Abstract
Purpose
Graves’ orbitopathy (GO) is a common extrathyroidal manifestation of Graves’ disease (GD) and can be triggered by radioactive iodine (RAI) treatment. However, the incidence or factors influencing GO development after RAI remain unclear. We aimed to investigate the incidence, risk factors, and clinical outcomes of GO following RAI treatment in GD patients.
Methods
A retrospective cohort study using data from Korea’s National Health Information System (2008–2020) included GD patients treated with RAI from 2008–2018, followed until 2020. GO was identified via diagnostic codes, and risk factors were assessed using Cox proportional hazards models. Clinical outcomes were compared between patients with GO who did and did not undergo RAI.
Results
Among 12,819 GD patients treated with RAI, 453 (3.5%) developed GO, compared to 4007 (0.3%) of 1,529,386 GD patients without RAI. The median time to GO onset post-RAI was 18.7 months. Younger age, shorter time from GD diagnosis to RAI, and smoking were significant risk factors, while post-RAI levothyroxine use reduced risk. Patients with GO after RAI required more intensive treatments, such as intravenous steroids and orbital radiotherapy, compared to those with GO not receiving RAI.
Conclusion
GO risk after RAI treatment was higher in younger patients, smokers, and those receiving earlier RAI. Levothyroxine treatment appeared protective. These findings emphasize the importance of patient selection, smoking cessation, timely hormone replacement, and close monitoring to mitigate GO risk and improve outcomes.
Keywords: Graves’ orbitopathy, Graves’ disease, Radioactive iodine, Risk factors
Introduction
Graves’ disease (GD), an autoimmune disorder characterized by hyperthyroidism, occurs primarily in women and is most prevalent between the ages of 20 and 50 years [1]. GD can be treated with antithyroid drugs. However, for patients who relapse after 1–2 years of antithyroid drug treatment or have severe side effects from the antithyroid drugs, removal of thyroid tissue by surgery or radioactive iodine (RAI) treatment becomes necessary [2].
RAI treatment has the advantage of being less invasive than surgery and does not require hospitalization, making it less expensive and easier treatment than surgery. After RAI treatment, hyperthyroid symptoms are relieved because the thyroid tissue is gradually destroyed, and thyroid hormone supplementation is required for severe hypothyroidism [3]. However, several studies have reported that RAI treatment is associated with the development and worsening of Graves’ ophthalmopathy (GO), an autoimmune disease that affects the eye muscles and orbital tissues [4, 5]. Risk factors that influence the development or worsening of GO after RAI treatment include smoking status, elevated TSH receptor antibody (TRAb) levels, pre-existing ophthalmopathy, and untreated hypothyroidism after RAI [6, 7].
To our knowledge, the precise incidence of new GO in patients with GD following RAI treatment remains unknown. Furthermore, the clinical outcomes of patients with GO after RAI treatment compared with those with GO unrelated to RAI treatment remain to be elucidated.
Therefore, this study investigated the incidence, risk factors, and clinical outcomes of GO following RAI treatment in a large-scale cohort with GD in the Republic of Korea.
Materials and methods
Data source and study population
This retrospective cohort study used data from the National Health Information System (NHIS) database of the Republic of Korea from 2008–2020. In the Republic of Korea, over 97% of the population is enrolled in the National Health Insurance program; hence, the NHIS offers information on the healthcare use of nearly the entire population. The NHIS data includes the insurance qualifications and premiums, medical examination results, medical history (International Classification of Diseases, 10th Revision diagnosis codes and test items, drug prescriptions, hospitalization, and outpatient clinic visits), and registration information for cancer and rare incurable diseases of the citizens. We obtained customized data on patients with GD from the NHIS database. Patients diagnosed with GD between 2008 and 2018 who were followed up until 2020 were included in this study.
This study protocol was approved by the Institutional Review Board (IRB) of Chung-Ang University Hospital (IRB number: 2205-023-19420), and the requirement for informed consent was waived because of the retrospective nature of this study. This study adhered to the principles outlined in the Declaration of Helsinki.
Operational definitions of disease, treatment, and clinical outcomes
The study population comprised patients diagnosed with GD, defined as those with at least one E05 diagnosis code. GO was defined as having an H06.2 diagnosis code more than once [8]. Diabetes (E10–14), hyperlipidemia (E78), and hypertension (I10–15), which were investigated as comorbidities, were also defined using their disease codes.
RAI treatment was defined using a treatment code (HD071, HD072, or HH140) for radioisotope application. Antithyroid drug treatment was defined by the presence of an antithyroid drug prescription code. The antithyroid drugs included methimazole, carbimazole, and propylthiouracil. The use of steroids before and after RAI treatment, as well as a levothyroxine prescription after RAI treatment, were also defined using their drug codes. Levothyroxine use was evaluated every 3 months after RAI treatment as a time-varying variable. A high-dose intravenous (IV) steroid treatment, external orbital radiation therapy, and orbital decompression surgery were investigated for their clinical outcomes after GO onset. IV steroid treatment was defined using its drug code, and radiation therapy (HD051–059, HD061, HD110, and HZ271) and orbital decompression surgery (S5182-4) were defined according to their treatment codes. The drug codes used in this study are listed in Supplementary Table S1.
Study population selection
Among the 1 594 851 patients with a diagnosis code for GD, those who received RAI treatment prior to their GD diagnosis were excluded. From the remaining patients, we excluded those who did not undergo RAI treatment, diagnosed with GO before RAI treatment, and without a prescription history of antithyroid drugs from 2008–2018. Additionally, to exclude patients with a high risk of GO or those with suspected mild GO prior to RAI treatment, we excluded individuals who were prescribed steroids from two weeks before to three months after RAI therapy. The remaining 12 819 patients were classified as patients with GD who underwent RAI treatment (Fig. 1). To select a group of patients who developed GO unrelated to RAI treatment, we excluded patients who received RAI treatment, diagnosed with GO within 3 months prior to the GD diagnosis, and received steroid or radiation treatment before the GO diagnosis. After these exclusions, the patients with a GO diagnosis code were selected.
Fig. 1.
Study flowchart. GD Graves’ disease, GO Graves’ ophthalmopathy, IV intravenous, RAI radioactive iodine
Statistical analyses
Baseline demographic and clinical characteristics are presented as means ± standard deviations for continuous variables and frequencies (percentages) for categorical variables. The t-test or Mann–Whitney U test was used to compare continuous variables, and the chi-square or Fisher’s exact test was used to compare categorical variables. Due to missing health examination data, multiple imputation by chained equations were employed to generate 10 imputed datasets. Predictive mean matching, logistic regression, and polytomous regression were performed for continuous, binary, and categorical data, respectively. Univariable and multivariable Cox regression models were used to estimate the hazard ratios (HRs) for the effects of risk factors on the occurrence of GO following RAI treatment in patients with GD. The factors analyzed included age at RAI initiation, sex, period from GD to RAI treatment (in months), smoking, alcohol consumption, steroid use, levothyroxine use, and comorbidities such as diabetes, hypertension, and hyperlipidemia. The HR estimate and its standard error were combined using Rubin’s rule [9]. To compare the clinical outcomes of patients with GO who had undergone RAI with those who had not, logistic regression analyses were performed using propensity score-matched data with a 1:3 matching ratio, adjusting for age at the time of GD diagnosis, sex, and history of autoimmune disease, diabetes mellitus, hypertension, and dyslipidemia.
Analyses were conducted using SAS software version 7.0 (SAS Institute, Cary, NC, USA) and R version 4.0.1 (R Foundation for Statistical Computing, Vienna, Austria). For all analyses, a P value < 0.05 was considered statistically significant.
Results
Rate and period of GO occurrence after RAI treatment
A total of 12 819 patients with GD received RAI treatment between 2008 and 2018. Of these, 453 (3.5%) patients developed GO (Fig. 1). Among the 1 529 386 patients with GD who did not receive RAI treatment, 4007 (0.3%) cases of GO were observed. The occurrence of GO after RAI treatment tended to decrease over time; 157 patients (36.2%) developed GO within 1 year after RAI treatment and 103 patients (22.7%) developed GO within 1–2 years. More than half of the cases occurred within 2 years and the median duration was 18.7 (interquartile range: 8.3–40.2) months.
Baseline characteristics of patients who did and did not develop GO after RAI treatment
The patients who developed GO after RAI treatment were younger than those who did not (42.5 ± 13.4 vs. 45.0 ± 13.8 years; P < 0.001) and had a shorter period from the GD diagnosis date to RAI treatment (27.9 ± 26.8 vs. 37.2 ± 34.8 months; P < 0.001) (Table 1). A higher percentage of ex-smokers and current smokers (38.1 vs. 30.1%; P = 0.001), and lower proportion of levothyroxine prescriptions (2.6 vs. 13.1%; P < 0.001) were observed in patients who developed GO after RAI treatment than those who did not. Moreover, a lower incidence of diabetes mellitus was observed in patients who developed GO after RAI treatment than those who did not (9.7 vs. 14.6%, P = 0.004).
Table 1.
Baseline characteristics of patients with GD treated with RAI
| Variable | Non-GO | GO | Total | P value | Missing (%) |
|---|---|---|---|---|---|
| (N = 12,366) | (N = 453) | (N = 12,819) | |||
| Age at RAI (years), mean ± SD | 45.0 ± 13.8 | 42.5 ± 13.4 | 44.9 ± 13.8 | <0.001 | 17 (0.1%) |
| Sex, n (%) | 0.35 | 17 (0.1%) | |||
| Male | 3882 (31.4%) | 152 (33.6%) | 4034 (31.5%) | ||
| Female | 8468 (68.6%) | 300 (66.4%) | 8768 (68.5%) | ||
| Period of GD-RAI (months), mean ± SD | 37.2 ± 34.8 | 27.9 ± 26.8 | 36.9 ± 34.6 | <0.001 | |
| Smoking, n (%) | 0.001 | 1802 (14.1%) | |||
| None | 7459 (70.2%) | 243 (61.8%) | 7702 (69.9%) | ||
| Current smoker | 1425 (13.4%) | 63 (16.0%) | 1488 (13.5%) | ||
| Ex-smoker | 1740 (16.4%) | 87 (22.1%) | 1827 (16.6%) | ||
| Alcohol consumption, n (%) | 0.382 | 1795 (14.1%) | |||
| None | 7700 (72.4%) | 286 (72.8%) | 7986 (72.4%) | ||
| Mild | 2853 (26.8%) | 102 (26.0%) | 2955 (26.8%) | ||
| Heavy | 78 (0.7%) | 5 (1.3%) | 83 (0.8%) | ||
| Antithyroid drug, n (%) | 0.248 | ||||
| carbimazole | 708 (5.7%) | 29 (6.4%) | 737 (5.7%) | ||
| propylthiouracil | 2673 (21.6%) | 111 (24.5%) | 2784 (21.7%) | ||
| methimazole | 8985 (72.7%) | 313 (69.1%) | 9298 (72.5%) | ||
| Levothyroxine use, n (%) | 1623 (13.1%) | 12 (2.6%) | 1635 (12.8%) | <0.001 | |
| Levothyroxine period (months), mean ± SD | 84.2 ± 44.0 | 26.5 ± 24.3 | 82.2 ± 44.7 | <0.001 | |
| Comorbidities | |||||
| Diabetes, n (%) | 1811 (14.6%) | 44 (9.7%) | 1855 (14.5%) | 0.004 | |
| Hypertension, n (%) | 4169 (33.7%) | 149 (32.9%) | 4318 (33.7%) | 0.754 | |
| Hyperlipidemia, n (%) | 2320 (18.8%) | 72 (15.9%) | 2392 (18.7%) | 0.14 | |
GD Graves’ disease, GO Graves’ orbitopathy, n number, RAI radioactive iodine, SD standard deviation
Analysis of risk factors for GO development after RAI treatment
A multivariate Cox proportional hazards model was used to identify the risk factors for GO development after RAI treatment (Table 2). The risk of developing GO decreased with age at RAI treatment (HR: 0.99; 95% confidence interval [CI]: 0.98–1; P = 0.023) and length of time from GD diagnosis to RAI treatment (HR: 0.99; 95% CI: 0.99–1; P < 0.001). The risk of developing GO was significantly elevated in smokers, with an HR of 1.73 (95% CI: 1.24–2.42; P = 0.001) for current smokers and an HR of 1.95 (95% CI: 1.44–2.65; P < 0.001) for ex-smokers. The GO risk was lower in patients with mild alcohol consumption (HR: 0.78; 95% CI 0.62–0.98; P = 0.032) and in those prescribed levothyroxine after RAI treatment (HR: 0.17; 95% CI: 0.09–0.3; P < 0.001).
Table 2.
Risk factors for GO occurrence following RAI treatment patients with GD
| Variable | Univariable | Multivariable | ||
|---|---|---|---|---|
| Crude HR (95% CI) | p-value | Adjusted HR (95% CI) | p-value | |
| Age at RAI | 0.99 (0.98,0.99) | <0.001 | 0.99 (0.98,1) | 0.023 |
| Sex | ||||
| Male | 1 (Reference) | 1 (Reference) | ||
| Female | 0.9 (0.74,1.1) | 0.301 | 1.3 (0.99,1.71) | 0.055 |
| Period of GD-RAI (months) | 0.99 (0.99,1) | 0.001 | 0.99 (0.99,1) | <0.001 |
| Smoking | ||||
| None | 1 (Reference) | 1 (Reference) | ||
| Current smoker | 1.35 (1.02,1.77) | 0.024 | 1.73 (1.24,2.42) | 0.001 |
| Ex-smoker | 1.55 (1.23,1.96) | <0.001 | 1.95 (1.44,2.65) | <0.001 |
| Alcohol consumption | ||||
| None | 1 (Reference) | 1 (Reference) | ||
| Mild | 0.98 (0.79,1.21) | 0.828 | 0.78 (0.62,0.98) | 0.032 |
| Heavy | 1.56 (0.64,3.81) | 0.291 | 1.12 (0.45,2.78) | 0.781 |
| Levothyroxine use | 0.17 (0.09,0.3) | <0.001 | 0.17 (0.09,0.3) | <0.001 |
| Comorbidities | ||||
| Diabetes | 0.67 (0.49,0.91) | 0.01 | 0.78 (0.56,1.08) | 0.128 |
| Hypertension | 0.98 (0.81,1.2) | 0.868 | 1.13 (0.92,1.39) | 0.24 |
| Hyperlipidemia | 0.94 (0.73,1.2) | 0.604 | 1.17 (0.89,1.52) | 0.264 |
GD Graves’ disease, GO Graves’ orbitopathy, HR hazard ratio, RAI radioactive iodine
Clinical outcome comparisons of patients with GO who had or had not undergone RAI treatment
Subsequently, we compared the clinical outcomes between patients with GO unrelated to RAI treatment (No RAI-GO) and those with GO after RAI treatment (RAI-GO). We performed 1:3 propensity score matching, matching each patient in the RAI-GO group to three No-RAI GO patients (n = 4007), based on age at GD diagnosis, sex, and comorbidities including autoimmune disease, diabetes mellitus, hypertension, and dyslipidemia. As presented in Table 3, the rate of IV steroid treatment for active GO was significantly higher for patients in the RAI-GO group (21.5 vs. 16.9%; P = 0.034), and the rate of orbital radiotherapy was also higher in the RAI-GO group (6.2 vs. 3.4%; P = 0.014), compared with the No RAI-GO group. The rate of orbital decompression surgery was higher for patients in the RAI-GO group than in the No RAI-GO group, although this difference was not statistically significant (6.0 vs. 3.9%; P = 0.086).
Table 3.
Comparison of the clinical outcomes of patients with GO who have or have not undergone RAI treatment
| Variable | No RAI-GO | RAI-GO | Total | p-value* | Multivariable adjusted OR (95% CI) |
|---|---|---|---|---|---|
| (N = 1356) | (N = 453) | (N = 1808) | |||
| IV Steroid, n (%) | 229 (16.9%) | 97 (21.5%) | 326 (18.0%) | 0.034 | 1.57 (1.11, 2.21)† |
| Orbital radiotherapy, n (%) | 46 (3.4%) | 28 (6.2%) | 74 (4.1%) | 0.014 | 2.84 (1.38, 5.83)† |
| Orbital decompression surgery, n (%) | 53 (3.9%) | 27 (6.0%) | 80 (4.4%) | 0.086 | 1.62 (0.87, 3.04) |
A multivariable logistic regression model was adjusted for age at GD, sex, autoimmune disease, diabetes, hypertension, and hyperlipidemia
GO Graves’ orbitopathy, IV intravenous, n number, RAI radioactive iodine, SD standard deviation, OR odds ratio
*Comparison of distribution differences based on the chi-squared test
†P < 0.05
Discussion
This study demonstrated that among the 12 819 patients with GD who received RAI treatment, 453 (3.5%) developed GO compared with 4007 (0.3%) GO cases among the 1 529 386 patients with GD who did not receive RAI treatment. The median time to GO onset post-RAI treatment was 18.7 months. The risk factors identified as associated with GO development were younger age, a shorter time from GD diagnosis to RAI treatment, and smoking status. The use of levothyroxine after RAI treatment was associated with a reduced GO risk. Furthermore, patients with GO following RAI treatment had more disease severity and required higher rates of IV steroid treatment and orbital radiotherapy.
In a previous randomized controlled trial, when steroids were not prescribed after RAI treatment, 15% of the patients experienced development or worsening GO; however, GO did not develop or worsen when steroids were administered concurrently with RAI treatment [4]. In our study, GO occurred in only 3.5% of patients who did not use steroids. A recently published retrospective cohort study of UK patients revealed a GO incidence following RAI treatment of approximately 5%; steroids were used in approximately 7.9% of this cohort [10]. In an Australian study with a 10-year retrospective follow-up, new GO occurred in 3% of patients after RAI treatment [11]. Therefore, the GO incidence after RAI treatment in patients with GD does not appear to be high. Alternatively, for mild GO cases, if the symptoms are not severe and transient but quickly resolved, underestimation may occur because these patients would not be diagnosed with GO.
Regarding the interval between RAI treatment and GO onset, the greatest risk of occurrence is within the 6 months following RAI treatment [10]. However, to our knowledge, no study has specifically addressed the number of patients who progress to GO during a specified period. Our study indicated that approximately one-third of patients developed GO within 1 year after RAI treatment, whereas approximately two-thirds developed GO within 2 years. The median time from RAI treatment to GO development was 18.7 (interquartile range: 8.3–40.2) months. Thus, it seems likely that regular follow-up to check for GO occurrence is necessary for at least 3 years after RAI treatment.
In our study, younger age and a shorter duration between the onset of GD and RAI therapy were associated with an increased risk of developing GO following RAI. One plausible explanation is the role of TRAb, which are known to be a key pathogenic factor in GO [12]. TRAb titers are typically highest at the time of GD diagnosis and tend to decrease gradually with antithyroid drug treatment [13, 14]. Therefore, patients who undergo RAI at an earlier stage of GD may have relatively higher TRAb levels at the time of treatment, which could contribute to the elevated risk of GO observed in this subgroup. In addition, previous studies described that TRAb levels begin to rise after RAI and reach their maximum at around three or six months [15, 16]. This transient elevation is thought to result from the release of thyroid antigens due to tissue destruction caused by RAI. In patients with already elevated baseline TRAb levels at the time of GD diagnosis, this additional post-RAI surge may further increase the risk of GO development. Therefore, patients who receive RAI soon after GD onset may be at greater risk due to both high initial TRAb titers and the subsequent post-treatment rise.
However, whether younger patients intrinsically exhibit higher TRAb titers than older individuals remains uncertain, as current evidence does not consistently support this association. Moreover, our study was limited by the lack of available TRAb titer data, preventing us from directly evaluating TRAb levels in these subgroups. Thus, while our findings suggest a potential mechanistic link between disease timing, patient age, and TRAb activity, this interpretation remains speculative and warrants further investigation in studies with comprehensive immunological profiling.
Interestingly, patients who developed GO after RAI were less likely to have received levothyroxine prescriptions, which may suggest preserved thyroid function and ongoing endogenous hormone production. This residual thyroid activity could contribute to continued autoimmune stimulation, potentially playing a more significant role in GO pathogenesis than RAI itself. Conversely, post-RAI levothyroxine treatment was associated with a significantly reduced GO risk, highlighting the potential protective effect of promptly correcting hypothyroidism after RAI. Because elevated TSH levels in the hypothyroid phase post-RAI can activate TSH receptors and induce an increased expression of thyroid/orbital antigens and worsening of autoimmune reactions [17]. Experimental studies have demonstrated that elevated TSH can stimulate TSH receptor-expressing orbital fibroblasts, promoting proinflammatory cytokine production and hyaluronic acid synthesis, which are central to the pathogenesis of GO [18, 19]. These findings support the hypothesis that prolonged hypothyroidism and elevated TSH levels post-RAI may exacerbate autoimmune orbital inflammation. In a cohort study by Stan et al., the risk of developing and worsening GO increased 3.3 times for individuals with confirmed hypothyroidism at the initial follow-up following RAI treatment [6]. Consequently, in patients at risk of developing GO, it is crucial to perform regular thyroid function tests following RAI treatment to maintain normal thyroid function.
The higher rates of IV steroid treatment and orbital radiotherapy among patients with RAI-GO suggest that GO following RAI treatment may be more severe or refractory, necessitating more aggressive therapeutic interventions. Although the rate of orbital decompression surgery was higher for patients with RAI-GO, the difference was not statistically significant, which could be due to the sample size or other confounding factors.
Our study provides valuable insights into the incidence, risk factors, and clinical outcomes of GO after RAI treatment in a large cohort of patients with GD in the Republic of Korea. However, a few limitations must be acknowledged to properly contextualize our findings. First, as a retrospective cohort study, selection bias may have occurred because the decision to treat patients with RAI and the subsequent follow-up may have been influenced by factors not captured in the database. Second, our GO definition relied solely on diagnostic codes (H06.2) and associated treatment codes, which may not have captured all GO cases. This reliance on coding could have led to an underestimation, particularly for milder cases that did not require aggressive treatment or were not accurately recorded in the database. Moreover, the use of ICD-10 code E05 to identify Graves’ disease may have introduced additional misclassification, as this code includes other thyrotoxic conditions such as toxic nodular goiter. However, given the very low prevalence of toxic nodular goiter in South Korea, which is an iodine-sufficient area, and the predominant use of radioactive iodine therapy for Graves’ disease, the likelihood of substantial inclusion of non-Graves’ patients is considered low. Third, the NHIS database lacks detailed clinical information, including the severity and activity of Graves’ orbitopathy (e.g., whether it was mild, moderate-to-severe, sight-threatening, or active), patient symptoms, and the specific clinical criteria used for GO diagnosis. Therefore, the ability to assess the full spectrum of GO manifestations and their clinical implications is limited. Fourth, the NHIS database does not include actual laboratory results, such as thyroid hormone levels or antibody titers (e.g., TRAb, anti-thyroglobulin, anti-thyroid peroxidase), which prevented analysis of thyrotoxicosis severity and the role of autoantibodies in GO development. Fifth, we did not evaluate the impact of hypercholesterolemia or statin treatment on the occurrence of GO following RAI therapy, despite recent evidence suggesting that hypercholesterolemia plays a role in GO development [20, 21]. Lastly, as our study was conducted using data exclusively from a Korean population, the generalizability of the findings may be limited. Ethnic, environmental differences could influence the risk of GO and treatment outcomes; thus, further studies in other populations are needed to validate our results. Despite these limitations, our study significantly contributes to understanding GO after RAI treatment in patients with GD. Future prospective cohort studies should focus on detailed clinical data to validate and extend our findings.
Although prior studies have suggested risk factors for GO after RAI therapy—such as smoking, early RAI intervention, and untreated hypothyroidism—many were limited by small sample sizes or single-center designs. Using a large, nationwide data set, our study supports these associations in a real-world context. However, given its observational design and reliance on diagnostic codes, the findings should be interpreted with caution. These results highlight the need for individualized risk assessment and careful monitoring, especially in patients with modifiable risk factors.
Supplementary information
Author contributions
H. Ahn contributed to the study methodology, interpretation, and writing of the original draft, as well as review and editing of the manuscript. J. Lee was involved in formal analysis and contributed to both the original draft writing and the review and editing process. M. Kim participated in the investigation and formal analysis of the study. J.K. Lee contributed to the review and editing of the manuscript.
Funding
This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (grant number: NRF-2021R1A2C1011351). The funding organization had no role in the design or conduct of this study. Open Access funding enabled and organized by Chung-Ang University Hospital.
Data availability
The Korean NHIS provides access to the data; however, confidential data are only available to researchers who meet specified conditions. The database is accessible to any researcher who has been authorized by the Korean NHIS.
Compliance with ethical standards
Ethical approval
This study protocol was approved by the Institutional Review Board (IRB) of Chung-Ang University Hospital (IRB number: 2205-023-19420).
Informed consent
The requirement for informed consent was waived because of the retrospective nature of this study.
Conflict of interest
The authors declare no competing interests.
Footnotes
These authors contributed equally: Hwa Young Ahn, Jooyoung Lee
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1007/s12020-025-04350-4.
References
- 1.T.F. Davies, S. Andersen, R. Latif, Y. Nagayama, G. Barbesino, M. Brito et al., Graves’ disease. Nat. Rev. Dis. Prim. 6, 52 (2020). 10.1038/s41572-020-0184-y. [DOI] [PubMed] [Google Scholar]
- 2.G.J. Kahaly, L. Bartalena, L. Hegedus, L. Leenhardt, K. Poppe, S.H. Pearce, 2018 European thyroid association guideline for the management of Graves’ hyperthyroidism. Eur. Thyroid. J. 7, 167–186 (2018). 10.1159/000490384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.H.B. Burch, D.S. Cooper, Management of Graves disease: a review. JAMA 314, 2544–2554 (2015). 10.1001/jama.2015.16535. [DOI] [PubMed] [Google Scholar]
- 4.L. Bartalena, C. Marcocci, F. Bogazzi, L. Manetti, M.L. Tanda, E. Dell’unto et al., Relation between therapy for hyperthyroidism and the course of Graves’ ophthalmopathy. N. Engl. J. Med. 338, 73–78 (1998). 10.1056/NEJM199801083380201. [DOI] [PubMed] [Google Scholar]
- 5.F. Träisk, L. Tallstedt, M. Abraham-Nordling, T. Andersson, G. Berg, J. Calissendorff et al., Thyroid-associated ophthalmopathy after treatment for Graves’ hyperthyroidism with antithyroid drugs or iodine-131. J. Clin. Endocrinol. Metab. 94, 3700–3707 (2009). 10.1210/jc.2009-0747. [DOI] [PubMed] [Google Scholar]
- 6.M.N. Stan, J.M. Durski, J.P. Brito, S. Bhagra, P. Thapa, R.S. Bahn, Cohort study on radioactive iodine-induced hypothyroidism: implications for Graves’ ophthalmopathy and optimal timing for thyroid hormone assessment. Thyroid 23, 620–625 (2013). 10.1089/thy.2012.0258. [DOI] [PubMed] [Google Scholar]
- 7.G. Vannucchi, I. Campi, D. Covelli, D. Dazzi, N. Currò, S. Simonetta et al., Graves’ orbitopathy activation after radioactive iodine therapy with and without steroid prophylaxis. J. Clin. Endocrinol. Metab. 94, 3381–3386 (2009). 10.1210/jc.2009-0506. [DOI] [PubMed] [Google Scholar]
- 8.S.W. Cho, J.H. Kim, H.S. Choi, H.Y. Ahn, M.K. Kim, E.J. Rhee, Big data research in the field of endocrine diseases using the Korean national health information database. Endocrinol. Metab. 38, 10–24 (2023). 10.3803/EnM.2023.102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.D.B. Rubin, N. Schenker, Multiple imputation in health-care databases: an overview and some applications. Stat. Med. 10, 585–598 (1991). 10.1002/sim.4780100410. [DOI] [PubMed] [Google Scholar]
- 10.N.Q.X. Quah, M.M. Sobti, A.M. Wren, R. Scawn, E. Kalogianni, J. Cleland et al., Radioactive iodine treatment for Graves’ hyperthyroidism: incidence of Graves orbitopathy. Nucl. Med. Commun. 45, 103–107 (2024). 10.1097/mnm.0000000000001791. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.E. Fanning, W.J. Inder, E. Mackenzie, Radioiodine treatment for graves’ disease: a 10-year Australian cohort study. BMC Endocr. Disord. 18, 94 (2018). 10.1186/s12902-018-0322-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.B. Shahida, K. Tsoumani, T. Planck, V. Modhukur, P. Asp, A. Sundlov et al., Increased risk of Graves ophthalmopathy in patients with increasing TRAb after radioiodine treatment and the impact of CTLA4 on TRAb titres. Endocrine 75, 856–864 (2022). 10.1007/s12020-021-02952-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.M. Jin, C.A. Kim, M.J. Jeon, W.B. Kim, T.Y. Kim, W.G. Kim, Dynamic risk model for the medical treatment of Graves’ hyperthyroidism according to treatment duration. Endocrinol. Metab. 39, 579–589 (2024). 10.3803/EnM.2024.1918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.S. Kalra, S. Selim, D. Shrestha, N. Somasundaram, S.A. Raza, M.P. Baruah et al., Best practices in the laboratory diagnosis, prognostication, prediction, and monitoring of Graves’ disease: role of TRAbs. BMC Endocr. Disord. 24, 274 (2024). 10.1186/s12902-024-01809-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.P. Laurberg, G. Wallin, L. Tallstedt, M. Abraham-Nordling, G. Lundell, O. Torring, TSH-receptor autoimmunity in Graves’ disease after therapy with anti-thyroid drugs, surgery, or radioiodine: a 5-year prospective randomized study. Eur. J. Endocrinol. 158, 69–75 (2008). 10.1530/EJE-07-0450. [DOI] [PubMed] [Google Scholar]
- 16.Q. Dong, X. Liu, F. Wang, Y. Xu, C. Liang, W. Du et al., Dynamic changes of TRAb and TPOAb after radioiodine therapy in Graves’ disease. Acta Endocrinol. 13, 72–76 (2017). 10.4183/aeb.2017.72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.L. Bartalena, M.L. Tanda, Current concepts regarding Graves’ orbitopathy. J. Intern. Med. 292, 692–716 (2022). 10.1111/joim.13524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.R.S. Bahn, Thyrotropin receptor expression in orbital adipose/connective tissues from patients with thyroid-associated ophthalmopathy. Thyroid 12, 193–195 (2002). 10.1089/105072502753600124. [DOI] [PubMed] [Google Scholar]
- 19.S. Kumar, R. Schiefer, M.J. Coenen, R.S. Bahn, A stimulatory thyrotropin receptor antibody (M22) and thyrotropin increase interleukin-6 expression and secretion in Graves’ orbital preadipocyte fibroblasts. Thyroid 20, 59–65 (2010). 10.1089/thy.2009.0278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.J. Cao, Y. Su, Z. Chen, C. Ma, W. Xiong, The risk factors for Graves’ ophthalmopathy. Graefes Arch. Clin. Exp. Ophthalmol. 260, 1043–1054 (2022). 10.1007/s00417-021-05456-x. [DOI] [PubMed] [Google Scholar]
- 21.E. Sabini, B. Mazzi, M.A. Profilo, T. Mautone, G. Casini, R. Rocchi et al., High serum cholesterol is a novel risk factor for Graves’ orbitopathy: results of a cross-sectional study. Thyroid 28, 386–394 (2018). 10.1089/thy.2017.0430. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The Korean NHIS provides access to the data; however, confidential data are only available to researchers who meet specified conditions. The database is accessible to any researcher who has been authorized by the Korean NHIS.