Bupropion hydrochloride (Zyban; Glaxo Wellcome Australia, Boronia, Vic.) was marketed in Australia in November 2000 in a sustained-release (SR) formulation as a short term aid in abstaining from smoking. It has previously been marketed as an antidepressant in other countries.1 Bupropion is structurally similar to stimulants such as the amphetamines, and selectively inhibits the neuronal reuptake of dopamine and noradrenalin.2

Previous reports of bupropion overdose almost all involve the immediate-release (IR) formulation.3-10 There are reports of the SR formulation in overdose,11-14 but there is limited information on its spectrum of toxicity in overdose. Information on paediatric safety of either formulation is also limited.15

Immediately after bupropion was introduced in Australia, we conducted a prospective study of deliberate self-poisoning in adults and accidental poisoning in chidren with SR bupropion to investigate the toxicity of this formulation.

Cases of bupropion poisoning were identified prospectively from calls to the New South Wales Poisons Information Centre (NSW PIC) from 1 November 2000 to 31 July 2001. NSW PIC is the only national 24-hour PIC in Australia providing poisons information to medical professionals and the general public. Cases were identified when hospitals requested advice regarding a patient with intentional or accidental bupropion ingestion.

At the time of the call, the poisons information specialist would record the patient's medical record number. A request was then sent to the medical records department of the respective hospital for a de-identified copy of the patient's record. Data collected included patient's age, sex and State where the overdose took place, relevant medical history, details of bupropion ingestion (time of ingestion, dose, co-ingestants), clinical features, investigations, decontamination therapy and treatment.

Information was extracted from the medical records by one author (C R B) using a standardised data collection form. It was assumed that there were no gastrointestinal symptoms, changes in mental status (specifically agitation, aggression or hallucinations), seizures or tremors if these were not recorded in the medical record.

Ethical approval was obtained from the ethics committee of the Children's Hospital at Westmead.

Median and interquartile range (IQR) are quoted for data not normally distributed. For comparison of two groups, the unpaired t test was used.

There were 69 cases of bupropion ingestion; 59 cases were intentional ingestions by adults and 10 were accidental ingestions by children.

Bupropion SR overdose caused significant effects, including seizures, in over a third of cases. Other common effects in the 19 patients who took bupropion alone included tachycardia (83%), hypertension (56%), agitation (32%) and gastrointestinal symptoms (37%). These findings are consistent with a previous study of the IR formulation in 58 patients,10 with the only significant differences from our findings being lower rates of tachycardia (43%) and gastrointestinal symptoms (14%) with the IR formulation.

Seizures occur in both therapeutic doses and overdose of bupropion. The risk of seizures associated with therapeutic use of the SR formulation is 0.1% with doses of no more than 300 mg per day.2 In overdose, the incidence of seizures has previously been reported to occur in a third of cases with the IR formulation.16 Our study suggests a dose-dependent relationship with seizures, where patients ingesting more than 30 tablets were more likely to have a seizure, and nearly all patients ingesting more than 60 tablets had a seizure. Most seizures resolved spontaneously, often before a benzodiazepine was given.

Early studies of bupropion suggested it was less cardiotoxic than other antidepressants, and this was supported in the study of overdoses with the IR formulation.10 More recently, two cases were reported suggesting cardiotoxicity based on a prolonged QTc.3,7 In our study the major cardiovascular effects were tachycardia and mild hypertension, with no patient having arrhythmias or hypotension, although one was treated for a prolonged QTc. There were no other reported cardiac effects in this patient, including arrhythmias other than tachycardia.

There is limited information on accidental paediatric ingestion of bupropion. A study of 114 children showed that 92% were asymptomatic, with most children taking less than two tablets.15 Our study supports this, with only one child becoming symptomatic.

Our study was limited by the need for clinical details to be extracted from medical records. Although the investigator was not blinded during this process, a standardised extraction process was used to collect predetermined data. Recruiting only patients where hospitals contacted the PIC for advice might introduce a bias towards severe cases. However, we began this study as soon as bupropion was introduced into Australia, so that information on the effects of overdose was not widely available. Sixty-three per cent of patients who experienced seizures had been reported to the NSW PIC before their seizure occurred, supporting our contention that cases were being reported regardless of severity.

The mainstay of treatment for bupropion overdose is supportive care.2 Patients should be given activated charcoal if they present soon after the ingestion.17 Our study demonstrates the importance of recognising the potential for deliberate self-poisoning, as well as accidental paediatric ingestions, of any drug before it is marketed.