N-Acetylcysteine IV Therapy: Mechanism, Evidence and the Repair Step the Vein Skips

In short: N-acetylcysteine is the one antioxidant precursor on an infusion menu with an approved intravenous product behind it, and the indication on that product is poisoning. The licensed course is 300 mg/kg — roughly 21,000 mg for a 70 kg adult, delivered over 20 or 21 hours, to prevent hepatic injury after a paracetamol overdose. A 600 mg bag is under three per cent of that, given once, for a different purpose, with no trial behind it in either direction. The route argument inverts for this molecule. NAC oxidises in solution to the disulfide dimer N,N-diacetylcystine; the small intestine reduces that dimer back and recovers the cysteine, while the bloodstream does not — so what has oxidised in the vial is recoverable orally and lost intravenously. The longevity evidence is contested and entirely oral: a 24-person trial reported reversed ageing markers, an independent 114-person trial missed its primary endpoint. The question a clinic can actually act on is not which route, but what is in the ampoule and how much of it is still the active form.

The route argument, and the one place it runs backwards

The first half of the usual case for infusing N-acetylcysteine is sound, and unusually well quantified. Oral NAC has a bioavailability of roughly 4 to 10 per cent — published pharmacokinetic work puts reduced NAC near 4 per cent and total NAC near 9 per cent — because the intestinal mucosa deacetylates a large share of the dose on first pass. A molecule that loses nine-tenths of itself before reaching the circulation is a fair candidate for a vein.

The second half is where supplier literature stops reading. NAC carries a free thiol group, and a free thiol in an aqueous solution exposed to oxygen does one thing reliably: it oxidises, pairing with another molecule of itself to form the disulfide dimer N,N-diacetylcystine. This is not a degradation curiosity. In stability work on standardised paediatric parenteral nutrition, the loss of NAC and the appearance of the dimer matched to within one per cent — the molecule does not vanish, it converts.

What matters commercially is what happens to that dimer next, and the answer differs by route. Taken by mouth, the dimer is reduced during passage through the small intestine and the cysteine is recovered; the gut is a repair step as well as a loss mechanism. Given intravenously, that step is absent, and the published parenteral nutrition literature describes the dimer as almost unconvertible to cysteine, with a correspondingly reduced delivery of the thing the product was bought for.

So the honest version of the route argument is narrower and considerably more useful to a buyer. The intravenous route delivers more of what remains intact in the vial — and none of what has already oxidised inside it. Oral dosing wastes most of the dose but forgives the oxidation. Intravenous dosing wastes none of it but forgives nothing. That trade has a practical consequence, and it is a procurement consequence rather than a clinical one.

Cysteine is the rate-limiting step, not glutathione

Glutathione is a tripeptide of glutamate, cysteine and glycine, assembled in two ATP-dependent steps. The first, catalysed by glutamate-cysteine ligase, is the controlled one, and the substrate that most often runs short is cysteine. That single fact is the whole mechanistic case for NAC: the acetyl group makes cysteine stable enough to formulate and transport, and intracellular deacetylation releases the amino acid where it is needed.

It is also the reason a precursor and the finished tripeptide are not interchangeable products. Infused glutathione raises plasma glutathione, but the enzymes that use it — glutathione-S-transferases in conjugation, glutathione peroxidase in peroxide disposal — work inside cells, and the cell largely builds its own supply rather than importing it intact. We set out that distinction in the glutathione article; NAC approaches the same compartment from the substrate side, which is a genuinely different pharmacological proposition even when the marketing copy is identical.

Two qualifications belong here rather than in a footnote. The first is that NAC's effects are not exclusively glutathione-mediated: it is also a direct thiol reductant and a source of sulfane sulfur species, and a mechanistic review published in Pharmacology & Therapeutics in 2021 argues that hydrogen sulfide chemistry accounts for part of what has historically been attributed to glutathione repletion. The second is that a precursor only helps where the substrate is limiting. In a person whose cysteine supply is adequate, adding more does not raise what is already saturated — which is why the trials below separate so cleanly into depleted populations where NAC works and replete populations where it does not.

What the licensed labels actually say

NAC belongs to the small group of molecules in this field where a buyer can read an approved indication rather than infer one, and the indication is narrow.

The FDA-approved label for Acetadote, acetylcysteine injection 200 mg/mL under NDA 021539 in the version revised July 2025, states that the product is indicated to prevent or lessen hepatic injury after ingestion of a potentially hepatotoxic quantity of paracetamol, in adults and in paediatric patients of 5 kg and above. The dose is 300 mg/kg in total: a three-bag regimen of 150 mg/kg over 1 hour, 50 mg/kg over 4 hours and 100 mg/kg over 16 hours; or, for patients of 41 kg and above, a two-bag regimen of 200 mg/kg over 4 hours followed by 100 mg/kg over 16 hours. The concentrate is hyperosmolar at 2,600 mOsmol/L and must be diluted before infusion, and the label requires the total volume to be adjusted below 40 kg because of the risk of hyponatraemia and seizures.

The UK licensed products say the same thing with different packaging arithmetic. The Summary of Product Characteristics for the 200 mg/ml concentrate sets out three consecutive infusions — 150 mg/kg in 200 mL over one hour, 50 mg/kg in 500 mL over four hours, 100 mg/kg in 1 litre over sixteen hours — with the indication tied to the paracetamol treatment line, staggered ingestion or uncertain timing. The World Health Organization lists acetylcysteine on its Model List of Essential Medicines as a specific antidote, in a 200 mg/mL, 10 mL ampoule.

There is a regulatory footnote that matters more to a distributor than to a clinician. In the United States NAC is excluded from the statutory definition of a dietary supplement, because it was approved as a new drug in 1963 before any supplement or food marketing; the FDA sent warning letters on that basis in 2020 and then, in final guidance issued in August 2022, stated it would exercise enforcement discretion for certain NAC products labelled as supplements — a discretion that does not extend to products intended to treat disease. In the United Kingdom and the European Union, injectable acetylcysteine is a prescription medicine. The same molecule, three legal categories, and the determining factor is often the claim attached to it rather than the chemistry.

The arithmetic a 600 mg bag has to answer

Put the licensed figure and the commercial figure in the same sentence and the gap is not subtle. A 70 kg adult treated under either approved regimen receives about 21,000 mg of acetylcysteine inside 21 hours. A wellness infusion carrying 600 mg delivers 2.9 per cent of that, once. At 1,200 mg it is 5.7 per cent.

The comparison is deliberately uncomfortable in both directions, and it is worth being precise about why. The licensed dose is not an antioxidant dose and was never calibrated as one. It is a substrate-replacement dose sized to a specific chemical insult: enough cysteine to let the liver rebuild glutathione faster than NAPQI, the reactive paracetamol metabolite, can consume it. Transplanting that figure into an elective menu would be as unexamined as the number it replaced. Equally, quoting 600 mg because that is the capsule size in the oral literature imports a dose from a route that behaves differently and from regimens that ran daily for weeks rather than once for an hour.

The defensible position is narrower than either. The oral trials that found anything used 600 to 1,800 mg per day over sustained periods; the intravenous evidence base exists only at 300 mg/kg for poisoning. Between those two sits a single infusion of a few hundred milligrams with no trial behind it in either direction — not shown to work, not shown to fail, simply untested. A clinic can defend that position if it states it. A protocol sheet that quotes a figure without naming where it came from is doing something else.

What the randomised evidence found

Sorted by claim, with the study that tested it and the result as published.

The shape will be familiar to readers of this series. Where NAC replaces a substrate that is genuinely depleted, it works so clearly that a placebo arm is no longer ethical. Where it was extended into populations that were not depleted, the properly powered test was eventually run and came back empty — PRESERVE with 5,177 patients being the cleanest example, and the high-dose vitamin C literature having followed the same arc a few years earlier.

GlyNAC: two randomised trials, two answers

The longevity case for NAC rests on supplying both ends of the glutathione tripeptide at once — glycine and cysteine — and it is the one place in this molecule's literature where the evidence has moved since we last wrote about it.

The amino acid article reported the Baylor College of Medicine trial: 24 older adults, 16 weeks, glycine and N-acetylcysteine at 100 mg/kg each against an isonitrogenous alanine placebo, with corrected glutathione deficiency, lower oxidative stress and inflammatory markers, better mitochondrial function, and improved gait speed and muscle strength. We noted then that it had not been replicated at scale. It since has been attempted, by an independent group, and the result pointed the other way.

A randomised controlled trial published in Frontiers in Aging in 2022 gave 114 adults aged 60 to 85 either 2.4, 4.8 or 7.2 g daily of a 1:1 glycine and NAC combination, or an isomaltulose placebo, for two weeks. The primary endpoint — total glutathione in whole blood — was not met: 959.6 mg/L on the highest dose against 903.5 mg/L on placebo, p=0.278, with no change in the reduced-to-oxidised ratio and none in malondialdehyde. A post-hoc subgroup defined by above-median oxidative stress and below-median baseline glutathione showed a 10.47 per cent rise on the medium and high doses. The authors' own summary is appropriately hedged: supplementation is safe, well tolerated, and may raise glutathione in older adults with high glutathione demand.

These two trials are not a clean contradiction and should not be presented as one. Two weeks against sixteen is a real difference, and the populations were selected differently. What they jointly establish is narrower and more honest than either headline: the effect, if it exists, is not large enough or fast enough to appear in a fortnight in an unselected older population, and it may depend on starting depleted. That is a hypothesis about who, not a protocol about how much. Both trials were oral. No randomised trial of intravenous NAC for this purpose exists.

One inconvenient direction deserves naming, because the category tends to treat antioxidant as a synonym for harmless. Antioxidant supplementation including NAC accelerated tumour progression in mouse models of lung cancer in work published in Science Translational Medicine in 2014, and increased lymph node metastases in a mouse melanoma model in the same journal in 2015. A paper published in International Journal of Molecular Sciences in 2026 reported that NAC at 150 mg/kg intraperitoneally every two days increased lung metastasis in immunocompetent mice by reprogramming neutrophils into an immunosuppressive phenotype, an effect that disappeared in T-cell-deficient animals. These are mice, and mouse tumour models do not transfer to humans. They are, however, the reason that more antioxidant is better is not a safe default assumption, and the reason the reasonable question remains whether a given person is depleted.

The infusion reaction is arithmetic, not allergy

NAC produces one of the better-documented adverse reactions in intravenous practice, and it is widely misdescribed in both directions — treated as a trivial flush by some and as a drug allergy by others.

The labels call it anaphylactoid: flushing, rash, pruritus, nausea and vomiting, and less often angioedema, bronchospasm and hypotension, typically starting 15 to 60 minutes after the loading infusion begins. The FDA-approved label records hypersensitivity reactions in 17 per cent of patients in clinical trials, with a postmarketing range of 1 to 19 per cent depending on regimen and institution, and notes that fatal and life-threatening reactions have occurred. The UK Summary of Product Characteristics describes the practical management: suspend the infusion, give supportive care, and restart at the slower rate — 50 mg/kg over four hours — once the reaction is controlled.

Two randomised results together identify the variable, and they are more instructive read as a pair than separately. Lengthening the loading infusion from 15 minutes to 60 minutes, tested in 180 patients in the Australasian trial published in 2005, did not reduce drug-related adverse outcomes; the UK label reports 18 and 14 per cent for the two arms. The SNAP trial published in the Lancet in 2014 did something different: it lowered the loading dose itself to 100 mg/kg over two hours — halving milligrams per kilogram per hour rather than merely stretching the clock — and added ondansetron. Adjusted odds ratios were 0.26 for vomiting and 0.23 for anaphylactoid reactions, both p<0.0001, with no difference in liver enzyme rise. A meta-analysis of two-bag against three-bag regimens published in Diseases in 2024, covering nine studies and 11,835 patients, found anaphylactoid reactions at an odds ratio of 0.33 with hepatotoxicity unchanged.

The pattern is consistent with direct, concentration-dependent histamine release rather than with immune sensitisation, which is why it can occur on first exposure and why the fix turned out to be arithmetic. For a clinic the operational content is simple and non-negotiable: this is not a molecule to push faster, and any presentation quoted without its intended dilution and infusion rate is incompletely specified.

The dimer, and what a certificate of analysis must separate

This is where the first section's chemistry becomes a purchasing specification, and it is the question most likely to be missed on a supplier call.

Because NAC converts to N,N-diacetylcystine rather than disappearing, a content assay that measures total NAC-related material can pass a batch in which a meaningful fraction is no longer useful by the intravenous route. The published numbers give a sense of the timescale. A stability study of parenteral N-acetylcysteine published in Pharmaceuticals in 2023 found that without additives the dimer stayed below 0.5 per cent for only one day at room temperature and four days under refrigeration, and that zinc gluconate at 62.5 micrograms per millilitre held a 25 mg/mL solution stable for at least eight days at 5 degrees Celsius. The paediatric parenteral nutrition work published in Nutrients in 2020 found that oxygen absorbers cut headspace oxygen by about 30 per cent within two and a half hours and substantially limited dimer formation — which makes this a container and headspace question before it is a refrigerator question, the same conclusion the cold chain and storage article reached from a different direction.

So the specification request is concrete. Ask for an assay method that resolves NAC from N,N-diacetylcystine, a stated limit for the dimer rather than a single content figure, and the container, headspace and storage condition under which the stability data were generated. A supplier who can produce those three things has characterised the product. One who answers with purity alone has described a mixture.

The excipient question is equally specific, and the two licensed products diverge on it. The UK Summary of Product Characteristics for Parvolex lists disodium edetate among its excipients. Acetadote has been free of EDTA and of any other stabilising or chelating agent since a reformulation first supplied in 2011, developed as a post-approval commitment after the FDA asked that EDTA be reduced or removed. Two approved products, the same active, opposite answers on a single excipient — which is sufficient reason to treat the phrase pharmaceutical grade NAC as describing nothing in particular until the full composition is on the page.

Before the order goes out

Four questions specific to this molecule, each of which a supplier should answer in writing.

One — which assay, and does it separate the dimer. A certificate of analysis reporting NAC content without a specified limit for N,N-diacetylcystine does not tell a buyer how much usable cysteine precursor the vial will deliver by the intravenous route. This is the one question on this list that has no equivalent for most other actives.

Two — the full composition, not the active. Chelating agent or not, pH, sodium load and osmolarity all belong on the sheet. The 200 mg/mL licensed concentrates declare substantial sodium content and an osmolarity around 2,600 mOsmol/L, which is why dilution is a label requirement rather than a preference, and why a concentration quoted without its intended diluent describes a solution nobody infuses.

Three — the storage condition the data were generated under. Given how quickly an unprotected thiol solution dimerises, a shelf life stated without the container, the headspace and the temperature it was measured in is not a shelf life. Ask what was measured, in what, for how long.

Four — who made it. For a parenteral thiol the specification, the stability data and the batch record sit with whoever ran the process and controlled the oxygen. EFBA manufactures the IVIXIR series in its own GMP-certified facilities, and the manufacturer and wholesale page sets out how that supply side is organised. N-acetylcysteine appears as a declared component at a stated concentration in the IVIXIR Anti Aging Repair Solution CLG concept, alongside L-lysine, L-proline, L-arginine, carnosine, magnesium chloride, selenium and vitamin C — the composition is set out on the corresponding protocol sheet rather than implied.

None of this argues against N-acetylcysteine. It argues that this molecule rewards reading the paperwork more than most, because it is the rare case where the active ingredient can be present, correctly labelled and partly unavailable at the same time. IVIXIR concepts are defined sets of molecules at stated concentrations for professional use — not medicinal products, and not treatments for any condition — and decisions about indication and patient belong to the practitioner. What can be specified can be assessed, and with NAC the specification has one more line in it than most buyers ask for.