We report on a patient with the clinical and biochemical features of the so-called neuroleptic malignant syndrome, occurring more than a decade after her last exposure to phenothiazines, but while taking the cyclopyrrolone zopiclone. We consider possible mechanisms underpinning the development of this clinical syndrome in the light of these findings and current models of basal ganglia dysfunction.

A 62 year old woman was found collapsed at home. On admission to hospital she was alert, eye opening and moving limbs spontaneously, but her affect was flat and she was unable to give any account of herself. She was pyrexial (38.9°C) and clinically dehydrated with a pulse rate of 110/min and a blood pressure of 90/60 mm Hg. She was mute and akinetic with a jaw tremor, but there was no muscle tenderness. There was profound axial and limb rigidity with opisthotonos, making it impossible to assess power. Psychiatric review found no evidence for a catatonic psychotic state. She was rehydrated with IV crystalloids, but on the day after admission had an episode of tachycardia (pulse rate 140/min) and profound hypotension (BP unrecordable) which responded to IV colloids. After this, she was noted to be less rousable, and her level of consciousness continued to fluctuate between alert and unrousable for the next three days.

Through relatives, friends, and her previous medical notes, a history of alcohol dependence came to light, dating back to 1984; she had been admitted on at least three occasions in the past for alcohol detoxification, but continued to indulge in occasional binges and was on long term nitrazepam (10 mg nightly) because of difficulty sleeping. However, there was no history of recent binge drinking, or of attempted withdrawal from benzodiazepine. Also, in 1984 the patient had presented with paranoid delusions and auditory hallucinations which were thought to be due to a psychosis secondary to alcohol withdrawal; brain CT at this time was normal. She was then seen by a psychiatrist and the diagnosis was revised to schizophrenia; she was treated initially with trifluoperazine (5 mg twice daily).

for less than 1 month, then with pimozide (6 mg daily) and procyclidine (5 mg thrice daily) for 5 months at which time medications were stopped. She was next seen in 1996 when she was admitted with an acute confusional state which was thought to be alcohol related. Brain CT was again normal, and an EEG showed diffuse beta activity consistent with her benzodiazepine use. At the time of discharge she was taking zopiclone (7.5 mg nightly) in addition to nitrazepam, but no neuroleptic medication was used.

At the time of her acute admission, investigations disclosed a mild neutrophil leucocytosis (10.1×10⁹/l), hypernatraemia (152 mmol/l), hypoalbuminaemia (22 g/l), and raised erythrocyte sedimentation rate (90 mm/h) and C-reactive protein (157 mg/l). Creatine kinase was raised (maximum 978 U/l, first measured 7 days after presentation; normal<190 U/l; 100% MM isoenzyme) as was lactate dehydrogenase (maximum 839 U/l, normal<450 U/l). Cultures of blood, urine, faeces, and CSF were negative. Brain CT was normal; MRI showed widespread ischaemic changes in the cerebral deep white matter but the basal ganglia were spared and no lesions were visible in the brainstem. Analysis of CSF showed a moderately raised protein (0.57 g/l) but was normal in all other respects; CSF polymerase chain reaction was negative for various infectious agents (herpes simplex virus, varicella zoster virus, adenovirus, enterovirus, mumps, Mycoplasma). Toxicology screen was negative for phenothiazines and enquiries with both her consultant psychiatrist and general practitioner afforded no evidence for recent neuroleptic prescription. Before the availability of this information, a working diagnosis of neuroleptic malignant syndrome had been made, based on the presence of the appropriate clinical features.1 Despite appropriate medical management, the rigidity had persisted for 10 days and therefore, as soon as the diagnosis of neuroleptic malignant syndrome was considered, the patient was treated with bromocriptine, initially 1 mg twice daily rising to a maximum of 2.5 mg twice daily. Over the first three days on this regime, the rigidity improved dramatically, tone returning to normal. Neurological examination at this time was normal; particularly, there were no abnormal brainstem signs, although there was some evidence of mild cognitive impairment.

The neuroleptic malignant syndrome, first named by Delay and Deniker,2 is characterised by hyperpyrexia, hypertonus, fluctuating level of consciousness, and autonomic disturbances; the presence of these features is thought to be essential to establish the diagnosis.1 Most cases have been associated with neuroleptic use, although levodopa withdrawal may also precipitate a similar state. Loss of dopaminergic drive in the basal ganglia has been suggested as a common feature in both of these situations. Similar phenomena, labelled as acute lethal catatonia or fatal catatonia, were reported in the pre-neuroleptic era.3 Hence, current neuroleptic use is not an absolute requirement for the occurrence of this clinical syndrome, as also shown in the reported case. It remains possible that medication may act as a trigger for some independent process, for which concurrent organic brain disease is recognised to be a predisposing factor.1 Perhaps previous neuroleptic use may also sensitise the basal ganglia in some way to the development of these features.

Could zopiclone have contributed to the pathogenesis of the syndrome? At first sight this would seem unlikely; It is a cyclopyrrolone with a pharmacological profile similar to that of short acting benzodiazepines, its actions mediated through increased GABA activity in the brain. Hence, it has muscle relaxant effects, as well as sedative, anxiolytic, and anticonvulsant properties.4However, GABA is an important neurotransmitter in the basal ganglia, present in striatopallidal, striatonigral, pallidothalamic, pallidosubthalamic, and nigrothalamic fibres. In movement disorders, it has been postulated that within the basal ganglia-thalamocortical motor circuit there are two separate projection systems from the putamen: a direct pathway of GABA/Substance P neurons to motor portions of the internal segment of the globus pallidus and substantia nigra pars reticulata, providing positive feedback to precentral motor fields; and an indirect pathway of GABA/enkephalinergic neurons influencing basal ganglia output through a sequence of connections involving the external segment of the globus pallidus and the subthalamic nucleus, providing a negative feedback to precentral cortex.5 Shifts in the balance of activity within these pathways may alter globus pallidus/substantia nigra pars reticulata output and hence result in hypokinesia or hyperkinesia; for example, hypokinesia is thought to result from increased pallidothalamic inhibition and enhanced conduction in the indirect pathway, whereas reduced conduction through the direct pathway, secondary to reduced pallidothalamic inhibition, results in hyperkinesia.5 For this model, it is of note that a selective loss of indirect pathway GABA/enkephalin neurons is found in Huntington’s disease,6 and that GABAmimetics (progabide) have been reported to iron out on-off fluctuations in Parkinson’s disease.7 Hence, in the light of this model, we suggest that our patient may have had increased brain GABA activity as a result of zopiclone (with or without nitrazepam) treatment, resulting in increased pallidothalamic inhibition and enhanced conduction in the indirect pathway, and thus producing profound hypokinesia in the absence of recent neuroleptic treatment.