Why this poison is interesting

Cyanide is the cleanest demonstration in medicine that oxygen delivery and oxygen use are different problems. Every measurement a resuscitation room makes of oxygenation is a measurement of delivery: the saturation probe reads the haemoglobin, the arterial gas reads the plasma, the chest is clear and the cardiac output may be preserved. All of them can be normal while the patient dies, because the lesion is at the very last step — the reaction in which oxygen finally accepts the electrons and becomes water.

The UKHSA overview states the chemistry directly: hydrogen cyanide "has a high affinity for the ferric moiety of cytochrome c oxidase in mitochondria, forming a stable but reversible complex", and binding to the cytochrome a-a3 complex "blocks the last stage in the electron transfer chain and thus blocks ATP production".¹ The word doing the clinical work in that sentence is reversible. A stable but reversible complex is precisely the kind of lesion an antidote can compete with — which is why cyanide, uniquely among the rapidly fatal poisons in this library, has antidotes that work in minutes.

The third reason is that the body already treats this poisoning continuously and quietly. Rhodanese converts cyanide to thiocyanate, which is "less acutely toxic" and readily excreted, and it handles "about 80%" of the cyanide the body encounters.¹ The enzyme is not the limitation. The sulphur is: the reaction requires sulphane sulphur, "with the supply of sulphur containing donor molecules being rate limiting".¹ That single clause explains an entire antidote.

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

The toxic principle

Cytochrome c oxidase — complex IV — is the terminal enzyme of the respiratory chain. It accepts electrons from cytochrome c and uses them to reduce molecular oxygen to water, and it does so at a heme a3 / copper centre containing iron in the ferric state. That ferric iron is what cyanide binds.¹ Everything upstream in the chain then backs up, because there is nowhere for the electrons to go.

The consequences follow in a fixed order and they are all simultaneous in every tissue at once. Oxidative phosphorylation stops, so ATP production stops. The cell switches to anaerobic glycolysis, which is why "anaerobic respiration increases to compensate, with a concomitant increase in plasma lactate levels".¹ Because oxygen is being delivered and not consumed, venous blood stays oxygenated — the arteriovenous oxygen difference collapses. The UKHSA overview names the resulting state "histiotoxic hypoxia" and notes that tissue damage results throughout the body "from the reduced cellular utilisation of oxygen".¹

One further point of chemistry matters clinically. Cyanide "may also inhibit other metalloenzymes"¹ — cytochrome c oxidase is the fastest and most lethal target but not the only one. This is why the toxicity is not fully described as a pure asphyxiant lesion, and it is a reasonable part of the explanation for the delayed neurological sequelae that follow survived poisonings, though this page makes no strong claim about that.

The route a UK emergency clinician is most likely to meet is a house fire

Cyanide is an important industrial reagent, but the exposure a UK emergency clinician is most likely to meet arrives in smoke. Hydrogen cyanide "may be released from a range of combustion process, particularly those that involve nitrogen-containing materials such as polyurethane and PVC"¹ — the foam in furniture and the insulation on cable — and "hydrogen cyanide intoxication is a contributing factor to morbidity and mortality arising from smoke inhalation, together with other toxicants such as carbon monoxide".¹ The two poisons arrive together, both produce a lactic acidosis, and only one of them is detected by the co-oximeter. That is the argument for treating a fire victim with a depressed conscious level and a high lactate on suspicion rather than on measurement.

Toxicokinetics

The kinetics are fast in both directions, and the symmetry is the reason survival is possible at all. A poison with a plasma half-life of "20 minutes to 1 hour"¹ that binds reversibly to its target is a poison a patient can be carried through — provided the interval is survived.

Cyanide — a fast poison with a fast antidote, rate-limited by a substrate
Parameter	Therapeutic	In overdose	Why it changes
Absorption	Inhaled hydrogen cyanide is absorbed essentially instantly across the alveolus; salts are absorbed from the gut	Inhalation gives no useful interval. Ingested salts and cyanogenic glycosides are slower	Route determines whether there is any time at all. Cyanogenic glycosides in plants "may liberate hydrogen cyanide when the plant is damaged... or enzymatically hydrolysed within the body"¹ — a bioactivation step that buys hours.
Distribution	Distributes rapidly to all perfused tissue	Unchanged	There is no compartment worth discussing and no reservoir. The poison is wherever the blood went, within a circulation time.
Metabolism	"About 80%" converted to thiocyanate by rhodanese; the conversion "is irreversible"¹	Same pathway, saturated at the substrate rather than at the enzyme	The single most important row on this page. The reaction "requires sulphane-sulphur as a co-factor" and is "typically rapid, with the supply of sulphur containing donor molecules being rate limiting".¹ The enzyme is not overwhelmed; its raw material is. Giving thiosulfate is not enhancing a pathway — it is restocking it.
Where the enzyme lives	Rhodanese is "widely distributed in the mitochondria of all tissues, with the highest concentrations found typically in the liver, kidney, brain, and muscle"¹	Unchanged	The detoxifying enzyme sits in the same organelle as the target. That is convenient and it is not sufficient, because the sulphur has to reach it — and an intravenous sulphur donor distributes far less readily into mitochondria than the cyanide ion does.
A second, minor route	Complexing "of cyanide with cobalt in hydroxocobalamin to form cyanocobalamin (vitamin B12)"¹	This minor endogenous route is what the antidote scales up	A rare case of an antidote that is a physiological pathway given at pharmacological dose. The body already does this with the trace cobalamin it has; Cyanokit supplies grams of it.
Spontaneous detoxification rate	—	"About 1 micrograms per kilograms per minute, which is far slower than that in rodents"¹	A number with a warning attached to it. Much of the animal literature on cyanide tolerance is generated in species that detoxify far faster than humans do, so extrapolation from rodent survival data systematically overestimates human resilience.
Elimination	"The majority of absorbed cyanide is excreted in the urine as thiocyanate"; small amounts unchanged in lungs, saliva, sweat or urine¹	Plasma half-life "20 minutes to 1 hour"¹	Short, and it is the reason resuscitation is worth attempting. Unlike paraquat, where the poison leaves and the injury proceeds, here the poison leaving is the recovery — provided the mitochondria were not stopped for too long.
Dialysability	—	Never assessed and irrelevant on the timescale. EXTRIP has published no recommendation covering cyanide³	An absence reflecting a question not worth asking. A poison with a 20-minute half-life and a two-minute lethal window is not a target for a therapy that takes an hour to set up — and effective antidotes exist. Thiocyanate accumulation in renal failure is a different question, and it is a chronic one.

Metabolism and the metabolites

This is one of the few pages in the library where metabolism is entirely protective. There is no oxon, no NAPQI, no formate. Every metabolic route available to cyanide makes something less toxic, and the clinical problem is that the principal route runs out of a raw material rather than out of capacity.

Cyanide — one lethal binding step, and three ways out of it

Cyanide ionNot bioactivated. From inhaled hydrogen cyanide, an ingested salt, or a cyanogenic glycoside hydrolysed in the body¹
Binds ferric iron of cytochrome a-a3Cytochrome c oxidase inhibited"A stable but reversible complex"¹ — reversible is what makes an antidote possible
Rhodanese + sulphane sulphur (~80% of the load)Thiocyanate"Irreversible", and readily excreted in urine.¹ Rate-limited by the sulphur donor, not by the enzyme¹
Complexing with cobalt of hydroxocobalaminCyanocobalaminA "lesser" endogenous pathway¹ — and the entire mechanism of the principal antidote²
ATP production stops; anaerobic glycolysis takes over"Cellular ATP depletion, lactic acidosis, and cell and tissue death"¹
Histotoxic hypoxia with a narrowed arteriovenous oxygen differenceNormal saturations, rising lactate. The measurements of delivery all read normal¹
Failure of the tissues that cannot run anaerobicallyCNS first, "owing to its high oxygen demand and limited capacity for anaerobic oxidation"¹, then the heart

Elimination and accumulation

Cyanide does not accumulate in any clinically useful sense — it is bound, metabolised or exhaled within an hour. What accumulates is the oxygen debt, and unlike almost every other page in this library the deficit is repaid quickly and completely if the interval is survived.

Two exposures worth knowing about because they are chronic and low-level rather than acute. Cigarette smoke is, "for the general population (excluding those exposed to high levels of cyanogenic glycosides in food)", "considered to be the greatest source of exposure to hydrogen cyanide", with mainstream smoke from one filter cigarette containing "about 100 μg".¹ And a number of edible plants contain cyanogenic glycosides — "the kernels of wild (bitter) almonds, apricots and black cherries, bamboo shoots, lima beans, and cassava".¹ Neither is a resuscitation-room problem, and both are the reason a low background thiocyanate is a normal finding rather than evidence of exposure.

Target organs — and why those

Cyanide reaches every cell and inhibits the same enzyme in all of them, so the organ question has an unusually pure answer: the targets are the tissues that cannot survive a short interruption of oxidative phosphorylation. The overview states the rule directly, naming "those with high oxygen demand or low detoxifying capacity".¹

Brain

TargetCortical and brainstem neurones

Why hereNamed in the source as particularly vulnerable "owing to its high oxygen demand and limited capacity for anaerobic oxidation".¹ Neurones have almost no capacity to generate ATP anaerobically and no substrate reserve, so the interval between enzyme inhibition and functional failure is measured in seconds. This is the same reason the brain fails first in every other cause of energy failure, which is what makes the selectivity so predictable here. Established

At the bedsideHeadache, anxiety and dizziness at low exposure; then rapid loss of consciousness, seizures and apnoea. Collapse can be the first sign, and a cardiac arrest in a fire victim who was walking a minute earlier should raise this diagnosis.

Heart

TargetCardiac myocytes

Why hereContinuously contracting tissue with a high and inflexible ATP demand and little tolerance of anaerobic metabolism. The cardiac failure is not a separate mechanism — it is the same enzyme inhibition in the tissue least able to absorb it, and it converts a utilisation failure into a delivery failure as the output drops. Established

At the bedsideInitial tachycardia and hypertension from a catecholamine surge, then bradycardia, hypotension, arrhythmia and arrest. The haemodynamic collapse is late and abrupt.

The whole body — as a lactate signal

TargetEvery cell that has switched to anaerobic glycolysis

Why hereNot an organ, and it belongs here because it is the only measurement that tracks the lesion. Anaerobic respiration increases to compensate "with a concomitant increase in plasma lactate levels"¹, and because the block is universal the lactate reflects total-body failure of oxygen utilisation rather than regional ischaemia. A very high lactate with a normal saturation and no obvious source of ischaemia is the signature. Established

At the bedsideSevere metabolic acidosis with a high lactate, disproportionate to the clinical state and to any visible cause. A narrow arteriovenous oxygen difference — arterialised venous blood — points the same way.

Lung — as a route, not a target

TargetAlveolar membrane

Why hereIncluded to make the negative point. Cyanide is not primarily a pulmonary irritant and the lung is a route of entry rather than a site of injury, which distinguishes it sharply from the rest of the industrial band — chlorine and ammonia injure the airway they pass through. A fire victim's respiratory failure is usually thermal injury, soot and irritant gases; the cyanide is causing the coma, not the cough. Inferred

At the bedsideIn smoke inhalation the airway findings belong to the other components of the smoke. Attributing them to cyanide misdirects the assessment, and attributing the coma to the airway findings misses the poisoning.

Timeline of effects

Cyanide — the shortest timeline in the library

Time

What you seeWhat is happening

SecondsInhalation of a high concentration
What you seeHeadache, dizziness, breathlessness, anxiety — then collapse, often without a recognisable warning phase.
What is happeningAlveolar absorption is essentially instantaneous and distribution takes one circulation time. Cytochrome c oxidase is inhibited throughout the body within a minute.¹
1–5 minHistotoxic hypoxia declares itself
What you seeNormal or near-normal oxygen saturations. Rapidly rising lactate. Hyperventilation, then depressed consciousness, seizures, apnoea.
What is happeningOxygen is delivered and not extracted; anaerobic glycolysis rises "with a concomitant increase in plasma lactate levels".¹ Every monitor of delivery reads normal.
MinutesCardiovascular collapse
What you seeTachycardia and hypertension give way to bradycardia, hypotension and arrest.
What is happeningThe myocardium runs out of ATP. A utilisation failure becomes a delivery failure, and from that point the poisoning behaves like any other arrest.
20 min – 1 hEndogenous detoxification
What you seeIn a survived exposure, gradual recovery of consciousness and clearance of the acidosis.
What is happeningPlasma half-life "20 minutes to 1 hour"¹; rhodanese converts about 80% to thiocyanate, as fast as sulphane sulphur is available.¹ The binding to cytochrome c oxidase is reversible¹, so the enzyme recovers as the ion leaves.
HoursThiocyanate excretion
What you seeAsymptomatic. Chromaturia if hydroxocobalamin was given — all patients show dark red urine, sometimes for up to 35 days.²
What is happening"The majority of absorbed cyanide is excreted in the urine as thiocyanate."¹ Cyanocobalamin formed by the antidote is "excreted in the urine" too.²
Days–weeksDelayed neurological sequelae in survivors
What you seeParkinsonism and other extrapyramidal syndromes are described after severe survived poisonings, particularly with basal ganglia injury.
What is happeningThe source attributes these findings to both "direct toxicity of cyanide" and cerebral hypoxia secondary to the intoxication¹ — an explicitly unadjudicated pair rather than a settled account. Compare carbon monoxide, whose delayed syndrome has a different and better-characterised basis. Inferred

What the mechanism predicts at the bedside

A normal oxygen saturation does not exclude it, and never will. The probe measures delivery and the lesion is in extraction. This is the single most important line on the page.
A high lactate with no visible cause is the signal. The block is universal, so the lactate reflects total-body failure of oxidative phosphorylation rather than regional ischaemia.¹
Suspect it in any fire victim with reduced consciousness. Combustion of "nitrogen-containing materials such as polyurethane and PVC" releases hydrogen cyanide¹ — the source's own list, though PVC contains no nitrogen and its characteristic combustion hazard is hydrogen chloride, and it contributes to morbidity and mortality in smoke inhalation alongside carbon monoxide.¹
Carbon monoxide and cyanide arrive together and only one is measured. Treating the measurable poison and stopping there is the predictable failure mode.
Venous blood that looks arterial supports the diagnosis — the arteriovenous oxygen difference collapses when oxygen is delivered and not used. This follows from the mechanism and is not stated by either source cited here. Inferred
Treatment is on suspicion. A poison that kills in minutes and whose confirmatory assay is not available in minutes leaves no other option.
Ingested salts and plant glycosides are the versions with an interval, because the ion has to be liberated first.¹ An asymptomatic patient after an ingestion is not necessarily out of danger.
Once the exposure stops, the poison leaves. Plasma half-life is 20 minutes to 1 hour¹ and the enzyme binding is reversible¹ — prolonged resuscitation is more justified here than in most poisonings.
Hydroxocobalamin turns the patient, the urine and the blood samples red, and interferes with laboratory tests, burn assessment and haemodialysis machines.² Expect it and interpret results accordingly.
Dialysis has no role acutely³ — the half-life is shorter than the setup time, and antidotes work.

The antidote, from the poison's side

Cyanide is the poison in this library best served by its antidotes, and the reason is written into the mechanism: the complex with cytochrome c oxidase is "stable but reversible".¹ A reversible complex is a competition, and a competition can be won by supplying something that binds the ion harder. Two independent strategies do so, and one of them is simply the body's own second-line pathway given at scale.

Hydroxocobalamin: "The action of hydroxocobalamin in the treatment of cyanide poisoning is based on its ability to tightly bind cyanide ions. Each hydroxocobalamin molecule can bind one cyanide ion by substituting the hydroxo ligand linked to the trivalent cobalt ion to form cyanocobalamin. Cyanocobalamin is a stable, non-toxic compound that is excreted in the urine."² One-to-one stoichiometry is why the dose is measured in grams — the antidote is consumed by the reaction, molecule for molecule, and the UK label's indication is the "treatment of known or suspected cyanide poisoning in all age ranges".² Established
Sodium thiosulfate: Restocks the rate-limiting substrate. Rhodanese handles about 80% of a cyanide load and converts it irreversibly to thiocyanate, but the reaction "requires sulphane-sulphur as a co-factor" and the "supply of sulphur containing donor molecules" is what limits it.¹ Thiosulfate is that donor. This is the fourth pattern's antidote in its purest form: give the cofactor back. Its limitation is speed rather than logic — it accelerates an enzymatic pathway rather than scavenging the ion directly.
Sodium nitrite: Creates methaemoglobin deliberately, so that ferric iron in the blood competes with the ferric iron of cytochrome c oxidase for the cyanide ion. This library has an entire page on why methaemoglobin is a problem — see methaemoglobin inducers — and here it is manufactured on purpose. In a fire victim who may also be carboxyhaemoglobinaemic, deliberately disabling further haemoglobin is a decision with an obvious cost, which is the mechanistic reason hydroxocobalamin displaced it as first choice in that setting.
Dicobalt edetate: Chelates cyanide through cobalt, on the same logic as hydroxocobalamin and without its safety margin: it is itself toxic when given to a patient who turns out not to be cyanide-poisoned. An antidote that punishes a wrong diagnosis is a poor fit for a poisoning that must be treated on suspicion, which is the mechanistic argument against it and for hydroxocobalamin.
Oxygen: Does not lift the block and is given anyway. Raising the dissolved oxygen fraction may support whatever cytochrome c oxidase remains unbound, and a fire victim needs it for carbon monoxide and for the airway regardless. Contrast paraquat, where oxygen feeds the mechanism — the two pages sit at opposite ends of the same question.
Extracorporeal removal: Never assessed³ and irrelevant acutely: the plasma half-life is 20 minutes to 1 hour¹, effective antidotes exist, and the lethal window is shorter than the time to cannulate.

The antidote that blinds the investigations

Hydroxocobalamin's own colour has consequences the label documents at length. "Because of its deep red colour, hydroxocobalamin has the potential to induce a red colouration of the skin and therefore may interfere with burn assessment" — though the label adds that "skin lesions, oedema, and pain are highly suggestive of burns".² It "may interfere with all urine colorimetric parameters", typically for 48 hours after a 5 g dose.² Most patients have red skin for up to 15 days and all patients have dark red urine, which may last up to 35 days.² And it "may interfere with the performance of haemodialysis machines".² An antidote given to a fire victim makes it harder to assess their burns, harder to interpret their bloods and harder to dialyse them — all of which is worth accepting, and none of which should be discovered afterwards.

Critical appraisal

No evidence-tier downgrade appears on this page, and that is a deliberate decision rather than an omission. The house rule requires a citation for the doubt before a claim is badged as traditional teaching, and no source found for this page contests the cytochrome c oxidase mechanism, the rhodanese pathway or hydroxocobalamin's binding chemistry. Not manufacturing a badge is as much a part of this discipline as applying one — the same decision was recorded for nitazene naloxone resistance in Band A.
The comparative antidote reasoning is mechanistic, not comparative-effectiveness evidence. The arguments for hydroxocobalamin over dicobalt edetate, and against nitrite in a fire victim, follow from the chemistry and from the risk of treating on suspicion. No trial cited on this page compares them, and a reader should not take the ordering as an evidence-based ranking.
The 1 µg/kg/min spontaneous detoxification rate is described in its source as an estimate¹, and it is printed with the source's own comparison to rodents attached, because the point being made is about extrapolation rather than about the number.
The delayed neurological sequelae are badged inferred and are the weakest content on the page. An earlier draft attributed them to hypoxic-ischaemic injury alone and stated that no cited source addressed the mechanism. The source does address it, and attributes the findings to both direct cyanide toxicity and secondary cerebral hypoxia¹ — so the page now reports the dual attribution rather than silently choosing one arm of it.
The thiocyanate accumulation point is inference, badged as such, and concerns a chronic exposure — most familiarly prolonged nitroprusside infusion — that this page otherwise does not cover.
The workplace exposure limit is an occupational hygiene standard and is labelled as one.¹ It is not a clinical threshold and cannot support any inference about an acute exposure.
No lethal dose, no lethal concentration and no time-to-death figure appears on this page. The UKHSA overview contains material of that kind and it was deliberately not printed. Descriptions of speed are qualitative throughout.
The 'normal saturations' teaching is stated strongly and is genuinely mechanism-derived, but note precisely what it claims: pulse oximetry cannot detect the lesion. It does not claim that saturations are always normal — a fire victim has many other reasons to be hypoxaemic, and a low saturation does nothing to exclude cyanide either.
EXTRIP's silence³ is uninformative here in the same way it is for nitrous oxide: the question is not merely unasked, it is not worth asking on this timescale.

How this page connects to the rest of the library

Its closest mechanistic neighbour is hydrogen sulphide, which inhibits the same enzyme and produces a syndrome that is faster still — the two pages whose entire mechanism is cytochrome c oxidase inhibition, and the comparison is the point of both pages. It is the mirror of carbon monoxide: both arrive in the same smoke, both produce a lactic acidosis, and one impairs delivery while the other impairs use. It shares its antidote with nitrous oxide and uses it for the opposite purpose — there hydroxocobalamin replaces a cobalt the poison has oxidised; here it offers a cobalt for the poison to bind — which is the neatest coincidence in the library. And it sits against paraquat on oxygen: the poison whose treatment includes oxygen freely, beside the poison for which oxygen is fuel.

References

1
Hydrogen cyanide: toxicological overview. UK Health Security Agency, Compendium of Chemical Hazards. ("Hydrogen cyanide has a high affinity for the ferric moiety of cytochrome c oxidase in mitochondria, forming a stable but reversible complex." "Binding of cyanide to cytochrome a-a3 complex blocks the last stage in the electron transfer chain and thus blocks ATP production." "Cellular ATP depletion, lactic acidosis, and cell and tissue death." "Anaerobic respiration increases to compensate, with a concomitant increase in plasma lactate levels." "Tissue damage (histiotoxic hypoxia) throughout the body results from the reduced cellular utilisation of oxygen, the most sensitive tissues being those with high oxygen demand or low detoxifying capacity." "The CNS is particularly vulnerable... owing to its high oxygen demand and limited capacity for anaerobic oxidation." "Cyanide may also inhibit other metalloenzymes." Metabolism: "about 80%" to thiocyanate by rhodanese, requiring sulphane-sulphur, "with the supply of sulphur containing donor molecules being rate limiting"; rhodanese "widely distributed in the mitochondria of all tissues, with the highest concentrations found typically in the liver, kidney, brain, and muscle"; lesser pathway "complexing of cyanide with cobalt in hydroxocobalamin to form cyanocobalamin"; spontaneous detoxification "about 1 micrograms per kilograms per minute, which is far slower than that in rodents"; plasma half-life "20 minutes to 1 hour"; combustion of "nitrogen-containing materials such as polyurethane and PVC"; cigarette smoke "about 100 μg" per filter cigarette; short-term workplace exposure limit 11 mg/m³ over 15 minutes.) gov.uk
2
Cyanokit 5 g powder for solution for infusion — Summary of Product Characteristics. emc product 4786. (§4.1 "Treatment of known or suspected cyanide poisoning in all age ranges." §5.1 "The action of hydroxocobalamin in the treatment of cyanide poisoning is based on its ability to tightly bind cyanide ions. Each hydroxocobalamin molecule can bind one cyanide ion by substituting the hydroxo ligand linked to the trivalent cobalt ion to form cyanocobalamin. Cyanocobalamin is a stable, non-toxic compound that is excreted in the urine." §4.4 interference with burn assessment, with laboratory tests and with haemodialysis machines; urine colorimetric interference typically 48 hours after a 5 g dose. §4.8 red colouration of skin and mucous membranes in most patients up to 15 days; chromaturia in all patients, "may last up to 35 days". §5.2 total urinary excretion of cobalamins-(III) "at least 60 to 70% of the administered dose".)
3
EXTRIP Workgroup — published recommendations index. Extracorporeal Treatments in Poisoning Workgroup. Cited for the absence of any recommendation covering cyanide. extrip-workgroup.org/recommendations

Cyanide

At a glance