Charcoal Hemoperfusion in the Treatment of Levothyroxine Intoxication

Introduction

T hyrotoxicosis due to ingestion of excess thyroid hormone may arise from an intentional, surreptitious ingestion of the hormone, known as thyrotoxicosis factitia (1 –3), or from accidental ingestion of excessive doses (4,5). We report a case of massive acute levothyroxine (l-thyroxine) intoxication due to an error in the preparation of capsules (with micrograms being read as milligrams). The resulting severe biochemical and clinical manifestations were dramatically reversed by the application of a charcoal hemoperfusion extractive technique. We discuss the risks of using preparations processed by pharmacists instead of using industrialized drugs. We also discuss current extractive techniques for the treatment of intentional or accidental l-thyroxine intoxication.

Clinical Case

A 61-year-old woman sought medical care at our hospital because of symptoms of psychomotor agitation, mental confusion, delirium, and weight loss, as observed by her family in the preceding days. According to family members, she had a history of goiter, which had been treated with l-thyroxine, 50 μg daily for the past 20 years. On admission, the patient presented with mental confusion, delirium, staring eyes, neck rigidity, normal body temperature, heart rate of 114 beats/min, and blood pressure of 160 mm Hg systolic over 90 mm Hg diastolic. Routine laboratory tests, including hemoglobin 12 g/L, hematocrit 36%, white-cell count 5.88 × 10⁹/L, platelet count 150 × 10⁹/L, sodium 137 mmol/L, potassium 4.3 mmol/L, total CO₂ 29.8 mmol/L, calcium 2.20 mmol/L, phosphorus 1.0 mmol/L, aspartate transaminase (AST) 42 U/L, alanine aminotransferase (ALT) 37 U/L, alkaline phosphatase 69 U/L, showed normal results. The patient underwent neurological assessment, including magnetic resonance imaging and evaluation of cerebrospinal fluid; there were no significant alterations and the neurological signs were considered likely to be metabolic in nature. The patient received interventions to alleviate her symptoms, such as acetaminophen for pain or discomfort, and intravenous fluid hydration, but 24 hours after hospital admission, she was transferred to the intensive care unit, because of an acute worsening of her general status and impending loss of consciousness. She was in a state of stupor, and she had a fever of 100.4°F (38°C), atrial fibrillation, tachypnea, and respiratory failure. The patient received mechanical ventilation and intensive care, including β-blocker treatment. Thyroid hormone levels showed surprisingly high values, with triiodothyronine (T3), the most powerful of thyroid hormones, of 101.84 nmol/dL or 6613.0 ng/dL (normal values: 1.2–3.2 nmol/L or 80.0–210.0 ng/dL); free thyroxine (T4), the major hormone secreted by the thyroid folicular cells, of 955.90 nmol/L or 74.1 ng/dL (normal values: 11.61–27.09 nmol/L or 0.9–2.1 ng/dL); and thyroid-stimulating hormone (TSH) < 0.01 mU/L or < 0.01 μU/mL (normal values: 0.35–5.5 mU/L or 0.35–5.5 μU/mL). These and other subsequent hormone measurements were performed by an automated chemiluminescence system.

The family revealed that during the last 10 days the patient had been taking l-thyroxine preparations that were processed by a pharmacist, instead of the commercially available medication. The concentration of l-thyroxine as listed on the label was 50 mg per capsule—1000-fold higher than the prescribed dose of 50 μg/day. The very high concentration of l-thyroxine in the capsules was confirmed by high-performance liquid chromatography. Serum thyroglobulin was undetectable and the level of thyrotropin receptor antibodies was 0.8 U/L (positive value > 10 U/L).

Metoprolol treatment was commenced to diminish sympathomimetic activity. Because of the absence of clinical improvement and the progressive hemodynamic instability, we decided that the patient should undergo charcoal hemoperfusion. Previous reports have shown that this technique may on occasion be successful (4 –7). The hemoperfusion circuit included blood lines, blood pump, air detector, venous air trap, and cartridges (Adsorba 300 g; Gambro do Brasil Ltda, São Paulo, Brazil). Initially, a cannula was inserted into a large central vein using a dialysis catheter, and the hemoperfusion circuit was primed in vertical position with 500 mL of 5% dextrose in water to load the charcoal with glucose to avoid hypoglycemia. Thereafter, the cartridge was rinsed with 2 L of 0.9% sodium chloride solution with heparin (2500 IU/L). The anticoagulation was performed using a bolus dose of heparin 2000 IU followed by 10 IU/(kg · h) to maintain partial thromboplastin time at about twice the reference value. The blood flow used was around 200 mL/min for about 2 hours. Hematocrit or albumin levels were not used to correct for possible fluid shifts because this procedure does not alter body fluids.

Three sessions of hemoperfusion were performed at 24-hour intervals. During the sessions, the patient showed hemodynamic stability as well as glycemic control, without thrombocytopenia or other coagulation events. Tables 1 –3 depict the levels of total T3 and T4 and serum-free l-thyroxine level (FT4) at different times during the three hemoperfusion sessions, showing decline over time. Clinical and laboratory tests showed dramatic improvement with hemodynamic stabilization, recovery of consciousness, and no need for ventilatory support approximately 32 hours after the last session of hemoperfusion. The patient remained in the hospital for 30 days because of her highly compromised general health status and a severe thyrotoxic myopathy leading to motor disability. At discharge the serum thyroid hormone levels were as follows: T3, 1.84 nmol/dL or 120 ng/dL; free T4, 11.58 nmol/L or 0.9 ng/dL; and TSH, 0.4 mU/L or 0.4 μU/mL. With intensive rehabilitation treatment, the patient was able to recover fully and resumed normal daily activities.

Discussion

Our patient developed severe clinical symptoms, equivalent to thyroid storm, because of an involuntary ingestion of massive doses of l-thyroxine. Her involuntary ingestion was due to pharmacist error in the preparation of capsules. For 9 days, she ingested an estimated dose of 50 mg of l-thyroxine per day, which was 1000-fold greater than the recommended daily prescription. She took the pharmacist-processed capsules of her own volition as an alternative to taking commercially available industrialized drugs.

In the absence of specific treatments for l-thyroxine intoxication and given the severity of the clinical presentation, charcoal hemoperfusion might be used (7). We performed hemoperfusions using filters with charcoal columns, in a similar manner as to that described by Burman et al. (5). They used a neutral Amberlite resin to bind and remove l-thyroxine in intoxicated dogs. Using this procedure, we observed dramatic improvements in laboratory tests and clinical assessments, and the patient was able to recover. Across the three sessions of hemoperfusion, serum levels of total T3, total T4, and free T4 decreased significantly. They decreased in reproducible patterns, of approximately 50% after each session, and the hormone levels reached nadirs 1–2 hours into each procedure, probably because of the saturation of the charcoal column. In addition, very low or negligible spontaneous declines in the levels of thyroid hormones were observed during the 24-hour intervals between each session. It is likely that performing charcoal hemoperfusion continuously and changing the filters every 2 hours would have speeded up the reduction of thyroid hormones and the patient's recovery. It is worth to mention that FT4 levels were higher than total T4 levels at the end of the third session of charcoal hemoperfusion probably because the serum levels of FT4 were already proportionally higher than total T4 at the baseline. This was probably due to the saturation of thyroid hormone transport proteins after the ingestion of such massive amounts of T4. During the charcoal hemoperfusion process, total T4 and FT4 presented a similar decrease.

Acute iatrogenic thyrotoxicosis may represent a potentially life-threatening endocrine emergency. Although thyroid hormones are very commonly prescribed, there are only a few reports of massive accidental or intentional intoxication with these compounds leading to emergencies (1,3,4,6).

According to de Luis et al. (8), adult exposure to l-thyroxine has a wide range of presentations, and most adults either do not develop symptoms or become only minimally symptomatic. They observed that even with a total dose of 5 mg of l-thyroxine, as ingested by a young woman, there were only mild symptoms of thyrotoxicosis with no arrhythmias or delusions. However, in another report, the same level of exposure led to a stuporous mental state and acute respiratory failure, recognized as a thyroid storm (1). In the report by Binimelis et al. (4), all six patients who had been admitted after an erroneous massive intake of l-thyroxine (70–1200 mg over an interval of 2–12 days) developed the classic symptoms of thyrotoxicosis; five presented with grade II–III coma and one presented with a stuporous mental state.

The most appropriate treatment after l-thyroxine intoxication has yet to be established (8). In the absence of specific treatments for iatrogenic thyrotoxicosis, extractive techniques such as peritoneal dialysis, plasmapheresis, and hemoperfusion through a resin bed or charcoal column have been used to remove excess circulating thyroid hormones (4,6,9,10).

In 1976, Burman et al. (5) reported significant reductions in T4 and T3 plasma levels in intoxicated dogs by hemoperfusion using a resin system. In 1987, Binimelis et al. (4) reported that they treated six patients with massive intakes of l-thyroxine and observed a much higher plasma disappearance rate of T4 with extractive techniques (plasmapheresis and/or hemoperfusion) than with standard medical treatments. Changes in T3 were greater with hemoperfusion than with plasmapheresis.

Henderson et al. (10) reported that plasmapheresis was not an effective treatment of l-thyroxine overload. More recently, Solá et al. (6) also reported an apparent failure of hemoperfusion to obtain faster reduction in the blood levels of free triiodothyronine. However, their patient had been accidentally intoxicated with triiodothyronine, which is associated with a more rapid spontaneous disappearance from plasma and with a shorter course of intoxication than that of l-thyroxine.

It should be mentioned that the extraction process is not effective in situations characterized by ongoing production of thyroid hormones such as excessive iodine exposure (9). In these cases a rebound of serum T4 levels was observed after the extraction process was stopped.

The use of l-thyroxine is much more widespread than that of triiodothyronine. The longer half-life of l-thyroxine represents an additional potential risk in the case of intoxication. The worldwide practice of using drugs processed by pharmacists instead of industrialized drugs requires special attention and extra care, because of the inherent risk of error. In the case of thyroid hormones, the change from micrograms to milligrams may result in catastrophic and life-threatening events, as experienced by our patient. Finally, the use of hemoperfusion with a charcoal filter appears to be a very important therapeutic tool for the treatment of acute and severe forms of thyrotoxicosis due to l-thyroxine intoxication.