Consequence if wrong would be reduced chemotherapy efficacy; mitigation is to separate infusion from dosing by at least 24 hours where feasible.

  • Anticoagulants (warfarin) — caution: Case reports describe reduced anticoagulant effect with high vitamin C intake, risking clot formation. Mitigation is to check the clotting time within one week of starting or changing infusion dose.

  • Deferoxamine and other iron chelators — caution: Ascorbate mobilises iron and has precipitated cardiac deterioration in iron-overloaded patients receiving chelation. Mitigation is to withhold high-dose vitamin C until iron burden is controlled.

  • Oral iron and heme-iron-rich meals — caution: Ascorbate multiplies non-heme iron absorption, adding catalytic iron. Mitigation is to separate iron supplements from infusion days and avoid routine iron unless deficiency is documented.

  • Dextrose-containing intravenous fluids — caution: Glucose competes with the oxidised form of vitamin C for the same cell transporters, blunting uptake. Mitigation is to dilute in sterile water or Ringer’s lactate, never dextrose.

  • Additive oxidant or oxalate supplements (alpha-lipoic acid, vitamin K3, green tea extract) — caution: These amplify ascorbate’s peroxide effect, raising tumour kill but also kidney oxalate exposure. Mitigation is to add one at a time with kidney monitoring.

  • Antioxidant supplements (N-acetylcysteine, vitamin E, glutathione, selenium) — caution: These quench the peroxide the infusion is designed to generate, plausibly cancelling its effect. Mitigation is to hold them on infusion days.

  • Over-the-counter analgesics (aspirin, acetaminophen) — monitor: Aspirin lowers plasma and white-cell vitamin C, while vitamin C may slow acetaminophen clearance. Consequence is altered drug exposure; separate dosing from infusion days.

  • Aluminium-containing antacids (aluminium hydroxide) — caution: Ascorbate increases aluminium absorption, a concern where kidney clearance is already reduced. Mitigation is to switch to a non-aluminium antacid during a course of infusions.

  • Radiation therapy — monitor: Preclinical data show sensitisation rather than protection, but the antioxidant concern is unresolved. Mitigation is to infuse within the window used in trials, immediately before or after the radiation fraction.

Populations who should avoid High-Dose Vitamin C:

  • Glucose-6-phosphate dehydrogenase deficiency of any severity — risk of severe haemolysis and death
  • Stage 4–5 chronic kidney disease (eGFR <30 mL/min/1.73 m², where eGFR is the estimated glomerular filtration rate, a calculated measure of kidney filtering capacity) or any dialysis dependence
  • Active oxalate nephropathy, primary hyperoxaluria, or recurrent calcium oxalate stones within 12 months
  • Hereditary haemochromatosis or transfusional iron overload with ferritin above 1,000 ng/mL
  • Active bortezomib- or ixazomib-based myeloma therapy
  • Decompensated heart failure (New York Heart Association Class IV) or uncontrolled ascites, because of the sodium and fluid load
  • Pregnancy, where no pharmacological-dose safety data exist

Risk Mitigation Strategies

  • Mandatory G6PD enzyme assay before the first infusion: A red-cell enzyme activity result, not a genotype or an ancestry assumption, prevents the haemolysis that has caused the only reported deaths from this intervention.

  • Kidney gate before every cycle: Protocols set an entry threshold of eGFR ≥60 mL/min/1.73 m², recheck creatinine before each cycle, and stop for a 25% rise. This is the primary defence against oxalate nephropathy.

  • Pre- and post-infusion hydration: 500–1,000 mL of isotonic fluid or equivalent oral intake around each infusion keeps urine flowing and dilutes tubular oxalate, the mechanism behind crystal injury.

  • Rate ceiling of 0.5 g per minute: Infusing 25–100 g over 1.5–4 hours rather than rapidly limits osmotic diuresis, vein irritation, nausea and tremor without lowering peak plasma concentration meaningfully.

  • Laboratory glucose confirmation for 8 hours after infusion: For anyone dosing insulin, a venous hexokinase result rather than a strip meter is the basis for dosing, which averts insulin overdose from falsely high readings.

  • Haemolysis surveillance after the first two infusions: Haemoglobin, reticulocytes (newly made red cells) and lactate dehydrogenase (LDH, an enzyme released when cells rupture) at 24–48 hours catch red-cell breakdown in people with a borderline enzyme result.

  • Ferritin and transferrin saturation before starting: Identifying iron overload before supplying catalytic iron prevents the oxidative tissue damage and cardiac deterioration reported in iron-loaded patients.

  • Medication reconciliation for proteasome inhibitors and antioxidants: An explicit check for bortezomib, ixazomib, N-acetylcysteine and high-dose vitamin E prevents both loss of myeloma therapy and cancellation of the infusion’s own mechanism.

  • Written commitment to continue standard oncology care: Because the dominant measurable harm is treatment displacement, agreeing in advance that infusions are additive removes the pathway to the two-fold mortality excess.

Therapeutic Protocol

  • Riordan Clinic integrative protocol: The most widely copied approach, developed at the Riordan Clinic, which bills for the infusions it advocates: 0.1–1.5 g/kg, escalating from a 15 g test dose to 25–100 g, two to three times weekly.

  • Academic pharmacological-ascorbate protocol: University of Iowa trials fix the dose by achieved concentration rather than body weight: 75 g three times weekly with gemcitabine and nab-paclitaxel, or escalation to 87.5 g to reach ≥20 mmol/L with radiation.

  • Weight-based cycle protocol: The Chinese phase 3 trial used 1.5 g/kg daily over 3 hours on days 1–3 of each chemotherapy cycle, a lower cumulative exposure than the thrice-weekly schedules and a possible reason for its null result.

  • Target plasma concentration, not target dose: All academic protocols titrate to ≥20 mmol/L mid-infusion. Fixed doses miss this in heavier patients and overshoot in lighter ones, which the weight-based schedules partly correct.

  • Best time of day: Morning or early afternoon, so the osmotic diuresis that follows a 25–100 g load resolves before sleep. Trials infused before or immediately after radiation fractions to exploit the sensitisation window.

  • Half-life and its consequence: Plasma half-life is roughly two hours at pharmacological doses and levels return to baseline within about four hours, so exposure is a series of brief spikes rather than a sustained level — the rationale for frequent repetition.

  • Single dose, not split: The whole dose is given as one infusion. Splitting it drops the peak below the millimolar threshold the mechanism requires, which is the pharmacological objection to oral maintenance dosing as a substitute.

  • Genetic factors in dose choice: G6PD variant class determines eligibility outright. HFE iron-loading variants lower the safe ceiling, and tumour RAS or BRAF mutation status is the only genotype so far linked to greater benefit.

  • Sex-based dosing: Weight-based dosing gives women higher peak concentrations because extracellular water is a smaller fraction of body mass. Concentration-guided titration removes the discrepancy; fixed-gram protocols do not.

  • Age adjustments: Past 70, protocols start at the low end (0.1–0.3 g/kg), extend infusion time, and set the dose by measured filtration rate rather than weight, because clearance and fluid tolerance both decline.

  • Baseline biomarkers that set the starting dose: Plasma ascorbate, creatinine-derived eGFR, ferritin and G6PD activity together determine both the ceiling and whether the target concentration is reachable safely.

  • Pre-existing conditions that reshape the protocol: Heart failure and ascites require the lowest practical sodium load and volume; diabetes requires laboratory glucose confirmation; myeloma on a proteasome inhibitor rules the protocol out entirely.

Discontinuation & Cycling

  • Not a lifelong intervention: Infusions are given in courses tied to active treatment, typically 8–24 weeks in trials. Some integrative clinics continue weekly maintenance for months or years while disease is controlled, without supporting outcome data.

  • No withdrawal syndrome: Stopping produces no recognised withdrawal state. Plasma returns to baseline within hours of any infusion regardless, so cessation is pharmacologically indistinguishable from a longer gap.

  • Rebound scurvy is not established: The concern that abrupt cessation after sustained megadosing triggers deficiency symptoms rests on isolated older reports and has not been reproduced with infusion protocols.

  • Tapering is optional, not pharmacological: Where used, clinics step down to 2–4 g oral daily rather than tapering the infusion itself, since the infusion leaves no accumulated depot to unwind.

  • Cycling is not shown to preserve efficacy: No human data support scheduled breaks. Laboratory work showing resistance can develop, with greater metastatic potential in resistant cells, argues for outcome-driven stopping rather than routine cycling.

  • Defined stopping triggers: Documented disease progression, a 25% creatinine rise, oxalate crystals on urinalysis, or any haemolysis ends the course rather than prompting a dose reduction.

Sourcing and Quality

  • Preservative-free injectable grade only: Retail oral ascorbic acid powder is not sterile, pyrogen-free or pH-adjusted, and infusing it has caused serious harm. Only ascorbic acid injection made to United States Pharmacopeia (USP) injectable standards is appropriate.

  • The one approved product: Ascor, from McGuff Pharmaceuticals, is the ascorbic acid injection approved by the Food and Drug Administration (FDA, the United States drug regulator) — approved for scurvy, not cancer, so oncology use is off-label but the material itself is regulated.

  • Compounded supply and its caveats: Registered outsourcing facilities (503B) such as Fagron Sterile Services and Merit Pharmaceutical supply large-volume vials. Compounded material is not FDA-approved, so a per-lot certificate of analysis matters more, not less.

  • What the certificate covers: Assay potency, sterility, bacterial endotoxin, particulate matter and heavy-metal content, plus confirmation the product is free of chelating preservatives such as EDTA that alter iron availability.

  • Buffered form and pH: Sodium ascorbate or sodium-buffered ascorbic acid at pH 5.5–7.0 avoids vein injury from acidic solutions, at the cost of the sodium load. Fully unbuffered ascorbic acid is not infused.

  • Diluent choice: Sterile water or Ringer’s lactate only, at an osmolarity the vein tolerates. Dextrose diluents compete with the oxidised form of vitamin C at cell transporters and undercut the mechanism.

  • Storage and visual check: Vials are light- and heat-sensitive, so they are refrigerated and light-protected, and darkened solution is discarded, since discoloration signals degradation to dehydroascorbic acid and oxalate.

  • Oral maintenance products between infusions: Liposomal and buffered mineral ascorbate formulations raise plasma more than plain ascorbic acid but remain orders of magnitude short of infusion levels; third-party tested products are the usual choice, and they function as repletion rather than therapy.

Practical Considerations

  • Time to effect: Symptom and fatigue changes, where they occur, appear within 2–4 weeks of starting a thrice-weekly schedule. Any effect on tumour burden is assessed on the oncology imaging schedule, typically at 8–12 weeks.

  • Pitfall — substituting oral for infused: The single most common error. Capsules cannot reach the concentration the mechanism requires, and the two negative Mayo trials tested exactly this substitution.

  • Pitfall — under-dosing: Doses below roughly 15 g rarely approach the millimolar range. A course of 10 g infusions tests nothing about the hypothesis while carrying the cost and inconvenience.

  • Pitfall — skipping the enzyme screen: Omitting the G6PD assay because a patient looks low-risk is how the reported fatal haemolysis cases happened. Ancestry is not a substitute for a measured result.

  • Pitfall — trusting a finger-prick glucose reading: Post-infusion strip readings are unreliable for hours, and insulin dosed from them has caused fatal hypoglycaemia.

  • Regulatory status: Injectable ascorbic acid is FDA-approved only for scurvy, so all oncology use is off-label. The FDA has issued warning letters to clinics marketing infusions as cancer treatments.

  • Cost and accessibility: Infusions typically run US$100–350 each and are almost never reimbursed, so a thrice-weekly course costs roughly US$15,000–50,000 annually out of pocket, plus 2–4 hours of chair time per session.

  • Payer incentives and structural bias: An off-patent infusion that displaced patented drugs would cut insurer and national-health-system spending, yet neither payers nor manufacturers fund the trials that could show it, leaving guideline formation and research funding tilted toward reimbursed patented agents.

  • Access route: Repeated large-volume infusions often require a central line or port, adding its own infection and thrombosis risk, and administration is largely confined to integrative clinics or trial sites.

Interaction with Foundational Habits

  • Sleep: Indirect. The osmotic diuresis from a 25–100 g load causes night-time urination if infused late, and the practical step is morning or early-afternoon scheduling. Where symptom relief occurs, better sleep follows secondarily; no direct effect on sleep architecture has been described at these doses.

  • Nutrition: Direct and two-sided. Large glucose loads around infusion compete with the oxidised form of vitamin C at the same cell transporters, so a low-carbohydrate window and a non-dextrose diluent are used. Iron-rich meals and iron supplements are separated from infusion days; adequate fluid and magnesium support oxalate handling.

  • Exercise: Blunting concern. Habitual high antioxidant intake can dampen the oxidative signalling that drives endurance adaptation, so hard training sessions are best kept off infusion days. Against that, less fatigue may preserve the capacity to train at all, which matters more during cancer treatment than adaptation does.

  • Stress management: Indirect. Ascorbate concentrates in the adrenal cortex and is consumed in catecholamine synthesis, so stress raises requirements, and infusions transiently flood that pathway. No clinical study has shown a change in cortisol or measured stress response at pharmacological doses.

Monitoring Protocol & Defining Success

Before the first infusion three results are non-negotiable: red-cell G6PD activity, kidney function, and a baseline plasma ascorbate level. A history of kidney stones, iron overload or heart failure changes the calculus and is established at the same visit, alongside a full blood count, serum electrolytes and whichever tumour markers apply to the specific cancer.

Once infusions begin, kidney function and electrolytes are rechecked before each cycle, with a full count and markers of red-cell breakdown 24–48 hours after each of the first two infusions. Thereafter the usual cadence is at 2 weeks, 4 weeks, then every 4–8 weeks while treatment continues, with plasma ascorbate measured mid-infusion whenever the dose changes. Imaging follows the oncology schedule for the underlying disease, not the infusion schedule.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
G6PD enzyme activity Normal activity, ≥60% of the laboratory’s reference mean Absent enzyme means infusion can rupture red cells Required before the first infusion; falsely normal during or after haemolysis, or after recent transfusion
Creatinine and eGFR eGFR ≥60 mL/min/1.73 m²; below 30 is a contraindication Kidney capacity to clear the oxalate load Before each cycle; fasting not required; a 25% rise from baseline stops the course
Plasma ascorbate Peak ≥20 mmol/L mid-infusion; trough 50–70 µmol/L Confirms the pharmacological threshold was actually reached Draw mid-infusion; light- and heat-sensitive, needs immediate stabilisation and freezing; few laboratories offer it
Urinary oxalate, 24-hour <40 mg per 24 hours Detects the oxalate accumulation that precipitates in tubules Optional; highest value with a stone history; collect on a non-infusion day for baseline
Haemoglobin and reticulocytes Stable within 1 g/dL of the individual’s own baseline Detects red-cell destruction after infusion 24–48 hours after each of the first two infusions; pair with LDH and bilirubin
Lactate dehydrogenase Within the laboratory reference range Rises when red cells or tumour cells rupture Haemolysed samples give false highs; interpret alongside haptoglobin
Ferritin and transferrin saturation Ferritin 50–100 ng/mL; saturation 25–35% Identifies the iron overload that amplifies oxidative damage Ferritin rises with inflammation and in cancer, so pair with CRP before interpreting
C-reactive protein No established target during active cancer; track the direction of change from the individual’s own baseline The inflammation marker that moved in the clinic series Conventional laboratories report <5 mg/L as normal, far looser than the <1 mg/L functional target used outside oncology
Sodium and potassium Sodium 135–142 mmol/L; potassium 4.0–4.5 mmol/L Tracks the salt load delivered with buffered ascorbate Draw before infusion; most important in heart failure, ascites and advanced kidney disease
Uric acid, phosphate, calcium Uric acid 3.5–5.5 mg/dL; phosphate 2.5–4.0 mg/dL; calcium 9.0–10.0 mg/dL Screens for the metabolic release of tumour lysis Check before and 24 hours after the first infusion in bulky or highly treatment-sensitive disease
Disease-specific tumour markers No universal target; track the trajectory against the individual’s own pre-treatment value The only direct read on whether disease burden is moving Conventional reference ranges are set for screening, not for tracking treated disease; interpret trends over three or more points

Qualitative markers are tracked alongside the laboratory panel, since symptom change is the most consistently reported effect:

  • Fatigue severity, scored the same way each week rather than recalled
  • Appetite and weight trajectory
  • Pain scores and analgesic requirement
  • Sleep continuity, particularly night-time urination on infusion days
  • Cognitive clarity, including the transient altered mental status reported after infusion
  • Nausea, constipation and mood
  • Tolerability of the infusion itself: vein discomfort, thirst, tremor, post-infusion lethargy

Success is defined as reaching the target plasma concentration without a kidney or haemolysis signal, plus a measurable improvement in symptom scores, while the underlying disease responds at least as well as expected on standard therapy alone. Failure to reach the concentration, or any kidney or red-cell signal, ends the attempt regardless of how the patient feels.

Emerging Research

  • Phase 2 ascorbate in glioblastoma: NCT02344355 enrolled 90 patients with overall survival as the primary endpoint, the direct follow-on to the first-in-human trial. Active and no longer recruiting, so it is the nearest thing to a definitive answer in brain tumours.

  • Ascorbate with chemoradiation in lung cancer: NCT02905591, a phase 2 study in 43 patients with non-small cell lung cancer, uses progression at completion of radiation and chemotherapy as its primary measure. Active and no longer recruiting.

  • Phase 3 pairing with immunotherapy: NCT04516681 is recruiting 400 patients with metastatic colorectal cancer to infused ascorbic acid plus adebrelimab, with objective response rate as primary endpoint — the first large test of the immune-modulation hypothesis.

  • High-dose ascorbate with a hypomethylating drug in myeloid disease: NCT07283900, a phase 2 trial in 38 adults with myelodysplastic syndrome (a bone-marrow disorder that can progress to leukaemia), pairs 75 g infusions with azacitidine, a drug that strips the same chemical tags off DNA, testing that mechanism directly with response rate as its endpoint.

  • Ascorbate added to azacitidine and venetoclax in leukaemia: NCT07177079 is a phase 1 study in 30 patients with acute myeloid leukaemia (a fast-growing blood cancer), reporting dose-limiting toxicities and complete remission rate.

  • Bladder-sparing chemotherapy substitution: NCT06493370 is recruiting 48 patients with muscle-invasive bladder cancer to ascorbate plus gemcitabine and carboplatin, with post-treatment pathological staging as the endpoint.

  • Dose–exposure optimisation: NCT07121036, a phase 1 pharmacokinetic study in 18 patients with advanced solid tumours, is designed to fix the target-concentration problem that plausibly explains why fixed-dose trials disagree.

  • Confirmation of the pancreatic survival signal: Bodeker et al., 2024 randomised only 36 patients, of whom 34 were treated. Replication at scale would strengthen the case; failure to replicate would leave the phase 3 colorectal null result as the best evidence.

  • Resistance and metastasis as a downside signal: Pope et al., 2025 showed ascorbate-resistant pancreatic cells gained metastatic potential. If that translates clinically, prolonged exposure could worsen outcomes, and this is the most credible route to weakening the case.

  • Mechanism-matched patient selection: Yun et al., 2015 predicted the RAS-mutant response later seen in the phase 3 subgroup. Prospective enrichment for that genotype is the untested design most likely to change the current picture.

Conclusion

High-dose vitamin C given through a vein is not the same intervention as a vitamin C tablet. Only the infused route reaches blood levels at which the molecule stops acting as a nutrient and starts producing a damaging chemical inside tissue, and tumour cells appear less able to clear it than healthy cells. That difference explains why the question was treated as closed for four decades on the strength of studies that used tablets, and why it reopened once the gap in blood levels was measured directly.

The human evidence supports most firmly that infusions reach the intended blood levels, that people on demanding cancer treatment tolerate them, and that symptoms and side effects were milder in small studies. Whether they lengthen life is unsettled. Summaries combining several studies, and one carefully controlled pancreatic trial, point one way; the single large controlled trial in bowel cancer points the other. The studies are small, concentrated in a few centres, and mostly ran without concealing who received the infusions.

The safety picture is narrow but sharp. A missing protective enzyme in red blood cells, existing kidney disease, iron overload, and one widely used blood-cancer drug turn a well-tolerated infusion into a serious hazard, and finger-prick sugar meters read falsely high for hours afterwards.

Interest cuts both ways. Clinics that bill for infusions produced much of the supportive case material, while the absence of a patent holder means no company has funded the large trials that could settle it.

Top - Benefits - Risks - Protocol