Intended for Healthcare Professionals
You are here
Home > Blog > Cardiology > Hyperthyroidism Treatment And Comorbid Outcomes

Hyperthyroidism Treatment And Comorbid Outcomes

Hyperthyroidism Treatment And Comorbid Outcomes


In this network meta-analysis, the objective was to assess the associations of different treatment modalities—anti-thyroid drugs (ATD), radioactive iodine (RAI), and thyroidectomy—with subsequent outcomes in individuals newly diagnosed with hyperthyroidism. The study conducted searches in Ovid Medline, Ovid Embase, and the Cochrane Library databases for relevant observational studies and randomized controlled trials published on or before May 1, 2022. The inclusion criteria required studies involving at least two of the treatments (ATD, RAI, and thyroidectomy) for hyperthyroidism.


The analysis encompassed 22 cohort studies involving 131,297 patients diagnosed with hyperthyroidism. The results of the Bayesian network meta-analysis revealed several key findings:


  1. Thyroidectomy vs. ATD and RAI:

   – Thyroidectomy was associated with lower risks of mortality compared to both ATD (Hazard Ratio [HR] = 0.54, 95% Credible Interval [CrI]: 0.31, 0.96) and RAI (HR = 0.62, 95% CrI: 0.41, 0.95).

   – Thyroidectomy was also linked to a lower risk of Graves’ ophthalmopathy (GO) compared to ATD (HR = 0.31, 95% CrI: 0.12, 0.64) and RAI (HR = 0.18, 95% CrI: 0.07, 0.35).


  1. RAI vs. ATD:

   – RAI treatment was associated with a higher risk of GO compared to ATD (HR = 1.70, 95% CrI: 1.02, 2.99).


In summary, the findings of this Bayesian network meta-analysis suggest that, in newly-diagnosed hyperthyroid patients, thyroidectomy is associated with lower risks of mortality and Graves’ ophthalmopathy when compared to both anti-thyroid drugs and radioactive iodine. Additionally, relative to anti-thyroid drugs, radioactive iodine therapy increased the risk of Graves’ ophthalmopathy. These results contribute valuable insights into the comparative effectiveness of different hyperthyroidism treatment modalities.


Hyperthyroidism, marked by the excessive production and release of thyroid hormone, affects 0.2% to 1.3% of individuals in iodine-sufficient regions. Graves’ disease (GD) is the primary cause, followed by toxic nodular goiter. Current therapies—anti-thyroid drugs (ATD), radioactive iodine (RAI), and thyroidectomy—prioritize symptomatic relief and achieving a euthyroid state. While surgery is seldom the initial choice, it serves as a definitive therapy in select cases. The aim of hyperthyroid treatment is to maintain normal thyroid function with minimal adverse effects.


The comparative efficacy and adverse effects of ATD, RAI, and thyroidectomy for hyperthyroid patients are contentious. For cardiovascular diseases (CVD), thyroidectomy has been associated with a lower risk of morbidity and mortality compared to ATD or RAI therapy. A dose-response link between RAI therapy and solid cancer mortality has been observed, though some studies find no significant difference in cancer risks among treatments. Importantly, RAI has been linked to Graves’ Ophthalmopathy (GO) development or exacerbation, whereas thyroidectomy may improve GO by reducing antibodies shared between the thyroid and orbit.


Recent extensive studies with large populations and extended follow-up periods prompted this Bayesian network meta-analysis. The objective is to comprehensively assess and compare the risks of CVD, cancer, mortality, and GO in hyperthyroidism patients treated with ATD, RAI, or thyroidectomy. This study aims to contribute valuable insights into the evolving landscape of hyperthyroidism management.


Our study meticulously adhered to the standards outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist (‐2385) and underwent evaluation using the Assessment of Multiple Systematic Reviews (AMSTAR) criteria for reporting systematic reviews of research (16,17). The meta-analysis was pre-registered at PROSPERO under the registration number CRD42020184492.


Data were sourced and screened independently by two reviewers, X.L and S.G., through a comprehensive search of electronic databases, including Ovid Medline, Ovid Embase, and Cochrane Library, from project initiation to May 1, 2022. Main keywords and detailed search strategies were documented in Table S1. Eligible studies were identified by reviewing full texts, eliminating duplicates, and excluding ineligible papers based on titles and abstracts. Discrepancies were resolved through consensus, consulting original research, and discussions with senior authors (C.K H.W. and B.H.H.L.).


For data collection, detailed Excel forms were developed to capture information such as the first author, publishing year, country, data source, study design, comparisons, follow-up time, recruitment period, sample size, age, gender, outcomes, and treatment modalities. Outcome data, including the total number of patients, the number of cases with outcome events, hazard ratios (HRs), and 95% confidence intervals (CIs), were collected for cardiovascular disease (CVD), cancer, mortality, and new-onset or worsening Graves’ Ophthalmopathy (GO). Specific classification criteria for CVD and cancer events were determined based on the reviewed studies. Additionally, information on the types and duration of anti-thyroid drugs (ATD), dose of radioactive iodine (RAI), and extent of thyroidectomy was collected.


The study’s network geometry was visually represented through network plots for each outcome, illustrating the treatment network across studies. Nodes represented different treatments (ATD, RAI, and surgery), while edges indicated head-to-head comparisons between treatments. Node size reflected the number of patients, and edge thickness was proportional to the number of studies.


To assess the risk of bias in individual studies, the Risk Of Bias In Non-randomized Studies-of Intervention (ROBINS-I) tool was used for non-randomized studies, and the Robvis visualization Risk-of-Bias Tool (RoB2) was employed for randomized trials ( Bias risk was summarized using traffic light and summary plots. Publication bias was estimated using funnel plots generated with the ‘netfunnel’ command in STATA.


Summary measures and statistical analyses involved estimating pooled hazard ratios (HRs) and 95% credible intervals (CrIs) to compare outcome risks between treatment groups, considering time-to-event outcomes. A random-effects pairwise meta-analysis was initially performed, followed by Bayesian network meta-analysis using R with packages ‘gemtc’ and ‘JAGS.’ The node-splitting method assessed consistency between direct and indirect evidence. Surface under the cumulative ranking curve (SUCRA) values were reported to indicate the comparative hierarchy of treatment groups’ efficacy and safety. Sensitivity and subgroup analyses were conducted to explore the impact of follow-up periods, extent of surgery, and pre-existing comorbidity on outcomes.


The validity of the methodology was confirmed, as Bayesian meta-analysis, with its enhanced ability to estimate between-study heterogeneity, was utilized. Graphical and statistical analyses were carried out using STATA version 16.0 and RStudio (Version 4.1.3), with a significance threshold set at p < 0.05 for two-tailed tests.

Eligibility Criteria

Our investigation focused on individuals newly diagnosed with hyperthyroidism. To be considered, studies had to assess and compare at least two treatment options among anti-thyroid drugs (ATD), radioactive iodine (RAI), and thyroidectomy. Patients were required to undergo a follow-up of six months or more post-treatment. This meta-analysis specifically incorporated studies involving the initial or first-line treatment for hyperthyroidism, excluding those with relapsed hyperthyroidism following prior therapy.


The study outcomes under scrutiny included the incidence of cardiovascular disease (CVD), cancer, mortality, and Graves’ Ophthalmopathy (GO) post-treatment. Eligible studies were expected to report on at least one of these predefined outcomes. The preferred study designs encompassed prospective or retrospective cohort studies and randomized controlled trials (RCTs). Acknowledging that observational studies provide insights within a real-life context, combining meta-analysis of observational data with RCTs was deemed advantageous. All included studies were required to be in English, without restrictions on the publication year. This meticulous selection criteria aimed to ensure the robustness and relevance of the data analyzed in our study.


In the study selection process, 22 studies involving 131,297 hyperthyroidism patients were included in the qualitative and quantitative synthesis out of 5,198 initially reviewed articles. The majority of patients received anti-thyroid drugs (ATD), radioactive iodine (RAI), or underwent thyroidectomy. The study design comprised five randomized controlled trials (RCTs), one prospective cohort study, and 16 retrospective cohort studies. Adjusted variables, including demographic factors and clinical parameters, were reported in 40.9% of the studies.


The primary outcomes, including cardiovascular disease (CVD), cancer, mortality, and Graves’ Ophthalmopathy (GO), were analyzed. The network meta-analysis revealed that surgery was associated with a lower risk of mortality compared to ATD and RAI. Additionally, thyroidectomy showed a lower risk of GO than ATD (HR = 0.31, 95% CrI: 0.12, 0.64) and RAI therapy (HR = 0.18, 95% CrI: 0.07, 0.35), while RAI was associated with a higher risk of GO (HR = 1.70, 95% CrI: 1.02, 2.99) compared to ATD.


SUCRA rankings indicated that ATD had the highest risk of CVD (83.47%), mortality (86.25%), and GO (51.03%), while surgery had the highest risk of cancer (78.83%). Sensitivity and subgroup analyses generally supported primary findings, with some variations in the association between surgery and the risk of CVD in studies with a 5–10 year follow-up and the lower risk of GO in patients undergoing total and subtotal thyroidectomy.


Risk of bias assessment showed that most cohort studies (96.15%) and randomized trials (80.00%) had a moderate risk of bias. Node-splitting analysis demonstrated consistency between direct and indirect comparisons for all outcomes.


In summary, the study suggests that thyroidectomy is associated with favorable outcomes, including lower mortality and GO risks, compared to ATD and RAI. The findings provide valuable insights into the comparative effectiveness of hyperthyroidism treatments.


The objective of this meta-analysis was to evaluate the association between initial treatments for hyperthyroidism—anti-thyroid drugs (ATD), radioactive iodine (RAI), and thyroidectomy—and subsequent patient outcomes, including cardiovascular disease (CVD), cancer, overall mortality, and Graves’ Ophthalmopathy (GO). The study found that thyroidectomy was associated with lower risks of GO and overall mortality compared to ATD and RAI treatments. While no significant differences in CVD and cancer risks were observed between treatment groups, the surgery group exhibited a significantly lower risk of mortality.


The analysis suggested that the lower risk of CVD in the surgery group might be attributed to the rapid improvement of cardiovascular manifestations following surgery, as observed in previous studies. Additionally, careful patient selection for thyroid surgery and the lower baseline comorbidity rate in surgical individuals may contribute to the observed lower risks of mortality and CVD in the surgery group.


Regarding cancer risk, the study’s findings aligned with a previous meta-analysis, indicating no significant difference in cancer risk between RAI and surgery. The study highlighted the importance of careful patient selection and the potential influence of decreased thyroid-stimulating hormone levels on mortality in hyperthyroidism patients.


The analysis also emphasized a higher risk of new or worsening GO in patients treated with RAI compared to ATD and surgery. This aligns with previous studies indicating RAI therapy’s association with an elevated risk of GO compared to ATD treatment. The occurrence rate of new and worsening GO was lowest in patients undergoing thyroidectomy, followed by ATD and RAI.


However, the study acknowledged certain limitations, including the observational nature of some studies, potential confounding biases, and the lack of uniformity in comorbidity reporting across studies. Despite these limitations, the findings suggested that thyroidectomy could offer favorable outcomes for newly-diagnosed hyperthyroid patients compared to ATD and RAI therapy, particularly in terms of mortality and GO risks.

Oncology Related Tools


Latest Research


About Author

Similar Articles

Leave a Reply