Discontinuation of thioridazine in patients with learning disabilities: balancing cardiovascular toxicity with adverse consequences of changing drugs

When evidence emerges on rare but potentially dangerous drug effects, the Committee on Safety of Medicines needs to act to protect the public. On the basis of clinical experience after guidelines relating to thioridazine were revised, Davies and colleagues argue that directives should take greater account of the risks associated with changing successfully established drug regimens, especially in vulnerable populations such as people with learning disabilities

When new evidence is unearthed of rare but dangerous side effects of drugs used in current practice, agencies such as the Committee on Safety of Medicines need to review and rewrite prescription guidelines. When new guidelines restrict the use of a commonly prescribed agent to a limited range of circumstances, doctors may be left in a dilemma. Patients who do not meet the criteria for continuing with a drug under revised guidelines, but who benefit from using it, may have to be exposed to the adverse consequences of changing their drugs. Influenced by the current culture of defensive medical practice, clinicians may be reluctant to continue prescribing drugs when guidelines state that the balance of risks and benefits is unfavourable in circumstances that apply to their patient. We present examples from our practice of the consequences of discontinuing established drugs in line with the Committee on Safety of Medicines′ advice.

Summary points

• The Committee on Safety of Medicines restricted the licensed indications for thioridazine after this antipsychotic drug was associated with prolonged QTc intervals

• Patients with learning disabilities and psychiatric illness have been placed at higher risk by stopping or substituting antipsychotic drugs than had the drug been continued

• Many people with learning disabilities lack the capacity to give their own consent to medical interventions, and psychiatrists may be reluctant to over-ride guidelines when responsibility for making decisions cannot be shared with the patient

• Directives should allow clinicians greater flexibility to continue established drug regimes

History

In December 2000 the Committee on Safety of Medicines restricted the licensed indications for thioridazine.¹,² This followed publication of data by Reilly et al,³ supported by earlier evidence.⁴ Psychiatric patients taking thioridazine and droperidol had significantly more prolonged QTc intervals (a marker for risk of ventricular arrhythmias and sudden death) than a healthy population.³ Although Reilly et al emphasised that QTc prolongation is only a surrogate endpoint for risk of ventricular arrhythmias and sudden death, the Committee on Safety of Medicines deemed that the risk was sufficient to limit the use of thioridazine to patients with a diagnosis of schizophrenia and then only as a second line drug. They also introduced a requirement for regular QTc monitoring for patients treated with thioridazine and made concomitant use of thioridazine and other drugs known to cause QTc prolongation a contraindication.¹,² These changes in the licensed indication for thioridazine not only affect patients requiring antipsychotic drugs for the first time but have also forced doctors to review the use of thioridazine in patients taking the drug for long term prophylaxis.

Problems associated with thioridazine withdrawal in learning disabilities

Patients with learning disabilities may have problems when their medication is stopped or changed. They may fail to understand the reasons for changing a tablet when they are feeling well and may react badly to being given unfamiliar tablets. Those who have been subject to prolonged institutionalisation may have particular difficulties in coping with the loss of any familiar routine. For these reasons, patients with learning disabilities may be especially vulnerable to relapse of psychotic illness or acute behavioural disturbance when a longstanding antipsychotic drug is stopped, even when an alternative is prescribed.

In our population of approximately 155 patients, 18 patients were regularly taking thioridazine at the time of the directive. Three of these patients had a diagnosis of schizophrenia; the remainder needed an antipsychotic for psychosis associated with mania, for marked behavioural disturbance, or for extreme anxiety. All 18 patients stopped thioridazine following the Committee on Safety of Medicines' advice, and seven (six females and one male) experienced moderate or severe difficulties in the following three months (table). In four cases the problems occurred despite immediate substitution of thioridazine with an alternative antipsychotic; in one case prescription of an alternative drug was not initially possible because of the difficult relationship between the patient and the service; and for two patients a trial without immediate substitution of thioridazine was attempted as doses had been low (12.5 mg/day and 50 mg/day) and the patients had been well for some time. Alternative psychotropic drugs were introduced at the next review, after both experienced acute behavioural disturbance.

Discussion

After thioridazine was stopped or substituted in our patients with learning disabilities and psychiatric illness, 39% had moderate or severe difficulties. We anticipate further consequences—in some cases the effects of changing antipsychotic drugs have led to disruption of the already fragile relationship between doctors, the patient, and carers. Such damage to these key relationships may take months or even years to heal.

Consequences of drug change

As a consequence of the Committee on Safety of Medicines' directive, some patients with learning disabilities have been exposed to the psychological and pharmacological impact of drug change, but their cardiovascular risk may not be reduced. Some recommendations, including those in our local area, identified promazine as a possible alternative conventional antipsychotic to thioridazine, but this drug was not included in Reilly's study on QTc prolongation.³ We identified no further data on literature search which could quantify risk of QTc prolongation with promazine. In a survey of medicolegal autopsies (coroner's cases) crude rates of sudden unexpected deaths per unit exposure to promazine and thioridazine were approximately equal.⁵ In one local district, promazine was prescribed as a substitute for thioridazine by general practitioners so frequently that the pharmacy was temporarily unable to meet the sudden upsurge in demand for a drug that had been little prescribed for many years, and for which robust data on cardiovascular toxicity is lacking.

The patient who became dehydrated and required emergency hospital admission with a diagnosis of neuroleptic malignant syndrome (table) provides the clearest example of the adverse consequences of drug change. Although she was no longer exposed to the risks of cardiovascular toxicity associated with thioridazine, her family believe that the consequences of its removal have outweighed the theoretical risks of continuation. They feel that stopping a drug which seemed to have made a strong contribution to her wellbeing over many years, and which had not provoked any detectable abnormalities on earlier electrocardiograms, has been a mistake. She has been subjected to the hazards of dehydration and electrolyte imbalance, a risk factor for arrhythmia in itself, and to the dangers of the neuroleptic malignant syndrome, which carries a mortality risk of about 12%—far greater than the theoretical risk of cardiovascular toxicity with thioridazine.⁶

CSM directives

Our experiences show that the Committee on Safety of Medicines should take account of the adverse consequences of drug change when making risk benefit judgments, especially in vulnerable populations such as patients with learning disabilities and psychiatric illness. The guidelines indicated that the balance of risks and benefits for use of thioridazine was unfavourable except where a diagnosis of schizophrenia had been made.¹ The committee does not seem to have considered that the risks associated with stopping and substituting thioridazine may outweigh the benefit in many patients with learning disabilities who have other psychiatric diagnoses, nor that making confident diagnoses of schizophrenia in patients with learning disabilities is often difficult—and may be impossible when disabilities are severe.

Of course, clinicians in any discipline have the option of prescribing drugs outside their licensed indications. Many people with learning disabilities lack the capacity to give their own consent to medical interventions, and psychiatrists may feel a particular reluctance to over-ride guidelines when responsibility for making decisions cannot be shared with the patient. In the culture in which we practise, it is difficult to continue prescribing thioridazine in the patient's “best interest” in circumstances where the Committee on Safety of Medicines has specifically stated that balance of risks and benefits is unfavourable.¹ Although the adverse consequences for the seven patients who stopped taking thioridazine could be viewed as being attributable to poor management of change, the clinicians involved felt they had little option but to follow the Committee on Safety of Medicines' advice.

LIANE PAYNE

A broader range of indications, perhaps including mania and impulsive behaviour, could have been retained as indications for thioridazine in this population, with the proviso of mandatory regular electrocardiograms. Indeed there is a precedent for broader licensed indications with another conventional antipsychotic, pimozide. Since an association with QTc prolongation was identified in the 1980s, patients with a wider range of diagnoses may be prescribed pimozide if they regularly have electrocardiograms.⁷ This approach allowed patients who had responded well to continue taking the drug but acted as a disincentive to new prescriptions. Similar flexibility may be appropriate to allow continuation of thioridazine in combination with other neuroleptics known to cause QTc prolongation, provided the total equivalent antipsychotic dose is below a reasonable limit.

Conclusion

Although the Committee on Safety of Medicines needs to act to protect the public when scientific evidence emerges on rare but potentially dangerous drug effects, directives should incorporate the flexibility required to accommodate the needs of patients who are already successfully established on these agents. Revision of indications for licensed drugs to minimise cardiovascular risk need to be balanced against the broader risks of the psychological and pharmacological impact of changing drugs. Guidelines that disseminate new information and provide advice are welcome, but they should not be couched so rigidly as to present clinicians who practise in the current defensive culture with insoluble dilemmas.

Table.

Adverse consequences of discontinuation of thioridazine among 18 patients taking the drug

Adverse consequence	No of patients affected (n=7)
Behavioural disturbances, including head banging, prolonged screaming, and physical or verbal aggression	5
Psychotic symptoms*	4
Unmasking of tardive dyskinesia† with teeth grinding (bruxism) prominent	1
Admission of over 2 months' duration to assessment and treatment unit under the Mental Health Act	3
Progressive dehydration, electrolyte imbalance, and emergency transfer to a general hospital with probable neuroleptic malignant syndrome four weeks after switching to an alternative antipsychotic drug regimen	1

^*None of the four patients who experienced psychotic symptoms were eligible to continue thioridazine under the revised licensed indications (three did not have a diagnosis of “schizophrenia” and the fourth had been prescribed thioridazine in combination with another drug capable of QTc prolongation). 

^†Symptoms not alleviated by an alternative conventional antipsychotic or by vitamin E and have only been partially relieved by an atypical antipsychotic.