Graves' Disease in Children: Long-term Outcomes of Medical Therapy

Shona Rabon; Amy M. Burton; Perrin C. White


Clin Endocrinol. 2016;85(4):632-635. 

In This Article


Determining appropriate initial management for children with Graves' disease is difficult given the limited treatment options, each with potential adverse effects. Many physicians offer medical therapy as first-line treatment as there is a chance of remission with prolonged treatment although optimal treatment duration is controversial. Recent debate has centred on whether certain prognostic factors can be identified that determine long-term remission (>2 years) or whether definitive therapy may be a better alternative for some patients.[1,3,6,12,14] Remission rates are generally higher in adults (40–75%) than in children (25%), although a remission rate of 46% has been reported in Japanese children.[6] Obviously, remission rates are dependent on the time since cessation of ATD treatment required to define a remission. Features of the treatment protocol that may also influence remission rates include whether there is a mandatory attempt to wean the ATD after a fixed period of time, and whether patients are managed by titration of ATD, or alternatively with use of a high dose of ATD and treatment of the resulting hypothyroidism with levothyroxine. In the latter case, it may be difficult to know when a patient could be trialed off the ATD. Endocrinologists in our centre almost always attempt to titrate the dose of ATD rather than adding levothyroxine, and generally recommend a trial off ATD if the patient is hypothyroid or euthyroid on a minimal dose of ATD.

In the present study, we defined a provisional remission as occurring after 3 months off ATDs, so that we could assess the likelihood of a more durable remission once 3 months had passed. Only 21% of our patients achieved provisional remission despite prolonged medical therapy. Twenty-eight per cent of patients achieving a provisional remission subsequently relapsed, almost always within 16 months; conversely, remissions lasting longer than 16 months were >95% likely to last indefinitely. These data suggest that the overall rate of long-term remission in our population is approximately 15%. Relapsed hyperthyroidism has been reported in 30–68% of patients in other studies.[10,15]

Rates of remission were not affected by ethnicity or gender. Previous studies, which have not been as ethnically diverse or included as many male subjects, have concluded that male subjects and ethnic minorities are less likely to experience remission. We found no difference in remission rates in these groups. Previous studies reported a decrease in remission rates after 8–10 years of ATD treatment.[11] Others have proposed that if remission is not achieved after 18 months of continuous medical therapy, it is unlikely to occur and patients should be offered definitive therapy.[14] Our data suggest a plateau in remissions after approximately 5·5 years.

Published reports have indicated adverse reactions to ATDs occur in 5–30% of patients being treated, but many of these studies also have included PTU.[10,12,16] More recent studies suggest adverse reactions occur in 19–30% of patients (compared with 21% of our patients) taking methimazole.[6,12–14] Of course, causality is difficult to infer in uncontrolled, retrospective studies such as ours.

Currently, routine monitoring is not recommended due to rarity of serious side effects, such as hepatotoxicity and agranulocytosis. Agranulocytosis occurs in less than 0·3% of adults treated with methimazole[3] and is rare in children.[1,2,6] We did not have any cases of severe liver injury, severe neutropenia (absolute neutrophil count <500 cells/μl), or Stevens–Johnson syndrome.

It is not known whether adverse reactions of methimazole resolve with time or whether it is safe to continue treatment once a reaction occurs in a patient. In our centre most patients experiencing adverse reactions were taken off medication and underwent definitive therapy, even with minor side effects such as rash.

This study highlights the importance of reviewing all treatment options for hyperthyroidism with the patient and family while setting realistic expectations for therapy. Families should be counselled regarding side effect profiles and long-term risks. Compliance with long-term medical therapy can be difficult in adolescents, and lapses in medical therapy can decrease likelihood of remission.[4] Behavioural concerns, palpitations, weight loss, sleep disturbance, poor concentration, and school performance are known symptoms of hyperthyroidism and can significantly affect quality of life. Patients who are noncompliant with methimazole or who are lost to follow-up are at significant risk for continuation of these symptoms which would likely resolve with definitive treatment. Conversely, those undergoing ablation or surgery are likely to develop hypothyroidism and its related symptoms. Studies in adults[17] and children[18] (and unpublished observations) suggest excessive weight gain is a concern with treated hyperthyroidism. Studies ascertaining whether treatment modalities affect quality of life (including undesirable weight gain) differently would provide families with more information regarding long-term outcomes.

Given the low remission rate, we believe it is entirely reasonable to offer definitive treatment with radioactive iodine upon diagnosis (discussion of the minimum age for such treatment is beyond the scope of this study[1,3,6,12,14]). However, many families are reluctant to choose this option because of fears about radioactivity, particularly given the paucity of data regarding very long-term risk of tumorigenesis with current treatment protocols. Moreover, families of lower socioeconomic status often live in crowded conditions and have other young children in the home, making it difficult to administer radioactive iodine on an outpatient basis and requiring an expensive hospitalization. Because the reported incidence of serious adverse events with methimazole treatment is low (there were none in our case series), it remains acceptable to treat children with Grave's disease medically. If remission has not occurred in 5 years, definitive treatment may be encouraged.

There were several limitations to this retrospective study. Goitre size could not be reliably compared as each practitioner utilized different terminology when describing the size of the goitre and ultrasound results were available on very few patients. Compliance with medical therapy could not be systematically assessed although there was often documentation regarding compliance concerns from the physician or a decision for definitive therapy. Such patients were considered to have failed on medical therapy. Finally, as the data were collected from a paediatric centre, patients who changed clinical status (remission or relapse) after transitioning to adult care may have been missed.