Isoniazid

Why this poison is interesting

Isoniazid is the overdose that breaks the anticonvulsant reflex. A seizing patient gets a benzodiazepine, and usually that is enough — but isoniazid's seizures resist it, and the reason is not that they are unusually violent. It is that benzodiazepines work through GABA, and isoniazid has starved the brain of GABA by depleting the vitamin needed to make it. The treatment that works is the vitamin, not more sedation.¹³

• Isoniazid depletes the active form of vitamin B6. The drug and its hydrazine metabolites inhibit the enzyme that activates pyridoxine and combine with pyridoxal-5-phosphate directly, so the active cofactor falls.¹³ Established
• That cofactor makes GABA. Pyridoxal-5-phosphate is required by glutamic acid decarboxylase, the enzyme that synthesises the brain's main inhibitory neurotransmitter. Deplete the cofactor and GABA synthesis falls — inhibition is lost and seizures follow.¹³ Established
• So the seizures are substrate-starved, and pyridoxine is the answer. Replacing pyridoxine refills the cofactor, restores GABA synthesis and terminates the seizures — a defined deficiency with a defined replacement.¹³ The seizures that would not stop on a benzodiazepine stop on the vitamin.

Isoniazid's seizures are not a sedation problem but a supply problem: the brake is intact, the brake fluid is gone.

The toxic principle

The toxic principle is a functional vitamin B6 deficiency, produced fast and deliberately by the drug, and everything clinical follows from the GABA it prevents the brain from making.

• Isoniazid inhibits the activation of pyridoxine and inactivates its active form. It reduces the enzyme activity that phosphorylates pyridoxine to pyridoxal-5-phosphate, and its hydrazine metabolites combine with pyridoxal-5-phosphate to form hydrazones that are inactive and more readily excreted — so the pool of active cofactor falls on two fronts.¹³ Established
• Pyridoxal-5-phosphate is the cofactor for GABA synthesis. Glutamic acid decarboxylase needs it to convert glutamate to GABA; without enough cofactor, GABA synthesis drops, cortical inhibition is lost, and seizures — often status epilepticus — result.¹³ Established
• A lactic acidosis accompanies the seizures. Repeated convulsions generate lactate, and isoniazid additionally interferes with the NAD-dependent conversion of lactate to pyruvate, so a high-anion-gap metabolic acidosis is characteristic and tends to resolve as the seizures are controlled.² Inferred

Toxicokinetics

Metabolism and the metabolites

Isoniazid's metabolism is unusual in that it is part of the toxicity rather than merely its clearance. Acetylation is the main route and is polymorphic, but the hydrazine-bearing metabolites formed along the way are among the species that inactivate pyridoxal-5-phosphate — so the drug is depleting the cofactor both as the parent molecule and through what it becomes.

What changes the answer

• The amount ingested sets how much cofactor is depleted and therefore how much pyridoxine is needed — the antidote requirement scales with the poison.³ Established
• Acetylator status alters how long the parent drug persists, and so the duration of the insult.² Inferred
• Pyridoxine availability is a real-world modifier: the antidote is effective but stocked in limited quantities, and a large overdose can outstrip local supplies.³
• Co-ingestants that lower the seizure threshold or add their own acidosis worsen the picture.

Elimination and accumulation

Isoniazid is cleared quickly — acetylated in the liver and excreted in the urine, with over 90% of a dose gone within 24 hours and most of it in the first 12.² The drug does not accumulate in the way a renally-dependent poison does; the danger is front-loaded, arriving fast and resolving as the drug clears and the cofactor is replaced.

The EXTRIP position

EXTRIP makes two recommendations, both weak and on very low quality evidence.¹ The first is the general case; the second is a narrow exception.

• Against dialysis in addition to standard care in severely poisoned patients — the antidote and supportive care suffice, and the drug self-clears within hours.¹
• For dialysis only in the rare circumstance where standard-dose pyridoxine cannot be administered and seizures are refractory to GABA-A receptor modulators.¹ This is a fallback for an antidote-supply failure, not a primary treatment.
• The reasoning is speed, not plausibility. Isoniazid is small and unbound and would dialyse; EXTRIP recommends against because pyridoxine acts immediately and the drug is largely gone within hours, so removal rarely earns its place.¹²

Target organs — and why those

Timeline of effects

Isoniazid's timeline is fast and front-loaded. There is a short gap while the drug is absorbed, then the triad arrives together — seizures, acidosis and coma — and the illness resolves as the drug clears and pyridoxine replaces the cofactor. There is no slow metabolic tail; the danger is early and sharp.

What the mechanism predicts at the bedside

Why the seizures are refractory — and what that tells you

The seizures resist standard anticonvulsants because those drugs potentiate GABA, and isoniazid has depleted the GABA there is to potentiate.¹³ So refractoriness is not a sign to keep escalating sedation — it is the diagnostic clue that the problem is a cofactor deficiency, and the prompt to give pyridoxine, which treats the cause. Established

Why pyridoxine is a real antidote, not a supplement

Pyridoxine is the precursor of the exact cofactor isoniazid depletes; replacing it restores GABA synthesis and terminates the seizures.¹³ Because the antidote requirement scales with the isoniazid ingested, a large overdose needs a large quantity of pyridoxine — enough that supply, not just dosing, can become the limiting factor.³ The amount and route are a TOXBASE and NPIS matter; the principle is that the dose answers the poison. Established

Why the acidosis is treated at its source

The high-anion-gap acidosis is largely lactic and is driven by the seizures and by isoniazid's interference with lactate handling,² so it tends to correct once the seizures are controlled. Bicarbonate addresses the acidaemia itself, but the durable fix is stopping the convulsions — which is, again, pyridoxine.

Why dialysis is usually unnecessary despite the favourable chemistry

Isoniazid is small and unbound, so it would dialyse — but it self-clears within hours and pyridoxine stops the seizures at once, so extracorporeal removal rarely adds anything and EXTRIP recommends against it.¹² The only scenario that reverses this is the one where the antidote cannot be given: pyridoxine unavailable and seizures refractory, where dialysis becomes a fallback.

The antidote, from the poison's side

Isoniazid poisoning has one of the cleanest antidote stories in toxicology: a drug that depletes a specific vitamin cofactor, and the vitamin itself as the treatment. Pyridoxine works at the precise point of the lesion, which is why it succeeds where sedation struggles.

• Pyridoxine replaces the depleted cofactor, restoring GABA synthesis and terminating the seizures — the mechanism-directed antidote for the central lesion.¹³ Established
• Benzodiazepines are adjuncts, not the answer, because they act through a GABA system the poison has starved; they may help but do not correct the deficiency.¹³
• Sodium bicarbonate treats the acidaemia, but the acidosis resolves mainly as the seizures are controlled, so pyridoxine addresses its source too.²
• Dialysis is a fallback for an antidote-supply failure, not a primary treatment — reserved for refractory seizures when pyridoxine cannot be given.¹

Critical appraisal

1. The pyridoxine-depletion account of the seizures is well established and clinically load-bearing.¹³ It explains the refractoriness, the choice of antidote, and why the antidote dose scales with the ingestion. It is the firmest element on the page and it drives the entire management.
2. The GABA-synthesis link rests on standard neurochemistry. Established Pyridoxal-5-phosphate is the accepted cofactor for glutamic acid decarboxylase, and its depletion lowering GABA is textbook — but the quantitative contribution of cofactor depletion versus other isoniazid effects in an individual seizure is not separable at the bedside, and the page states the mechanism rather than measuring it.
3. The lactic-acidosis mechanism is partly inferred. Inferred Seizure-generated lactate is uncontroversial; isoniazid's specific interference with lactate-to-pyruvate conversion is the standard additional explanation but is not quantified here, so the acidosis is attributed to both the seizures and the drug without apportioning them.
4. The antidote evidence is strong in direction but not trial-grade. The supporting citation is a case report illustrating both the effectiveness of intravenous pyridoxine in status epilepticus and the real-world supply constraint;³ there is no randomised trial, and there will not be one. The mechanism and consistent clinical experience carry the recommendation.
5. The EXTRIP recommendations are weak and on very low quality evidence,¹ and the page presents them as such. The against recommendation is an argument from speed (antidote faster, drug self-clearing), not from an inability to dialyse — a distinction an auditor should preserve, because it is the reason the narrow for exception exists.
6. Acetylator status is real but secondary in acute overdose. It alters clearance and duration,² but in a large acute ingestion the amount taken dominates, so the page treats it as a modifier rather than a determinant.

References

1. 1
   EXTRIP Workgroup. Extracorporeal treatment for isoniazid poisoning. Recommendation set at extrip-workgroup.org/isoniazid. Source of the weak recommendation, on very low quality evidence, against performing extracorporeal treatment in addition to standard care in severe isoniazid poisoning, and of the narrow exception — a weak suggestion to perform extracorporeal treatment only in patients with seizures refractory to GABA-A receptor modulators in the rare circumstance where standard-dose pyridoxine cannot be administered. Read from the workgroup's published recommendation page; verified 13 Sep 2026.
2. 2
   Isoniazid 100 mg Tablets BP — Summary of Product Characteristics. emc product 9139. §4.9 and §5.2 fetched and read in full. Source of the overdose features (nausea, vomiting, CNS toxicity including vertigo, seizures and coma), the treatment with anticonvulsants, large doses of intravenous pyridoxine and sodium bicarbonate for acidosis, the complete oral absorption, the free distribution into tissues and CSF, the absence of protein binding, the acetylation metabolism with its named metabolites and rapid/slow acetylator polymorphism, the 1.2 h (rapid) and 3.5 h (slow) half-lives, and the >90% urinary excretion within 24 hours. Verified 13 Sep 2026.
3. 3
   Morrow LE, Wear RE, Schuller D, Malesker M. Acute isoniazid toxicity and the need for adequate pyridoxine supplies. Pharmacotherapy 2006;26(10):1529–32. PubMed 16999664. Case report of status epilepticus after a large isoniazid ingestion, managed with intravenous pyridoxine. Source of the account that intravenous pyridoxine is the effective antidote for isoniazid-induced status epilepticus, that the antidote is required in large quantity, and that a single overdose can exhaust the regional intravenous pyridoxine supply, making collaborative stockholding necessary. Verified 13 Sep 2026 from the abstract.
4. 4
   TOXBASE — isoniazid. National Poisons Information Service. toxbase.org (NHS login required. NPIS: 0344 892 0111. The authoritative UK source for pyridoxine indications and dosing, anticonvulsant choice and management of the acidosis. Login-gated, so not quoted here.)