Tricyclic antidepressants have a narrow therapeutic index, and significant CNS and cardiovascular toxicity may be seen with ingestions that exceed therapeutic doses by as little as three-fold.

Mental status on presentation has been shown to be predictive of patient outcomes, and most life-threatening arrhythmias occur within the first 24 h after ingestion [2]. Owing to rapid GI absorption, patients may initially appear well and then deteriorate rapidly. Serum drug assays that measure specific serum tricyclic antidepressant levels are not widely available and should not be relied on when managing suspected overdose. Because these drugs are highly lipophilic and have high volumes of distribution, concentrations in brain and myocardium may exceed serum concentrations by as much as 40 to 200-fold, and serum drug levels have not been shown to predict the clinical course [1, 3].

The single most useful test in identifying tricyclic antidepressant toxicity is the electrocardiogram. The most common early EKG finding in overdoses is sinus tachycardia, due to anticholinergic effects and initial transient catecholamine excess. As toxicity progresses, tricyclic antidepressants block fast sodium channels in the cardiac conduction system thereby slowing depolarization and decreasing contractility. This results in a widened QRS complex. Prolongation of the QRS complex >100 ms has been shown to predict risk for seizures, and prolongation >160 ms predicts a risk for ventricular arrhythmias [4]. Owing to the greater susceptibility of the right ventricular conduction system to the effects of sodium channel blockade, a rightward axis shift of the terminal 40 ms of the QRS axis in the frontal plane will develop; but this finding may be difficult to detect [3]. A more practical means for identifying this rightward axis shift is by noting the presence of a larger than expected R wave amplitude in aVR (Fig. 6.2). An amplitude of the R wave in this lead in excess of 3 mm has been shown to predict risk for seizures and arrhythmias [5].

Activated charcoal has been shown to effectively bind tricyclic antidepressants. In a study of healthy volunteers activated charcoal given 5 min after oral administration of therapeutic doses of amitriptyline decreased GI absorption by 99 % [6]. Activated charcoal probably has the most benefit when administered within 1 h of a toxic ingestion, but may be considered for up to 2 h since the anticholinergic effects of the drug may delay gastric emptying [7].

Gastric lavage carries risks such as aspiration, and is not routinely recommended following tricyclic antidepressant ingestion. One randomized study demonstrated no significant difference in patient outcomes when gastric lavage was combined with the use of activated charcoal [8].

Whole bowel irrigation is typically reserved for overdoses involving sustained-release preparations or in cases where the drug binds poorly to activated charcoal. Therefore, its use is unlikely to provide any benefit [7].

Although no randomized-controlled clinical trials exist in support of sodium bicarbonate therapy in the setting of tricyclic antidepressant poisoning, numerous animal studies and human case reports support its use as a first-line treatment [9–11]. Sodium bicarbonate has been shown to both prevent and terminate ventricular dysrhythmias. The benefits of serum alkalinization through the use of sodium bicarbonate are likely multifactorial. The fraction of drug that is protein-bound increases with alkalinity, and therefore the use of sodium bicarbonate would be expected to decrease the amount of free drug available for inhibition of sodium channels [2]. Additionally, increasing serum sodium concentrations may overwhelm the sodium channel blockade by increasing the gradient between the intracellular and extracellular space [2]. The net effect is improved myocardial contractility, narrowing of the QRS complex, and improved electrical stability.

Dosing and titration guidelines for sodium bicarbonate have not been well-studied, but most recommend initiating a bolus of 1–2 meq/kg followed by a continuous infusion to maintain a goal serum pH within 7.45–7.55 and titrating to a goal QRS duration of less than 100 ms [2, 10]. The prophylactic use of sodium bicarbonate is not supported when EKG abnormalities are absent [10]. Serum pH must be monitored closely, and while optimum pH has not been studied, it is reasonable to not exceed a pH of 7.6. Risks of marked alkalosis may include impaired tissue oxygenation due to shifts in the oxygen dissociation curve, cerebral vasoconstriction, reduction in ionized calcium concentration, intracellular shift of potassium, and an increased myocardial sensitivity to catecholamines [10].

Hypotension is a common feature of tricyclic antidepressant toxicity. These medications decrease systemic vascular resistance through alpha-adrenergic blockade. Relative catecholamine depletion may occur through their effects on norepinephrine reuptake. Additionally, hypotension may occur as a direct result of decreased cardiac inotropy [2]. Profound hypotension may be refractory to fluid resuscitation and alkalinization therapy, and vasopressors may be indicated. There are very few studies evaluating the benefit of one such agent over another [12]. One retrospective case series suggested norepinephrine may be superior to dopamine in these instances [13]. A single case report showed that vasopressin stabilized blood pressure in a patient with hypotension that failed to respond to high-dose norepinephrine [14].