Abstract
There are many reasons why electroconvulsive treatment (ECT) may be useful in the management of patients who develop neuroleptic malignant syndrome (NMS). First, NMS may be difficult to differentiate from lethal catatonia, and the life-saving role of ECT in the treatment of lethal catatonia is well described [1]. Second, ECT may be a suitable alternative to neuroleptics in patients who are psychotic but have had a previous epidosde of NMS [2], especially if a rechallenge with neuroleptics is not advisable. Third, ECT may in fact be an effective treatment of NMS, as has been reported by many investigators [3,4].
The use of ECT in the treatment of NMS per se is of practical and theoretical importance. It would be useful to know how ECT compares with the other commonly used treatments of NMS such as bromocriptine, dantrolene and benzodiazepines. As a comparative study of this kind is unlikely to be carried out given the rarity of the syndrome, important information can be gleaned from detailed anecdotal accounts of ECT in NMS. If ECT is indeed an effective treatment for NMS, the question that should be addressed is whether it poses any additional medical risks in NMS patients who usually have significant autonomic dysfunction [4]. We need to know whether the recovery of NMS is complete with ECT, or if residual symptoms persist. It has also been suggested that ECT may change the subsequent reaction to neuroleptics, thereby reducing the risk of a relapse of NMS upon neuroleptic rechallenge [5]). From an anaesthesiological perspective, it is important to know if NMS patients are vulnerable to the development of malignant hyperthermia (MH), a genetic disorder characterised by fever and muscle rigidity that occurs upon exposure to certain triggering agents, and whether special precautions are necessary during the anaesthetic procedure. The efficacy of ECT, if established, will have implications for our understanding of the pathogenesis of NMS.
In this paper, we review the published literature on the use of ECT in the treatment of NMS and describe nine patients of our own who were prescribed ECT for the management of episodes of NMS.
Method
Abibliographic search was performed on Medline dating back to 1966 using NMS, catatonia and ECT as key words. In addition, previously published reviews of this topic were examined [3,4,6,7], and cross-references checked for the sake of completion. The selection of articles was limited to those in the English literature. The case reported by Chen et al. [8] was omitted as it was not available to us for review. Cases were excluded if they did not meet at least one of five published sets of diagnostic criteria for NMS [9–13], and the syndrome was not clearly attributable to neuroleptic medication. This led to the exclusion of a number of cases previously included in other reviews where, in our opinion, a diagnosis of lethal catatonia [14–17] may have been more appropriate. Cases in which ECT was given following the resolution of NMS were also excluded [18–20].
For selection of our own cases, the charts of patients at a number of teaching hospitals in Sydney, whose psychiatrists had consulted the second author for treatment of NMS between 1988 and 1994, were reviewed in detail. Only those cases which met the research criteria for NMS as proposed in DSM-IV were included. Of the 30 patients diagnosed to have NMS, nine were treated with ECT and their case histories and the rationale for the use of ECT are described below. Seven of these were from the Prince Henry and Prince of Wales Hospitals with which the authors are directly affiliated.
Electroconvulsive therapy and anaesthesia procedures used in our cases
The procedure for ECT used in the new cases reported here was as follows. No pre-medication was used and the patient was preoxygenated for 2 min. The anaesthetic agents used for induction were thiopentone (n = 2), propofol (n = 6) or methohexitone (n = 1). The muscle relaxants used were suxamethonium (n = 3), about 0.5 mg/kg, or atracurium (n = 6), about 20 mg. Seizures were induced using a Kabtronics Model 271e ECT apparatus, which delivered a fixed stimulus of 300 mC using the following parameters: current of 800 mA; square wave pulse of 1.0 ms duration at 125 pulses per second for 3 s. Electrode placement was bifrontal for bilateral and over the right hemisphere (d'Elia position) for unilateral ECT. The frequency of treatment was three per week.
Case reports
Patient 1
P1, a 55-year-old man, was first diagnosed with schizoaffective disorder at the age of 25 and had had several admissions to hospital since then. He also suffered from maturity-onset diabetes mellitus controlled with diet. He was admitted involuntarily in an agitated state, with grandiose delusions, an expansive affect, auditory hallucinations, disinhibition and incoherent speech of about 2 weeks' duration. He was treated with haloperidol (up to 60 mg/day) and diazepam (up to 80 mg/day) orally for the first 6 days with only partial control of his behaviour, resulting in frequent seclusion and restraint. The medication was changed to chlorpromazine 600–800 mg/day while diazepam was continued. Owing to lack of improvement in his agitation and psychosis, he was given one bilateral ECT using thiopental sodium and suxamethonium with no intraoperative complications. On the following day, he was noted to have fever (38°C), muscle rigidity, tachycardia, labile blood pressure with intermittent hypertension, sweating and altered consciousness. A diagnosis of NMS was made, the neuroleptic drug and ECT were ceased, and he was investigated in a medical ward for other possible causes of fever with negative results. His condition worsened over the next 3 days and then fluctuated for about a week, with continued mutism, negativistic behaviour, extrapyramidal rigidity and intermittent fever. The plasma creatine kinase (CK) level rose to a peak of 4093 Units/L (normal range = 25–125 U/L) on the 12th day, and then reduced gradually.
Owing to his catatonic state and lack of spontaneous improvement, he was commenced on ECT for the treatment of NMS. Using propofol for general anaesthesia and atracurium for muscular relaxation, three unilateral treatments were given. The patient's physical health deteriorated at this stage, with the development of pneumonitis, decubitus ulcers and possible septicemia. Electroconvulsive therapy was temporarily ceased and bromocriptine at a dose of 5 mg b.i.d was administered. After 3 weeks, his physical condition improved to a sufficient degree to recommence ECT. During the course of the next 16 bilateral treatments, his mental and physical status continued to improve gradually. Features of NMS had reversed by the 10th bilateral ECT. Upon completion of the course of ECT, he continued to be disinhibited, hyperactive and occasionally incontinent of urine and faeces. He was treated with clonazepam and carbamazepine and admitted to a psychiatric hospital for a further 6 months in which he gradually recovered. Deficits in memory and frontal-executive functioning remained.
Patient 2
P2 was a 19-year-old man who had recently been admitted to the psychiatry ward for 1 month for the treatment of first onset schizophreniform disorder with haloperidol up to 15 mg/day. Two weeks after discharge, while being maintained on haloperidol 5 mg/day and benztropine 2 mg/day, he developed agitation, muscular rigidity, mild fever (37.3°C per axilla), tachycardia, sweating, tremor of hands and tongue, and hypertension (160/100 mmHg), and a plasma CK level of 1223 U/L. He was readmitted to hospital with the diagnosis of NMS, the neuroleptic drug was ceased, and he was treated with diazepam (15 mg/day) over the next few days.
A week after admission, his condition deteriorated with the development of mutism, posturing, negativism and waxy flexibility. He was noted to be febrile (up to 38.3°C), sweaty, with fluctuating blood pressure, and incontinent of urine and faeces. The plasma CK level was 109 U/L. It was uncertain whether the catatonic picture was a presentation of the primary psychiatric problem or an exacerbation of the NMS. In view of this, ECT was considered the most appropriate treatment, and this was administered using propofol and atracurium. His mental state improved considerably after the third unilateral ECT, and his temperature and autonomic symptoms returned to normal after the eighth treatment. A further three treatments were necessary to achieve a satisfactory improvement in his psychiatric status. No complications of ECT were noted.
Patient 3
P3, a 36-year-old man, presented with an overdose of 50 mg trifluoperazine following a recurrence of psychotic symptoms for 2 weeks prior to hospitalisation. He suffered from schizophrenic disorder, chronic, and had been treated intermittently for 14 years, and maintained on regular neuroleptic medication for the previous 3 years. On the day of admission, he was given intravenous droperidol 10 mg and diazepam 10 mg for increasingly agitated behaviour. On the second day of admission, he developed fever (37.9°C), mild rigidity, and tachycardia, and had a serum CK level of 1180 U/L. His agitation was treated with further intravenous diazepam (15 mg/day). His mental status deteriorated further and he developed marked agitation, increasing thought disorder, prominent hallucinations, polydipsia, tachycardia and labile blood pressure. His CK level rose further (2574 U/L) and he developed leucocytosis (10.33 × 109/L, with baseline value at 6.75 × 109/L; normal range = 3.0–10.0). A medical review did not suggest any focus of infection or other medical cause, and a diagnosis of possible NMS was made. In view of his deteriorating physical and mental health, ECT was prescribed on an emergency basis. Eight unilateral treatments were performed using propofol and atracurium. His clinical condition improved markedly, his behaviour was settled, the CK level returned to normal but mild thought disorder persisted.
Within 1 week of the cessation of ECT, his psychotic symptoms returned, and he was therefore initiated on thioridazine 25 mg/day; increasing by 25 mg every second day. A week later, thioridazine was ceased and trifluoperazine was reintroduced at 5 mg/day. Over the following days he continued to deteriorate, with prominent catatonic symptoms, with mutism, posturing, and stereotypic movements. Muscle tone increased, CK level rose to 804 U/L, leucocytosis (10.46 × 109/L) was present but temperature was normal. A recurrence of NMS was suspected and the trifluoperazine was ceased. Initial conservative management was attempted with bromocriptine (2.5 mg t.i.d.). However, because his condition deteriorated further, ECT was reintroduced and bromocriptine was continued. His NMS and catatonia improved considerably with six unilateral ECTs, but some psychotic symptoms persisted.
P3 was subsequently commenced on clozapine and went on to have another episode of NMS which we have reported elsewhere [21]. This last episode was treated with bromocriptine (5 mg t.i.d.) resulting in gradual recovery from NMS. The patient was maintained off neuroleptic medication subsequently.
Patient 4
P4 developed a schizophreniform disorder at the age of 14 years and was treated with various neuroleptic drugs over the next 8 months. He developed an episode of NMS, characterised by fever, rigidity, altered consciousness and autonomic instability, when being treated with remoxipride and haloperidol. This resolved within a week of cessation of the neuroleptic drugs. He was given a 4-week trial of thioridazine without improvement in the psychosis. He was then treated with clozapine at a starting dose of 25 mg/day which was gradually built up to 375 mg/day over 4 weeks. The patient developed tachycardia, autonomic instability, disorientation, fever (37.5°C), raised CK (513 U/L) and leucocytosis. Clozapine was ceased at this point and the patient was treated with bromocriptine 2.5 mg b.i.d., with the later addition of dantrolene 50 mg/day. Because abnormalities such as fever, tachycardia, fluctuating level of consciousness, intermittent muteness and raised CK levels persisted over the next 7 weeks, a decision was made to commence unilateral ECT. The physical abnormalities returned to normal after five treatments, and a further five ECT sessions were administered for the treatment of the psychosis. Thiopentone and suxamethonium were used for ECT administration without any complications.
Patient 5
P5, a 59-year-old woman, presented with a history of mood swings over many years, and a 6-month history of manic symptoms and stereotypical movements, leading to a diagnosis of bipolar disorder. Treatment with multiple neuroleptics, mood-stabilising drugs and ECT failed to produce significant improvement. During the course of the management of her manic symptoms with thioridazine (800 mg/day), she developed a fluctuating level of consciousness, fever, muscle rigidity, mutism, a CK level of 3468 U/L and leucocytosis. After the cessation of the neuroleptic drug, she was treated with bromocriptine and dantrolene for over 1 month, with some gradual improvement in her physical state. She continued to be markedly restless, demonstrating stereotypic behaviours and posturing, echolalia, disorganisation and fluctuations in her level of alertness. An EEG demonstrated increased slow activity consistent with an ongoing encephalopathy. In view of her slow improvement, and the difficulties entailed in providing supportive care without the use of neuroleptics, she was prescribed ECT and went on to have 14 unilateral treatments using propofol and atracurium. Her sensorium and catatonic symptoms improved after eight treatments, but some agitation and disorganised behaviour persisted. A re-challenge with 1 mg/day haloperidol led to a rapid recurrence of features suggestive of NMS such that the drug had to be ceased. She was subsequently managed with clonazepam and intensive supportive care and needed to remain in hospital for 3 months before the eventual resolution of her psychiatric disorder.
Patient 6
P6, a 62-year-old woman with a history of bipolar disorder, was admitted with the diagnosis of delirium attributable to lithium toxicity (serum level 1.6 mmol/L) and recent withdrawal from benzodiazepines. Her delirium improved over the next 2 weeks, but she became agitated and aggressive and developed pressure of speech, suggesting a relapse of mania. She was prescribed haloperidol (20 mg/day), and lithium (250 mg b.i.d.) was reintroduced. On the following day, she was noted to be confused and developed extrapyramidal rigidity and hypersalivation, with the further development of fever (up to 38.4°C), negativism and mutism, and a serum CK level of 2519 U/L. A diagnosis of NMS was made after investigations had ruled out other causes. She was treated with lorazepam and bromocriptine (15 mg/day) with partial improvement in her condition, but she continued to be mute, agitated and disorientated, suggesting a continuation of the NMS.
The patient was given a course of 10 unilateral ECTs using methohexitone and suxamethonium. She showed considerable improvement after the sixth ECT, with a reversal of the indices of NMS, and had reverted to her premorbid level at the end of the course. She was reinstated on lithium and was well for the follow-up period of 6 months.
Patient 7
A 17-year-old youth with a diagnosis of schizophrenic disorder of 3 years' duration was admitted to the hospital for the treatment of an exacerbation of his psychosis. After an increase in his regular dose of chlorpromazine from 300 to 600 mg/day failed to improve his condition over 5 weeks, and because of the severity of his symptoms, unilateral ECT was prescribed, with considerable improvement in his psychosis over six treatments. He was given an injection of flupenthixol decanoate with a view to maintenance medication postdischarge. Four days later, he developed fever (37.8°C per axilla), confusion, rigidity, marked agitation, tachycardia, hypertension and diaphoresis, and a diagnosis of NMS was made. Mild leucocytosis and a rise in CK level (254 U/L) were noted. He was treated with bromocriptine (2.5 mg t.i.d.) and dantrolene (50 mg t.i.d.) with only partial improvement over 3 weeks. Because of continuing agitation, fluctuating level of consciousness and autonomic dysfunction, unilateral ECT was introduced, with good improvement after the sixth ECT and recovery to pre-hospitalisation levels after eight treatments. Methohexitone and succinylcholine were used for anaesthesia and muscular relaxation, respectively. He was subsequently rechallenged with haloperidol without a recurrence of NMS.
Patient 8
P8, a 62-year-old woman, had a history of recurrent major depression of over 20 years' duration. She was admitted on this occasion with a diagnosis of psychotic depression which was resistant to treatment with adequate doses of tricyclic antidepressants and neuroleptics. Cognitive impairment was noted and formal assessment revealed deficits in memory and construction, suggestive of an early dementia. An incidental slowly growing right sphenoidal wing meningioma was excised without change in her clinical status. A subsequent course of 23 unilateral ECTs resulted in only partial improvement. After the ECT was ceased, she developed recurrence of psychotic symptoms in the context of worsening depression. She was given escalating doses of trifluoperazine (20 mg/day) and lithium was added to the regimen. She developed an episode characterised by marked rigidity, confusion, catatonic withdrawal, mutism, agitation, polydipsia, tachycardia, hypertension, diaphoresis and incontinence of urine. Leucocytosis (up to 21.2 × 109/L; normal range 3.0–10.0), raised serum CK level (up to 5120 U/L), and mildly elevated liver enzymes were noted. A diagnosis of NMS was made.
The episode was treated with the cessation of trifluoperazine and lithium and the use of diazepam (up to 25 mg/day) to manage agitation. Partial recovery was noted over the next 4 weeks, but owing to the continuation of catatonic features, rigidity and marked agitation, considered to be continuing features of NMS, ECT was introduced. The patient was administered 12 unilateral treatments, followed by 12 bilateral treatments, using thiopentone and succinylcholine. Improvement was gradual, with the symptoms of NMS having reversed after eight treatments, and further recovery of psychotic depression with the subsequent treatments. Mild memory impairment persisted but was consistent with the deficits noted prior to ECT. Pimozide was introduced at 2–6 mg/day without a recurrence of NMS, and the patient was subsequently also maintained on carbamazepine.
Patient 9
P9 was a previously well 20-year-old man who first presented to his local psychiatric service with emotional lability, agitation and bizarre behaviour several weeks after the break-up of a relationship. After the failure of a brief trial of outpatient treatment with moclobemide and chlorpromazine, he was admitted to his local psychiatric hospital. Here he required oral and parenteral sedation with haloperidol and diazepam for control of aggressive, delusionally motivated behaviour. A diagnosis of schizophreniform disorder was made.
Seven days after admission, the patient developed fever (reaching 42°C), diaphoresis, muscular rigidity and confusion. A diagnosis of NMS was made, neuroleptic medication withdrawn and the patient transferred to the local general hospital. The initial CK level was 1600 U/L, peaking at 31000 U/L. Other abnormalities included a leucocytosis (16.56 ° 109/L) and a persistently elevated calcium (2.87 mmol/L; normal range 2.10–2.50). Routine work-up to exclude an infective agent, including examination of the cerebrospinal fluid and cerebral computerised tomographic scan, was normal. He was treated with bromocriptine (5 mg t.i.d.) and dantrolene sodium (50 mg t.i.d.), and parenteral diazepam (up to 20 mg/day) was used to control agitated behaviour. Because of a poor response to medication, and continuing rhabdomyolysis and encephalopathy, the patient was transferred to a tertiary hospital and eventually to our ward for further management. On arrival, the patient demonstrated intermittent agitation, fluctuating level of consciousness, response to apparent visual hallucinations, verbal and motor perseveration and generalised muscular rigidity. After three unilateral ECTs failed to produce adequate seizures, the patient was given eight bilateral treatments. Gradual improvement was noted with improvement in sensorium, reduction in muscular rigidity and improved verbal communication. Psychotic phenomena and some catatonic features persisted. The patient was returned to the referring hospital for the continuation of ECT treatment, where further improvement was noted after a further nine bilateral ECTs.
A summary of the reported cases
We gleaned 46 cases from the published literatures in which ECT was used for the treatment of patients with NMS. (A detailed reference list is available from the authors upon request.) Including our own, this yields a total of 55 cases. The most frequent primary diagnosis was schizophrenia (18), followed by other psychoses of acute onset (15), bipolar disorder (10), major depression (5) and schizoaffective disorder (4). There were single cases of organic psychosis, personality disorder and another of unspecified diagnosis. Catatonia was a prominent manifestation in 42 (76%) cases. Concomitant treatment with lithium carbonate occurred in 10 (18%) cases.
In 31 (56%) cases, ECT was initiated after various drug treatments had failed to produce resolution of NMS, and in the remaining 24 (44%) cases, ECTwas used after supportive treatment alone had failed. The medications used prior to ECT included dantrolene (18 cases), bromocriptine (17 cases), benzodiazepines (12 cases), amantadine (four cases), L-Dopa (three cases) and anticholinergic drugs (two cases). Trials of several different drugs alone or in combination was a common feature.
In 40 (73%) cases, ECT was primarily used for treatment of NMS, in 10 (18%) cases it was used for the combined management of NMS and psychosis and in five (9%) cases its use was primarily to treat psychiatric symptoms during active or resolving NMS. Of those which specified ECT type, unilateral was the more common (16 unilateral, 13 bilateral). The average number of ECTs given was 10, and there was little difference between the average number of unilateral treatments (11.6) compared to bilateral treatments (9.7).
An examination of the time course of improvement with ECT is instructive. Response was measured by improvement in fever, rigidity, altered consciousness, catatonia, and laboratory indices (CK and white blood cell count). Onset of response to ECT was, on average, after 4.1 treatments. Nineteen patients had a significant improvement in the symptoms of NMS after one to three treatments. In a further eight cases, improvement occurred gradually over four to seven treatments. Since some of these patients had previously been treated for a few weeks with drug therapy, it would suggest that the improvement was indeed due to the ECT and not a spontaneous improvement. The dramatic reversal of NMS in seven reported cases [3,22–27] would argue against it being a placebo effect. The reports also suggest that the majority of improvement in NMS tends to occur early in the course of ECT, and the treatment was often continued subsequently for the management of the primary psychiatric disorder. In our own cases, the response was definitely obvious by the sixth treatment.
Where ECT was given as a treatment primarily for NMS (40 cases), complete recovery was observed in 25 (63%) cases, with partial recovery in another 11 (28%) cases. Thus, in this subgroup, there was benefit from the treatment in a total of 36 (90%) cases. This conclusion must be tempered by the possibility of the bias of more positive reports being published. The positive outcome was shared by those patients given ECT for treatment of combination of NMS and psychiatric symptoms, and those given ECT primarily for psychiatric symptoms. Age, gender, psychiatric diagnosis, and any particular features of NMS were not suggested as predictors of a good response. While catatonic symptoms such as mutism and negativism were commonly reported, their presence was not essential for a good outcome.
The safety of ECT for NMS has come under scrutiny for two reasons: the autonomic dysfunction in NMS increasing the likelihood of cardiovascular problems with ECT, and the clinical similarity between NMS and malignant hyperthermia (MH) raising the issue of the safety of certain anaesthetic agents in an NMS patient. We encountered four cases in which cardiovascular complications appeared to be associated with ECT. The first [28] suffered ventricular fibrillation after the sixth ECT and subsequently was comatose presumably due to anoxic brain damage at 7-month follow-up. Importantly, the patient suffered significant medical complications of NMS prior to ECT, including massive pulmonary emboli, pneumonia, pleural effusion and supraventricular dysrhythmia, and was thus at a high anaesthetic risk. In addition, it would appear from the report that neuroleptic medication was continued despite the severity of NMS. The type of anaesthetic and muscle relaxant agents used is unspecified. We would therefore agree with Scheftner and Shulman [7] that the outcome in this case cannot be definitively linked with the use of ECT. Prakash and Leavell [29] described an episode of trigeminy responding to lignocaine in a patient who received thiopentone and succinylcholine during ECT. This patient also had an episode of status epilepticus responding to diazepam during a subsequent ECT. Status epilepticus is considered a rare complication of ECT [30] which is more likely to occur in the setting of pre-existing brain abnormality, various metabolic disturbances and during concomitant treatment with various medication such as theophylline or lithium. It is possible that encephalopathy in the course of NMS may have been a risk factor for status epilepticus in this case; however, the presence of an additional risk factor (mild mental retardation) may also be aetiologically important. As in this case, ECT has been continued safely in patients who have had an episode of status during their treatment [31,32]. The patient of Hughes [33] had a cardiac arrest during her sixth ECT from which she was successfully resuscitated. The patient reported by Lazarus [24] developed a brief atrial dysrhythmia between treatments 1 and 2. George and Wood [34] reported a patient who developed hyperkalemia in response to succinylcholine when used in ECT for NMS. None of our nine patients had any ECT-related complications.
The cases with a negative response deserve examination in more detail, in particular the context in which ECT was given. As stated above, ECT was often given after conservative treatment had failed, and usually at a time when the patient was seriously medically ill. Among the cases reviewed, there was one case [35] in which the authors noted ‘no benefit’ after 21 bilateral ECTs. It is unclear from the report whether this statement referred to a lack of improvement in NMS, psychosis or both. This patient eventually responded to cautious reintroduction of neuroleptics, suggesting that it was the lack of improvement in psychosis with ECT that the authors were possibly referring to. In two cases [28,36], the patient's condition was considerably worse following ECT. In the case documented by Grigg [36], the return of fever and elevation of CK was suggested by the author to be the result of an exacerbation of NMS, or the induction of MH, after the administration of a single ECT using thiopental sodium and pancuronium bromide. Electroconvulsive therapy was ceased and the patient had an uneventful recovery. Given that the patient had been comatose and febrile as recently as 36 h prior to ECT, and symptoms of NMS do tend to fluctuate, it is conceivable that this exacerbation may have represented a natural fluctuation in the patient's condition rather than a specific adverse response to ECT. Furthermore, the majority of the evidence points against a relationship between NMS and MH [37], making MH an unlikely explanation for this deterioration. There are two reports from Kish et al. [8] of death due to NMS in which ECT had been used for treatment. Death in both cases was due to cardiac failure and was not associated with administration of ECT. Importantly, in both these cases, the neuroleptic drug had been continued during the episodes of NMS.
It is equally important to note that ECT has been used safely and effectively in NMS patients who were medically quite unwell (e.g. in the presence of bilateral pneumonia [27], acute renal failure [39] and a combination of aspiration pneumonia, urinary tract infection, cellulitis, candida sepsis and fluid and electrolyte disturbances [26]). Electroconvulsive therapy has also been used safely for NMS in the setting of recent cardiac conduction disturbance [35] and recent cardiac arrest [40].
With regard to anaesthetic agents, while many clinicians avoid depolarising muscle relaxants, those who have used them have not reported any adverse consequences. In this series, the muscle relaxant used was specified in 27 (50%) cases. Succinylcholine was used often (16 cases), followed by the non-depolarising muscle relaxant atracurium (10 cases). Succinylcholine did not result in MH, or a worsening of the symptoms or laboratory abnormalities of NMS, in any patient.
Discussion
The place of electroconvulsive therapy in the treatment of neuroleptic malignant syndrome
Many patients with NMS improve spontaneously with the cessation of the neuroleptic, or respond to treatment with a dopamine agonist and a muscle relaxant [41]. A proportion of patients show an incomplete response and should receive a course of ECT, with improvement likely to occur after the first few (up to six in our series) treatments. Although no direct comparison exists, the response to ECT appears to be faster than that to drugs. This, along with the fact that ECT is useful for NMS as well as the baseline psychiatric disorder, makes it our preferred primary treatment in the following situations: (i) when NMS is severe and the risk of complications is high if left untreated; (ii) when the differential diagnosis between NMS and lethal catatonia is uncertain [42]; (iii) when psychotic depression was the primary psychiatric disorder for which the neuroleptic drug was prescribed; and (iv) when a catatonic syndrome is a major feature of NMS. In the first two situations, ECT can be considered to be life-saving.
Electroconvulsive therapy is relatively safe in the presence of even severe NMS when performed appropriately, with only a few adverse reactions reported as listed previously. Caution is necessary, however, because the possibility of under-reporting of negative outcome cannot be dismissed. A pre-anaesthetic review is essential and vigilance is necessary for cardiovascular complications. Patients should also be monitored for an increase in muscle injury due to ECT, and for hyperkalemia. As the majority of patients given ECT for NMS are incapable of an informed consent, the issue of proxy consent and its ethical implications should be considered.
Neuroleptic malignant syndrome, catatonia and electroconvulsive therapy
The relationship between NMS and catatonia has received much comment recently [43,44]. Catatonia is often seen in the course of NMS, and it has been argued, in fact, that NMS represents a form of druginduced catatonia [43,45]. The syndrome of catatonia has been observed to occur in a variety of neurologic and metabolic disorders [46], of which NMS may be one. The syndrome of lethal catatonia (LC) [47] is very similar in its description to NMS, and it has been argued that the two are indistinguishable except for the history of recent exposure to neuroleptics [43,44]. For this review, however, we reclassified some reported cases of LC as NMS [28,40,48], and vice versa [14–16,49], depending upon whether a neuroleptic had been introduced prior to the development of the syndrome or not. The literature suggests that neuroleptics may be inadequate for the treatment of LC, and ECT may be a more appropriate treatment [47]. This further supports the use of ECT in NMS. The potential benefit of adenocorticotropin hormone and corticosteroids in some cases of LC is of interest [50] and hormonal treatment should be investigated in NMS. It is also possible that patients with prodromal catatonic syndromes may be more likely to develop NMS, but no systematic support for this possibility is available [51].
Neuroleptic malignant syndrome, malignant hyperthermia and electroconvulsive therapy
The clinical similarity between NMS and MH has raised the concern that NMS patients may be at an increased risk for the development of MH when exposed to depolarising muscle relaxants during ECT [52]. There are empirical and theoretical arguments to be considered in evaluating this concern. Our review indicates that many patients with past or current NMS have been exposed to depolarising agents without any definite reports of MH occurring in these patients. Patients with a genetic susceptibility to MH have also not been reported to be at greater risk for NMS. Malignant hyperthermia is a familial disorder in which there is an underlying muscle disease, possibly a deficiency of inositol 1,4,5,-triphosphate phosphatase [53], which makes the muscle hypercontractile in response to certain agents such as halothane, caffeine and depolarising muscle relaxants, but not neuroleptic drugs. Neuroleptic malignant syndrome is, on the other hand, a central nervous system disorder precipitated by neuroleptic drugs whose aetiology is not known but is speculated to be a profound hypodopaminergic state [37]. The two are therefore quite distinct disorders pharmacologically. There is some controversy in the literature about whether the muscles of NMS patients have an abnormality similar to that seen in MH. Published studies are divided on whether the in vitro response of skeletal muscle to halothane is abnormal in NMS, with studies for [52,54,55] and against [56–58] having been published. Some authors have considered this to suggest that there is indeed an MH susceptibility in NMS patients [52], and the reasons that NMS patients seem to tolerate anaesthesia may be that their susceptibility is lower than that seen in MH, and MH does not uniformly occur even in individuals with well-documented subsequent episodes. The in vitro susceptibility in some cases may, however, be a non-specific response of injured muscle to ‘strong’ triggers such as halothane.
The MH-susceptibility of NMS patients is therefore inconclusive, but the evidence suggests that the use of succinylcholine is safe. A single dose of succinylcholine, in the absence of a volatile anaesthetic such as halothane, is a weak trigger even in MH patients [59], and there are animal data to support this [60]. We therefore support the use of succinylcholine for ECT in NMS patients. The use of a non-depolarising muscle relaxant, as was done in five of our cases, considerably prolongs the duration of anaesthesia, with considerably more barbiturate or other anaesthetic being used, thereby increasing the risk of complications. If the patient has a personal or family history of complications with anaesthetic agents, one would of course treat it as a high risk of potential MH and the usual precautions will apply.
In conclusion, our review and case material suggests that ECT is an effective treatment in many cases of NMS, even when drug therapy has failed, and is the preferred treatment in severe NMS when a prompt response is necessary, or if lethal catatonia cannot be ruled out, or if psychotic depression or catatonia were the primary psychiatric conditions. The response is usually apparent after the first few treatments, generally up to six. It is a relatively safe treatment although the increased risk of cardiovascular complications and hyperkalemia should be considered. The occurrence of MH is extremely unlikely, and succinylcholine is demonstrably safe. The theoretical issues in relation to ECT and NMS, and its relation to catatonia, need further exploration.
Footnotes
Acknowledgements
This work was supported by a New South Wales Institute of Psychiatry Fellowship to Dr Trollor. We thank all the psychiatrists who agreed for their patients to be included in this report, and Jane Kruk for assistance with the literature search.