Neuroleptic Malignant Syndrome
Pathophysiology, recognition, atypical variants, treatment, and the bedside findings that separate NMS from serotonin syndrome. A provider reference for adult practice.
Scope
What this review covers
Adult neuroleptic malignant syndrome (NMS): its mechanism, the agents that cause it, predisposing factors, the classic and atypical clinical spectrum, evidence-based treatment, and the clinical and laboratory features that separate it from serotonin syndrome (SS) and other hyperthermic toxidromes. A recurring theme is that both entities are clinical diagnoses without a pathognomonic test, and that the physical examination, particularly the neuromuscular findings, carries more discriminating weight than any laboratory value.
For the broader medical differential in a first presentation, see our medical workup of first-episode psychosis and the psychosis triage protocol.
Mechanism
Pathophysiology
NMS is a rare, idiosyncratic reaction to dopamine-blocking or dopamine-depleting agents driven by central D2 receptor blockade across several circuits. Hypothalamic D2 blockade raises the thermoregulatory set point and impairs heat dissipation; nigrostriatal and spinal D2 antagonism produce rigidity and tremor; and loss of tonic dopaminergic inhibition of the sympathetic nervous system produces autonomic instability. At the muscle level, sustained rigidity drives excess heat generation and calcium release from the sarcoplasmic reticulum, causing rhabdomyolysis and CK elevation, but there is no primary skeletal muscle defect, which distinguishes NMS mechanistically from malignant hyperthermia.
Supporting evidence includes reduced CSF homovanillic acid during acute episodes and the fact that abrupt dopamine-agonist withdrawal reproduces the syndrome. Because clozapine, which has weak D2 affinity, can still cause NMS, pure D2 blockade is an incomplete explanation, and many experts now regard NMS as a form of malignant catatonia precipitated by dopamine antagonism, given shared features and shared response to benzodiazepines and ECT.
0.01 to 0.04%
Incidence among antipsychotic-treated patients.
Under 10 to 15%
Mortality with modern recognition and intensive care, down from historical rates near 76%, but still 10 to 20% when the diagnosis is missed.
Exposure
Causative agents
High-potency first-generation antipsychotics
Haloperidol, fluphenazine, and pimozide carry the greatest risk, with an incidence around 0.2% and mortality of 10 to 20%.
Second-generation antipsychotics
Clozapine, olanzapine, risperidone, quetiapine, and aripiprazole cause NMS less often (around 0.006%) and with lower mortality (3 to 5.5%).
Non-psychiatric dopamine blockers, frequently overlooked
Metoclopramide, prochlorperazine, droperidol, promethazine, and domperidone, plus the dopamine-depleter tetrabenazine.
Dopaminergic withdrawal
Abrupt withdrawal of levodopa or amantadine precipitates a clinically identical picture termed parkinsonism-hyperpyrexia syndrome.
Recognition
Clinical features and diagnostic criteria
The classic tetrad is hyperthermia, lead-pipe rigidity, autonomic instability, and altered mental status. Temperature commonly exceeds 38°C and can surpass 41°C; rigidity is uniform and typically unresponsive to antiparkinsonian agents; autonomic signs include tachycardia, labile blood pressure, diaphoresis, and sialorrhea; and mental status ranges from agitated delirium to alert mutism and stupor.
The temporal sequence matters clinically: altered consciousness and extrapyramidal signs are often the earliest manifestations, with fever, CK elevation, and dysautonomia following, and severity peaking over 48 to 72 hours. Onset is typically 1 to 3 days after drug initiation or dose change (median about 4 days), with nearly all cases within 30 days.
Laboratory findings are supportive but nonspecific: CK often above 4x ULN and frequently above 16,000 IU/L, leukocytosis of 15,000 to 30,000/mm3, elevated transaminases and LDH, metabolic acidosis, and characteristically low serum iron. Complications include rhabdomyolysis with acute kidney injury in up to 30%, aspiration, respiratory failure, DIC, and venous thromboembolism.
DSM-5 criteria
Dopamine-blocker exposure plus rigidity and fever, with at least two additional features: diaphoresis, dysphagia, tremor, incontinence, altered consciousness, mutism, tachycardia, labile blood pressure, leukocytosis, or elevated CK.
2011 international Delphi consensus
Weighted priority points out of 100. No single feature is mandatory.
- Dopamine antagonist exposure or agonist withdrawal20
- Hyperthermia18
- Rigidity17
- Mental status change13
- Creatine kinase at least 4x ULN10
- Sympathetic nervous system lability10
- Hypermetabolism (tachycardia and tachypnea)5
- Negative workup for alternative causes7
Who is at risk
Predisposing and risk factors
Risk is multiplied by both pharmacologic and patient-related factors. The strongest and most consistent modifiable factor is dehydration. Notably, NMS occurs within the therapeutic dose range and is not related to overdose or duration of exposure, and prior episodes recur in 15 to 20%, suggesting individual vulnerability.
Pharmacologic
High-potency first-generation antipsychotics; high doses and rapid titration; parenteral or intramuscular administration; depot formulations; antipsychotic polypharmacy; concomitant lithium.
Patient-related
Dehydration (the strongest modifiable factor); psychomotor agitation and exhaustion; preexisting catatonia; male sex (roughly 2:1); age under 40; organic brain disease or dementia; iron deficiency; prior NMS episode; postpartum state.
Heightened sensitivity states
Dementia with Lewy bodies, intellectual disability, and anti-NMDA receptor encephalitis, where neuroleptic intolerance occurs in roughly 47%.
Genetic
DRD2 TaqI A1 allele and the -141C Del polymorphism in Japanese cohorts; CYP2D6 variants inconsistently associated, and the 2024 NEJM review concluded slow-metabolizer status does not confer increased risk overall.
Differential
Hyperthermic toxidromes side by side
NMS and serotonin syndrome overlap in hyperthermia, autonomic instability, and altered mentation, but differ fundamentally in mechanism, tempo, and, most usefully at the bedside, the neuromuscular examination. SS reflects excess serotonergic activity, principally 5-HT2A stimulation, and is dose-dependent and predictable, whereas NMS is idiosyncratic.
| Feature | NMS | Serotonin syndrome | Malignant hyperthermia | Anticholinergic toxicity |
|---|---|---|---|---|
| Trigger | Dopamine blocker, or dopamine agonist withdrawal | Serotonergic agent(s) | Volatile anesthetic or succinylcholine | Antimuscarinic agent |
| Onset | 1 to 3 days | Minutes to hours (under 24 h) | Minutes to hours | Under 12 h |
| Muscle tone | Lead-pipe rigidity, all muscle groups | Increased tone, lower-limb predominant | Rigor-like rigidity | Normal |
| Reflexes | Normal or decreased | Hyperreflexia, clonus | Hyporeflexia | Normal |
| Pupils | Normal | Mydriasis | Normal | Mydriasis |
| Bowel sounds | Normal or decreased | Hyperactive | Decreased | Absent |
| Skin | Pallor, diaphoresis | Diaphoresis | Mottled | Hot, dry, red |
| CK and leukocytosis | Markedly elevated | Less prominent | Elevated | Normal |
Hunter Serotonin Toxicity Criteria
Sensitivity 84%, specificity 97%. Requires a serotonergic agent plus any of: spontaneous clonus; inducible clonus with agitation or diaphoresis; ocular clonus with agitation or diaphoresis; tremor with hyperreflexia; or hypertonia with temperature above 38°C plus ocular or inducible clonus.
Two caveats
In severe SS, extreme hypertonicity can mask clonus and hyperreflexia, rendering it clinically indistinguishable from NMS. And some agents, metoclopramide among them, have both antidopaminergic and serotonergic activity; simultaneous NMS and SS has been reported. When a patient takes both classes, pronounced CK elevation, leukocytosis, transaminitis, and low serum iron favor NMS.
Where the diagnosis is missed
Atypical presentations
NMS is best conceptualized as a dimensional spectrum rather than an all-or-none diagnosis. Roughly half the cases in the AMSP pharmacovigilance program followed an abortive course not meeting full ICD-10 or DSM-IV criteria, and DSM-5 deliberately softened the criteria to capture early or partial forms.
Second-generation antipsychotics
SGAs produce a largely similar picture with less rigidity and fewer extrapyramidal signs, driven predominantly by clozapine, and considerably lower mortality (3.0% versus 16.3%). Clozapine-associated NMS characteristically presents with tachycardia, mental status change, and diaphoresis while fever, rigidity, and CK elevation are attenuated or absent. In an FAERS analysis, ziprasidone carried the strongest signal and lurasidone the weakest.
Afebrile and rigidity-free NMS
Afebrile NMS accounts for roughly 8% of cases, and NMS without rigidity or with a normal CK is well described, particularly with SGAs, because CK may be normal at onset and rise only days into the illness. CK is inherently nonspecific: up to 70% of psychosis inpatients have an elevated CK without NMS. Do not prematurely exclude NMS when severe rigidity or hyperthermia is absent, since these may be early or impending forms.
Parkinsonism-hyperpyrexia syndrome
The mirror-image mechanism: an NMS-identical crisis triggered by abrupt reduction or cessation of dopaminergic therapy in Parkinson disease (non-adherence, perioperative withholding, dysphagia or malabsorption, DBS battery failure, amantadine withdrawal). It occurs in 0.3 to 3.6% of Parkinson patients with 4 to 15% mortality, and its treatment is rapid reintroduction of dopaminergic therapy. In one dementia clinic series, levodopa discontinuation caused 12 of 19 NMS episodes.
Antiemetic-induced NMS
Metoclopramide, prochlorperazine, droperidol, and promethazine cases are especially treacherous because the offending drug is not being used for psychosis and clinicians may not recognize its dopamine-blocking properties. These cases occur at standard doses and often lack typical laboratory abnormalities. Older adults with dementia are highly vulnerable, with atypical presentations in roughly 67% and frequent concurrent infection that confounds a diagnosis of exclusion.
Management
Severity-tiered treatment
No randomized controlled trials exist for any NMS-specific therapy; all recommendations derive from case series, retrospective cohorts, and expert consensus. The foundation is immediate withdrawal of the offending agent plus aggressive supportive care: stop the dopamine blocker, or in parkinsonism-hyperpyrexia syndrome reinstate the dopaminergic agent; give IV fluids targeting urine output of roughly 200 to 300 mL/h for rhabdomyolysis; cool the patient and give acetaminophen; correct electrolytes and acidosis; provide DVT and stress-ulcer prophylaxis; and deliver ICU-level airway and autonomic support using clonidine, calcium-channel blockers, or dexmedetomidine rather than reintroducing antipsychotics.
| Agent | Role and severity | Dosing | Notes |
|---|---|---|---|
| Lorazepam | First-line, mild to moderate | 1 to 2 mg IM or IV every 4 to 6 hours | Also treats catatonia; favorable safety profile. |
| Dantrolene | Moderate to severe rigidity and hyperthermia | 1 mg/kg IV, then 0.25 to 0.75 mg/kg every 6 hours; maximum 10 mg/kg/day | Hepatotoxic; mortality data are mixed and confounded by indication. |
| Bromocriptine | Dopamine agonist, moderate to severe | 2.5 mg every 6 to 8 hours, up to 40 mg/day | Contraindicated in serotonin syndrome. |
| Amantadine | Dopaminergic alternative | 100 mg PO, up to 200 mg every 12 hours | Also anticholinergic. |
| ECT | Refractory or life-threatening disease | Bitemporal, daily when severe | Associated with the lowest mortality in severe NMS series. |
Evidence for specific pharmacotherapy is genuinely conflicting. A 405-patient case-series analysis found no overall mortality benefit for dantrolene but a significant benefit for specific therapy (dantrolene, bromocriptine, or ECT) in severe NMS, 10% versus 30% mortality, whereas a nationwide Japanese cohort and a VA analysis found higher mortality with dantrolene, almost certainly confounding by indication.
Rechallenge
After recovery, resume antipsychotics only after at least 2 weeks, using a different, lower-potency agent at low dose with slow titration. Premature rechallenge is the principal recurrence risk, though documented recurrence rates are low, around 4%.
Treatment diverges
NMS versus serotonin syndrome
For serotonin syndrome, discontinue serotonergic agents, give benzodiazepines at every severity level, and add cyproheptadine, a 5-HT2A antagonist, for moderate to severe cases. Severe SS with temperature above 41.1°C requires sedation, nondepolarizing neuromuscular paralysis (avoid succinylcholine), and intubation, because the hyperthermia is muscle-generated and antipyretics have no role. Avoid physical restraints, which worsen isometric muscle activity and lactic acidosis. Several NMS therapies are contraindicated or unhelpful in SS: bromocriptine, dantrolene, and propranolol.
| Feature | NMS | Serotonin syndrome |
|---|---|---|
| Onset and resolution | Days to weeks; resolves over 7 to 11 days | Hours; usually resolves in under 24 hours |
| Benzodiazepines | First-line for mild to moderate disease | Essential at every severity level |
| Specific therapy | Dantrolene, bromocriptine, or amantadine | Cyproheptadine 12 mg, then 2 mg every 2 hours as needed, maintenance 8 mg every 6 hours |
| ECT | For refractory cases | Not indicated |
| Antipyretics | May help, hyperthermia is partly central | No role, heat is muscle-generated |
| Mortality | 5 to 15% | Roughly 5% |
Bottom line
Summary and evidence gaps
NMS is a dopamine-blockade-driven hyperthermic emergency defined by lead-pipe rigidity, fever, dysautonomia, and altered mentation evolving over days, with markedly elevated CK. It is most reliably separated from serotonin syndrome by its slower tempo and its hypokinetic, hyporeflexic motor picture, versus the rapid-onset clonus, hyperreflexia, and myoclonus of SS. Maintain a low threshold for the diagnosis in atypical presentations: clozapine or aripiprazole cases lacking rigidity or fever, antiemetic-induced cases, and parkinsonism-hyperpyrexia from dopaminergic withdrawal, because these are where recognition most often fails.
Gaps remain. The low recurrence rate on rechallenge and the occurrence of NMS with low-D2-affinity clozapine both argue against a simple D2-blockade model. No RCTs guide NMS pharmacotherapy, optimal dosing and duration are undefined, and observational mortality data on dantrolene are heavily confounded. Evidence for cyproheptadine in SS is very low quality. Whether atypical or abortive NMS represents early disease or a distinct entity, and precisely where NMS ends and malignant catatonia begins, remain unresolved.
Sources
Selected references
- Wijdicks, E. F. M., & Ropper, A. H. (2024). Neuroleptic malignant syndrome. New England Journal of Medicine, 391(12), 1130-1138.
- Boyer, E. W., & Shannon, M. (2005). The serotonin syndrome. New England Journal of Medicine, 352(11), 1112-1120.
- Gurrera, R. J., et al. (2011). An international consensus study of neuroleptic malignant syndrome diagnostic criteria using the Delphi method.
- Dunkley, E. J., et al. (2003). The Hunter Serotonin Toxicity Criteria.
- Trollor, J. N., Chen, X., Chitty, K., & Sachdev, P. S. (2012). Comparison of neuroleptic malignant syndrome induced by first- and second-generation antipsychotics. British Journal of Psychiatry, 201(1), 52-56.
- Picard, L. S., et al. (2008). Atypical neuroleptic malignant syndrome: Diagnostic controversies and considerations. Pharmacotherapy, 28(4), 530-535.
- Oruch, R., et al. (2017). Neuroleptic malignant syndrome: An easily overlooked neurologic emergency. Neuropsychiatric Disease and Treatment, 13, 161-175.
- Perry, P. J., & Wilborn, C. A. (2012). Serotonin syndrome vs neuroleptic malignant syndrome. Annals of Clinical Psychiatry, 24(2), 155-162.
- He, J., et al. (2025). Investigation into neuroleptic malignant syndrome triggered by atypical antipsychotics: Insights from the FDA Adverse Event Reporting System Database. Journal of Affective Disorders, 386, 119481.
- Wilson, J. E., Oldham, M. A., Francis, A., et al. (2025). Resource document on catatonia. American Psychiatric Association.
Return to provider resources.
Keep reading
All clinician tools pages →- Clinician toolsMedical workup in first-episode psychosisLabs, imaging, and the organic causes that must be excluded.For clinicians
- Clinician toolsPsychosis triage protocolA decision pathway for choosing the right level of care when psychosis is suspected.For everyone
- Clinician toolsClinical cheat sheetsOne-page diagnostic checklists for busy clinicians.For clinicians
- Clinician toolsProvider resourcesReferral information, tools, and materials for clinicians.For clinicians
- Clinician toolsScreening questionnairesPQ-16, PQ-B and other downloadable screening instruments.For clinicians