Why this poison is interesting

Cocaine's reputation is as a stimulant, and the sympathomimetic half of its pharmacology is genuinely responsible for most of what happens: tachycardia, hypertension, mydriasis, agitation, hyperthermia. But cocaine was introduced into medicine as a local anaesthetic, and it never stopped being one. A local anaesthetic is a sodium-channel blocker, and a sodium-channel blocker delivered to the whole myocardium produces exactly what the sodium-channel blockade page describes: a widening QRS, and a wide-complex dysrhythmia that responds to intravenous bicarbonate.¹

That is why cocaine is the only entry in this library whose ECG must be read against two different pages. The tachycardia belongs to the sympathomimetic story. The QRS belongs to the tricyclic one. Kerns and colleagues reported three cases of wide-complex dysrhythmia in acute cocaine intoxication in the absence of myocardial infarction — one resolved untreated, two after intravenous bicarbonate directed at reversing sodium-channel blockade — and noted that this presentation is rare and its optimum management undefined.¹

The third reason is the beta-blocker question, which is the most confidently taught rule in the whole of recreational-drug toxicology and the one with the least secure foundation. The authors of the original studies went back and examined their own evidence, and what they concluded is set out in the appraisal section rather than summarised here, because the wording matters.

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

The toxic principle

Cocaine does three separable things, and the clinical picture is the sum. Keeping them separate is the whole of the analysis.

The first and third mechanisms are not independent — the vasoconstriction is a consequence of the reuptake blockade acting at a particular receptor in a particular vascular bed — but separating them is useful, because they respond to different interventions. Benzodiazepines reduce the central drive that generates the excess noradrenaline. Alpha blockade acts at the receptor where that noradrenaline lands. Bicarbonate does neither and treats the channel.

Cocaethylene deserves separate treatment because it is a genuine bioactivation and it is entirely optional. When cocaine and ethanol are taken together, the liver transesterifies cocaine to cocaethylene — a pharmacologically active compound with a longer half-life than cocaine itself. It is the closest cocaine comes to the pattern described on the paracetamol page, and it is unique in this library in being a metabolite that exists only if the patient has also been drinking. Cocaine plus alcohol is not two drugs; it is three. Established

Toxicokinetics

Cocaine has no UK summary of product characteristics as a recreational preparation, so the figures below are the well-established pharmacology of the molecule rather than label-derived numbers, and the table says where each row's confidence comes from. What the kinetics mostly establish is that they explain very little: the drug is gone before most of its consequences are.

Cocaine — a short-acting drug with long-lasting consequences
Parameter	Therapeutic	In overdose	Why it changes
Absorption — route dependent	Smoked (crack) and intravenous: effect within seconds to a minute. Insufflated: slower, over tens of minutes. Swallowed: slower still	The same, but the ingested route is the one that produces prolonged, escalating toxicity	Route is the most important single variable in a cocaine presentation. A smoked exposure has already peaked by the time the patient arrives; a swallowed one has not. This is the entire reason body stuffers and body packers are managed as a separate problem.
Time to peak	Minutes by the smoked and intravenous routes	Unchanged	The peak is usually behind the patient at first contact, which is why a deteriorating cocaine patient should raise the question of an ingested source, a co-ingestant, or a complication rather than more absorption.
Half-life	Short — of the order of an hour	Unchanged; cocaethylene's is longer	A short half-life is not reassurance here. The drug's dangerous events — dysrhythmia, infarction, dissection, intracranial haemorrhage, hyperthermia, seizure — are injuries rather than drug effects, and they do not resolve when the drug does.
Metabolism	Hydrolysed to ecgonine methyl ester by plasma butyrylcholinesterase, and — largely non-enzymatically, plus by hepatic carboxylesterase — to benzoylecgonine. Both largely inactive	With ethanol: transesterification to cocaethylene, which is active and longer-lived	The only optional bioactivation in the library. Whether cocaine produces an active metabolite depends on whether the patient has been drinking — and the two are very commonly taken together.
Benzoylecgonine	Inactive metabolite; the substance urine screens detect	Detectable for days	A positive urine screen means exposure within days, not intoxication now. This is the commonest misreading of a cocaine test result, and the kinetics are the reason: the parent has a half-life of about an hour, the metabolite the screen finds does not.
Sodium-channel effect	Not apparent at anaesthetic mucosal doses	Myocardial sodium-channel antagonism similar to class I antidysrhythmic drugs, producing wide-complex dysrhythmia¹	The concentration-dependent second mechanism. Kerns and colleagues describe it as rare in the absence of myocardial infarction, with optimum management undefined¹ — a caution this page preserves rather than flattening into a rule.
Dialysability	—	No. EXTRIP has published no recommendation covering cocaine³	The absence is uninformative. A drug with a one-hour half-life is cleared faster by the patient than by any dialyser, and nobody has proposed otherwise. This is one of the EXTRIP absences that reflects a question not worth asking.

Metabolism and the metabolites

Cocaine's ordinary metabolism is a detoxification — esterases hydrolyse it to inactive products, which is why the drug is so short-lived. The interesting pathway is the one that only opens when there is alcohol in the blood.

Cocaine — an inactivating metabolism, and one branch that is not

CocaineAlready the active species. Reuptake inhibition, sodium-channel blockade¹ and vasoconstriction, all from the parent
Plasma and hepatic esterase hydrolysisBenzoylecgonine + ecgonine methyl esterLargely inactive. Benzoylecgonine arises mostly by spontaneous hydrolysis rather than enzymatically, and is what urine immunoassays detect for days after the drug has gone
Transesterification — ONLY in the presence of ethanolCocaethyleneActive, and longer-lived than cocaine — a comparison this page states without a citation, since no source here quantifies either half-life. An optional bioactivation created by the co-ingestant
Sympathomimetic toxidromeTachycardia, hypertension, mydriasis, agitation, hyperthermia, psychosis
Wide QRS, wide-complex dysrhythmiaTreated with sodium bicarbonate, not with a sympatholytic¹ — see sodium-channel blockade

Elimination and accumulation

Cocaine does not accumulate in any pharmacologically meaningful way in the routine presentation — it is hydrolysed within hours. There are exactly two situations in which accumulation is the whole clinical problem, and both are anatomical rather than metabolic.

The second reason ingestion changes the analysis is kinetic. Every other route front-loads the exposure: it peaks and then falls. Swallowed cocaine, whether stuffed or packed, delivers continuously from a reservoir the clinician cannot see or measure. That converts a one-hour drug into an exposure of indefinite duration, and it is the same structural problem the calcium-channel blocker page describes for modified-release tablets — a formulation, or in this case a wrapping, dictating the timeline instead of the molecule.

Target organs — and why those

Heart — myocardium and coronary arteries

TargetAlpha-1 receptors on coronary smooth muscle; myocardial sodium channels; increased demand

Why hereThe organ where all three of cocaine's mechanisms converge, which is why it is the organ that kills. Reuptake blockade raises noradrenaline, which raises rate, blood pressure and therefore oxygen demand; alpha-1 stimulation narrows the vessels supplying it; and sodium-channel blockade destabilises conduction independently of either.¹ Convergence of three independent mechanisms on one organ is unusual — on the digoxin and calcium-channel blocker pages the myocardium is reached by one route each. Established

At the bedsideChest pain, myocardial infarction in young people with normal or near-normal arteries, tachydysrhythmia, and — separately — a widening QRS.¹ The ECG must be read for both the rate and the QRS duration; they belong to different mechanisms and need different treatments.

Brain — cerebral vasculature

TargetVasoconstriction plus an abrupt hypertensive surge

Why hereThe same alpha-mediated vasoconstriction that narrows coronary arteries acts on cerebral vessels, and the sympathetic surge produces a sudden rise in blood pressure in a vascular bed that has no time to adapt. Both haemorrhagic and ischaemic stroke are described, which is unusual — most vascular insults favour one mechanism. Inferred

At the bedsideSudden headache, focal deficit or seizure in a young person. Cocaine is one of the few toxicological causes that belongs on the differential for both stroke types, and its relevance to the ischaemic stroke pathway is precisely that it does not respect the usual age assumptions.

Brain — dopaminergic pathways

TargetDopamine transporter blockade in mesolimbic and mesocortical projections

Why hereThe euphoria and the psychosis are the same mechanism at different intensities. Blocking dopamine reuptake raises synaptic dopamine; sustained excess produces the paranoid, hallucinatory state that is clinically indistinguishable from an amphetamine psychosis and often from a primary one. This is the pharmacology the antipsychotics page describes from the opposite direction — one drug class blocks the receptor, this one floods it. Established

At the bedsideAgitation, paranoia, tactile hallucinations, and the severe agitated state that carries a high risk of hyperthermia and rhabdomyolysis. Benzodiazepines rather than antipsychotics are the conventional first line, and the reason is thermoregulatory as much as pharmacological.

Skeletal muscle and kidney

TargetSustained muscle activity, hyperthermia and vasoconstriction acting together

Why hereThree separate insults converge on the same endpoint. Agitation produces continuous muscle work; hyperthermia damages myocytes directly; and alpha-mediated vasoconstriction reduces renal perfusion at the moment the kidney is asked to clear myoglobin. Rhabdomyolysis in cocaine toxicity is not a complication of one thing — it is the intersection of three. Inferred

At the bedsideRaised creatine kinase, myoglobinuria, acute kidney injury. Check creatine kinase in any agitated or hyperthermic cocaine presentation; the renal injury is preventable and the muscle injury is already done.

Aorta and large vessels

TargetAbrupt shear stress from a hypertensive surge

Why hereNot a metabolic lesion at all — a mechanical one. A sudden, severe rise in blood pressure and contractility increases wall stress in a vessel that may already be abnormal. This card is included because it is the cocaine complication most often missed, and because the mechanism explains why it presents at the moment of the surge rather than following the drug's half-life. Inferred

At the bedsideAortic dissection in a young patient with chest pain and a cocaine history — a presentation in which the reflex diagnosis of cocaine-associated chest pain is not merely incomplete but points treatment in the wrong direction.

Nose, lung and skin — route-specific injury

TargetDirect tissue effects of vasoconstriction and of the smoked preparation

Why hereCocaine's vasoconstriction acts wherever it is applied. Repeated insufflation produces mucosal ischaemia and eventually septal perforation; smoked crack cocaine produces an acute pulmonary syndrome with cough, dyspnoea, hypoxia and pulmonary haemorrhage. These are the only organ effects on this page that follow from the route rather than the systemic concentration. Established

At the bedsideNasal septal perforation in a chronic insufflator; 'crack lung' — acute respiratory symptoms with infiltrates and sometimes haemoptysis — within hours of smoking.

Timeline of effects

Cocaine — a drug that outlives itself only through the injuries it causes

Time

What you seeWhat is happening

Seconds to 1 minSmoked or intravenous
What you seeEuphoria, tachycardia, hypertension, mydriasis.
What is happeningReuptake blockade at all three monoamine transporters. Onset limited by circulation time — there is no absorption step to speak of.
10–60 minInsufflated
What you seeThe same effects, arriving more gradually.
What is happeningMucosal absorption, slowed by the drug's own vasoconstriction of the vessels it must cross. Cocaine limits its own absorption, which is a genuinely unusual property.
0–1 hThe window in which the dangerous events happen
What you seeChest pain, dysrhythmia, seizure, stroke, dissection, severe agitation and hyperthermia.
What is happeningPeak concentration of all three mechanisms simultaneously. Nearly everything that kills in cocaine toxicity happens here, while demand, vasoconstriction and sodium-channel blockade coincide.¹
1–3 hDrug falling, patient not necessarily improving
What you seeAgitation may persist; an infarct, a haemorrhage or a dissection does not resolve with the drug.
What is happeningEsterase hydrolysis to inactive benzoylecgonine. The gap between drug elimination and clinical recovery is the point of this phase — unlike almost every other page here, the poison leaving does not end the problem.
Hours, if alcohol was takenCocaethylene
What you seeA longer, flatter tail of sympathomimetic effect.
What is happeningTransesterification in the presence of ethanol produces an active metabolite with a longer half-life than cocaine. A poisoning extended by a metabolite the patient created by drinking.
Hours to daysIngested — packers and stuffers
What you seeDeceptively well, then abruptly not. The gap drawn here is a reservoir, not a metabolic delay.
What is happeningContinuous or catastrophic release from swallowed packages. The timeline is set by wrapping integrity, which cannot be measured — the same structural problem as a modified-release tablet on the calcium-channel blocker page.
Weeks to monthsLevamisole
What you seeAgranulocytosis; retiform purpura of the ears and extremities.
What is happeningAn adulterant, not the drug. A separate toxicology on a separate clock, reached through the same history.

What the mechanism predicts at the bedside

Read the QRS, not just the rate. Cocaine is a sodium-channel blocker¹, and a wide QRS in cocaine toxicity is treated as sodium-channel blockade — see sodium-channel blockade.
Ask about alcohol, because it changes the metabolism. Cocaethylene is active and longer-lived, and it exists only if the patient has been drinking.
A positive urine screen means exposure within days, not intoxication now. The assay detects benzoylecgonine; the parent drug's half-life is about an hour.
A deteriorating cocaine patient is absorbing from somewhere. By the smoked or intravenous route the peak is already past — deterioration should raise ingestion, a co-ingestant, or a complication.
Benzodiazepines first, because they reduce the central drive that generates the noradrenaline, the muscle activity and the heat all at once.
Check the temperature and the creatine kinase. Hyperthermia with agitation is where cocaine becomes a renal disease.
Chest pain in a young cocaine user is not automatically demand ischaemia — infarction with clean arteries, aortic dissection and pneumothorax all occur, and the last two are not treated by treating the cocaine.
Both stroke types are on the differential, which is unusual and is a direct consequence of combining vasoconstriction with a hypertensive surge.
Check the full blood count in a regular user. Levamisole adulteration causes agranulocytosis on a timescale unrelated to the last dose.
Dialysis has no role and has never been assessed³; a one-hour half-life makes the question moot.

The antidote, from the poison's side

There is no antidote to cocaine, and there could not be one: the drug acts at three transporters and one ion channel, and no single molecule reverses that combination. What exists is a set of treatments aimed at each mechanism separately, and one long-standing prohibition that has been re-examined by the people whose work created it.

Benzodiazepines: The conventional first-line intervention and the closest thing to a mechanism-directed treatment. They reduce central sympathetic outflow, reduce agitation, reduce muscle activity and therefore heat production, and raise the seizure threshold — four of cocaine's problems addressed by one drug acting upstream of all of them.
Sodium bicarbonate: For QRS widening. Kerns and colleagues describe intravenous bicarbonate as the current accepted treatment for the wide-complex dysrhythmia of myocardial sodium-channel poisoning, and report two cases in which it resolved after bicarbonate directed at reversing sodium-channel blockade.¹ It treats the channel, not the cocaine — the same principle as on the sodium-channel blockade page.
Phentolamine: Alpha blockade, acting at the receptor where cocaine's excess noradrenaline lands. Mechanistically direct for vasoconstriction and hypertension, and the intervention that best fits the coronary vasospasm account.
Active cooling: For hyperthermia. As on the SSRI page, the heat is largely produced by muscle, so sedation is part of cooling rather than separate from it.
Beta-blockade: The contested question. See the callout below; the review that examines it is by the authors of the original evidence.²
Extracorporeal removal: No role and never assessed.³ Cocaine is cleared by hydrolysis faster than by any machine.

The unopposed alpha stimulation phenomenon — what the original authors concluded

The teaching is absolute: never give a beta-blocker to a cocaine-positive patient, because blocking beta-2-mediated vasodilatation leaves alpha-mediated vasoconstriction unopposed and worsens coronary spasm. Richards, Hollander and Ramoska — who state that they are the authors of the original studies, case reports and systematic review in which unopposed alpha stimulation was believed to be a factor — went back and investigated the pathophysiology, pharmacology and published evidence behind it. Their review records that for the past three decades the phenomenon has been cited as an absolute contraindication, despite limited and inconsistent clinical evidence, and examines other potential explanations including the unique and deleterious pharmacologic properties of cocaine in the absence of β-blockers, together with the safety and efficacy of the mixed beta/alpha blockers labetalol and carvedilol.² A prohibition described as an absolute contraindication by the very authors who then call the evidence limited and inconsistent is exactly the shape of claim this library badges. Traditional teaching

The badge is on the mechanism and the absoluteness, not on the practice. Nothing here recommends giving a beta-blocker to a cocaine-toxic patient; that decision belongs to TOXBASE, NPIS and the treating team, and the conservative position remains widely held for reasons that include but are not limited to the mechanism. What is downgraded is the confident teaching that unopposed alpha stimulation is a demonstrated phenomenon establishing an absolute contraindication — a claim the authors of its evidence base describe as resting on limited and inconsistent clinical evidence.² This is the same distinction the clonidine page draws about naloxone: a badge on a mechanism is not a recommendation about a treatment.

Critical appraisal

The unopposed-alpha badge rests on a citation that contests the claim, and the contest comes from an unusual source: the authors of the original evidence. Richards, Hollander and Ramoska describe the phenomenon as having been cited as an absolute contraindication ... despite limited and inconsistent clinical evidence, and re-examine it explicitly.² They are not neutral parties — a review by the authors of the work under review has an obvious interest — and this page treats that as a reason to state the finding precisely rather than to amplify it. It is a citation for the doubt, and it is not a demonstration that beta-blockers are safe.
Kerns and colleagues report three cases.¹ They describe cocaine-induced wide-complex dysrhythmia in the absence of infarction as rare, and state that optimum management is undefined. Both qualifiers are preserved on this page. The mechanism — that cocaine is a local anaesthetic with myocardial sodium-channel antagonism similar to class I antidysrhythmics — is secure pharmacology; the frequency and the management are not established by three cases and this page does not claim they are.
The coronary vasospasm mechanism is deliberately not downgraded, and it is worth recording why. Alpha-mediated coronary vasoconstriction is established pharmacology, and no citation was found that contests it — only sources that place it alongside demand ischaemia and thrombosis as contributors. An absence of a citation for the claim would not justify the badge, and neither does a claim that the mechanism is merely one of several. House rule applied in the direction that resists the badge.
No half-life figure on this page carries a label citation, because there is no label. The 'of the order of an hour' phrasing is deliberately imprecise rather than falsely precise, and every clinical inference drawn from it — that the peak is past, that dialysis is pointless, that the urine screen is misleading — holds across the range in which the true figure sits.
Cocaethylene's formation and longer half-life are well established; no figure for it is quoted here for the same reason, and the claim is kept qualitative.
The rhabdomyolysis and aortic cards are inference from mechanism, marked as such. Both describe real, reported complications; the three-way convergence account for rhabdomyolysis and the shear-stress account for dissection are reconstructions from physiology rather than findings from cocaine-specific studies.
Levamisole's association with agranulocytosis is well described, but no prevalence figure for adulteration appears here — seizure-sample composition varies by place and year, and a number taken from one jurisdiction's data would be quoted as though it were general.
No dose, threshold or lethality figure appears anywhere on this page. For a drug of unknown purity, taken by variable routes, any such number would be meaningless as well as out of scope.
EXTRIP's silence is uninformative³ and is recorded only for consistency with the rest of the library.

References

1
Kerns W 2nd, Garvey L, Owens J. Cocaine-induced wide complex dysrhythmia. Journal of Emergency Medicine 1997 May–Jun;15(3):321–9. PMID 9258782. (Three cases; cocaine described as a local anaesthetic with myocardial sodium channel antagonism similar to class I antidysrhythmic drugs; wide complex dysrhythmia in the absence of myocardial infarction described as rare with optimum management undefined; two cases resolved after intravenous bicarbonate directed at reversing sodium channel blockade.)
2
Richards JR, Hollander JE, Ramoska EA, et al. β-Blockers, cocaine, and the unopposed α-stimulation phenomenon. Journal of Cardiovascular Pharmacology and Therapeutics 2017 May;22(3):239–49. PMID 28399647. doi:10.1177/1074248416681644. (Review by the authors of the original studies, case reports and systematic review in which unopposed alpha stimulation was believed to be a factor; records that for three decades it has been cited as an absolute contraindication despite limited and inconsistent clinical evidence, and examines alternative explanations and the mixed beta/alpha blockers labetalol and carvedilol.)
3
EXTRIP Workgroup — published recommendations index. Extracorporeal Treatments in Poisoning Workgroup. Cited for the absence of any recommendation covering cocaine. extrip-workgroup.org/recommendations

Cocaine

At a glance