TRANEXAMIC acid (trans-4-aminomethyl cyclohexane carboxylic acid) is a synthetic amino acid commonly used in treating various kinds of bleeding disorders. 1,2It produces an antifibrinolytic effect by competitively inhibiting the activation of plasminogen to plasmin. As a result of this inhibition, fibrin is not broken down, and this allows the formation of a more stable clot, thereby reducing the risk of recurrent bleeding. We report a case of accidental intrathecal administration of a large dose of tranexamic acid with a fatal outcome despite aggressive resuscitation.
The patient was a 49-yr-old, 52-kg woman with rectal adenocarcinoma who had undergone resection of the colon. She was admitted for management of sacrococcygeal dysthesia and left sciatica, which had developed gradually over 3 yr despite several courses of adjuvant chemotherapy. A bone scan showed local recurrence of tumor in the lower lumbar spine and in both sacroiliac joints. Neither oral morphine, 300 mg/day, nor intravenous morphine for patient-controlled analgesia (infusion of 7 mg/h plus bolus dose of 8 mg) provided sufficient pain relief. After obtaining satisfactory analgesia with a trial of epidural morphine, we implanted an indwelling intrathecal access device, with the intrathecal catheter tunneled subcutaneously and inserted 12 cm into the L3–L4 interspace. The patient had received a bolus injection of 0.5 mg morphine diluted in 2 ml normal saline through an angled Huber needle with an extension line twice daily for 3 days. The attached line was used because she could not tolerate repetitive transcutaneous punctures. The patient also received tranexamic acid, 500 mg diluted in 10 ml normal saline, intravenously four times daily for persistent hematuria; no adverse effect associated with the tranexamic acid was reported by the patient. By accident, an intern injected the third dose of tranexamic acid into the line connected to the intrathecal device. Immediately after the injection, the patient complained of severe pain in the back and the gluteal region. Her blood pressure increased to 200/130 mmHg (her baseline blood pressure was 130/80 mmHg). Generalized convulsions developed 2 min later. The seizure abated after intravenous injection of 10 mg diazepam. However, the patient became pulseless and cyanotic. The resuscitation team instituted closed-chest cardiac massage and ventilation through a mask immediately. An electrocardiographic monitor was attached, and ventricular fibrillation was noted. The patient was treated with direct-current electrical shocks of 300 J, 360 J, and 360 J consecutively, which failed to convert the rhythm into sinus rhythm. The patient was then intubated and mechanically ventilated. However, the ventricular fibrillation was refractory to subsequent resuscitation attempts, including the administration of drugs and more electrical shocks with continued closed-chest cardiac massage. The drugs used during the procedure included cumulative doses of 400 mg lidocaine, 8 mg epinephrine, 1 mg atropine, and 300 mg amiodarone. The ventricular fibrillation persisted despite the resuscitative measures, and the patient was pronounced dead after 1.5 h of resuscitation attempts. The patient's family refused to allow an autopsy to be performed. The patient did not have any other major systemic diseases other than rectal cancer. Her chemotherapy did not involve any cardiac-toxic agents. The patient's laboratory data on the day before the fatal episode were as follows: hemoglobin, 12.5 mg/dl; leukocyte count, 7,520/mm3; glucose, 86 mg/dl; sodium, 134 mEq/l; potassium, 4.5 mEq/l; calcium, 4.8 mEq/l; chloride, 105 mEq/l; and normal electrocardiographic results, with regular sinus rhythm.
Tranexamic acid has a low acute toxicity at therapeutic doses. The lethal oral dose exceeds 3–10 g/kg body weight in all species studied, and the 50% lethal dose after intravenous injection is approximately 1–1.5 g/kg body weight in mice, rats, rabbits, and dogs. 3However, when applied topically to the cerebral cortex in animal studies, tranexamic acid produced powerful seizures. 4–6
Little is known about the effect of direct intrathecal administration of tranexamic acid in humans. Wong et al. 7reported a case of inadvertent intrathecal injection of 75 mg tranexamic acid in a healthy, 18-yr-old man. The patient had satisfactory anesthesia with sensory block level to T10 for a scheduled appendectomy. When the patient returned to the ward 4 h after the spinal injection, he experienced persistent motor and sensory block of both lower extremities, restlessness, and urinary incontinence. He developed clonic convulsions that progressed to a generalized seizure with hyperthermia of 40.5°C 5.5 h after the injection. His seizure and fever gradually subsided over the next 5 h after treatment with intravenous diazepam and diclofenac. He recovered completely, without any sequelae, the next day. de Leede-Van der Maarl et al. 8also reported a case of a 68-yr-old man who accidentally received an intrathecal injection of 50 mg tranexamic acid instead of the intended anesthetics. He developed status epilepticus immediately after administration of the drug, which had to be treated with midazolam, phenytoin, and thiopental infusion for 2 days. His medical course was complicated, with multiorgan dysfunction and hypotonic paresis of all four extremities, which eventually resolved but resulted in residual bilateral peroneal palsy. In our case, the patient reported severe back pain, and convulsions developed immediately after intrathecal injection of 500 mg tranexamic acid, which was a much larger dose compared with those in the former two case reports. The seizure was controlled with prompt administration of intravenous diazepam. However, the patient quickly developed ventricular fibrillation, which was not amenable to aggressive resuscitative efforts. It has been reported that intravenous injection with 1 g tranexamic acid results in a concentration of 5–20 mg/l in the plasma 9–11and 2–5 mg/l in the cerebrospinal fluid. 12Assuming that the drug was thoroughly diluted within the thecal space containing 500 ml of cerebrospinal fluid, 500 mg tranexamic acid would produce a concentration of 1,000 mg/l in the cerebrospinal fluid, which is 500 times greater than the therapeutic level (2–5 mg/l). 3Yamamura et al. 5reported that intracisternal injection of tranexamic acid at a dose of 5 mg/kg in cats causes seizure activities within 45 to 60 s. In the present case, the patient received an intrathecal injection of tranexamic acid at a dose of 10 mg/kg, which far exceeded the epileptogenic dose of 5 mg/kg in the report by Yamamura et al. 5The most disturbing aspect of this case is the rapid development of ventricular fibrillation despite successful management of the patient's seizure. The ventricular fibrillation was refractory to all resuscitative efforts.
We do not know the mechanism by which tranexamic acid induces convulsions or ventricular fibrillation. One may postulate that very high doses of tranexamic acid would cause massive sympathetic discharge, as evidenced by the initial hypertensive response and the subsequent ventricular arrhythmia in this patient. The concomitant administration of 15 mg tetracaine in the case reported by Wong et al. 7seemed to ameliorate the central nervous system effects of the 75 mg tranexamic acid when compared with that reported by de Leede-van der Maarl et al. 8with 50 mg tranexamic acid.
An intrathecal drug administration system is a well-accepted method of delivering morphine. On rare occasions, the line attached to the intrathecal access device could be misused as an intravenous line. Fatal or near-fatal catastrophes after inadvertent intrathecal injection of contrast media (amidotrizoate), 13penicillin, 14muscle relaxants (gallamine triethiodide), 15and chemotherapy agents (vincristine) 16have been reported. Among the patients who were successfully resuscitated, the treatments included administration of anticonvulsants, cerebrospinal fluid lavage, and subsequent intensive monitoring. 13–16In the present case, we found that the convulsions were responsive to benzodiazepine. However, refractory ventricular fibrillation and subsequent cardiovascular collapse precluded the possibility of cerebrospinal fluid lavage therapy. To prevent this iatrogenic complication, we have since begun labeling lines with different colors so that the intrathecal drug administration line is clearly identified, and we strongly recommend taking these precautionary measures.