Naloxone-responsive acute dystonia and parkinsonism following general anaesthesia


Correspondence to: Dr Iselin-Chaves


Various movement disorders such as dystonia may acutely develop during or at emergence from general anaesthesia in patients with or without pre-existing Parkinson disease. These movements are triggered by a variety of drugs including propofol, sevoflurane, anti-emetics, antipsychotics and opioids. The postulated mechanism involves an imbalance between dopaminergic and cholinergic neurotransmitters in the basal ganglia. We report an acute, severe and generalised dystonic reaction in an otherwise healthy woman at emergence from general anaesthesia, dramatically reversed by the administration of naloxone, pointing to a potential role of the fentanyl and morphine that the patient had received. Recent literature on the mechanisms of abnormal movements induced by opioids are discussed. The severity of the reaction with usual doses of opioids, in a patient with no prior history of parkinsonism, led to further investigation that demonstrated the possibility of an enhanced susceptibility to opioids, involving a genetically determined abnormal function of glycoproteine-P and catechol-O-methyltransferase.

Various movement disorders may acutely develop during the peri-operative period in patients undergoing surgical procedures under either local or general anaesthesia. These abnormal movements have been reported following the use of a variety of drugs including propofol, sevoflurane, anti-emetics, antipsychotics and opioids [1–7]. Patients with basal ganglia dysfunction such as Parkinson’s disease appear to be at particular risk, although dystonic movements have also been reported in patients with no prior history of movement disorders. We report an acute, severe and generalised dystonic reaction associated with features of parkinsonism in an otherwise healthy woman following general anaesthesia. This reaction stopped after administration of naloxone, pointing to a potential role of the fentanyl and morphine the patient had received.


A 35-year-old female (163 cm, 52 kg, ASA physical status 2) was scheduled for a diagnostic hysteroscopy, intra-uterine contraceptive device insertion and laparoscopic myomectomy. Her birth had been complicated by coiling of the umbilical cord and she needed resuscitation but there were not any features in her later development that suggested any degree of brain damage. Her past medical history was unremarkable except for migraine, hay fever and an episode of skin rash following penicillin. She was a non-smoker and reported no drinking habits or recreational drug abuse. Over the last 3 months, she had been taking St John’s wort for depression. She had undergone general anaesthesia on three previous occasions. Two were uneventful including one for tonsillectomy in childhood (anaesthetic technique unknown) and another for termination of pregnancy performed 4 months earlier (duration of anaesthesia 20 min; anaesthetic regimen alfentanil 500 μg, propofol 180 mg, ocytocin 5 U and paracetamol 1 g). One month after this procedure, she underwent a uterine curettage for retained products of conception, and reported abnormal movements when emerging from anaesthesia, that resolved spontaneously after 2 h (duration of anaesthesia 40 min; anaesthetic regimen: alfentanil 1000 μg, propofol 500 mg, ocytocin 5 U, paracetamol 1 g, and ketorolac 30 mg). For the present procedure, she received an oral combination of misoprostol 200 μg and midazolam 7.5 mg before surgery. Anaesthesia was induced with propofol 150 mg, lidocaine 50 mg, fentanyl 150 μg and rocuronium 35 mg, and maintained with sevoflurane in oxygen and air, bolus doses of fentanyl (total 200 μg) and rocuronium (total 20 mg). The surgery lasted 2 h and was uneventful. She also received clindamycin 600 mg pre-operatively and ketorolac 30 mg during the procedure. Upon emergence from anaesthesia, she acutely developed a severe and painful dystonia involving all four limbs. Dystonia evolved into a relapsing-remitting fashion, with dystonic episodes coming every 5 min and lasting between 1 and 2 min. Because of the pain, she received paracetamol (1 g) and morphine (2 mg); this was followed by an increase in the severity of abnormal movements and was accompanied by marked hypomimia with abolished blink rate, hypophonia, diffuse rigidity and severe bradykinesia. At this stage dystonic movements were still present in her hands but parkinsonian signs predominated. She was conscious, able to talk, distressed and there was tachycardia (100–140 beats.min−1) during the episodes, but she remained normotensive and well oxygenated. Several bolus doses of intravenous clonidine (total 105 μg) and midazolam 1mg failed to produce any improvement in her dystonia. Two hours after the onset of symptoms, a consulting neurologist administered 40 μg of iv naloxone, which immediately and completely relieved the dystonic movements.

Four hours later, she received paracetamol 1 g, ketorolac 30 mg and three carefully titrated doses of of 25 μg fentanyl to relieve pain, with no recurrence of abnormal movement. She was kept overnight for observation and remained stable. She was discharged symptom-free the following day. Three weeks later, a detailed neurological examination and MRI scan of the brain were entirely normal, in particular excluding any structural abnormality of the basal ganglia. Polymorphisms of glycoprotein-P (P-gp) and catechol-O-methyltransferase (COMT) were determined to identify potential increased susceptibility to opioids. The patient was heterozygous for the two tested mutations (C3435T: C/T and G2677T/A:G/T) of the P-gp, which could be associated with a low expression and reduced activity of P-gp. She was also found to be homozygous for the Met/Met genotype of COMT (Val158Met), a polymorphism known to be associated with a four-fold reduction of COMT activity [8]. These results were consistent with an increased sensibility to normal adult doses of opioids.


Acute dystonia and parkinsonism associated with general anaesthesia have been reported in patients with or without pre-existing movement disorders. These cases were attributed to various anaesthetic agents (propofol [4], sevoflurane [3, 5]), and to other drugs frequently associated with anaesthesia (anti-emetic, antipsychotic [2], gabapentin [9]). The postulated mechanism by which dystonic reactions are triggered involves an imbalance between dopaminergic and cholinergic neurotransmitters in the basal ganglia[4], for example as a result of dopamine receptor blockade by neuroleptics and related compounds. There have also been several reports of dystonic movements attributed to opioids [6, 7]. These abnormal movements can take the form of limb or neck flexion, opisthotonos, masseter spasm, and with large doses or fast administration of potent opioids, chest wall and generalised muscle rigidity, the latter symptoms being well known in the anaesthesia community [6]. Abnormal movements can occur at induction of anaesthesia [7] or at emergence [6].

Mechanisms underlying the production of abnormal movements apparently induced by opioids are not fully understood. The striatum, the main input nucleus of the basal ganglia, is known to be enriched with various opioid receptors that are believed to play a crucial role not only in reward and addiction, but also in motor control [10]. Postsynaptic opioid δ, κ and μ receptors in the striatum regulate the function of the medium spinal neurons through a negative feedback mechanism, and presynaptic opioid receptors of the nigrostriatal neuron terminals appear to inhibit dopamine release [11]; together these lead to an inhibitory effect upon dopaminergic transmission [12]. Accordingly, morphine, a μ receptor agonist, has been shown to inhibit the release of striatal dopamine in vitro [13] and experiments using superfused rat brain tissue have shown that the inhibitory effects of opioids on the release of dopamine from nigral neurons is mediated by presynaptic κ receptors [14]. Finally, in Parkinson’s disease, taken as an example of altered basal dopaminergic transmission, drugs such as fentanyl acting on opioid receptors may trigger severe worsening of tremor, rigidity and bradykinesia [6, 15, 16] and induce dystonia [7].

Dystonia during or following general anaesthesia has been reported to be decreased or relieved by anticholinergic drugs such as benztropine [4] or dyphenydramine [17]. On the other hand, in our patient, the role of opioids is mainly suggested by the unequivocal effect of naloxone administration. However, because of the severity of the reaction with such small doses of opioids, in a patient with no prior history of parkinsonism, we explored the possibility that she exhibited an enhanced susceptibility to opioids, involving a genetically determined abnormal function of P-gp and COMT.

Glycoprotein-P is an efflux transporter whose expression is particularly high in the intestine, kidney and blood-brain barrier. The function of P-gp is to reduce intracellular or intra-organ drug accumulation through active eflux, therefore limiting the entry in the central nervous system. The expression and function of P-gp show an inter-individual variability partly explained by a genetic polymorphism [18]. The single-nucleotide polymorphism 3435CT is of interest as the genotype 3435TT has been associated with a two-fold lower duodenal expression of the P-gp than the 3435CT or 3435CC genotypes. In knock-out mice and rats, morphine is a P-gp substrate. In human, the influence of the P-gp polymorphism on morphine pharmacokinetics and pharmacodynamics is suggested by higher plasma concentrations in patients whose P-gp is inhibited by quinidine. Moreover, higher cerebrospinal fluid morphine concentrations have been noted in homozygote 3435TT. Data about fentanyl are contradictory but some authors mention fentanyl as a P-gp substrate [19]. The P-gp genotypisation of this patient showed that she was indeed heterozygous for the two tested mutations (C3435:C/T and G2677T/A:G/T). Therefore, one possible explanation underlying her dystonic movements relies on an increased sensitivity to opioids owing to a facilitated entry of these compounds into the central nervous system through the blood-brain barrier.

Genotyping of COMT in this patient showed that she belonged to a population group associated with weak COMT activity. The enzyme COMT catabolises catecholamines, and genetic polymorphism can reduce the activity of COMT. Human volunteers with a particular genotype (the Met/Met genotype) are more sensible to pain and have a higher density of μ opioid receptors than volunteers lacking this polymorphism [20]. A recent study showed that patients with this genotype need lower doses of morphine compared to control patients. This mechanism may also be involved in the motor striatum and contribute to the motor complications, despite peripheral reduction of COMT-mediated levodopa O-methylation increasing the levodopa delivery to the brain.

Altogether, the genetic constellation demonstrated in this patient and the dramatic response to naloxone strongly support a key role of opioids at the basis of the acute movement disorders reported here, and reasonably rule out a causal or contributing effect of the propofol and sevoflurane that she also received. Administration of morphine following that of fentanyl further worsened dystonic movements and induced parkinsonian signs, which may suggest some dose-dependency. This is further supported by the fact that small doses of fentanyl (25 μg) given for pain relief later in the recovery room did not elicit abnormal movements. Similarly, the small dose of alfentanil (500 μg) given during the second anaesthetic in the preceding months was not associated with motor complications, whereas a larger dose (1000 μg) given during the subsequent general anaesthetic was complicated by an acute dystonic reaction. During induction of anaesthesia, no abnormal movement was noticed immediately after administration of fentanyl, which may be the consequence of the neuromuscular blocker (rocuronium) given concomitantly. Dystonic movements only appeared immediately on emergence, probably owing to a cumulative effect of repeated boluses administered during maintenance of anaesthesia.

In conclusion, we describe a patient with no known basal ganglia disorder but carrying a potential genetic susceptibility to opioid effects and toxicity, who developed an acute, opioid-mediated dystonia-parkinsonism reaction at emergence of general anaesthesia, that resolved dramatically with naloxone. Anaesthetists and neurologists who face patients exhibiting rare, anaesthesia-related movement disorders should be aware of this uncommon cause, and consider the use of opioid antagonists when anticholinergic treatment have failed, especially if no anti-emetic and antipsychotic agents have been used.