Atrial flutter is an uncommon arrhythmia in the pediatric population. It is principally encountered in the newborn period or after atrial repair of congenital heart disease. Although hyperthyroidism is clearly known to be a cause of atrial flutter in older adults,^1,2 there are no reports of children presenting with atrial flutter as the initial manifestation of thyrotoxicosis.

When children present with atrial flutter, the diagnosis is often difficult, attributable to rapid atrioventricular conduction and superimposition of flutter waves on QRS and T waves. The presence of hyperthyroidism would be expected to further affect these relationships. We report here an interesting case of thyrotoxicosis in a 3-year-old presenting with atrial flutter in whom the response to adenosine administration and to cardioversion were unusual and ultimately helpful in suggesting the diagnosis of thyrotoxicosis.

CASE REPORT

On physical examination, she was 17.3 kg and had a regular heart rate of 275 beats/minute, respiratory rate of 66 breaths/minute, blood pressure of 111/78 mm Hg, and a axillary temperature of 37.7°C. She had moderate retractions. There was a nontender, enlarged thyroid gland. The lung fields were clear. The precordium was hyperdynamic with normal first and second heart sounds, and easily audible third and fourth heart sounds, but no murmurs. The liver was palpable 5 cm below the right costal margin, nontender, and extended slightly left of midline. The peripheral pulses were normal and equal without radial-femoral delay, and there was normal peripheral perfusion.

An electrocardiogram (ECG) showed a narrow QRS complex tachycardia with a QRS duration of 60 ms and a rate of 280 beats/minute (Fig 1). There were typical atrial flutter waves, best seen in leads II, V1, and V2, with 1:1 atrioventricular conduction. Laboratory studies included an arterial blood gas determination which showed a moderate metabolic acidosis, normal electrolytes with the exception of a serum bicarbonate of 15 mEq/L, and mild elevation in liver transaminases (AST, 179; ALT, 174). Chest radiograph demonstrated moderate cardiomegaly with a small right pleural effusion.

Adenosine at doses of 100, 200, 400, and 600 µg/kg was given by rapid peripheral intravenous infusion without a change in heart rate. The patient underwent successful direct current cardioversion using 10 Joules, to a sinus tachycardia with a rate of 180 beats/minute. (Fig 2). On echocardiogram, there was diminished ventricular function with a percent fractional shortening of 27%. There was left atrial enlargement (left atrial diameter of 29 mm), left ventricular dilatation (end diastolic dimension of 36 mm), and mild mitral regurgitation.

After successful cardioversion, the patient was started on maintenance digoxin, and a slowly decreasing sinus rate was observed. Laboratory studies at presentation confirmed the clinical suspicion of hyperthyroidism (normal values in parentheses): thyroid-stimulating hormone (high sensitivity), .002 mU/L (0.5 to 3.0); thyroxine (T₄), 12.1 (5.5 to 12); and free T₄ index, 19.5 (5.5 to 12). Repeat thyroid function tests done several days later indicated persistently severe thyrotoxicosis: free T₄, 3.1 ng/dL (0.6 to 2.0); free thyroxine index, 20.8; and triiodothyronine (T₃), 385 ng/dL (100 to 275). Antithyroglobulin antibody was not present but antithyroid microsomal antibody titer was elevated 1 to 6400 (normal, 0 to 100). Ultrasound imaging showed a homogeneously enlarged thyroid. Supersaturated potassium iodide therapy was begun. Her family declined antithyroid medication or radioactive iodine and she subsequently underwent surgical resection of her thyroid gland. All other viral and bacterial studies were negative. Pathologic evaluation was remarkable for a diffusely enlarged thyroid gland weighing 6 g with histologic evidence of chronic lymphocytic thyroiditis and diffuse lymphocytic follicles, consistent with Hashimoto's thyroiditis. Her postoperative course was complicated by transient hypocalcemia. Follow-up echocardiograms showed a left atrial diameter of 20 mm, improved myocardial function with a percent fractional shortening of 33%, and trace mitral regurgitation. Atrial flutter did not recur, and digoxin was successfully discontinued 3 months after institution.

DISCUSSION

Atrial flutter as a manifestation of hyperthyroidism, to the best of our knowledge, has not been previously reported in a child. Previous studies evaluating children with hyperthyroidism have reported that sinus tachycardia, increased ECG voltages, left atrial hypertrophy, left ventricular hypertrophy, or combined ventricular hypertrophy are the most commonly seen electrocardiographic effects.³ Very rarely, atrial fibrillation, ventricular fibrillation, or complete atrioventricular block have also been documented.⁴ In the largest published study evaluating 380 children and young adults with atrial flutter, none were attributable to hyperthyroidism.⁵ From a different perspective, review of several series of thyrotoxicosis in childhood encompassing 279 patients failed to identify any with atrial flutter.⁶

Serum thyroid hormone has direct and indirect effects on the heart. T₃ and T₄ both stimulate the Ca-ATPase of the myocardial sarcolemma resulting in decreased extracellular calcium, accelerated diastolic relaxation, and enhanced sinoatrial automaticity. T₄ influences automaticity of the sinoatrial node and atrial tissue by increasing the rate of phase four depolarization resulting in a decreased action potential duration and increased frequency of sinoatrial node discharge. The latter effects when coupled with the enhanced catecholamine response also seen in thyrotoxicosis, leads to shortening of the atrial effective refractory period as well as decreasing the atrial stimulation threshold in late diastole.^11,12 Shortening of atrial refractoriness may act to promote atrial flutter, by allowing areas of block to recover quickly enough to allow reentry.¹³

Thyroid hormone has also been shown by Goel et al⁴ in 1972 to increase atrioventricular conduction in hyperthyroid dogs. In the latter study hyperthyroid denervated dog hearts were shown to have a decreased atrioventricular node functional refractory period and conduction time allowing for faster ventricular rates independent of the sympathetic nervous system. This may help to explain why patients with thyrotoxicosis and supraventricular tachycardia tend to have faster ventricular rates despite normal to subnormal levels of serum catecholamines.¹⁴ Others have shown that the increased catecholamime sensitivity may be due to either increased adrenergic receptor density or G-protein synthesis and not attributable to increased levels of serum catecholamines.¹⁸

In normal doses, the endogenous nucleoside adenosine inhibits atrioventricular conduction and, when given to patients with atrial flutter, the brief interruption of atrioventricular conduction reveals the flutter waves. However, it failed to have any such effect in this patient despite very high doses. The lack of a response to adenosine was unusual, but explainable when one considers the known effects of thyroid hormone. Adenosine acts by competing for binding at the adenosine 1 receptor which seems to mediate the inhibition of atrioventricular node conduction. In this case however, it may be that thyroid hormone also competitively binds, induces increased inactivation, or negatively influences the interaction between adenosine and its receptor.

Finally, the significant sinus tachycardia seen well after cardioversion was also unusual for patients with isolated atrial flutter, but entirely consistent with a diagnosis of hyperthyroidism, and this provided a clue that the arrhythmia was not the primary problem. This was confirmed after the laboratory studies demonstrated a thyrotoxic profile. Although one might ascribe such tachycardia to the possible presence of tachycardia-induced cardiomyopathy, in this patient, the shortening fraction was in fact in the normal range immediately after cardioversion, and so a severe cardiomyopathy was unlikely.¹⁹ The question of whether this patient had an inherent predilection to develop atrial flutter which was brought out by hyperthyroidism, or whether the patient developed atrial flutter secondary to electrophysiologic changes precipitated by hyperthyroidism has not been resolved. If atrial flutter were to recur, an electrophysiologic study may be warranted.

In summary, when atrial flutter is encountered in a pediatric patient in whom there is 1:1 atrioventricular conduction, a lack of a response to adenosine, and persistent sinus tachycardia after cardioversion, the clinician should be alert to the possibility of thyrotoxicosis.