Sepsis and thyroid storm in a patient with methimazole-induced agranulocytosis

Abstract

Paediatric hyperthyroidism cases are mostly caused by Grave’s disease. Thyroid storm is a life-threatening condition seen rarely, in severe thyrotoxicosis, occurring in about 1%–2% of patients with hyperthyroidism. Antithyroid medications and beta-blockers are typically the first-line management of thyroid storm. We report a challenging case of a 15-year-old girl who presented with thyroid storm in the setting of septic shock and methimazole-induced agranulocytosis. Since the first-line agents were contraindicated, plasmapheresis was used to control the thyroid storm and as a bridging therapy to the definitive therapy of early thyroidectomy. This is the first paediatric case report that outlines the use of plasmapheresis in the management of complicated thyrotoxicosis in a setting of septic shock.

Background

In children, the most common cause of hyperthyroidism is Graves’ disease.¹ The incidence of Graves’ disease increases during puberty and about 80% of paediatric cases occur after 11 years of age.² Thyroid storm is a rare clinical condition of exaggerated hyperthyroidism with a mortality rate of about 20%–30%.^{3 4} Conventionally, first-line medical management for children and adolescents includes antithyroid drugs like thioamides, beta-blockers and radioactive iodine (RAI).⁵ Agranulocytosis is an extremely rare complication of thionamides class of medications, occurring in 0.2%–0.5% of the patients on methimazole therapy.^{6 7} This, in turn, predisposes patients to infection and sepsis, complicating the management of thyrotoxicosis. In our patient, thionamides and beta-blockers were contraindicated, prompting us to initiate plasmapheresis to stabilise the patient and bridge her to definitive therapy of thyroidectomy.

Case presentation

A 15-year-old girl with recently diagnosed hyperthyroidism secondary to Graves’ disease presented to the emergency department (ED) with a 5-day history of fever, dysphagia and a 2-day history of dyspnoea and profuse sweating. A month ago, she was diagnosed with Graves’ disease [high T3 high free T4 undetectably low TSH (Thyroid Stimulating Hormone) and elevated TSI (Thyroid Stimulating Immunoglobulin), positive TSH receptor antibodies] and started on methimazole and atenolol. Since the development of dysphagia 5 days ago, she had not been reported her methimazole and atenolol.

On arrival to the ED, the patient was febrile (temperature 103°F), tachycardic (heart rate 180 beats/min), tachypneic (respiratory rate 50/min), blood pressure 110/40 mm Hg and oxygen saturation 98% on room air. Physical examination was significant for erythematous swollen tonsils, thyromegaly, exophthalmos, increased work of breathing and delayed capillary refill. She had chills and rigours with clammy hands. Labs were significant for a white cell count of 4.9 x 10⁹/L, potassium of 2.3 mmol/L, TSH of 0.01 µIU/mL and an absolute neutrophil count of 20 (0.02 k/µL). Pharyngeal throat swab was positive for beta-haemolytic streptococcus A. X-ray of the lateral neck was concerning for a retropharyngeal abscess, therefore a CT of the neck was done which revealed bilateral tonsillar abscesses/phlegmon (right measuring 22×15 mm and the left one 11×5 mm), significant swelling of the epiglottis and narrowing of the airway, and thyroid goitre with enlargement of both thyroid lobes (left side 72×44 mm and right side 67×34 mm). In the ED, the patient was started on ceftriaxone for bacterial coverage and given boluses of lactated Ringer. She was also given dexamethasone for airway swelling and ibuprofen for fever. On reassessment, the patient was less tachycardic but continued to be tachypneic requiring 2 L of oxygen via nasal cannula from hypoxia. Supportive care in the form of fluid and electrolyte replacement was continued.

On arrival to the paediatric intensive care unit (PICU), she continued to demonstrate signs of worsening septic shock with blood pressures dropping to a systolic value of 80 mm Hg. She was in moderate respiratory distress with hoarse raspy breaths without stridor. Eyes were proptotic, and on an examination of the oropharynx, tonsils were touching with purulent exudate over grey-coloured tonsillar tissue. The neck appeared swollen, tender to palpation with several lymph nodes palpable with difficulty in distinguishing thyroid from other structures like lymphatic tissue due to severe tenderness. The respiratory examination was significant for transmitted rhonchi from upper airways. The abdomen was soft without any organomegaly. The neurological examination was benign. Since the sepsis was unmasked completely by the continuation of propranolol, the esmolol drip was never initiated in the PICU. Hence, aggressive fluid hydration was continued. Renal function panel drawn at the PICU was significant for hypokalemia (potassium was 2.3) and hyponatremia (sodium was 131), and liver function tests were within normal limits. Antibiotics were changed to Unasyn with a dose of vancomycin given as well.

Methimazole treatment was discontinued as she was neutropenic. Hydrocortisone and Lugol’s iodine were started on admission to the PICU to prevent the peripheral conversion of T4 to T3.⁸ She progressed to respiratory failure requiring mechanical intubation within 24 hours of admission to the PICU. The paediatric anaesthetists performed the intubation as it was a difficult airway due to narrowing caused by the thyromegaly and epiglottitis. The patient was also given GCSF to shorten the recovery time of methimazole-induced agranulocytosis.⁶ Blood cultures drawn on admission were positive for group A beta-haemolytic streptococcus. Beta-blockers could not be used in the management of her thyrotoxicosis as she was in septic shock. Due to a lack of adequate clinical improvement and inability to do conventional management of her thyrotoxicosis, plasmapheresis (therapeutic plasma exchange (TPE)) was considered to rapidly lower thyroid hormone levels. TPE was considered and discussed from a multidisciplinary standpoint, involving the paediatric nephrology and paediatric endocrinology teams. Paediatric endocrinology team had recommended obtaining a central line on admission to the PICU, with the idea of considering TPE early, given the patient’s clinical picture of septic shock, active decompensations and labs indicating thyroid storm. The goal of TPE was to stabilise the patient to undergo thyroidectomy.⁸ Five cycles of TPE were done. Each type of TPE was performed with Prismaflex using a 12F dialysis catheter. The blood flow rate was 200 mL/min. TPE exchange was performed with 3000 mL, 5% albumin and 500 mL of fresh frozen plasma. Anticoagulation with 2000 units of heparin. Three grams of calcium gluconate was given. Premeds of benadryl 25 mg intravenous and hydrocortisone were given. Three cycles were done on consecutive days after which two cycles were done with an alternate day in between. There was a clinical improvement after the TPE and repeated labs were performed 2 days after the last cycle of TPE, indicated improved thyroid status (TSH was 0.10 µIU/mL and free T4 was 0.6 ng/dL). Three days after completion of TPE successfully, our patient underwent total thyroidectomy and reimplantation of the parathyroid glands. After thyroidectomy, the patient was discharged home after her hospital stay of 2 weeks (after a 14-day course of antibiotics), on thyroxine replacement therapy.

Treatment

Thyroid storm management involves multiple aims including supportive management, new hormone synthesis inhibition, thyroid hormone release inhibition, peripheral beta-adrenergic receptor blockade, prevention of the peripheral conversion of T4 to T3 and treatment and identification of precipitating factors.⁹ Antithyroid drugs are traditionally used for inhibition of thyroid hormone production. Methimazole and propylthiouracil (PTU) are the drugs generally used from the thionamides class of medications for the management of thyrotoxicosis. Thionamides lower the production of thyroid hormone production and also have potential immunosuppressive effects.¹⁰ Soon after (1–2 hours) thionamides administration, medications like Lugol solution, potassium iodide, lithium carbonate or serum lithium can be given for inhibition of thyroid hormone release. Beta-adrenergic receptor blockade with propranolol or esmolol is an important component of the management of a thyroid storm.⁹ Beta-blockers are contraindicated in adults for other medical conditions like CCF, thyrotoxic cardiomyopathy, tachydysrhythmia and congestive heart failure, whereas, in our paediatric patient, we were not able to use it due to our patient’s septic shock. The peripheral conversion of T4 to T3 causes about 80% of T3 present in the circulation. Peripheral conversion is prevented partially by PTU and propranolol and significantly by glucocorticoids.⁹ Glucocorticoids like hydrocortisone or dexamethasone are one of the important modalities in the management of thyrotoxicosis. The use of glucocorticoids in the management of thyroid storm has shown to improve mortality rates.^4–11 Definitive therapy involves either thyroidectomy or RAI ablation therapy, depending on which one is suitable for each patient.⁹ Supportive care must be provided for fluid losses, since significant dehydration can result from the patient being febrile, having emesis, diarrhoea and diaphoresis.⁹

For patients like ours, where the first-line agents (beta-blockers and thionamides) are contraindicated or not tolerated, we need to explore other options to bridge to early thyroidectomy. Plasmapheresis was the option that was considered for our patient in order to manage her thyrotoxicosis.

Plasmapheresis is an extracorporeal blood purification technique that extracts larger molecular weight substances from the blood.¹² One of the types of plasmapheresis is TPE when the patient’s plasma is removed and replaced with another substance like colloids, crystalloids or donor plasma.¹² The recommendations for the practice of TPE are based on guidelines that were formed based on case reports, research studies and expert opinions.¹² There are different categories of TPE classified according to 2016 clinical practice guidelines, by The American Society for Apheresis. Treatment of thyroid storm falls under category III, indicating that the decision to use TPE for this condition should be individualised as the optimal role of TPE in thyroid storm is not established.¹³ These guidelines suggest that we use albumin or fresh frozen plasma as the replacement fluid when using plasmapheresis to treat thyroid storm.^{12 13} The advantage of FFP (Fresh Frozen Plasma) is that it increases the TBG concentration, which specifically binds to both thyroxine and triiodothyronine, and albumin has a higher capacity to bind to thyroid hormone.^{12 13} It is typically recommended to be done everyday to every 3 days, depending on when clinical improvement is noted. In cases of failure of first-line medical therapy for thyroid storm, or if treatment is not tolerated without adverse effects, TPE should be considered.^{12 13}

Early consultation with paediatric nephrology team and paediatric endocrinology team is important to coordinate treatment. Per standard guidelines for the use of plasmapheresis, our patient’s thyrotoxicosis was category III, which dictates the number and frequency of TPE sessions.¹³ Our patient had five cycles, the first three were done daily consecutively for the first 3 days and then had two cycles on alternate days. In between cycles, response to TPE was monitored via serum levels of TSH and free T4.

Outcome and follow-up

After five cycles of plasmapheresis, a significant decline was seen in serum levels of free T4 and free T3. Vital signs quickly stabilised and the patient successfully underwent thyroidectomy. She was discharged home on thyroxine supplementation. On a 6-month follow-up visit, she was clinically and biochemically euthyroid and reported her medications.

Discussion

The exact mechanisms that precipitate thyroid storm are incompletely understood, and the sudden release of hormones may play an important role. It has also been shown that the total T4 and T3 concentrations in patients in crisis are not significantly higher than those of severely thyrotoxic patients not in crisis.¹⁴

The diagnosis of thyroid storm separate from thyrotoxicosis is subjective.¹⁵ Diagnosis of thyroid storm is clinical, based on the history along with the presence of hyperthermia, acute mental status changes and cardiovascular and gastrointestinal dysfunction. The diagnosis criteria were standardised by Burch and Wartofsky who developed a scoring system that quantifies the likelihood of thyroid storm by assessing the degree of dysfunction in the aforementioned organ systems.¹¹ A score of 45 or more is highly suggestive of thyroid storm.

Conventional management of thyrotoxicosis or a patient presenting in thyroid storm includes thionamides and beta-blockers.⁸ In general, toxicities due to thionamides are uncommon, but sometimes patients are unable to take them due to the rare side effects of agranulocytosis or hepatotoxicity.⁸ In this patient, we were unable to use thionamides as she had agranulocytosis as a side effect of the methimazole. We were also unable to use beta-blockers in this patient as she had presented in septic shock. We used Lugol’s iodine and corticosteroids to prevent the peripheral conversion of T4 to T3. She eventually needed thyroidectomy, to get her thyrotoxicosis under control. To stabilise the patient before thyroidectomy, plasmapheresis (TPE) was considered.¹⁶ Early use of TPE was beneficial with rapid clinical improvement observed to get the patient to a euthyroid state appropriate for thyroidectomy.¹⁶

There exists no clear consensus regarding the use of plasmapheresis in thyroid storm.¹⁷ During this procedure, along with the patient’s plasma, TBG, with bound thyroid hormones, is also removed from the blood. After the colloid replacement (eg, albumin), the free thyroid hormone then has more binding sites available.^{18 19} Although albumin binds to thyroid hormone less avidly than TBG, it has a greater capacity for a low-affinity binding which may contribute to decreased free thyroid hormone levels. TPE is a safe procedure, with a recommendation of grade IIc and category III in the latest American Society for Apheresis document.²⁰ Approximately, the total incidence of side effects of TPE, which are mostly reversible, is about 5%. The common adverse effects are citrate-related nausea and vomiting, anaphylaxis due to transfusions, hypotensive or vasovagal reactions, respiratory distress and tetany or seizures.²¹ Death is unusual and mostly caused by the presence of an underlying disease.²¹

In this case, this patient’s presentation could be easily diagnosed as both septic shock as well as thyrotoxicosis. The initial presentation of fever, tachycardia and respiratory distress with a retropharyngeal abscess paint a picture or both sepsis and thyrotoxicosis. This patient had thyrotoxicosis with severely abnormal thyroid function tests on presentation with hypotension and the above-mentioned features that could be easily signs of septic shock. Evidence of positive blood cultures on admission, with the history of a retropharyngeal abscess being the initial focus of infection, confirmed that this was sepsis. The infection caused dysphagia which led to questionable medication report was thought to represent possible precipitants for thyroid storm. The patient’s symptoms of sepsis, including tachycardia and hypotension, could all be explained by thyroid storm as well. Using clinical decision rules suggested by Burch and Wartofsky, our patient would have scored 60 (for fever, agitation, tachycardia and positive pertinent history) calculated retrospectively, which indicates probable thyrotoxicosis.^6–11 All of the above suggest a vicious cycle of agranulocytosis leading to infection leading to non-compliance causing a thyroid storm, meanwhile progressing to septic shock. The challenge, in this case, was twofold; diagnosing the simultaneous presentation of thyroid storm and septic shock as well as managing this patient with both these life-threatening conditions.

To the author’s best knowledge, this report reflects the first depiction of sepsis and agranulocytosis in a paediatric patient on methimazole presenting with thyroid storm and septic shock requiring plasmapheresis used as a bridge to early thyroidectomy for management of thyrotoxicosis. Such cases serve as an important reminder to physicians that, despite its ubiquity, the management of thyrotoxicosis can be challenging.

Learning points.

• Serious side effects of thionamides, like agranulocytosis, are rare but can be fatal.

• Management of thyroid storm in the setting of septic shock is challenging.

• Conventional management includes beta-blockers and thioamides.

• Plasmapheresis as a bridge to early thyroidectomy should be considered as an adjunct therapy in severe thyroid storm where standard therapies are contraindicated or ineffective.