Letters to the Editor

Dear Sir,

We read with interest Dr Tam's article on "A young man with asymptomatic sinus tachycardia".¹ Tachycardia is a well recognized feature in chronic use of tricyclic antidepressants (TCA) and we would congratulate Dr Tam's successful recognition of the problem and treatment by switching to another antidepressant.

However, we would like to clarify several issues concerning the management of acute poisoning from TCA as stated in the article. Dr Tam suggested that "seizures are best managed by diazepam or phenytoin, rather than barbiturates, as they are less likely to aggravate respiratory depression". In fact, the use of phenytoin is no longer recommended for seizure in TCA overdose as data not only demonstrate a failure to terminate seizures, but also suggested enhanced cardiovascular toxicity because of the sodium channel blocking effects in common with both phenytoin and TCA.² Benzodiazepines, for example diazepam is considered first line medication for recurrent or prolonged seizures in TCA overdose. Barbiturates or propofol should be instituted if benzodiazepines failed to control the seizure,³ and the patient should be considered to have the airway protected if second line therapy for seizure is required.

The usage of TCA level as the indicator of the clinical severity of TCA poisoning is limited compared with the electrocardiographic features (e.g. QRS complex duration of more than 100 msec) in TCA overdose. TCAS have large volume of distribution, prolonged absorption phase, long distribution halflives, pH-dependent protein binding and the terminal elimination half-lives between patients is widely variable. Measurement of the concentration in the blood does not reflect the TCA concentration in the affected tissue (namely the heart and the brain) and the use of TCA concentration is not useful for clinical management of an acutely poisoned patient.^3-4

Authors' reply

Dear Editor,

We are grateful for the interest and discussion from Dr Ng and his team on acute management of tricyclic antidepressant (TCA) overdose.

We are aware of the scarcity of evidence regarding the choice of drugs for management of seizures related to TCA toxicity. Generally, benzodiazepine is used as first line medication; alternatives include phenytoin and barbiturate. There had been suggestions that phenytoin was preferred over barbiturate, partly because barbiturate may further aggravate the already suppressed respiratory efforts in TCA overdosed patients.¹ From Dr Ng's comments, his concern about use of phenytoin was probably based on an experimental study investigating the cardiovascular effect of prophylactic phenytoin on twenty anesthetized dogs being constantly infused with increasing doses of amitriptyline till death. The results of the experiment showed that phenytoin group did have significantly less adverse effect of QRS prolongation; but concomitantly there was much greater increase in number of ventricular tachycardia than the control group. From these findings, the authors postulated that prophylactic phenytoin in that specific animal model with unusually toxic TCA level provided no benefit in cardiac protection and might even increase the severity of ventricular tachycardia and hypotension. ²

Our attention is also drawn to the newly released "2011 Guideline in Emergency Medline Network on management of tricyclic antidepressant overdose", which was released few months after our case report was published.³ The Guideline Working Group had completed the difficult task of consolidating recommendations from assimilation of the rather limited evidence available on management of TCA overdose. Concerning drug treatment for seizures, this 2011 guideline recommended that benzodiazepines remained the first line therapy (Grade E recommendation; based on expert opinions). Notably, we are made aware that there was change in recommendation from previous guidelines ⁴ that phenytoin was no longer recommended in the current guideline (Grade D recommendation; based on experimental studies). Essentially, there had been no studies ever comparing phenytoin with benzodiazepine in patients with TCA overdose.³ As evidence for the benefit of phenytoin came only from sporadic case studies,^5,6 and that there were safety concerns about the potential drug interactions between phenytoin and TCA, and the narrow therapeutic range of phenytoin; the Guideline Working Group currently advised that phenytoin should be avoided.³

In summary, the current thoughts on the choice of anti-convulsants in the context of TCA overdose are still subject to controversies and further challenges by future studies. Nonetheless, we would like to sincerely thank Dr Ng et al again for generously sharing their invaluable comments and experience on this topic. These have undoubtedly increased our clinical knowledge; enriched our understanding on the limitations behind authoritative guidelines and recommendations, hence enabling more effective and individualized approach in patient management.

As for the measurement of plasma TCA level, we have already emphasized its uncertain value in clinical management and have no further comments to add on this issue.

Yours truly,