Cyproheptadine

Why this drug is interesting

Serotonin toxicity is one of the few toxidromes with a named receptor. The Hunter Serotonin Toxicity Criteria were built around 5-HT2A, and cyproheptadine antagonises it — so there is an obvious antidote-shaped drug sitting exactly where the pathology is. That neatness is why cyproheptadine appears in every textbook account and on every teaching slide.

It is also, on the best current reading of the evidence, probably not doing what people think it is doing. This page exists to give the dose people will look for and, in the same breath, the reason to be sceptical of it.

Pharmacology

Mechanism

Cyproheptadine is a first-generation histamine H₁ receptor blocker with non-specific antagonist properties at 5-HT1A and 5-HT2 receptors.¹ The therapeutic claim rests on the 5-HT2A component: serotonin toxicity is driven by excess synaptic serotonin, and 5-HT2A activation is thought to mediate the hyperthermia and neuromuscular features that make it dangerous.

It also carries the rest of the first-generation antihistamine profile — sedation and antimuscarinic effects — and that turns out to matter for interpreting what it appears to do.

What the animal data show

In the Nisijima rat model, treatments were given intraperitoneally 15 minutes before the serotonergic agent:¹

• Ritanserin and pipamperone — potent, selective 5-HT2A antagonists — prevented death and hyperthermia.
• Cyproheptadine 5 mg/kg, chlorpromazine 20 mg/kg, dantrolene 20 mg/kg and propranolol did not prevent death. All those rats developed hyperthermia to 40 °C.
• Only at higher doses — cyproheptadine 10 mg/kg, chlorpromazine 40 mg/kg — did the rats survive.

Dosing

What the literature reports

• "Initial doses administered vary greatly, typically from 4 to 12 mg, and even higher doses have been given."¹
• There are reports of mild to moderate toxicity resolving 1–2 hours after cyproheptadine.¹
• There are also "cases receiving 16 mg with no response" and "cases of fatality despite cyproheptadine".¹
• The widely reproduced 12 mg loading dose followed by 2 mg every two hours derives from a 2005 narrative review,³ not from trial data, and this page does not reproduce it as a protocol.

Practical constraints

Tablets are 4 mg. In an agitated patient with clonus and tremor who can still swallow, a dose is deliverable. In a patient with rigidity and a temperature above 39 °C, it is not — and that patient needs the airway, the cooling and the paralysis, not a tablet.

Diagnosis — the Hunter criteria

The receptor that justifies cyproheptadine is the one the diagnostic rule was built around, so the two belong on the same page. Dunkley and colleagues studied 2,222 patients admitted after overdose of a serotonergic drug, using a learning dataset of 473 SSRI-only overdoses.²

Of the many features associated with serotonin toxicity, only these predicted it accurately:²

• Clonus — inducible, spontaneous or ocular
• Agitation
• Diaphoresis
• Tremor
• Hyperreflexia

Because the learning dataset contained no life-threatening cases, hypertonicity and a maximum temperature above 38 °C were added, both being universal in such patients.² The resulting Hunter criteria were more sensitive (84% vs 75%) and more specific (97% vs 96%) than Sternbach's.²

What actually treats serotonin toxicity

The 2025 review is unusually direct about where the effort should go, and it is not the antidote:¹

Hence, the cornerstone of treating severe cases does not lie in 'antidote' administration or even diagnosis but in effective early resuscitative and supportive care.

Chiew and Isbister, British Journal of Clinical Pharmacology, 2025¹

1. Stop the serotonergic agent. Serotonin toxicity "will often resolve within 24 hours of discontinuing the serotonergic agent and initiating supportive care" — though drugs with long half-lives or active metabolites take longer.¹
2. Benzodiazepines — see midazolam. These address agitation and the muscle activity generating the heat.
3. Identify severe toxicity early. The review directs attention to "a rapidly rising temperature >39 °C and muscle rigidity".¹
4. In severe cases: intubation, paralysis and active cooling.¹ Paralysis stops heat production at source; this is the intervention that changes outcome in hyperthermic serotonin toxicity.
5. Do not rely on antipyretics. The hyperthermia is muscular, not hypothalamic.

Other serotonin antagonists — chlorpromazine and olanzapine — have been explored, with evidence limited to case reports.¹ Chlorpromazine additionally lowers the seizure threshold and drops the blood pressure, which is a poor trade in an unstable patient.

Safety and supply

UK availability

Expected effects

As a first-generation antihistamine, cyproheptadine causes sedation and antimuscarinic effects — dry mouth, urinary retention, blurred vision, tachycardia. In serotonin toxicity these matter in two directions: the sedation may be the whole of the observed benefit, and the antimuscarinic load is unhelpful in a hyperthermic patient who needs to sweat.

It is also an appetite stimulant and is used off-label for that purpose, which is worth knowing when it appears unexpectedly on a drug chart.

In children

Cyproheptadine has a long paediatric history as an antihistamine and appetite stimulant, but its use in paediatric serotonin toxicity is unlicensed, unstudied and should be a conversation with NPIS and paediatric intensive care.

Critical appraisal

Cyproheptadine is the clearest example on this site of an antidote sustained by mechanistic elegance rather than by evidence. The receptor is right, the diagnostic criteria are built on the same receptor, and the drug is cheap and nearly harmless. Every incentive points towards giving it, and almost none towards testing it — which is why, twenty years on, the evidence is still "case reports and series".

The 2025 Chiew and Isbister review is the most useful corrective, and its position is worth stating in full rather than softening: serotonin antagonists' "clinical effectiveness remains uncertain"; cyproheptadine specifically "lacks conclusive evidence of efficacy"; the apparent benefit is "likely due to their sedative actions"; and the cornerstone of treatment is resuscitative and supportive care.¹

One genuine gap: the selective 5-HT2A antagonists that did work in the animal model — ritanserin, pipamperone — are not available for clinical use in this indication. If serotonin toxicity ever gets a real antidote, it will probably not be this one.

References

1. 1
   Chiew AL, Isbister GK. Management of serotonin syndrome (toxicity). British Journal of Clinical Pharmacology 2025;91(3):654–661. PubMed 38926083 · PMC11862804. Open access; full text read 4 September 2026. Source of the oral-only statement, the 4–12 mg reported dose range, the 16 mg non-response and fatality reports, the 85–90% receptor occupancy figure, the Nisijima animal results, the "likely due to their sedative actions" and "never been shown" statements, the >39 °C and rigidity severity markers, and the conclusion on supportive care. The authors report no conflicts of interest.
2. 2
   Dunkley EJC, Isbister GK, Sibbritt D, Dawson AH, Whyte IM. The Hunter Serotonin Toxicity Criteria: simple and accurate diagnostic decision rules for serotonin toxicity. QJM 2003;96(9):635–642. PubMed 12925718. n = 2,222; learning dataset 473 SSRI-alone overdoses. Source of the five predictive features, the added hypertonicity and >38 °C criteria, and the sensitivity and specificity figures against Sternbach's criteria.
3. 3
   Boyer EW, Shannon M. The serotonin syndrome. New England Journal of Medicine 2005;352(11):1112–1120. PubMed 15784664. The narrative review from which the widely quoted 12 mg then 2 mg two-hourly regimen derives. Cited for provenance only — the full text was not obtained for this page, and the regimen is deliberately not reproduced here as a protocol.
4. 4
   National Poisons Information Service. TOXBASE · NPIS 0344 892 0111. Required: cyproheptadine has no emc entry and no UK licence for this indication (searched 4 September 2026).