Correspondence *University of California San Francisco, Fresno, University Center Specialty Medical Center, 2828 Fresno Street, Fresno, CA 93721, USA

Email twarwick@fresno.ucsf.edu

The case

A 78-year-old man presented to the emergency department with a 5-day history of fatigue, generalized weakness, lethargy and confusion, and a 1-day history of progressive muscle stiffness. He had been at baseline with no antecedent illness before onset of these symptoms. The patient had a 5-year history of progressive memory problems, and a Mini Mental State Examination score of 26 out of 30 was documented by his primary care physician one week before admission. The patient had been started on donepezil 5 mg daily one week before presentation to treat his dementia. He had an unclassified psychiatric illness, which had started 22 years earlier. Further questioning of his family identified symptoms that met diagnostic criteria for schizoaffective disorder (Diagnostic and Statistical Manual of Mental Disorders 4^th edition). The patient was first treated with haloperidol, but was switched to olanzapine 10 mg daily 10 years before presentation and did not experience adverse effects. His mother had Parkinson's disease (PD). He lived at home with his wife and had a 40 pack-year history of smoking. The patient had no history of drug or alcohol use, and no known drug allergies.

On presentation, the patient was diaphoretic, tachycardic (heart rate 110 beats/min), hypertensive (blood pressure 165/90 mmHg), mildly tachypneic (respirations 18 breaths/min), and normothermic (temperature 36.5 °C). His general physical exam was unremarkable. On neurological exam the patient was lethargic and minimally arousable to noxious stimulation—he grimaced to noxious stimulation but was unable to withdraw. He was nonverbal and did not follow commands. Cranial nerve examination revealed bilateral facial dystonia, blepharospasm, tonic jaw deviation to the left, and torticollis to the right. His pupils were pinpoint and minimally reactive. The patient was unable to cooperate with additional cranial nerve examination. Motor exam revealed severe diffuse rigidity in all four extremities, with flexor posturing in the upper extremities and extensor posturing of the lower extremities. Deep tendon reflexes were difficult to assess because of rigidity, but the patient had flexor plantar responses. Gait and coordination testing were deferred.

Laboratory evaluation revealed normal serum electrolyte levels and normal CBC results. Levels of serum aspartate aminotransferase and creatine kinase were mildly elevated (60 units/l and 510 units/l, respectively), with negative troponin results and normal cardiac enzymes. Blood cultures, thyroid function, and cerebrospinal fluid analyses were normal. Urinalysis showed no sign of infection but revealed myoglobinuria. Chest X-ray was normal. Electrocardiogram revealed sinus tachycardia and brain CT scan was normal.

A neurologist diagnosed the patient with a variant of neuroleptic malignant syndrome (NMS). He was admitted to the intensive care unit and started on bromocriptine 2.5 mg every 8 h via a nasogastric tube, intravenous dantrolene 2.0 mg every 6 h, and intravenous hydration. Donepezil and olanzapine were discontinued. The patient's oxygen saturation level was normal and his airway was stable, so intubation was deferred. Sedation was held in an attempt to maintain the airway, as the family was reluctant to intubate. Pain management was attempted with intravenous morphine. Electroencephalography (EEG) showed severe generalized slowing of cortical activity consistent with severe generalized encephalopathy, ruling out tonic seizures. Despite medical therapy and improvement in rigidity and mental status, on day three the patient went into respiratory distress requiring intubation and mechanical support. Chest X-ray revealed right lower-lobe lung infiltrate consistent with aspiration pneumonia, and the patient was started on antibiotics. By day four the patient's level of serum creatine kinase decreased to normal (73 units/l). On day five the patient was extubated and transferred to an inpatient ward, where he was slowly weaned off dantrolene. He remained mildly parkinsonistic, and, according to his family, might have had some parkinsonian traits before this illness, including slow walking, talking and generalized movements. He was started on carbidopa–levodopa tablets of 25 mg/100 mg daily, titrated up to one tablet three times a day, and transferred to inpatient rehabilitation. One month after discharge the patient was seen in follow-up neurology clinic and his functioning had returned to near baseline.

Discussion of diagnosis

NMS is a rare, life-threatening drug reaction (Box 1) characterized by fever, muscle rigidity, autonomic dysfunction and altered mental status.¹ Although the diagnostic criteria for NMS are very specific (Box 2), they can have broad clinical presentations. Altered mental status can range from confusion to obtundation and coma.¹ Muscle rigidity can range from hypertonicity to frank 'lead pipe' rigidity, opisthotonos, dystonia, blepharospasm, trismus, and oculogyric crisis.¹ Autonomic dysfunction can range from mild tachycardia and hypertension to fulminant cardiovascular collapse. Fever has always been considered a classic feature of NMS but was not seen in this patient. Some clinicians have suggested that NMS variants, where patients meet the majority of the diagnostic criteria but not all, are part of a spectrum of disorders or formes frustes. This theory of disease variants or a spectrum of disorders has been the focus of debate since the identification of NMS.

NMS is thought to be caused by CNS dopamine depletion or blockade leading to increased muscle tone and rigidity.² Dopamine alterations alone, though, cannot account for the symptomatology of NMS, suggesting that other neurotransmitters such as acetylcholine or serotonin might have a role in the condition.

Donepezil is an acetylcholinesterase inhibitor that inhibits the breakdown of the neurotransmitter acetylcholine and increases its availability. The question then becomes "How did this medication illicit the described symptoms in this patient?" Given that the patient was already taking an atypical neuroleptic, could there have been some alteration of metabolism of that agent, resulting in increased plasma levels or potency that led to the subsequent syndrome?

Both donepezil and olanzapine are known to be metabolized by the cytochrome P450 (CYP) enzyme system and by glucuronidation. Donepezil is metabolized mainly by CYP3A4 and to a lesser extent by CYP2D6 and glucuronidation. Olanzapine is metabolized by CYP1A2, CYP2D6 and the flavin-containing monooxygenase system, as well as by glucuronidation. Neither of these drugs is known to inhibit or induce CYP enzymes. Although both drugs are highly protein bound, neither has been associated with displacement of other drugs from their protein binding sites. It is unlikely that the pharmacokinetics of olanzapine were altered in this patient following initiation of donepezil. After being metabolized, donepezil is broken into four major metabolites, two of which are active. The effects of these metabolites on the dopaminergic system are unknown.

Alternatively, donepezil may cause a dopamine–acetylcholine imbalance within the striatum. There have been several reports of patients with pre-existing PD or dementia with Lewy bodies whose condition worsened following treatment with donepezil.^{3, 4, 5} Perhaps these patients suffer from a central dopaminergic abnormality that predisposes them to drug reaction. It is known that patients with PD are sensitive to neuroleptics and susceptible to their adverse effects, especially NMS.¹ In the patient discussed in this Case Study, it is possible that an increase in acetylcholine levels led to a dopamine–acetylcholine imbalance in his already compromised dopaminergic system. It is unclear if his parkinsonian symptoms resulted from PD or from years of neuroleptic use.

Differential diagnosis

When evaluating patients with possible NMS, other conditions should be considered and ruled out (Box 3). Given that this patient was receiving an antipsychotic agent and had been started on a new medication just two days before the onset of symptoms, we strongly suspected that a drug interaction was the cause of his symptoms. EEG ruled out status epilepticus and confirmed that the patient had CNS dysfunction. A normal electrolyte panel eliminated tetany, and the normal lumbar puncture and blood cultures ruled out infection. The patient's toxicology screen was negative, and he had no recent exposure to anesthesia or heat. He had not been on any primary serotonin agents.

It can often be difficult, if not impossible, to differentiate between lethal catatonia and NMS⁶—the clinical presentations can be the same. Given this patient's medication regimen and the emergent need for treatment of his NMS symptoms, we proceeded with a treatment protocol for NMS. Medication-induced dystonic reaction is a condition that also must be considered in the differential diagnosis, although the reaction usually starts soon after administration of medication and is rarely life-threatening. The one feature that distinguishes between NMS and drug-induced dystonia is altered mental status. In this patient EEG confirmed the observed generalized encephalopathy.

Treatment and management

There have been no controlled clinical trials to establish a standard of care for patients with NMS, and there are many opinions regarding which medications are best. There are, however, certain universal approaches that all clinicians must take when treating patients with suspected NMS. Successful management of NMS requires rapid diagnosis followed by supportive care, stabilizing measures, and disease-specific treatment. As in any neurological emergency, airway, breathing and circulation should be stabilized before any other interventions are started. As with any patient with altered mental status, patients with NMS should receive thiamine, dextrose and naloxone to treat possible hypoglycemia and drug overdose. As hyperthermia considerably increases morbidity and mortality, patients should be cooled aggressively with acetaminophen, cooling blankets, cooled intravenous fluids, and ice packs if needed. The patient in this Case Study was afebrile, so temperature management was not an issue. All neuroleptic, anti-emetic and new medications should be discontinued, and the patient should be admitted to a closely monitored unit and put on continuous cardiac monitoring. In the case of this patient, olanzapine and donepezil were discontinued immediately upon admission. Volume depletion and hypotension should be treated with intravenous fluids. Patients with acute NMS can also experience considerable pain, so adequate levels of analgesic medications should be used. This patient's pain was managed with intravenous morphine, titrated to control pain without oversedation.

There are a variety of disease-specific medications available to treat NMS. In the early phases of NMS, most sources suggest the use of benzodiazepines such as lorazepam.⁷ If these treatments are ineffective, medications such as bromocriptine, dantrolene, dopamine agonists, levodopa and amantadine can be used (Box 4).⁷ The patient in this Case Study was immediately started on bromocriptine and dantrolene. High-dose benzodiazepines were avoided as the family was hesitant to allow intubation at the time of admission and we were concerned that oversedation would lead to loss of airway control. Electroconvulsive therapy has been shown to be effective in patients with severe NMS; case reports suggest that a response is usually apparent after a few treatments.⁸ Electroconvulsive therapy was not available for the treatment of this patient. Methylprednisolone has been shown to be effective in patients with PD who have NMS as a result of medication withdrawal.⁹ The present patient's symptoms resulted from a drug interaction, not drug withdrawal, so he was not a candidate for steroid therapy. For those patients whose rigidity and rhabdomyolysis is so severe that renal failure is eminent, nondepolarizing paralytic agents can be used. NMS can take two weeks or longer to fully resolve, and treatment should be continued throughout that period.^{1, 2}

After the emergent phase of his illness, the patient in this Case Study continued to have some mild parkinsonian symptoms. Given his family history of PD, treatment with carbidopa–levodopa was initiated, resulting in considerable alleviation of his parkinsonian symptoms.

Conclusions

Drug interactions can cause morbidity and mortality in psychiatric and geriatric populations. The patient described in this Case Study, who had a pre-existing parkinsonian syndrome, experienced a variant of NMS secondary to an interaction between olanzapine and donepezil. This report and others suggest that caution should be used when starting donepezil in patients who are on an anti-psychotic medication or who have a previous history of Parkinson-like symptoms. The risk-to-benefit ratio should be carefully considered and alternative medications considered in such patients.

References

1. Pelonero A et al. (1998) Neuroleptic malignant syndrome: a review. Psychiatr Serv 49: 1163–1172 | PubMed | ChemPort |
2. Bottoni TN (2002) Neuroleptic malignant syndrome: a brief review. Hosp Physician 38: 58–63
3. Rozzini L et al. (2004) Severe worsening of parkinsonism in Lewy body dementia due to donepezil. Neurology 63: 1543–1544 | PubMed |
4. Bourke D and Druckenbrod RW (1998) Possible association between donepezil and worsening Parkinson's disease. Ann Pharmacother 32: 610–611 | Article | PubMed | ChemPort |
5. Morita S et al. (2004) A patient with probable dementia with Lewy bodies, who showed catatonia induced by donepezil: a case report [Japanese]. No To Shinkei 56: 881–884 | PubMed |
6. Lee JW (2007) Catatonic variants, hyperthermic extrapyramidal reactions, and subtypes of neuroleptic malignant syndrome. Ann Clin Psychiatry 19: 9–16 | Article | PubMed |
7. Hu S and Frucht S (2007) Emergency treatment of movement disorders. Curr Treat Options Neurol 9: 103–114 | PubMed |
8. Trollor JN and Sachdev PS (1999) Electroconvulsive treatment of neuroleptic malignant syndrome: a review and report of cases. Aust N Z J Psychiatry 33: 650–659 | Article | PubMed | ChemPort |
9. Sato Y et al. (2004) Efficacy of methylprednisolone pulse therapy on neuroleptic malignant syndrome in Parkinson's disease. J Neurol Neurosurg Psychiatry 75: 510–515

Contact the journal about this article

Subject areas under which this article appears: Injury, repair and rehabilitation