Why this poison is interesting

Gabapentin and pregabalin are named after a neurotransmitter neither of them binds. They were designed as GABA analogues, and the shape survived into the marketed molecules while the pharmacology did not: both act at the alpha-2-delta auxiliary subunit of voltage-gated calcium channels, reducing the trafficking of those channels to the nerve terminal and the calcium-dependent release of transmitter that follows. The names are a fossil of a hypothesis that failed.

That would be a footnote if the two drugs then behaved alike. They do not, and the difference is the most useful thing on this page. Pregabalin is absorbed almost completely and in proportion to the dose swallowed. Gabapentin is not — its bioavailability tends to decrease with increasing dose², because it depends on a saturable transporter. One of these drugs has a built-in ceiling on how much of an overdose can reach the blood. The other has none.

The second reason this page exists is that the EXTRIP workgroup has published a real recommendation for it rather than an absence — which most pages in this library cannot say, and which puts it in the small company of lithium, thallium and the other assessed poisons. The recommendation is also unusual in its own right. It does not turn on how much was taken. It turns on whether the patient's kidneys work, and it points in opposite directions depending on the answer.³

A poison is a drug whose kinetics have escaped its pharmacology.

The toxic principle

There is no toxic metabolite, no bioactivation and no reactive species anywhere in this poisoning. The toxic principle is the drug, doing exactly what it does therapeutically, to more of the nervous system than intended. Pregabalin undergoes negligible metabolism in humans, with about 98% of the recovered radioactivity in urine being unchanged pregabalin and the N-methylated derivative accounting for 0.9% of the dose.¹

The clinical picture in overdose follows from a diffuse reduction in excitatory transmitter release rather than from any single receptor. The pregabalin label lists somnolence, confusional state, agitation and restlessness as the commonest post-marketing overdose reactions, with seizures also reported and coma in rare occasions.¹ Gabapentin's list is similar in kind — dizziness, double vision, slurred speech, drowsiness, loss of consciousness, lethargy and mild diarrhoea — and it adds the observation that in overdoses up to 49 g acute, life-threatening toxicity has not been observed, with all patients recovering fully on supportive care.²

The two agents diverge on the only property that matters for how much drug arrives. Pregabalin's oral bioavailability is estimated to be ≥ 90% and is independent of dose, with peak concentrations within 1 hour and low inter-subject variability (< 20%).¹ Gabapentin's absolute bioavailability is approximately 60% for a 300 mg capsule and tends to decrease with increasing dose.² A tenfold gabapentin ingestion does not produce a tenfold exposure; a tenfold pregabalin ingestion very nearly does. Established

Toxicokinetics

Pregabalin is used as the reference agent here because it is the one with no protective mechanism, the one whose label carries the dialysis figure, and the one implicated in the UK misuse problem.

Pregabalin — a drug with almost no kinetic complexity, and one dominant variable
Parameter	Therapeutic	In overdose	Why it changes
Absorption	Rapid when fasted; peak within 1 hour; bioavailability ≥ 90%, independent of dose¹	Unchanged in kind — there is no transporter to saturate	The absence of a ceiling is the whole difference from gabapentin. Food reduces Cmax by approximately 25–30% and delays tmax to about 2.5 hours¹, which blunts a peak without reducing the total dose absorbed.
Distribution	Apparent Vd about 0.56 L/kg; not bound to plasma proteins¹	Unchanged	A small, water-like volume with no binding is the profile of a molecule an extracorporeal circuit can reach. That is why the dialysis figure below is as good as it is, and why it is still not a reason to dialyse most patients.
Metabolism	Negligible; about 98% of the recovered radioactivity in urine is unchanged pregabalin; the N-methyl derivative is 0.9% of the dose¹	Unchanged — there is nothing to overwhelm	No bioactivation, no enzyme polymorphism, no metabolic interaction. The label notes hepatic impairment would not be expected to alter concentrations significantly.¹ A poison with no metabolism has no latent phase, and this page draws none.
Elimination	Renal, as unchanged drug; clearance directly proportional to creatinine clearance¹	Falls in exact proportion to the fall in GFR	This is the sentence the rest of the page hangs on. Renal function is not a modifier of this poisoning; it is the poisoning's only independent variable.
Half-life	Mean 6.3 hours¹	Prolonged in renal impairment, and only there	Six hours is short. A patient with working kidneys is on a predictable downward curve from the moment of peak; a patient without them may have no meaningful curve at all until dialysis provides one.
Linearity	Linear across the recommended dose range, inter-subject variability < 20%¹	Predictable	The label states that multiple-dose kinetics are predictable from single-dose data and that routine concentration monitoring is unnecessary.¹ Predictability is a therapeutic virtue that becomes a toxicological liability: nothing limits what a large ingestion delivers.
Dialysability	—	Yes. Four hours of haemodialysis reduces plasma pregabalin by approximately 50%¹; dose supplementation after dialysis is required in routine practice¹	EXTRIP suggests against ECTR in severe poisoning with normal kidney function, and for it where kidney impairment coexists — especially with coma requiring mechanical ventilation. Both are weak recommendations on very low quality evidence; where ECTR is used, intermittent haemodialysis is recommended over any other modality (strong recommendation), stopping on clinical improvement (strong recommendation).³ The removal is real. The argument is about whether removal changes anything in a patient whose own kidneys would clear the drug in a day.

Metabolism and the metabolites

This is the shortest section of any page in the library, and the brevity is the point. There are no metabolites worth naming. Pregabalin's single named human metabolite accounts for 0.9% of the dose.¹ Gabapentin is not metabolised in humans to any appreciable extent and is eliminated unchanged by the kidney.

Pregabalin — a pathway that is almost entirely a straight line

Pregabalin, swallowedAlready the toxic species. Bioavailability ≥ 90%, independent of dose¹
Absorption — no saturable stepPeak within 1 hour fasted; food delays tmax to ~2.5 h and cuts Cmax by 25–30% without reducing the total absorbed¹
Central nervous systemThe alpha-2-delta subunit. Somnolence, confusion, agitation, restlessness; seizures reported; coma rarely¹
Unchanged pregabalin in urine — about 98% of what is recoveredThe whole elimination story. Clearance directly proportional to creatinine clearance¹
N-methyl pregabalin — 0.9% of the doseThe major urinary metabolite, and toxicologically irrelevant at that share¹

Elimination and accumulation

Both drugs leave by one route, and both accumulate whenever that route narrows. This produces a second, quieter kind of gabapentinoid toxicity that has nothing to do with overdose: the patient taking a licensed dose whose kidney function has fallen.

This is why EXTRIP's recommendation is split the way it is. In a patient with normal kidneys, the drug's own half-life of about 6 hours¹ does the work within a day, and an extracorporeal circuit adds risk to a process that is already happening. In a patient with kidney impairment there may be no such process, and the workgroup suggests dialysis especially in the presence of associated coma requiring mechanical ventilation.³ The intervention is aimed at replacing an absent organ, not at treating a large ingestion.

Gabapentin's label reaches the same destination by a plainer route: although gabapentin can be removed by haemodialysis, based on prior experience it is usually not required. However, in patients with severe renal impairment, haemodialysis may be indicated.² Two documents written decades apart, for two different molecules, converge on the same conditional.

Target organs — and why those

Central nervous system

TargetThe alpha-2-delta auxiliary subunit of voltage-gated calcium channels on presynaptic terminals

Why hereThe target is expressed on neurons throughout the central nervous system, and the drugs reach it because they are small, unbound and freely distributed — pregabalin has been shown to cross the blood-brain barrier in mice, rats and monkeys.¹ This organ is targeted because it is the only place the target is, and because nothing keeps the drug out of it. There is no selectivity to explain: the same action that produces analgesia at a therapeutic concentration produces obtundation at a higher one. Established

At the bedsideSomnolence, confusional state, agitation and restlessness are the commonest reported overdose reactions; seizures also reported; coma in rare occasions.¹ Gabapentin adds dizziness, diplopia and slurred speech.²

The respiratory centre — in company

TargetNot a distinct receptor; an additive depression of central respiratory drive

Why hereThis card exists because the label puts it there. There have been reports of severe respiratory depression in relation to pregabalin use, with compromised respiratory function, respiratory or neurological disease, renal impairment, concomitant CNS depressants and the elderly all named as higher-risk groups.¹ Four of those five risk factors are properties of the patient rather than of the dose, which is the clearest statement anywhere on this page that the drug is not the whole poisoning. Established

At the bedsideHypoventilation and apnoea, characteristically in a patient who has also taken an opioid or a benzodiazepine. Naloxone will reverse the opioid component and leave the gabapentinoid untouched — see opioids.

The kidney — as the injured party, not the target

TargetNone. The kidney is not damaged by these drugs; it is the organ whose failure creates the poisoning

Why hereIncluded deliberately as a non-target, because the temptation on a page like this is to build an organ list out of an adverse-effect table. Neither drug is nephrotoxic. The kidney appears here because renal clearance is the only determinant of exposure that varies between patients, and because EXTRIP's recommendation is written entirely in terms of it.³ An organ card that cannot answer why here should not exist; this one answers a different question and says so. Established

At the bedsideNo renal syndrome to look for. The clinical act is to measure the function, not to watch for injury — the creatinine is a toxicokinetic measurement in this poisoning, not a screening test.

The gut

TargetReduced lower gastrointestinal tract motility, additive with opioids

Why hereA minor card kept because it has a mechanism and a consequence. Post-marketing reports describe intestinal obstruction, paralytic ileus and constipation when pregabalin is co-administered with medications that produce constipation, such as opioid analgesics.¹ A slowed gut in a patient who has swallowed tablets is also a pharmacokinetic event, prolonging absorption of everything taken alongside. Inferred

At the bedsideIleus and constipation in the therapeutic setting. In overdose the relevance is indirect: co-ingested drug may continue to be absorbed for longer than expected.

Timeline of effects

Pregabalin taken alone, with working kidneys — the uncomplicated case, shown so the complicated one is legible

Time

What you seeWhat is happening

0–1 hAbsorption
What you seeIncreasing drowsiness. Often little else.
What is happeningPeak plasma concentration within 1 hour in the fasted state; bioavailability ≥ 90% and independent of dose.¹ No latent phase, because there is nothing to manufacture.
1–6 hThe peak
What you seeSomnolence, confusion, agitation or restlessness.¹ Coma is possible but rare on the drug alone.¹
What is happeningMaximal alpha-2-delta occupancy. This is where a co-ingested opioid or benzodiazepine turns a sedated patient into an apnoeic one — the label names concomitant CNS depressants as a risk factor for severe respiratory depression.¹
6–24 hDecline
What you seeSteady improvement, tracking the concentration.
What is happeningHalf-life about 6.3 hours¹, so most of the drug has gone within a day provided the kidneys are working. Nothing here needs an antidote.
24 h onwards — impaired kidneys onlyThe curve that does not fall
What you seePersisting obtundation, out of proportion to the reported dose and to the time elapsed.
What is happeningClearance is directly proportional to creatinine clearance.¹ In anuria there is effectively no elimination, and this is the arm in which EXTRIP suggests dialysis, especially with coma requiring ventilation.³ Four hours of haemodialysis removes about 50% of plasma pregabalin.¹
DaysWithdrawal, in the dependent patient
What you seeInsomnia, anxiety, sweating, pain, agitation — and convulsion.¹
What is happeningDependence may occur at therapeutic doses¹, and EXTRIP explicitly warns to watch for withdrawal after ECTR.³ The end of the poisoning is not the end of the problem in a dependent patient.

What the mechanism predicts at the bedside

Ask what else was taken, and treat that. The pregabalin label's own case-control data give an adjusted odds ratio of 1.68 (95% CI 1.19–2.36) for opioid-related death with concomitant pregabalin, rising to 2.51 (1.24–5.06) above 300 mg and still present at 1.52 (1.04–2.22) at 300 mg or less.¹ The gabapentinoid is rarely the thing that stops the breathing; it is reliably the thing that made the opioid enough.
Measure renal function early, because it is the only variable. Pregabalin clearance is directly proportional to creatinine clearance.¹ The same ingestion in two patients is two different poisonings.
Expect a large gabapentin ingestion to underperform its number. Absorption saturates; the label reports no life-threatening toxicity at up to 49 g taken alone, with full recovery on supportive care.² This is a reason to be less impressed by the dose, not less careful about the patient.
Do not expect the same reassurance from pregabalin. Its bioavailability is ≥ 90% and independent of dose¹ — there is no absorptive ceiling to rely on.
A confused patient on a therapeutic dose is a real presentation. Falling GFR with an unchanged prescription produces gabapentinoid toxicity without an overdose, and this is the patient EXTRIP is actually talking about.³
Dialysis works and is usually pointless. Four hours removes about half the plasma pregabalin¹; six hours of working kidneys does the same thing without a catheter. EXTRIP suggests against ECTR with normal kidney function and for it with kidney impairment.³
Naloxone does nothing to a gabapentinoid. It is given for the co-ingestant and its failure to wake the patient fully should not be read as evidence against opioid involvement — see opioids.
Watch for withdrawal afterwards, particularly after dialysis. Dependence occurs at therapeutic doses and the withdrawal syndrome includes convulsion.¹³
A seizure has three possible authors here — the drug, its withdrawal, or the epilepsy it was prescribed for. The distinction changes what happens next.

The antidote, from the poison's side

There is no antidote and there is no plausible candidate for one. An antidote has to have something to do: an enzyme to block, a receptor to compete at, a cofactor to replace, a molecule to bind. A gabapentinoid presents none of these. It is not metabolised, so nothing can be blocked; its target is a structural subunit rather than a signalling receptor with a ready antagonist; it depletes nothing.

Supportive care: The whole of the treatment for the drug itself. Gabapentin's label reports full recovery with supportive care after ingestions up to 49 g²; pregabalin's specifies general supportive measures.¹
Haemodialysis: Effective at removal — about 50% of plasma pregabalin in four hours¹ — and recommended by EXTRIP only in the presence of kidney impairment, especially with coma requiring mechanical ventilation. Intermittent haemodialysis is preferred over other modalities (strong recommendation), stopping on clinical improvement (strong recommendation).³
Activated charcoal: Considered on general early-presentation principles. Neither label makes a gabapentinoid-specific claim for it, and this page does not invent one.
Naloxone: For the co-ingested opioid only. It has no action at the alpha-2-delta subunit, and expecting it to reverse the gabapentinoid component is a category error — see opioids.
Flumazenil: Not indicated, and mentioned only to be dismissed: the sedation is not benzodiazepine-mediated. The benzodiazepine page sets out why flumazenil is a poor idea even where it would work.

Critical appraisal

The alpha-2-delta account of these drugs' action is standard pharmacology and is not cited to a primary source here. Neither UK label is quoted for it in this page's prose, and no attempt is made to explain analgesia or anticonvulsant efficacy from it. It is used only to make the negative point that the target is not a GABA receptor, which the drugs' names imply and their pharmacology does not support.
The saturable-absorption argument for gabapentin is the label's own inference, not a measured outcome. The label states that reduced absorption at higher doses may limit absorption and, hence, minimize toxicity² — a mechanism offered to explain an observation, using the word may. The observation (no life-threatening toxicity up to 49 g) is what should be relied on; the explanation is plausible and is badged accordingly.
The 49 g figure is a reported experience, not a safe dose, and this page prints no threshold of any kind. No lethal dose, no toxic dose per kilogram and no concentration appears here — those are risk-assessment decisions belonging to TOXBASE and NPIS.
The opioid odds ratios are the numbers on this page most worth re-deriving. They are quoted from the Lyrica SmPC section 4.4, which describes them as coming from a case-control study of opioid users.¹ They are adjusted odds ratios from an observational design, and confounding by indication — sicker patients receiving both drugs — is not excluded by the label's summary. The direction is credible and the causal step is not established from what is cited here. Inferred
No evidence-tier downgrade is applied anywhere on this page, and that is a deliberate decision rather than an oversight. The obvious candidate would be the claim that these drugs are comparatively safe in overdose. It is not badged, because no source here contests it — both labels support it and EXTRIP's recommendation against dialysis in the intact patient is consistent with it. An absence of contrary evidence is not a citation for the doubt, which is the house rule, and the same reasoning that withheld a badge from iron applies here.
The claim that the co-ingestant is usually the killer is an inference from the label's risk-factor list and its odds ratios, not from a case series presented here. It is the page's central clinical message and it rests on ¹ alone. An auditor should test it hardest.
Neither drug's UK label describes an antidote, and this page invents none. The charcoal entry is explicitly framed as a general principle rather than a gabapentinoid-specific recommendation.

References

1
Lyrica 100 mg hard capsules (pregabalin) — Summary of Product Characteristics. electronic medicines compendium, product 10303. Sections 4.4 (Special warnings and precautions for use), 4.9 (Overdose) and 5.2 (Pharmacokinetic properties). medicines.org.uk/emc/product/10303
2
Neurontin 300 mg Hard Capsules (gabapentin) — Summary of Product Characteristics. electronic medicines compendium, product 3195. Sections 4.9 (Overdose) and 5.2 (Pharmacokinetic properties). medicines.org.uk/emc/product/3195
3
EXTRIP Workgroup. Gabapentin / pregabalin — recommendations. Extracorporeal Treatments in Poisoning Workgroup. Quoted here with the recommendation strength and evidence quality attached to each statement. extrip-workgroup.org/gabapentin-pregabalin

Gabapentinoids

At a glance