Relapse of Graves' disease in Chinese children: a retrospective cohort study

Introduction

Graves' disease (GD) in children has attracted worldwide attention. Antithyroid drug (ATD) has been recommended by 2022 European Thyroid Association Guideline as the first-line therapy for GD children (1). Since propylthiouracil (PTU) has a shorter half-life and may cause serious liver failure, ANCA-related (anti-neutrophil cytoplasmic antibody) vasculitis as well as other adverse events, methimazole (MMI) has been the first choice for GD children (2). Nevertheless, studies have shown that the long-term remission rate of GD children after ATD therapy is about 17–33% (3, 4), which is lower than that of adults, and they are more likely to relapse after ATD withdrawal. Relapse after ATD therapy may lead to prolonged exposure to thyrotoxicosis, resulting in adverse health outcomes.

The factors that might influence the relapse of GD children are still controversial. Rapid achievement of euthyroid status after initiation of ATD therapy, lower initial triiodothyronine, and older age were considered predictors of remission among GD children (5). Thyrotrophin receptor antibody (TRAb) level and normalization time have been reported to be associated with GD remission (3). One study indicated that a longer course of ATD therapy would lead to a higher remission rate in GD children (6). Conversely, Hamburger and colleagues found that extended treatment did not improve the remission rate (7). According to Ohye and colleagues the cumulative remission rate increased with the duration of ATD treatment, while no significant predictors of GD remission were identified (8). A systematic review showed an overall response rate of 28.8% in GD children treated with MMI. Although some small studies have shown that longer treatments can improve remission rates, the evidence is limited, and further studies are needed to investigate the efficacy of longer therapy durations (9).

There are little data on long-term follow-up of GD children, especially in China. Therefore, this study followed up on the diagnosis and treatment process of GD children in our center to explore the efficacy of ATD and related factors that may affect the prognosis, in order to provide a scientific basis for the diagnosis and treatment of GD children.

Population and method

The retrospective cohort study included 204 newly diagnosed GD children with positive TRAb by the Endocrinology Department of Children's Hospital of Nanjing Medical University from January 2013 to August 2019. This study was approved by the Ethics Committee of Children's Hospital of Nanjing Medical University (2022011003-1).

Results

The mean duration of follow-up was 5 years (range 2–8 years). At the end of observation, two cases developed central precocious puberty and were treated with GnRHα (gonadotropin-releasing hormone analogs). Both cases were in girls, whose secondary sexual characteristics appeared before the age of 8. In the luteinizing hormone-releasing hormone stimulation test, the peak value of luteinizing hormone (LH)/follicle-stimulating hormone was >0.6 and the peak value of LH was ≥5.0U/L. In another case of a GD child, which complicated with allergic purpura during treatment, no recurrence of allergic purpura was observed at the cutoff point after treatment.

At the end of follow-up, among the 204 patients included, 79 discontinued ATD medication and 40 cases relapsed after drug withdrawal (1 case underwent surgery and 39 cases continued ATD orally). One hundred and twenty-four cases were on ATD medication since they have not met the withdrawal criteria. One case received radioactive ¹³¹I therapy because of two relapses after ATD withdrawal. The average duration of ATD therapy was 4 years (ranging from 2 to 8 years). The differences between the 125 patients who had continuous therapy and 79 patients who had withdrawal therapy are presented in Supplementary Table 1 and Table 2 (see section on supplementary materials given at the end of this article).

Table 1 describes the outcomes of GD children with different therapy durations. The remission rates of groups of 2 years, 3 years, 4 years, and ≥5 years were 9.1% (3/33), 23.9% (16/67), 32.1% (9/28), and 14.5% (11/76), respectively. Forty cases relapsed after drug withdrawal. Among them, 10 cases relapsed within 3 months, 15 cases relapsed within 6 months, 11 cases relapsed within 1 year, and 4 cases relapsed within 2 years after ATD withdrawal.

Table 2 shows the general conditions between the remission and relapse groups. No significant differences were found in gender, GD family history, heart rate of preliminary diagnosis, level of goiter, tremor, proptosis, and passive smoking between the remission and relapse groups (P > 0.05), while significant differences were observed between the two groups in terms of age, Tanner stage, sleeping duration, non-iodized diet during medication and drug withdrawal period, as well as stress (P < 0.05).

We compared laboratory parameters of GD children on admission between the remission and relapse groups. Among them, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), white blood cell, neutrophil, red blood cells, hemoglobin, platelets, TSH, anti-thyroglobulin antibodies, and thyroid peroxidase antibody did not show significant differences between the two groups (P > 0.05). In the relapse group, the levels of direct bilirubin (DBIL), FT3, FT4, and the time until normalization of FT4 were higher than those in the remission group (P < 0.05). There were marginal differences in total bilirubin (TBIL) and TRAb between the relapse and remission groups (P = 0.05) (Table 3).

Table 4 shows factors related to the prognosis of GD children. After adjusting for age, gender, and GD family history, the time until normalization of FT4 (adjusted OR = 2.510, 95% CI: 1.561–4.034) was associated with increased risk of relapse of GD, while sleeping duration (adjusted OR = 0.452, 95% CI: 0.232–0.879) was associated with reduced risk of relapse of GD.

We performed subgroup analysis by stratifying cases according to pubertal development. In pubertal GD children, after adjusting for the same covariates, the time until normalization of FT4 was associated with increased relapse risk (adjusted OR = 2.293, 95% CI: 1.040–5.055), while sleeping duration (adjusted OR = 0.181, 95% CI: 0.035–0.948) was associated with decreased relapse risk. In prepubertal GD children, the time until normalization of FT4 (adjusted OR = 2.762, 95% CI: 1.404–5.432) and stress (adjusted OR = 7.510, 95% CI: 1.285–43.880) were associated with increased relapse risk of GD (Table 5).

Discussion

The treatment of GD in children aims to achieve permanent immunological remission and thus reduce the relapse rate. Although ATD is associated with adverse drug reactions (11), it is considered the recommended treatment for GD children, especially MMI. In this study, 204 newly diagnosed GD children were treated with MMI initially. Among them, four cases changed medication to PTU due to significant complications (severe granulocytopenia and systemic allergic reaction). Comparing with adults, the remission rate among GD children after ATD treatment was lower. In our study, the average duration of follow-up for all involved cases was 5 years (ranging from 2 to 8 years), and the overall remission rate was 19.1% (39/204).

Previous study showed that longer ATD treatment duration may lead to higher remission rate, and each additional 2 years of ATD treatment increased the remission rate among GD children by 25% (12). The study by Kaguelidou and colleagues also reported that relapse risk decreased with the duration of the first course of ATD treatment (13). However, in another study, no significant increase in remission rate was found in cases treated for more than 4 years compared with those treated for 2 years (7). In our study, 104 cases (37.3%) had treatment for more than 4 years. The highest remission rate of 32.1% was observed in cases with 4 years of treatment, and the remission rate no longer increased with a longer treatment duration. Therefore, for children with ATD treatment for more than 4 years without remission, it is necessary to carefully consider whether to select ¹³¹I or surgery therapy.

About 50.6% (40/79) of patients who were able to stop ATD had a relapse of hyperthyroidism, and 90% (36/40) of them occurred within 1 year of ATD withdrawal. Therefore, thyroid function should be reexamined regularly after ATD withdrawal. There has been a concern about the factors affecting the remission of GD children. A previous retrospective study revealed that the long-term remission rate was higher in GD children with smaller goiter, high BMI, and low TRAb levels (14). Shulman and colleagues observed that GD children in prepubescence were less likely to achieve remission and required a longer course of therapy (4). Another study discovered that male gender and high FT4 levels were the risk factors for GD relapse (15). In contrast, Ohye and colleagues found no predictors for remission in their research (8). In our study, fewer children in the remission group were in puberty and were diagnosed at a younger age than those in the relapse group, which is consistent with the results of a recent Portuguese study (16). Nevertheless, Jiang and colleagues and Glaser and colleagues observed a higher remission rate in older GD children (5, 17). Another study found no relationship between puberty and the remission of GD children (18). The different results might be due to differences in research demographics, relapse criteria, and length of follow-up. A study with a larger sample size could be conducted to explore the relationship between the age of onset, puberty, and prognosis of GD children.

The relationship between genetic factors and prognosis of GD is controversial. Chiang and colleagues followed up 300 GD children and discovered that children with GD family history had a lower remission rate (19). Mao and colleaguese found that children with a family history of GD were more likely to relapse after withdrawal of medication (20). In contrast, Ohye and colleagues discovered no association between GD family history and prognosis (8). In the present study, there was no significant difference in the positive rate of family history between the remission group and the relapse group. It is necessary to expand the sample size to verify the results.

Iodine, as an important component of thyroid hormone synthesis, is closely related to the occurrence and development of thyroid diseases (21). In our center, health education on a non-iodized diet had been conducted for GD children, while individuals showed variable compliance. According to the statistics of whether the GD children implemented a non-iodized diet during ATD therapy and after ATD withdrawal, we found that the proportion of children in the remission group with a non-iodized diet during ATD therapy and after ATD withdrawal was significantly higher than that in the relapse group. This confirmed the importance of avoiding iodine during ATD therapy and after withdrawal. Hence, we need to strengthen the education on avoiding an iodine-rich diet to improve the prognosis of GD children.

Lifestyle also plays an important role in GD (22). With changes in the environment and dietary structure, the incidence of GD in children increases. Currently, few research have been conducted on the effects of stress and sleep on the prognosis of GD children. Our study investigated stress and sleep duration during ATD therapy of GD children in remission and relapse groups. The results showed that children in the relapse group had significantly more stress and less sleep duration than the remission group, suggesting that environmental factors such as sleep and stress may affect the prognosis of GD children after ATD therapy. After stratified analysis of the pubertal development of GD children, we found that stress was a risk factor for prognosis in prepubertal GD children, and sleep duration was a protective factor of prognosis in pubertal GD children. This may be due to small sample size and the instability of the model. Further research is required with a larger sample size. For prepubertal GD children, parents might focus on relieving their study and life stress. Whereas for pubertal GD children, it is necessary to increase sleep time in order to improve the prognosis of GD. However, the disadvantage is that this study only involved sleep duration and did not analyze the sleep quality of the included cases. The effect of sleep quality on the prognosis of GD children needs further study.

The levels of FT3 and FT4 in the relapse group at first diagnosis were higher than those in the remission group, and the time until normalization of FT4 was longer than that in the remission group. After adjusting for age, gender, and GD family history, the time until normalization of FT4 was associated with an increased risk of relapse of GD. This is consistent with the research results of Glaser and colleagues and Jiang (5, 17). The prognosis of GD children with a lower level of thyroid hormone at an early stage and better response to ATD is relatively good. To assess the severity of GD at the onset of the disease and the response to ATD might help to evaluate the prognosis of children.

Increased thyroid volume has been found to be a prognostic risk factor in GD children (17, 23), while another study presented that the prognosis of GD children was not significantly correlated with the degree of goiter (24). Our study did not confirm differences in the distribution of the degree of goiter and proptosis by different prognoses. This might be due to the lower accuracy of the degree of goiter assessed by palpation than thyroid color Doppler ultrasound. Individual palpation skills, neck muscle, and fat mass can also affect the palpation results.

In this study, the TRAb level of newly diagnosed GD cases in the relapse group was marginally higher than that in the remission group. The study of Chiang and colleagues also found that low TRAb level at the initial diagnosed stage was a protective factor for remission among GD children (19). The study by Gastaldi and colleagues showed that children with high TRAb levels at drug withdrawal had a high relapse rate (3). In this study, all GD children were negative for the TRAb test at the time of drug withdrawal, and so the relationship between TRAb level at the time of drug withdrawal and relapse was not studied.

We compared the early liver function between the remission and relapse groups and found that the early TBIL and DBIL levels of the relapse group were higher than those of the remission group, but there was no significant difference in the levels of ALT, AST, and ALP between the two groups.

This study was one of the few cohort studies focusing on GD children in China. The prognosis of GD children in our center was described and followed up for a long time, which made our data more convincing. We discussed various factors that may affect the prognosis of GD children and provided clues for the management of GD children. However, this was a single-center study, and the sample size limited the power of the data. Our study lacked some confounding factors, for example, we did not measure urinary iodine concentration, but we recorded the diet of GD children, which might compensate for the lack of urinary iodine detection.

Conclusion

In our study, GD children with 4 years of therapy had the highest remission rate. The relapse rate of GD children after remission is high, and most of them occur within 1 year after ATD withdrawal. The response of GD children to ATD, diet, and lifestyle may affect the prognosis. Therefore, it is necessary for clinicians to improve guardians' awareness of GD relapse. Early measures to help GD children cope with adverse lifestyle might reduce the relapse.

Supplementary materials

This is linked to the online version of the paper at https://doi.org/10.1530/ETJ-23-0018.

Declaration of interest

There is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This study was funded by Nanjing Medical science and technology development key project (ZKX21043).

Author contribution statement

Dr Wang designed the study, supervised the data collection process, checked the final analysis results, and revised the manuscript. Dr Cui collected the data, carried out the initial analysis, and drafted the initial manuscript. Dr Chen and Dr Guo collected the data, completed the analysis and interpretation of the data, and participated in the drafting of initial manuscript. Dr Bigambo, Dr Yang, and Dr Gu checked the analysis results and reviewed the manuscript.