Tricyclic Antidepressant (TCA) Poisoning

Overview

Tricyclic antidepressants (drugs such as amitriptyline, imipramine, clomipramine, nortriptyline) are rapidly absorbed from the intestine and in overdose can cause a severe and rapidly progressing toxicity within hours. The main mechanism of toxicity is the blockade of cardiac fast sodium channels; this blockade slows intercellular communication, prolongs the QRS duration on the electrocardiogram, causes right axis deviation (manifested as prominent terminal R wave in lead aVR), and increases the tendency for ventricular arrhythmia.

In addition to this sodium-channel effect, tricyclic antidepressants block anticholinergic receptors, leading to tachycardia, mydriasis, dry and warm skin, decreased bowel sounds/ileus, and delirium; alpha-1 adrenergic blockade causes peripheral vasodilation, leading to hypotension; GABA-A receptor blockade lowers the seizure threshold. When a seizure occurs, the resulting acidosis increases the drug’s unbound fraction in the plasma, further exacerbating sodium-channel blockade, and this cycle can rapidly drive the patient to cardiovascular collapse.

The clinical course can change within minutes; a patient who appears awake and hemodynamically stable upon arrival can have a seizure, develop an arrhythmia, or arrest within 1-2 hours; therefore, a single ECG at the initial assessment is not reliable, and serial monitoring is essential. Diagnosis is made by evaluating the history of ingestion, the presence of anticholinergic syndrome, and characteristic sodium-channel blockade findings on the ECG together.

Quantitative serum tricyclic antidepressant levels, as emphasized by StatPearls and GEMNet guidelines, do not reliably correlate with the severity of clinical toxicity and are not used to guide treatment decisions; treatment is guided entirely by clinical findings and ECG. The predictive value of the QRS duration was numerically demonstrated in the 1985 study by Boehnert and Lovejoy: patients with a QRS duration of 100 milliseconds or more had approximately a 34% chance of developing seizures and about a 14% chance of developing life-threatening arrhythmia; when QRS duration exceeds 160 milliseconds, the likelihood of ventricular arrhythmia increases to about 50%.

Qualitative tricyclic antidepressant immunoassay screening in urine provides rapid bedside results but has low specificity; other drugs with a tricyclic structure such as quetiapine, carbamazepine, cyclobenzaprine, diphenhydramine, and hydroxyzine can also cross-react in the test and lead to false-positive results, therefore urine testing alone should not be considered diagnostic and treatment decisions should still be based on the clinical picture and ECG.

Key considerations

• Other sodium-channel blocker poisonings (agents such as carbamazepine, cocaine, flecainide, diphenhydramine) similarly cause QRS widening and a terminal R wave in aVR; differential diagnosis is made based on history and the substance taken because the ECG pattern alone is not distinctive.
• Pure anticholinergic poisoning (antihistaminic, antipsychotic, or certain herbal products) causes tachycardia and delirium, but since there is no sodium-channel blockade, the QRS duration is generally normal; this point is a key finding in distinguishing from tricyclic poisoning.
• Hyperkalemia can cause QRS widening without directly affecting the sodium channel; differentiation is provided by accompanying findings such as serum potassium level and peaked T waves on ECG.
• Acute cardiac ischemia or spontaneously occurring ventricular arrhythmia can be confused with wide QRS in a patient without a history of drug intake; troponin and history are decisive for differentiation.
• Mixed drug intake (combined with paracetamol, opioids, or benzodiazepines) can alter the clinical picture; in the presence of opioids, respiratory depression may be prominent, which can mask anticholinergic findings.
• Metabolic conditions causing altered consciousness and hypotension, such as sepsis or hypoglycemia, are differentiated by the absence of QRS widening on ECG and the presence of blood sugar/infection findings.
• Serotonin syndrome can accompany due to the serotonergic effect of some tricyclic antidepressants; neuromuscular findings such as clonus and hyperreflexia distinguish it from a pure anticholinergic picture.
• Malignant neuroleptic syndrome is highlighted in a patient with a history of antipsychotic use by muscle rigidity and elevated creatine kinase; the sodium-channel blockade ECG findings seen in tricyclic poisoning are typically absent here.
• Lithium or other psychiatric drug poisonings present with various neurological findings such as tremor and ataxia; QRS widening is not expected and differentiation can be made based on drug levels.

Why it matters

Sodium bicarbonate is the main antidote for tricyclic antidepressant cardiotoxicity and works through two separate mechanisms: it facilitates passage through blocked sodium channels by increasing extracellular sodium load and keeps the drug in a protein-bound, less toxic form by alkalinizing the serum. Therefore, in patients with a QRS duration of 100 milliseconds or more, ventricular arrhythmia, or hypotension, bicarbonate is started without delay and the target serum pH is maintained between 7.45 and 7.55; these thresholds are consistently recommended in the GEMNet 2011 guideline and the StatPearls compilation.

The basis for QRS thresholds in risk classification is the 1985 New England Journal of Medicine study by Boehnert and Lovejoy: this study showed that QRS duration is a more reliable predictive tool than serum drug level and identified thresholds of 100 milliseconds and 160 milliseconds. The 1995 study by Liebelt and colleagues suggested that the terminal R wave in the aVR lead could be a stronger predictor than QRS duration: when the R wave in aVR was 3 mm or more, sensitivity in predicting seizure or arrhythmia was 81%, specificity was 73%, and the positive predictive value (43%) of this finding was higher than the positive predictive value of the QRS 100-millisecond threshold (35%); in multivariate logistic regression analysis, the terminal R wave in aVR was the only independent ECG variable that remained significant.

Benzodiazepines are chosen as the first-line treatment for seizures because phenytoin, as a sodium channel blocker, can worsen the cardiotoxic effects of tricyclic antidepressants; therefore, the GEMNet guideline explicitly recommends avoiding phenytoin. In the management of hypotension, isotonic fluid resuscitation is recommended first; if there is no response, adding a vasopressor and continuing bicarbonate therapy is advised.

Hypertonic saline, while shown to narrow the QRS in pig models, is considered only as a secondary option in resistant cases since there is insufficient evidence in human studies — this point remains unclear and controversial among guidelines. Physostigmine is CONTRAINDICATED; it is known that when given with tricyclic antidepressants, it can trigger bradyarrhythmia, asystole, and seizures, and this contraindication is emphasized in both the StatPearls and GEMNet guidelines.

Lipid emulsion therapy is applied as a rescue treatment in cases of cardiotoxicity or cardiac arrest resistant to standard therapy (bicarbonate, fluids, vasopressors) according to the 2017 position statement of the ACMT (American College of Medical Toxicology); although the ACMT does not define a definitive cumulative dose limit, the literature recommends a 10 mL/kg limit. The common point of all these recommendations is that maintaining a low threshold by starting bicarbonate early and early transfer to intensive care are the most important factors determining survival.

Critical complications

• QT prolongation and associated torsades de pointes can occur; this is due to the potassium channel blocking effect of tricyclic antidepressants, and if QT/QTc remains prolonged despite sodium bicarbonate administration, 2 grams of intravenous magnesium sulfate is recommended.
• Seizure-related hyperthermia, rhabdomyolysis, and metabolic acidosis can develop; rapid termination of the seizure with benzodiazepines is decisive in preventing these triple complications, and developed rhabdomyolysis is managed with aggressive fluid therapy and monitoring of urine output.
• Severe anticholinergic hyperthermia develops due to central and peripheral muscarinic receptor blockade; in severe hyperthermia, if external cooling is insufficient, neuromuscular paralysis and endotracheal intubation may be necessary.
• Aspiration pneumonia can occur in patients with low consciousness when activated charcoal is given or during vomiting; therefore, activated charcoal should not be given without assessing airway protective reflexes, and if necessary, intubation should be ensured first.
• In cardiotoxicity resistant to fluids and bicarbonate, lipid emulsion therapy itself can also be a source of complications; if the infusion rate exceeds the clearance rate, fat overload syndrome (respiratory distress, fever, jaundice, hepatosplenomegaly, cytopenias, and coagulopathy) may develop, and therefore adherence to the cumulative dose limit is recommended.
• Bicarbonate therapy itself can lead to hypernatremia, excessive metabolic alkalosis, and hypokalemia; therefore, after each bolus, serum sodium, potassium, and blood gas levels should be monitored, and care should be taken not to exceed the target pH range (7.45–7.55).
• Prolonged hypotension and low cardiac output can lead to multiorgan hypoperfusion and acute kidney injury; in these patients, early vasopressor support and intensive care monitoring may limit organ damage.

Special considerations

• Management of tricyclic antidepressant overdose during pregnancy is the same as in adults (sodium bicarbonate, airway-respiratory-circulation support should not be delayed); however, in pregnant patients, reducing aortocaval compression by lying on the left side and requesting obstetric/perinatology consultation as early as possible is necessary because maternal hemodynamic compromise directly affects fetal perfusion.
• In pediatric patients, even a single adult dose in tablet form can be lethal as the toxic dose threshold per kilogram is lower than in adults; therefore, if there is a history of tricyclic antidepressant ingestion in a small child, hospital admission and prolonged monitoring are recommended even if the ECG is normal and the patient is asymptomatic.
• In geriatric patients, due to underlying conduction system disease and polypharmacy (especially concurrent use of other QT-prolonging or sodium-channel-affecting drugs), cardiotoxicity may occur earlier and more severely even at the same overdose amount; in these patients, obtaining a baseline ECG early for comparison and reviewing the list of concomitant medications is recommended.
• In patients with renal failure, elimination of tricyclic antidepressants and their active metabolites is delayed, so toxicity may persist longer; moreover, during sodium bicarbonate therapy, the risks of fluid overload and hyperosmolarity should be monitored more closely compared to patients with normal kidney function.
• In patients with liver failure, the hepatic metabolism of tricyclic antidepressants is slowed, so serum levels may remain higher than expected for a prolonged period, and clinical deterioration may appear later; therefore, in patients with liver failure, extending the observation period beyond the standard 6 hours should be considered.
• In obese patients, lipid emulsion therapy should be calculated based on lean (ideal) body weight rather than actual body weight; calculating based on actual weight can lead to overdosing and an increased risk of fat overload syndrome.
• During breastfeeding, after tricyclic antidepressant overdose, breastfeeding should be paused until the mother stabilizes, and the choice of medication should be reassessed following psychiatric evaluation; decisions about breastfeeding during the acute poisoning period should be made once the mother's clinical status has stabilized.

Follow-up plan

Every symptomatic patient or patient with widened QRS on ECG should be referred to a center with continuous ECG monitoring, intensive care, and toxicology support. Even in asymptomatic patients with a normal initial ECG, at least 6 hours of observation is required, including serial ECGs and repeated clinical assessments during this period; if no new findings develop during this time, QRS duration remains within normal limits, and the ingested amount is considered low, discharge can be considered after confirmation with the local toxicology service and psychiatric evaluation.

In patients receiving sodium bicarbonate treatment, serum sodium, potassium, blood gas pH, and QRS duration should be closely monitored after each bolus; hypernatremia and excessive alkalosis should be prevented. In patients receiving lipid emulsion, triglyceride levels and liver/pancreas function tests should be monitored for fat overload syndrome.

Every patient being discharged should be assessed for suicide risk or intentional self-harm, and psychiatric evaluation and safety measures (limiting access to medications, close follow-up planning) should be completed before discharge. In every case, if there is any doubt regarding the dose or drug interactions, it is recommended to contact the regional Poison Information Center (Turkey 114).

In the long term, follow-up appointments should be planned after discharge for neurological and cognitive assessment in patients who have experienced severe seizures or prolonged hypoxia/hypotension.