3.4.3 Myxoedema coma

Myxoedema coma is the extreme expression of severe hypothyroidism and fortunately is quite rare. The first reported case appears to have been in 1879 by Ord from St Thomas’s Hospital, London. Two other patients who died in a hypothyroid coma were reported in 1888 in the proceedings of the Clinical Society of London (1). The next cases in the literature appeared in 1953 (2, 3), and some 300 cases have since been reported. Epidemiological data indicate an incidence rate of 0.22/1 000 000 per year (4). The most common presentation of the syndrome is in hospitalized elderly women with long-standing hypothyroidism, with 80% of cases occurring in women over 60 years of age. However, myxoedema coma occurs in younger patients as well, with 36 documented cases occurring during pregnancy (5, 6). In spite of early diagnosis and treatment, the mortality rate may be as high as 40–60%.

Patients with myxoedema coma generally present in the winter months, suggesting that external cold may be an aggravating factor. Other precipitating events include pulmonary infections, congestive heart failure, and cerebrovascular accidents (Box 3.4.3.1). The comatose and hypoventilating patient is also at risk for pulmonary infection or aspiration pneumonia as a secondary event. Similarly, other abnormalities frequently accompanying myxoedema coma, such as hypoglycaemia, hypercalcaemia, hyponatraemia, hypercapnia, and hypoxaemia, may be either precipitating factors or secondary consequences. In hospitalized patients, drugs such as anaesthetics, narcotics, sedatives, antidepressants, and tranquillizers may depress respiratory drive and thereby either cause or compound the deterioration of the hypothyroid patient into coma.

Hypothermia (often profound to 80 °F (26.7 °C)) and unconsciousness constitute two of the cardinal features of myxoedema coma. The syndrome will typically present in a patient who develops an infection or other systemic disease superimposed upon previously undiagnosed hypothyroidism. Sometimes a history of antecedent thyroid disease, thyroidectomy, treatment with radioactive iodine, or thyroxine (T₄) replacement therapy that was discontinued for no apparent reason can be elicited. Other clues to the presence of underlying thyroid disease may be seen on examination of the neck, such as a surgical thyroidectomy scar, goiter, or even the absence of palpable thyroid tissue as may be seen in chronic Hashimoto’s thyroiditis. A pituitary or hypothalamic basis for hypothyroidism is encountered in less than 10–15% of patients. In one large series (7) that identified 12 patients with myxoedema coma, the findings on presentation included hypoxaemia in 80%, hypercapnia in 54%, and hypothermia with a temperature below 94 °F (34.4 °C) in 88%. Six patients died despite treatment with thyroid hormone. A dreaded aspect of the usual clinical course is progression into respiratory failure and CO₂ retention which is heralded by hypoventilation with lethargy progressing to stupor and then coma. Because of the delayed metabolism of drugs in hypothyroidism, the deterioration may be hastened by the use of sedative hypnotics or narcotics.

In myxoedema coma, typical findings of hypothyroid heart disease may include bradycardia, decreased quality and intensity of the heart sounds, enlarged cardiac silhouette, and minor ECG abnormalities such as varying degrees of block, low voltage, flattened or inverted T waves, and prolonged Q–T interval which can result in torsades de pointe ventricular tachycardia (8). Myocardial infarction should be ruled out by the usual diagnostic procedures. The lactate dehydrogenase isoenzyme pattern in severe hypothyroidism may mimic that of myocardial infarction (9), and creatine kinase levels also are elevated (10, 11). Moreover, aggressive or injudicious T₄ replacement may increase the risk of myocardial infarction (see below). The enlarged cardiac silhouette may be due, in part, to ventricular dilatation or a pericardial effusion which can be confirmed by echocardiography. This fluid is rich in mucopolysaccharide and tends to accumulate slowly over time, only rarely causing cardiac tamponade.

Cardiac contractility is impaired, leading to reduced stroke volume and cardiac output, but congestive heart failure is rare. T₄ replacement therapy will slowly reverse the abnormalities in left ventricular function; although the pericardial effusion may also gradually diminish, reduced cardiac output with hypotension secondary to the effusion must be borne in mind. Patients should be admitted to an intensive care unit because of the propensity for shock and potentially fatal arrhythmias. Hypotension may occur in spite of increases in total body water and extracellular fluid volume because of reduction in intravascular volume. Although blood pressure may be normalized with T₄ replacement, severe hypotension or shock may supervene acutely before the T₄ effect is seen, necessitating the use of pressor drugs.

The mechanism for hypoventilation in profound myxoedema is a combination of a depressed hypoxic respiratory drive and a depressed ventilatory response to hypercapnia (12). CO₂ narcosis results from the reduction in alveolar ventilation with the hypoventilation compounded by impairment in respiratory muscle function ultimately leading to coma. The central factor in the pathophysiology of coma appears to be a depressed ventilatory response to CO₂ (13–15). When present, obesity may impair the bellows action of the chest. Improvement in the response to CO₂ after T₄ therapy has been seen in some (13, 15, 16) but not all (12) studies. Irrespective of the underlying pathophysiology, the mechanical function of the chest in myxoedema coma usually is reduced sufficiently to require mechanically assisted ventilation. Tidal volume may be reduced by other factors such as pleural effusion or ascites. Upper airway partial obstruction may also play a role, caused by oedema or swelling of the tongue, or laryngeal obstruction due to marked oedema of the vocal cords. Hypothyroid patients may be predisposed to increased airway hyper-responsiveness and chronic inflammation (17). Even with appropriate and adequate therapy, the complexity of the pathophysiology of respiratory failure means that ultimate recovery may be prolonged.

The gastrointestinal tract in myxoedema may be marked by mucopolysaccharide infiltration and oedema of the muscularis and neuropathic changes leading to impaired peristalsis, obstipation, and potential paralytic ileus. Given the risks of anaesthesia in the profoundly hypothyroid patient, surgical intervention should be temporized for apparent obstruction by conservative management with decompression until the therapeutic response to thyroid hormone might occur. Initially, parenteral administration of T₄ or triiodothyronine (T₃) may be preferable because absorption of oral medications could be impaired due to the gastric atony often present in myxoedema coma. Ascites has been documented in 51 cases (18) and gastrointestinal bleeding can occur secondary to a coagulopathy.

Alterations in mineral metabolism and renal clearance in severe hypothyroidism may include decreases in plasma volume, serum sodium and osmolality, glomerular filtration rate, and renal plasma flow, and increases in total body water, urine sodium, and urine osmolality. Atony of the urinary bladder with retention of large residual urine volumes is commonly seen. High creatine kinase levels are typical of hypothyroidism, but unusually high values may be a clue to underlying rhabdomyolysis. Increased serum antidiuretic hormone levels (19) and impaired water diuresis caused by reduced delivery of water to the distal nephron (20) are likely to account for the hyponatraemia. Depending upon its duration and severity, hyponatraemia will add to altered mental status, and when severe may be largely responsible for precipitating the comatose state. T₄ treatment promotes water diuresis resulting in an increase in serum sodium and a decrease in oedema and total body water.

Although coma is the predominant clinical presentation in myxoedema coma, a history of disorientation, depression, paranoia, or hallucinations (‘myxoedema madness’) may often be elicited. Other findings present either just before entering the comatose state or early during recovery include cerebellar signs, such as poorly coordinated purposeful movements of the hands and feet, ataxia, adiadochokinesia, poor memory and recall, or even frank amnesia. Abnormal findings on electroencephalography are few and include low amplitude and a decreased rate of α-wave activity. Status epilepticus has been described (21) and up to 25% of patients with myxoedema coma may experience minor to major seizures possibly related to hyponatraemia, hypoglycaemia, or hypoxaemia due to reduced cerebrovascular perfusion from low cardiac output and atherosclerotic vessels in elderly patients. T₄ treatment will generally lead to improved perfusion.

A microcytic anaemia may be seen secondary to gastrointestinal haemorrhage, or a macrocytic anaemia due to vitamin B₁₂ deficiency which may also worsen the neurological state. Granulocytopenia with a decreased cell-mediated immunological response may contribute to a higher risk of severe infection. In contrast to the tendency to thrombosis seen in mild hypothyroidism, severe hypothyroidism is associated with a higher risk of bleeding due to coagulopathy related to an acquired von Willebrand’s syndrome (type 1) and decreases in factors V, VII, VIII, IX, and X (22). The von Willebrand syndrome is reversible with T₄ therapy (23). Another cause of bleeding may be disseminated intravascular coagulation associated with sepsis.

The first clinical clue to the diagnosis of myxoedema coma may be hypothermia which occurs in approximately 75% of patients and may be dramatic (below 80°F (26.7°C)), with temperatures of less than 90°F (32.2°C) being associated with the worst prognosis. Because patients with myxoedema and infection may not mount a febrile response, a diagnosis of profound hypothyroidism should be entertained in any unconscious patient with a known infection but no fever. In view of the latter and because undiscovered infection might lead inexorably to vascular collapse and death, some authors have advocated the routine use of antibiotics in patients with myxoedema coma. Underlying hypoglycaemia may further compound the decrement in body temperature. With T₄ therapy, the hypothermia gradually improves in parallel with the fall in serum thyroid-stimulating hormone (TSH) and increments in serum T₄ and T₃ levels.

The typical patient presenting with myxoedema coma is a woman in the later decades of her life who may have a history of thyroid disease and who is admitted to hospital during the winter months, possibly with pneumonitis. Physical findings could include bradycardia, macroglossia, hoarseness, delayed reflexes, dry skin, general cachexia, hypoventilation, and hypothermia, commonly without shivering. Laboratory evaluation may indicate anaemia, hyponatraemia, hypercholesterolaemia, and increased serum lactate dehydrogenase and creatine kinase. On lumbar puncture there is increased pressure and the cerebrospinal fluid will have a high protein content. The electrocardiogram and chest radiograph may demonstrate the characteristic findings described above. If hypothyroidism is suspected in a comatose patient, blood should be obtained for thyroid function testing but treatment should not be delayed to await laboratory confirmation of the diagnosis. On the other hand, a correct diagnosis is particularly important because the unnecessary administration of large doses of T₄ or T₃ to an elderly euthyroid patient could induce a fatal arrhythmia or coronary event. In addition to routine thyroid function tests, ancillary studies should be performed to determine whether CO₂ retention, hypoxia, hyponatraemia, or infection are present. Indeed, in many patients the clinical features may be so notable as to render the measurement of thyroid function tests necessary only for confirmation of the diagnosis. The urgency of the diagnosis should be stressed to the laboratory, which often can perform a serum T₄ and TSH determination in 3–4 h. Although an elevated serum TSH concentration is the most important laboratory evidence of the diagnosis, the presence of severe complicating systemic illness or treatment with drugs such as dopamine, dobutamine, or corticosteroids may serve to reduce the elevation in TSH levels (24, 25). There may also be a pituitary cause for the hypothyroidism, in which case an increased TSH would not be found. Until the presence of pituitary disease is ruled out, corticosteroid therapy is recommended in addition to T₄.

Myxoedema coma is a true medical emergency, and treatment must be instituted in a critical care setting with modern electronic monitoring equipment as soon as the diagnosis is made in view of the extremely high mortality anticipated in these patients when treatment is delayed. As outlined below, a multifaceted approach is required because of the multiple metabolic derangements derived from, or affecting, several organ systems which may be contributing to the comatose state.

The patient’s comatose state is perpetuated by hypoventilation with CO₂ retention and respiratory acidosis. Appropriate diagnostic and therapeutic measures must be instituted for any suspicious infiltrate seen on chest radiographs. The high mortality rate is often related to inexorable respiratory failure, and hence maintenance of an adequate airway and prevention of hypoxaemia is the single most important supportive measure required to avoid a disastrous outcome. Mechanical ventilation is usually required during the first 36–48 h, particularly if the hypoventilation is related in part to drug-related respiratory depression. Although the patient may become alert by the second or third day of treatment, it may be necessary to continue assisted ventilation for as long as 2–3 weeks in some patients.

Intubation may be necessary initially or with worsening of hypoxaemia or hypercarbia, and arterial blood gases need to be monitored regularly until the patient is fully recovered. The hypercapnia may be rapidly relieved with mechanical ventilation, but the hypoxia may tend to persist possibly due to shunting in nonaerated lung areas (26). Moreover, the physician should guard against extubating the patient prematurely; some case reports have cited the danger of relapse, and it should not be attempted until the patient is fully conscious.

Total body sodium is believed to be normal to increased, but it is the impaired excretion of water that causes hyponatraemia. Low serum sodium may cause a semicomatose state or seizures even in euthyroid patients, and the very severe hyponatraemia (105–120 mmol/l) in profound myxoedema is likely to contribute substantially to the coma in these patients. With such severe hyponatraemia, it may be appropriate to administer a small amount of hypertonic saline (50–100 ml 3% sodium chloride), enough to increase sodium concentration by about 2 mmol/l early in the course of treatment, and this can be followed by an intravenous bolus dose of 40–120 mg furosemide to promote a water diuresis (27). A small quick increase in the serum sodium concentration (2–4 mmol/l) is effective in acute hyponatraemia because even a slight reduction in brain swelling results in a substantial decrease in intracerebral pressure (28). On the other hand, too rapid correction of hyponatraemia can cause a very dangerous complication, the osmotic demyelination syndrome. In patients with chronic hyponatraemia this complication is avoided by limiting the sodium correction to less than 10–12 mmol/l in 24 h and to less than 18 mmol/l in 48 h.

After achieving a sodium level of more than 120 mmol/l, no further hypertonic saline infusion should be required, and restriction of fluids may be all that is necessary to correct hyponatraemia, especially if it is mild (120–130 mmol/l). Because of the likelihood of decreased cardiac reserve, therapy with saline or other intravenous fluids must be approached cautiously. If hypoglycaemia is present, dextrose in 0.5 N sodium chloride may be used to correct the low blood glucose. With regard to fluid or saline therapy, careful monitoring of volume status based on clinical parameters and central venous pressure measurements is essential in patients with significant cardiovascular decompensation.

A new vasopressin antagonist, intravenous conivaptan, has been approved by the FDA for the treatment of hospitalized patients with euvolaemic and hypervolaemic hyponatraemia. This treatment could be attempted in this clinical setting in view of the high vasopressin levels observed in myxoedema coma. Current dosing recommendations are for a 20-mg loading dose to be infused over 30 min followed by 20 mg/day continuous infusion for up to 4 days. No data are available on the use of conivaptan in severe hyponatraemia (<115 mmol/L) in hypothyroid patients, or whether sole therapy with conivaptan without hypertonic saline would be effective (29).

Restoration of body temperature to normal will require administration of T₄ or T₃. Blankets or increasing room temperature can be used to keep the patient warm until the thyroid hormone effect is achieved, but caution must be exercised in the use of more vigorous electric warming blankets. Too aggressive warming may cause peripheral vasodilatation, a precipitous fall in peripheral vascular resistance with increased peripheral blood flow, and increased oxygen consumption, which may then lead to hypotension or shock.

Hypotension should also be correctable by treatment with T₄; this may take several days and the hypotensive patient may require additional therapy. Fluids may be administered cautiously as 5–10% glucose in 0.5 N sodium chloride initially, or as isotonic normal saline if hyponatraemia is present. It is wise to administer hydrocortisone (100 mg intravenously every 8 h) until the hypotension is corrected. Pressors are only very rarely required, and the possibility of an adverse cardiac event needs to be kept in mind, especially in patients with suspected underlying ischaemic heart disease. An agent such as dopamine might be employed to maintain coronary blood flow, but patients should be weaned off the pressor as soon as possible. The physician must weigh the risk of a pressor-induced ischaemic event against the known high mortality of poorly managed hypotension in myxoedema coma.

A rising urea nitrogen, hypotension, hypothermia, hypoglycaemia, hyponatraemia, and hyperkalaemia may signal the coexistence of adrenal insufficiency. Indeed, decreased adrenal reserve has been found in 5–10% of patients on the basis of either hypopituitarism or primary adrenal failure accompanying Hashimoto’s disease (Schmidt’s syndrome). Otherwise, plasma total and free cortisol levels and the adrenal response to adrenocorticotropic hormone (ACTH) infusion should be normal in hypothyroidism or myxoedema coma. However, ACTH reserve or the ACTH response to stress may be impaired in myxoedema coma. There should be no reluctance to administer short-term corticosteroids until the patient is stable and the integrity of the pituitary–adrenal axis can be determined. On theoretical grounds, one should also administer corticosteroids when first instituting thyroid hormone therapy, in view of the potential risk of precipitating acute adrenal insufficiency due to the accelerated metabolism of cortisol that follows T₄ therapy. The typical dosage of hydrocortisone is 50–100 mg every 6–8 h during the first 7–10 days with tapering of the dosage thereafter based upon clinical response and any plans for further diagnostic evaluation.

One of the most controversial aspects of the management of myxoedema coma is which thyroid hormone medication to give and how to give it (dose, frequency, and route of administration). Because of the relative rarity of this condition, the paucity of reported treatment results, and the difficulties inherent in performing a controlled investigation, the optimum treatment remains uncertain, and several approaches will be discussed. Some of the differences of opinion relate to whether to administer T₄ and rely on the patient to convert it to the more active T₃, or to give T₃ itself. One must balance the need for quickly attaining physiologically effective thyroid hormone levels against the risk of precipitating a fatal tachyarrhythmia or myocardial infarction. T₄ provides a steady smooth onset of action with a lower risk of adverse effects.

Parenteral preparations of either T₄ or T₃ are available for intravenous administration. Although oral forms of either T₃ or T₄ can be given by nasogastric tube in the comatose patient, this route is fraught with risks of aspiration and uncertain absorption, particularly in the presence of gastric atony or ileus. Parenteral preparations of T₄ are available in ampoules of 100 and 500 μg. The latter dose, as a single intravenous bolus, was popularized by reports (30) suggesting that replacement of the entire estimated pool of extrathyroidal T₄ (usually 300–600 μg) was desirable to restore near-normal hormonal status. After this initial ‘loading’ dose, a maintenance dose of 50–100 μg is given daily (either intravenously or by mouth if the patient is adequately alert). This method may be attended by increases in serum T₄ to within the normal range within 24 h and by significant decrements in serum TSH. Larger doses of T₄ probably have no advantage and may, in fact, be more dangerous (31). Due to its conversion from T₄, a progressive increase in serum T₃ is seen after 300- to 600-μg doses of T₄, as has been described by Ridgway et al. (31).

The approach to therapy employing an initial large intravenous bolus dose of T₄ followed by maintenance therapy has been considered optimal (30, 31), but other evidence suggests improved outcomes with lower doses of thyroid hormone (32). This was also indicated in a prospective trial in which patients were randomized to receive either a 500-μg loading dose of intravenous T₄ followed by a 100-μg daily maintenance dose, or only the maintenance dose (34). The overall mortality rate was 36.4% with a lower mortality rate in the high dose group (17%) versus the low dose group (60%). Although suggestive, the difference was not statistically significant. Factors associated with a worse outcome included a decreased level of consciousness, lower Glasgow coma score, and increased severity of illness on entry as determined by an APACHE II score of more than 20.

T₄ treatment has been generally considered effective, but there is one important drawback to total reliance on T₃ generation from T₄. The rate of conversion of T₄ to T₃ is reduced in many systemic illnesses (the euthyroid sick or low T₃ syndrome) (25) and hence T₃ generation may be reduced in myxoedema coma as a consequence of any associated illness (27). Theoretically then, one might administer T₄, see increases in serum T₄ levels confirming adequate absorption, but fail to witness any significant fall in TSH or dramatic clinical improvement. As a consequence, small supplements of T₃ should be given along with T₄ during the initial few days of treatment, especially if obvious associated illness is present. Irrespective of the type of treatment selected, all patients should have continuous ECG monitoring with reduction in thyroid hormone dosage should arrhythmias or ischaemic changes be detected.

T₃ is available for intravenous use (Triostat) in 1 ml vials containing 10 μg/ml. When therapy is approached with T₃ alone, it may be given as a 10- to 20-μg bolus followed by 10 μg every 4 h for the first 24 h, dropping to 10 μg every 6 h for days 2–3, by which time oral administration should be feasible. T₃ has a much quicker onset of action than T₄ and increases in body temperature and oxygen consumption may occur 2–3 h after intravenous T₃, compared to 8–14 h after intravenous T₄. A patient with profound secondary myxoedema believed due to postpartum pituitary necrosis has been reported who presented with cardiogenic shock which responded to T₃ but not T₄ therapy (33). Because of the high mortality rate in myxoedema coma, advocates for T₃ therapy argue that the more rapid onset of action could make the difference between life and death. But the benefits of the more rapid onset of action need to be weighed against the greater risk of complications. As a consequence, it is difficult to justify the high risk/benefit ratio of a regimen that uses rapid replacement with relatively large doses of intravenous T₃ alone. Such treatment would be marked by large and unpredictable fluctuations in serum T₃ levels, and high serum T₃ levels during treatment with thyroid hormone have been associated with fatal outcomes (34).

A more conservative but seemingly rational course of management is to provide combined therapy with both T₄ and T₃. Rather than administer 300–500 μg T₄ intravenously initially, a dose of 4 μg/kg lean body weight (or about 200–300 μg) is given, and an additional 100 μg is given 24 h later. By the third day, the dose is reduced to a daily maintenance dose of 50 μg, which can be given by mouth as soon as the patient is conscious. Simultaneously with the initial dose of T₄, a bolus of 10 μg T₃ is given and intravenous T₃ is continued at a dosage of 10 μg every 8–12 h until the patient is conscious and taking maintenance T₄. Sensitivity to thyroid hormone in terms of cardiac risk varies, depending on age, cardiac medications, and the presence of underlying hypoxaemia, coronary artery disease, congestive failure, and electrolyte imbalance. Clinical improvement has been seen with even a single dose of only 2.5 μg T₃ (35). It is wise to monitor the patient for any untoward effects of therapy before administering each dose of thyroid hormone.

Clearly, given their fragile clinical state, nonemergent surgery should be deferred in a patient with myxoedema coma. However, in the patient with myxoedema coma requiring emergent surgery, the same general management principles prevail (36) with particular attention to careful monitoring of intraoperative and postoperative respiratory and cardiovascular status. Postoperatively, close monitoring for maintenance of the airway is essential.

In addition to the specific therapies outlined, other treatments will be indicated as in the management of any other elderly patient with multisystemic problems. This might include the treatment of underlying problems such as infectious processes, congestive heart failure, diabetes, or hypertension. The dosage of specific medications (e.g. digoxin for congestive heart failure) may need to be modified based on their altered distribution and slowed metabolism in myxoedema. Even with this vigorous therapy, the prognosis for myxoedema coma remains grim, and patients with severe hypothermia and hypotension seem to do the worst. Several prognostic factors may be associated with a fatal outcome (32, 34, 37, 38) and include: older age, persistent hypothermia or bradycardia, lower degree of consciousness by Glasgow Coma Scale, multiorgan impairment indicated by an APACHE II score of more than 20, or SOFA score ≥6. The most common causes of death are respiratory failure, sepsis, and gastrointestinal bleeding. Early diagnosis and prompt treatment, with meticulous attention to the details of management during the first 48 h, remain critical for the avoidance of a fatal outcome.