Neonatal thyroid storm: the importance of an accurate antenatal history

Background

Neonatal thyroid storm is a rare but life-threatening emergency. Infants with neonatal thyrotoxicosis are usually born to mothers with autoimmune thyroid disease, usually Graves’ disease, where thyroid stimulating immunoglobulins (TSI) crosses the placenta.1 TSIs may be produced even after thyroidectomy or radioiodine ablation of the thyroid gland.2 In most cases, neonatal thyrotoxicosis is a transient disorder, limited by the neonate’s clearance of maternal antibodies. Thyroid storm is an acute aggravation of hyperthyroid symptoms. In the neonate, signs and symptoms of thyrotoxicosis are usually apparent by 10 days of life.1 This delay can be due to maternal antithyroid medication or coexisting blocking antibodies. Typical features of neonatal thyrotoxicosis include tachycardia, arrhythmias, cardiac failure, systemic and pulmonary hypertension. Typical hyperthyroid features such as a goitre, periorbital oedema, lid retraction and exophthalmos can be present in the neonate. Among the features of fetal thyrotoxicosis are fetal growth retardation and increased risk of intrauterine death. Mortality in neonatal cases is 12% to 20%, the majority due to cardiac failure (12% to 20%).3

In this paper, we report a case of a neonate presenting with symptoms of neonatal thyroid storm less than an hour after birth, and highlights the importance of prompt recognition and treatment.

Case presentation

A male infant was born at 35+3 weeks gestation, birth weight 2.28 kg (25^th centile), by spontaneous delivery with grade 3 meconium. Intrapartum monitoring was reassuring and there was a fetal heart rate documented at 140 beats per minute. This was after an uncomplicated pregnancy with normal antenatal scans where the fetus was regularly growing along the 30^th centile. Apgars were 9¹ and 9⁵ and while initially self-ventilating in room air the baby developed respiratory distress at 10 min with tachypnoeic and increased work of breathing requiring continuous positive airway pressure (CPAP) and admission to the neonatal unit.

On arrival to the neonatal unit, he appeared unwell with cool peripheries, heart rate up to 220 beats per minute (bpm) and required up to 80% Fio2 on CPAP. Following stabilisation and minimal handling this weaned to 21% with intermittent tachycardia with a maximum rate of 220 bpm. On consultant review by 6 hours of age baby had ongoing tachycardia up to 200 bpm and increasing oxygen requirements. He was electively intubated, received a dose of surfactant and mechanically ventilated. His mother was a 42-year-old, gravida 2 para 1, serology negative, rubella immune, blood group A positive with no history of Group B Streptococcus. She was noted to have a history of hypothyroidism and was on levothyroxine. This was her second child and there were no complications after the delivery of her first.

Investigations

Initial baby blood gas showed a mixed acidosis with pH 7.13, pO2 6.16, pCO2 8.78, HCO3- 21.3 and BE −9. The baby received 1 bolus of 10 mL/kg 0.9% saline with improvement noted on repeat gas at 1.5 hours of age. A septic work up was performed (full blood count (FBC) and blood culture) and empiric antibiotics of benzylpenicillin and gentamicin were started. The FBC showed polycythaemia haematocrit 0.72, the remaining parameters were within normal limits.

A chest X-ray showed a large opacification on the right side overlying lung fields with no apparent pneumothorax or consolidation subsequently reported as a very large thymus overlying lung. This had not been noticed antenatally (figure 1).

ECG rhythm strip showed sinus tachycardia—no evidence of supraventricular tachycardia, with P waves present throughout.

An echo showed significant tachycardia, pulmonary hypertension and poor left ventricular systolic and diastolic function.

Thyroid function tests (TFTs) were checked and results showed thyroid stimulating hormone (TSH) <0.01 (5 to 40 mIU/L) free T4 (FT4) 78.4 (10 to 36 pmol/L) and subsequently thyroid peroxidase antibody <5 (0 to 34 kunits/L) and thyroid receptor antibody (TRAB) 90.1 (0 to 1.8 IU/L).

Differential diagnosis

Sepsis was initially considered given the most likely cause of cardiorespiratory instability in a neonate hence the septic work up and antibiotic cover. Respiratory distress syndrome was felt to be unlikely as severity not typical of a late preterm baby of 35+3 weeks gestation; however surfactant was given once intubation and ventilation was required and in the setting of a higher oxygen requirement.

Given the extent of the thick meconium at delivery, meconium aspiration syndrome was another differential however there was no typical findings of this on the chest X-ray. Following the findings of the echocardiogram, thyrotoxic storm was suspected and this prompted further questioning of the mother. At booking she had reported suffering from hypothyroidism and was on levothyroxine for same. She had a degree of proptosis and a neck scar and it transpired that she had Graves’ disease since her first pregnancy and had a thyroidectomy rendering her hypothyroid.

Treatment

The echo and working diagnosis was discussed with the paediatric cardiology team. An intravenous esmolol infusion was started initially at 50 mcg/kg/min then reduced and weaned, according to heart rate and blood pressure monitoring, slowly over 20 hours and discontinued without subsequent requirement for beta blockers. An echo was repeated prior to discharge which showed complete resolution of initial findings and no further follow-up was required.

The case was discussed with the paediatric endocrinology team who advised to start Lugol’s iodine 1 drop every 8 hours and carbimazole 250 mcg/kg three times a day. He was extubated at 36 hours of age after his cardiovascular instability had resolved and was self-ventilating on room air by day of life 4. The TFTs were monitored closely and the Lugol’s iodine was stopped after 4 days of treatment. The carbimazole was weaned to two times a day then one time a day as the FT4 decreased but required to increase to two times a day predischarge from hospital. (78.4 ->75 ->43.8 ->20.6 ->29.7 pmol/L). The TSH remained undetectable throughout this period.

Outcome and follow-up

This infant was discharged home from our NICU (neonatal intensive care unit) on carbimazole once his care had fully normalised at day 9 of life with a plan for neonatal outpatient monitoring of TFTs and adjusting medication in conjunction with the endocrinology service

The full blood count was within normal limits, which is reassuring due to the risk of the potential side effect of agranulocytosis with carbimazole. It was checked on one further occasion, however, routine checking in an asymptomatic patient is not required. As there was a need for frequent dose changes transfer of outpatient care to the endocrinology team was arranged. At his 6-week check he was found to be thriving, exclusively breast feeding and had increased on his centiles, from 25^th at birth to 40^th at term and to the 70^th centile at 6 weeks corrected. He remained on carbimazole for 8 weeks and his TFTs were being regularly monitored. His mother understood that the natural history of the condition in that as her antibodies clear in time, he would become euthyroid and that he would require close monitoring of his thyroid function. When he had a final review in neonatal outpatient department at 5 months of age he was well and there were no developmental concerns. He had subsequently become hypothyroid and was on replacement therapy, which was weaning on a monthly basis with stabilisation of his TFTs.