Fatality After Cardiac Arrest in Thyrotoxic Periodic Paralysis Due to Profound Hypokalemia Resulting from Intravenous Glucose Administration and Inadequate Potassium Replacement

Introduction

T hyrotoxic periodic paralysis (TPP) is a common complication of hyperthyroidism in Asian men. In Western countries, familial hypokalemic periodic paralysis is more common (1 –3). TPP is becoming increasingly common in the Western world (4). It primarily occurs in persons aged 20–40 years. In China, TPP accounts for 80% of cases of hypokalemia-related paralysis. The overall prevalence of TPP in thyrotoxic patients is 3% (2). Although most patients with thyrotoxicosis are women, men with TPP outnumber women by ∼20:1. The hypokalemia of TPP results from a sudden intracellular shift of potassium, not to total body potassium deficiency. It is an alarming and potentially lethal complication of hyperthyroidism. Hypokalemia can lead to near-fatal or fatal ventricular arrhythmias or even respiratory insufficiency. Immediate therapy with potassium chloride (KCl) supplementation is very important in hyperthyroid patients when TPP occurs.

Patient

A 30-year-old otherwise seemingly healthy Chinese patient received ablative ¹³¹I treatment for Graves' disease (GD) in our hospital. Four months earlier, he had presented with hyperhidrosis, tachycardia, finger tremor, anxiety, and weight loss despite having a good appetite. A diffusely enlarged goiter was noted. Cardiac and lung examination were normal. The pulse rate was 100 beats per minute (bpm). Laboratory tests showed the following serum values: total triiodothyronine (T3): 7.29 nmol/L (normal: 1.15–2.79); total thyroxine (T4): 490 nmol/L (normal: 83–167); free T3: 57.4 pmol/L (normal: 2.95–9.35); free T4: 15.8 pmol/L (normal: 7.50–19.5); and thyroid stimulating hormone: 0.1 mIU/L (normal: 0.03–5.0). The thyroid ¹³¹I uptake was 36% at 2 hours and 58% at 24 hours. He was treated with 148 MBq (4 mCi) of ¹³¹I and 10 mg of propranolol thrice daily.

Three days later, after he had eaten supper, he went out to eat again late at night and had a heavy carbohydrate intake. At 2:00 a.m. the next morning, he presented with generalized weakness, flaccid limbs, inability to walk, complete flaccid paralysis, abdominal distension, and inability to urinate. By 9:00 a.m., he was sent to a small clinic by his family members. In the rural area, where such small clinics feature only poor equipment and no laboratory tests, he was diagnosed with TPP based on his clinical presentation. Vital signs were as follows: temperature, 36.8°C; blood pressure, 105/60 mmHg; pulse rate, 92 bpm; and respiratory rate, 22 breaths per minute. Tests for serum potassium were not available. In the span of 5.5 hours, he was given 1250 mL of 5% glucose (= 62.5 g glucose) and 25 mL of 10% KCl (= 33.56 mEq) by the intravenous route. Vitamin C 1.0 mg, vitamin B6 100 mg, ATP 40 mg, and cobalamin 200 μg were also included. A bladder catheter drained 500 mL of urine.

There was no improvement after this treatment, and he was transferred by an emergency medical ambulance to a larger city hospital. On arrival at 2:25 p.m., he was conscious with a temperature of 37.5°C, a blood pressure of 120/60 mmHg, a pulse rate of 96 bpm, and a respiratory rate of 22 breaths per minute. Breath sounds were decreased, and there were no audible rales. The neurological examination showed a muscle strength of grade 1, severe facial and bulbar muscle weakness, and hypoactive reflexes in the upper and lower limbs. The sensory examination was normal. Five minutes later, he developed rolling eyes, limb convulsions, and cyanotic lips followed by a cardiac heart arrest. He died, as resuscitation efforts, including electrical cardioversion, were unsuccessful. Blood samples taken shortly before his death showed a potassium level of 1.15 mEq/L (normal range: 3.5–5.0 mEq/L); serum magnesium, sodium, chloride (NaCl), calcium, and creatinine were normal.

Discussion

It is very likely that our patient had TPP rather than another form of periodic paralysis. Lower serum potassium levels reduce the nerve–muscle excitability. Patients experience limb weakness, decreased muscle tone, reduced or absent deep tendon reflexes, decreased bowel movements, diminished bowel sounds, and abdominal distension. The symptoms may develop into severe flaccid paralysis, paralytic ileus, respiratory muscle paralysis, and even a fatal embolism. Hypokalemia causes various types of arrhythmias (such as sinus tachycardia, extrasystoles, and atrioventricular block), including serious ventricular fibrillation. It can also cause alkalosis. Respiratory muscles are rarely involved, but a complete paralysis of the respiratory, bulbar, and ocular muscles have been reported in severe attacks. There may also be prodromal symptoms of aches, cramps, and stiffness of the affected muscles. In the majority of patients, deep tendon reflexes are either markedly diminished or absent. Some patients may have either brisk or normal reflexes, even during paralysis.

The factors that may be associated with acute paralysis include thyrotoxicosis, carbohydrate load, hyperinsulinemia, glucose intake, hypokalemia, and beta-adrenergic blockage (5 –7). Attacks are commonly precipitated either by the ingestion of carbohydrate-rich meals or sweet snacks, a diet high in salts, and alcohol, or by strenuous exercise.

TPP may present either as an initial feature of GD or during a relapse of GD, or it may occur after radioactive iodine therapy. Patients with TTP do not develop hypokalemia and paralysis after they become euthyroid, which is an additional reason for controlling hyperthyroidism in these patients (8). TPP in our patient occurred 4 days after ¹³¹I had been administered, with the early onset of beta-ray effects on the tissue accompanied by a significant release of excess amounts of the thyroid hormone. Therefore, ¹³¹I may have been a factor causing the attack in this TPP patient. While total T3 and free T3 were consistently elevated, in contrast to total T4, free T4 was in the normal range, which was perhaps due to a lab error or a rare negative antibody effect. Such results may lead to an underestimation of the possible complications after ¹³¹I treatment if only one of these laboratory tests is performed. Thyroid hormones have been shown to increase Na/K-ATPase activity (9), whereas there is no association between thyroid hormone levels and the degree of hypokalemia reported in the literature. However, thyroid hormone values could not be estimated in the rural environment.

Paralysis in TPP is triggered by heavy carbohydrate intake and a high sodium load (5). There is evidence that TPP subjects may have an underlying genetic predisposition to the activation of Na/K-ATPase activity (10). Insulin resistance (8) and hyperinsulinemia (5,11) were suggested to be key factors in the pathogenesis of TPP. Insulin activates the Na⁺/K⁺-ATPase pump, thus explaining the precipitating effect of a high carbohydrate meal or a glucose load. Patients with TPP usually experience the paralytic attack either a few hours after a heavy meal or early in the morning on waking. Significant decreases of plasma potassium levels induced by high glucose disposal rates have been reported. The results of a glucose tolerance test in patients with a history of periodic paralysis, either familial or thyrotoxic, suggested that attacks can be provoked by inducing insulin-mediated potassium serum decreases (12,13). The initial serum potassium level is usually less than 3.0 mEq/L and can be dangerously low at 0.95 mEq/L (13) or 1.1 mEq/L (6). Hypokalemic alkalosis may be accompanied by low chloride. Since the isotonic saline concentration (154 mEq/L) is higher than the plasma concentration (96–104 mEq/L) and its neutral pH is lower than the plasma pH, it is controversial whether 0.9% NaCl may correct the hypochloremic alkalosis or whether the potassium normalization will auto-correct the hypochloremic alkalosis. The patient's potassium requirements were unknown, but the low potassium level persisted even after the administration of 33.56 mEq of potassium. The additional glucose load could have stimulated the release of insulin, causing the uptake of potassium from the extracellular fluid into the cells. It has been reported that intravenous (IV) administration of 1 L of 5% glucose containing 20 mEq potassium (1.5 g) could reduce the serum potassium by 0.2–1.4 mEq/L (14). The glucose loading may have enhanced the beta-cell function and increased the insulin level, which might have been accompanied by an increase in catecholamine-mediated potassium influx and a resulting decrease in its extracellular concentration; in addition, the digestion of the carbohydrate-rich diet was inhibited due to a paralytic ileus. In one textbook from China, it is suggested that severe TPP should be treated with the IV administration of 3–5 g of KCl+1000–1500 mL of a 5% glucose solution or a 0.9% NaCl solution (2). However, the administration of IV KCl in 5% glucose (50 mEq/L) was reported to be associated with decreasing strength without an increase in potassium level, whereas intravenous potassium in 5% mannitol was associated with an increase in potassium and an improvement in strength, confirming the hazard of using glucose-containing solutions (15). There is no consensus among experts regarding whether NaCl has an adverse effect on potassium correction (16,17). Lu et al. (18) recommended an immediate potassium replacement with 10 mEq/hour of KCl IV and/or 2 g of KCl orally every 2 hours, with close monitoring of the serum potassium level to avoid rebound hyperkalemia. The required KCl dose varied between 40 and 200 mEq. The recovery time was reported to be significantly shorter in patients who were administered an IV KCl infusion compared with those who were administered a saline infusion alone. It was reported that 40%–70% of patients who were administered IV KCl at a rate of 10 mEq/hour developed rebound hyperkalemia with potassium levels greater than 5.5 mEq/L (7,18), which indicates that the possibility of rebound hyperkalemia should be carefully considered. Rebound hyperkalemia was rarely observed with potassium supplementation of ≤50 mEq. Ahmed and Chilimuri (7) reported that their patient had received ∼240 mEq of potassium (oral and IV) over 8 hours; this value was most likely based on a net loss estimation of 400–800 mEq rather than on intracellular shifts. The potassium level in their patient ranged from 2.3 to 10.1 mEq/L, the latter being the fatal concentration due to an intracellular shift rather than a net deficiency.

Nonselective beta-adrenergic blockers decrease insulin secretion and are highly effective in preventing hypokalemic paralysis (8,17). Our patient received 10 mg of propranolol thrice a day after ¹³¹I treatment. Due to a lack of experience, the dose was not increased in the first hospital. In emergency therapy, the administration of 3–4 mg/kg of oral or IV propranolol has also been proposed as being an effective alternative treatment for alleviating the resulting paralysis without risking rebound hyperkalemia and increasing the serum potassium level. At high doses of 40 mg four times a day, propranolol inhibits the activity of Na/K-ATPase in carbohydrate-induced TPP (12).

Although a number of questions surrounding the management of TPP remain unanswered (17), the administration of 0.5–1.0 mEq/kg of potassium at the beginning and during attacks, preferably by the oral route, is recommended with extra caution in patients with TPP (3,17). Accordingly, the rate at which potassium is administered is dictated by the clinical scenario. Respiratory distress and hypokalemic arrhythmias warrant aggressive infusion initially in order to prevent a fatal outcome. However, periodic monitoring of serum potassium and watchful waiting for 30 minutes after each infusion of 10–20 mEq KCl is the best approach. If serum potassium monitoring is not available in rural areas, then KCl supplementation should be based on the clinical monitoring of muscle jerks and muscle strength. If electrocardiogram (ECG) monitoring is available, then one should observe classical ECG changes due to low potassium. Mannitol has been recommended rather than a glucose solution to avoid hyperinsulinemia (15); modern anesthesia practice suggests using a saline solution, which is also effective for correcting a concomitant alkalosis (19). In rural areas, safe prophylaxis should entail an oral potassium dose not exceeding 100 mEq per day in the absence of an attack (17). In case of an attack, the patient should be immediately transferred to a hospital that is equipped to perform basic laboratory tests. If the situation becomes life-threatening rebound hyperkalemia, then immediate access to hemodialysis should be available. In addition, a non-selective beta-adrenergic blocker, such as 40 mg propanolol, administered four times is justified as preventing and/or relieving paralysis (20).

Conclusion

Of the factors discussed, the combination of thyrotoxicosis, carbohydrate load, and a combination of inadequate potassium replacement compounded by the administration of glucose-containing solutions likely contributed to the fatal outcome. A saline infusion in small amounts rather than a glucose solution should be used for the IV administration of potassium. Oral potassium is considered safe and effective treatment, especially if the patient does not have immediate access to IV potassium. Propanolol can prevent paralytic attacks.