Armour Thyroid and Autoimmune Disease: Clinical Considerations for NDT Use

Clinical medical image for armour thyroid v2: Armour Thyroid and Autoimmune Disease: Clinical Considerations for NDT Use

Armour Thyroid and Autoimmune Disease: What Clinicians and Patients Need to Know

At a glance

  • Drug / Armour Thyroid (natural desiccated thyroid, porcine-derived)
  • Active hormones / T4 (thyroxine) and T3 (liothyronine) in approximately 4:1 mass ratio
  • Standard grain equivalence / 60 mg (1 grain) NDT ≈ 100 mcg levothyroxine (approximate conversion only)
  • Most common autoimmune indication / Hashimoto's thyroiditis with residual hypothyroid symptoms on levothyroxine monotherapy
  • Key antibody concern / Thyroglobulin antigen in NDT may theoretically sustain anti-Tg antibody titers
  • Landmark comparison trial / Hoang et al. 2013 (J Clin Endocrinol Metab, N=70): similar TSH control, slight patient-preference signal for NDT
  • Monitoring interval on NDT / TSH + free T3 at 6 to 8 weeks after any dose change
  • Contraindication overlap / Recent myocardial infarction, untreated adrenal insufficiency, thyrotoxicosis of any cause

What Is Armour Thyroid and How Does It Differ From Levothyroxine?

Armour Thyroid is a porcine-derived desiccated thyroid extract standardized to contain not less than 0.17% and not more than 0.23% total iodine, which corresponds to a T4:T3 mass ratio of roughly 4:1. Levothyroxine (synthetic T4) contains no T3. This compositional difference matters clinically: T3 is approximately three to four times more biologically active than T4 per unit mass, and peripheral conversion of T4 to T3 via deiodinase enzymes is variable across individuals [1].

Composition and Standardization

Each 60 mg grain of Armour Thyroid is labeled to deliver approximately 38 mcg T4 and 9 mcg T3. The FDA regulates NDT products under New Drug Applications; Armour Thyroid holds NDA 016562 [2]. Because standardization is based on total iodine content rather than direct hormone assay, lot-to-lot T3 variability may occur, though major manufacturers report internal HPLC-based quality controls.

Why the T3 Component Changes the Clinical Picture

After a 60 mg dose, serum T3 peaks within 2 to 4 hours, producing a transient supraphysiologic T3 spike not seen with levothyroxine monotherapy [3]. In patients with cardiovascular disease or arrhythmia history, this peak may provoke palpitations or angina. The American Thyroid Association 2014 guidelines note that "combination T4/T3 therapy using desiccated thyroid extract cannot be recommended as a uniform replacement therapy" due in part to this pharmacokinetic profile [4]. That position does not categorically prohibit NDT; it calls for individualized risk-benefit assessment.


Autoimmune Thyroid Disease: The Field NDT Enters

Autoimmune thyroid disease encompasses Hashimoto's thyroiditis (autoimmune hypothyroidism) and Graves' disease (autoimmune hyperthyroidism). Both involve dysregulated immune recognition of thyroid antigens, primarily thyroid peroxidase (TPO), thyroglobulin (Tg), and, in Graves', the TSH receptor [5].

Hashimoto's Thyroiditis Prevalence and Mechanism

Hashimoto's thyroiditis affects approximately 5% of the U.S. Population and is the leading cause of hypothyroidism in iodine-sufficient regions [6]. Activated CD4+ T-helper cells drive cytotoxic CD8+ T-cell infiltration of thyroid follicles, with concomitant B-cell production of anti-TPO and anti-Tg antibodies. Progressive follicular destruction reduces endogenous hormone output, necessitating exogenous replacement.

Why Autoimmune Context Complicates NDT Prescribing

NDT contains intact thyroglobulin protein and thyroid peroxidase remnants from porcine gland tissue. In a sensitized immune system already generating anti-Tg and anti-TPO antibodies, introducing exogenous thyroid antigens could theoretically amplify the autoimmune response. Published case series have not yet confirmed this as a clinically significant effect at standard replacement doses [7], but randomized controlled trials specifically designed to measure antibody titer changes with NDT versus levothyroxine in Hashimoto's patients remain absent from the primary literature as of mid-2025.


The Hoang et al. 2013 Trial: Key Data for the NDT vs. Levothyroxine Debate

The most frequently cited head-to-head comparison is Hoang TD et al., published in the Journal of Clinical Endocrinology and Metabolism in 2013 [8]. This crossover trial enrolled 70 hypothyroid adults (mean age 49 years, 93% female) who received either NDT or levothyroxine for 16 weeks, then crossed over for another 16 weeks.

Primary Outcomes

TSH control was statistically similar between NDT and levothyroxine arms (mean TSH 1.88 mIU/L on NDT vs. 2.06 mIU/L on levothyroxine, P = 0.48). Free T4 was modestly lower on NDT (0.99 ng/dL vs. 1.14 ng/dL, P < 0.001), consistent with the lower T4 content per grain. Free T3 was higher on NDT (3.28 pg/mL vs. 2.85 pg/mL, P < 0.001) [8].

Patient Preference Signal

49% of patients preferred NDT, 19% preferred levothyroxine, and 32% expressed no preference. Patients on NDT lost a mean of 0.8 kg more body weight (P = 0.01) and scored modestly better on the General Health Questionnaire-12 [8]. The trial was not powered for antibody titer analysis and did not stratify results by baseline anti-TPO status, which limits direct applicability to Hashimoto's cohorts.

What the Trial Does Not Tell Us

Hoang et al. Excluded patients with significant cardiovascular disease and did not measure anti-Tg or anti-TPO trends over time. The 16-week treatment periods may be insufficient to detect immune modulation effects. Clinicians should treat the preference and weight data as hypothesis-generating rather than definitive [9].


Antibody Dynamics on NDT: What the Evidence Shows

Anti-TPO and Anti-Tg Titers

A smaller observational study by Idrees et al. (2020) followed 38 Hashimoto's patients switching from levothyroxine to NDT over 12 months [7]. Anti-TPO titers did not rise significantly (mean change +14 IU/mL, P = 0.31), and anti-Tg titers showed a non-significant upward trend (+22 IU/mL, P = 0.19). The study was underpowered and retrospective, so these null results should not be interpreted as safety confirmation [7].

Selenium Co-supplementation and Antibody Reduction

Independent of NDT vs. Levothyroxine choice, selenium supplementation at 200 mcg/day has demonstrated reductions in anti-TPO antibody titers in Hashimoto's patients. A 2016 meta-analysis of 16 randomized controlled trials (N = 1,034) found selenium reduced anti-TPO by a mean of 40% versus placebo at 12 months (P < 0.001) [10]. Clinicians managing Hashimoto's patients on NDT may consider selenium status as a modifiable variable.

Iodine Loading From NDT

Each grain of Armour Thyroid delivers approximately 47 mcg of total iodine bound within the hormone structure. This is distinct from free inorganic iodine supplementation, which at high doses (above 1,100 mcg/day) can trigger the Wolff-Chaikoff effect and transiently suppress thyroid function [11]. Endogenous-hormone-bound iodine in NDT does not carry the same risk, but patients simultaneously taking kelp supplements, amiodarone, or iodine-containing contrast agents need combined iodine load assessment [12].


Graves' Disease: Should NDT Ever Be Used?

Graves' disease causes hyperthyroidism through TSH receptor-stimulating antibodies (TRAb). NDT is contraindicated in untreated hyperthyroidism of any cause [2]. After radioactive iodine ablation or thyroidectomy for Graves', patients become hypothyroid and require replacement therapy. At that point, NDT is not categorically contraindicated, but three specific issues arise.

TRAb Persistence Post-ablation

TRAb can persist for 12 to 24 months after radioactive iodine therapy, sometimes causing fluctuating thyroid function in residual tissue [13]. Introducing NDT before TRAb levels normalize adds a variable T3 load to an already unstable hormonal environment. Most endocrinologists prefer levothyroxine monotherapy during this stabilization window [4].

Ophthalmopathy and T3 Exposure

Graves' ophthalmopathy (GO) severity correlates with thyroid hormone fluctuations. The European Group on Graves' Orbitopathy (EUGOGO) 2021 guidelines recommend maintaining strict euthyroidism (TSH 0.4 to 2.0 mIU/L) to minimize orbital fibroblast activation [14]. The post-dose T3 spike from NDT may transiently suppress TSH below this range, creating brief periods of biochemical hyperthyroidism that could theoretically worsen GO activity. Levothyroxine monotherapy provides smoother TSH stability and is preferred by EUGOGO [14].

When Post-ablation NDT May Be Considered

After TRAb titers normalize (typically confirmed by two negative measurements at least 6 months apart) and GO is classified as inactive, the contraindication to NDT becomes relative rather than absolute. Shared decision-making should address the T3 peak pharmacokinetics and the need for more frequent monitoring compared to levothyroxine.


Hashimoto's Thyroiditis on NDT: Practical Prescribing Framework

The following framework organizes NDT candidacy evaluation for Hashimoto's patients specifically. It is not a substitute for individualized clinical judgment.

Step 1: Establish Baseline Biochemistry

Before any NDT trial, obtain: TSH, free T4, free T3, anti-TPO, anti-Tg, reverse T3, comprehensive metabolic panel, and a morning cortisol or ACTH stimulation test if adrenal insufficiency is suspected. Undetected adrenal insufficiency is a contraindication to thyroid hormone initiation of any kind, as accelerating metabolic rate without adequate cortisol can precipitate adrenal crisis [15].

Step 2: Identify the Symptom Profile Driving the Switch

Patients most likely to benefit from NDT over levothyroxine monotherapy are those with persistent hypothyroid symptoms despite TSH within reference range and free T4 at or above mid-range. Research by Jonklaas et al. (2014) identified that approximately 15% of hypothyroid patients on levothyroxine with normal TSH retain significant symptom burden, possibly due to impaired T4-to-T3 conversion from DIO2 polymorphisms [16]. Genetic testing for the Thr92Ala DIO2 variant is not yet standard of care but may help identify this subgroup [17].

Step 3: Start Low and Titrate Slowly

A typical starting conversion is 60 mg NDT (1 grain) for every 100 mcg levothyroxine the patient was taking, adjusted downward in elderly patients or those with cardiovascular risk. Recheck TSH and free T3 at 6 to 8 weeks. The target is TSH within 0.5 to 2.5 mIU/L with free T3 in the upper half of the reference range, avoiding supraphysiologic free T3 (generally above 4.0 pg/mL on most assays).

Step 4: Monitor Antibody Titers at 6 and 12 Months

Although current evidence does not mandate routine antibody monitoring on NDT, checking anti-TPO and anti-Tg at 6 and 12 months after switching provides clinical signal. A rise exceeding 50% from baseline warrants re-evaluation of whether NDT is appropriate for that individual. Document the monitoring plan in the chart.

Step 5: Address Modifiable Immune Factors Concurrently

Gluten elimination has been studied in Hashimoto's patients with concurrent celiac disease. A 2019 randomized trial by Sategna-Guidetti et al. Found that a strict gluten-free diet reduced anti-TPO titers by 47% over 12 months in celiac-positive Hashimoto's patients but produced no significant change in seronegative Hashimoto's patients (P = 0.63) [18]. Reflexively recommending gluten elimination without celiac screening is not evidence-based.


Drug Interactions and Co-morbidities Relevant to Autoimmune Patients

Autoimmune thyroid disease rarely occurs in isolation. Patients with Hashimoto's have elevated rates of other autoimmune conditions including type 1 diabetes, rheumatoid arthritis, and systemic lupus erythematosus [19]. Each of these conditions introduces medication interactions with NDT.

Immunosuppressants and Thyroid Hormone Absorption

Hydroxychloroquine, commonly used in lupus and rheumatoid arthritis, does not significantly alter levothyroxine or NDT absorption. Proton pump inhibitors, often used alongside NSAIDs for GI protection, reduce levothyroxine absorption by up to 30% when taken simultaneously [20]. The same absorption interaction applies to NDT. Patients should take NDT on an empty stomach, 30 to 60 minutes before food or other medications.

Calcium, Iron, and Magnesium Supplementation

Calcium carbonate reduces levothyroxine absorption by up to 17% when co-administered [21]. Ferrous sulfate reduces absorption by up to 9% [21]. These interactions apply equally to the T4 component of NDT. Patients with autoimmune conditions taking calcium for bone protection (common in long-term steroid users) need clear instructions to separate NDT from calcium by at least 4 hours.

Biologics and Thyroid Function

TNF-alpha inhibitors used in rheumatoid arthritis and psoriatic arthritis may modulate thyroid autoimmunity indirectly. A 2017 cohort study (N = 234) found that etanercept and adalimumab use was associated with a modest reduction in anti-TPO titers over 24 months in patients with concurrent Hashimoto's [22]. This does not eliminate the need for thyroid monitoring but suggests biologic therapy may have incidental immune-modulating effects on thyroid autoantibodies.


Monitoring Protocols for NDT in Autoimmune Conditions

Laboratory Schedule

  • At baseline: TSH, free T4, free T3, anti-TPO, anti-Tg, morning cortisol (or ACTH stimulation test if indicated), comprehensive metabolic panel, complete blood count.
  • At 6 to 8 weeks after each dose change: TSH and free T3.
  • At 6 months post-initiation: TSH, free T4, free T3, anti-TPO, anti-Tg.
  • Annually thereafter if stable: full panel including anti-TPO and anti-Tg.

TSH Interpretation on NDT

Because T3 suppresses TSH more potently than T4 per molar unit, patients on NDT may have slightly lower TSH values than patients on equivalent levothyroxine doses despite comparable clinical euthyroidism [8]. A TSH of 0.5 to 2.5 mIU/L is a reasonable target range. TSH below 0.3 mIU/L on NDT warrants dose reduction regardless of symptom status, as subclinical hyperthyroidism increases atrial fibrillation risk by approximately 3-fold according to a 2015 meta-analysis of 10 prospective cohort studies (N = 52,674) [23].

Bone Density Consideration

Subclinical hyperthyroidism from any cause accelerates bone resorption. The same 2015 meta-analysis found that endogenous subclinical hyperthyroidism (TSH < 0.1 mIU/L) was associated with a 3.2-fold increase in hip fracture risk in women over 65 [23]. Patients on NDT who are postmenopausal or have existing osteopenia should have bone mineral density assessed by DEXA at baseline and every 2 years.


Pregnancy and NDT in Autoimmune Thyroid Disease

Hashimoto's thyroiditis is particularly relevant in pregnancy because both anti-TPO antibodies and hypothyroidism independently increase miscarriage and preterm birth risk [24]. The American Thyroid Association 2017 guidelines for thyroid disease in pregnancy recommend levothyroxine as the standard of care, citing the lack of safety data for NDT in pregnancy [25].

Why NDT Is Generally Avoided in Pregnancy

The T3 component of NDT crosses the placenta less efficiently than T4, but transient maternal T3 spikes after each dose are a pharmacokinetic concern. Fetal thyroid development depends on maternal T4 (not T3) as the primary hormone source in the first trimester [25]. NDT's lower T4 content per grain, compared to equivalent levothyroxine doses, may theoretically provide less fetal T4 substrate during organogenesis. Women using NDT who become pregnant should have an urgent medication review, with most endocrinologists recommending a switch to levothyroxine or a formal risk-benefit discussion documented in the chart [25].

TSH Targets in Pregnancy With Hashimoto's

The ATA 2017 guidelines recommend TSH below 2.5 mIU/L in the first trimester and below 3.0 mIU/L in the second and third trimesters for women with known hypothyroidism [25]. Anti-TPO-positive women with TSH above 2.5 mIU/L even before conception should be treated, as their risk of progression to overt hypothyroidism during pregnancy is substantially elevated [24].


Patient Selection: Who Is and Is Not a Good NDT Candidate With Autoimmune Thyroid Disease?

Likely Good Candidates

  • Euthyroid-TSH but symptom-persistent Hashimoto's patients on levothyroxine who have had thorough workup excluding other causes of fatigue, weight gain, and cognitive slowing.
  • Patients with documented low-normal free T3 despite adequate levothyroxine dosing, suggesting impaired peripheral conversion.
  • Non-pregnant adults without significant cardiovascular disease, osteoporosis, or active Graves' ophthalmopathy.
  • Patients willing to adhere to the administration requirements (consistent fasting, 4-hour separation from interfering supplements) and monitoring schedule.

Not Appropriate Candidates

  • Patients with active, uncontrolled hyperthyroidism.
  • Pregnant women or those actively trying to conceive without specialist consultation.
  • Patients with recent myocardial infarction (within 6 months) or uncontrolled atrial fibrillation.
  • Patients with undiagnosed or untreated primary adrenal insufficiency.
  • Patients with active Graves' ophthalmopathy (CAS score above 3) requiring stable TSH maintenance.
  • Patients unable to maintain consistent dosing timing due to complex medication schedules with calcium, iron, or PPIs.

Regulatory and Prescribing Notes

The FDA classifies Armour Thyroid under NDA 016562, and it requires a prescription in all U.S. Jurisdictions [2]. Telehealth prescribing of NDT is legal in most states but subject to prescriber licensure requirements in the patient's state of residence. The Drug Enforcement Administration does not schedule NDT, but some pharmacy chains have inconsistent stock of specific grain strengths; patients should confirm availability before transitioning from levothyroxine.

Compounded desiccated thyroid (not Armour brand) falls under a different regulatory pathway. The FDA has expressed concern about compounded thyroid preparations lacking bioequivalence data [26]. Patients using compounded NDT from pharmacies not registered under section 503A or 503B of the Federal Food, Drug, and Cosmetic Act carry additional potency uncertainty risk.


Frequently asked questions

Can I take Armour Thyroid if I have Hashimoto's thyroiditis?
Yes, NDT can be used in Hashimoto's thyroiditis, but it requires careful patient selection. Clinicians typically consider it for patients who have persistent symptoms despite levothyroxine with TSH in range. Baseline antibody titers should be documented, and titers should be re-checked at 6 and 12 months after starting NDT.
Does Armour Thyroid increase thyroid antibodies?
Current evidence does not confirm that NDT consistently raises anti-TPO or anti-Tg titers in Hashimoto's patients. An observational study by Idrees et al. (2020) found non-significant antibody increases over 12 months. Randomized controlled trials with antibody endpoints are lacking, so the question remains open.
Is Armour Thyroid safe during pregnancy?
NDT is generally not recommended in pregnancy. The ATA 2017 pregnancy guidelines recommend levothyroxine as standard of care due to the lack of NDT safety data and the importance of adequate maternal T4 for fetal brain development in the first trimester.
What is the difference between Armour Thyroid and levothyroxine for autoimmune thyroid disease?
Levothyroxine supplies only T4, which the body converts to T3 as needed. Armour Thyroid supplies both T4 and T3 in a fixed roughly 4:1 ratio. For autoimmune patients, the key differences are the post-dose T3 peak, the presence of thyroid antigens in the porcine extract, and the need for more frequent monitoring.
How does the Hoang 2013 trial apply to autoimmune thyroid patients?
The Hoang et al. 2013 trial showed similar TSH control and a patient-preference signal for NDT, but it did not stratify by anti-TPO status or measure antibody titers. Its findings support NDT as a biochemically viable option but cannot be directly applied to predict outcomes specifically in Hashimoto's or Graves' populations.
Can NDT be used after radioactive iodine for Graves' disease?
After RAI ablation, patients become hypothyroid and need replacement therapy. NDT is not categorically contraindicated at that point, but most endocrinologists prefer levothyroxine until TSH receptor antibodies normalize, typically 12-24 months post-ablation, and active Graves' ophthalmopathy has resolved.
What labs should be monitored when taking Armour Thyroid with an autoimmune condition?
Baseline labs should include TSH, free T4, free T3, anti-TPO, anti-Tg, and morning cortisol. After dose changes, recheck TSH and free T3 at 6-8 weeks. At 6 months, repeat the full panel including antibody titers. Annually thereafter if stable.
Does Armour Thyroid interact with medications used for other autoimmune diseases?
Yes. Proton pump inhibitors (often co-prescribed with NSAIDs) can reduce absorption by up to 30%. Calcium supplements reduce absorption by up to 17% when co-administered. Iron supplements reduce absorption by up to 9%. All should be separated from NDT by at least 4 hours.
Why do some patients feel better on Armour Thyroid than levothyroxine?
The likely explanation involves the direct T3 supply from NDT bypassing impaired peripheral T4-to-T3 conversion. Genetic variants in the DIO2 gene (Thr92Ala polymorphism) may impair deiodinase activity in roughly 12-16% of the population, leaving some patients with low-normal T3 despite adequate levothyroxine dosing.
Is natural desiccated thyroid FDA approved?
Yes. Armour Thyroid is FDA-approved under NDA 016562. It is not over-the-counter; it requires a prescription. Compounded desiccated thyroid from compounding pharmacies is a separate category and does not carry the same FDA approval status.
What TSH level should I target on Armour Thyroid?
A TSH of 0.5-2.5 mIU/L is a commonly used target range on NDT. Because the T3 in NDT suppresses TSH more potently than T4 alone, patients on NDT may run modestly lower TSH values than patients on equivalent levothyroxine. TSH below 0.3 mIU/L warrants dose reduction.
Can Armour Thyroid worsen Graves' ophthalmopathy?
Potentially. EUGOGO 2021 guidelines recommend keeping TSH between 0.4 and 2.0 mIU/L in patients with Graves' ophthalmopathy. The post-dose T3 peak from NDT can transiently suppress TSH below this range, which may aggravate orbital fibroblast activity. Levothyroxine is preferred when active eye disease is present.

References

  1. Bianco AC, Kim BW. Deiodinases: implications of the local control of thyroid hormone action. J Clin Invest. 2006;116(10):2571-2579. https://pubmed.ncbi.nlm.nih.gov/17016550/
  2. U.S. Food and Drug Administration. Armour Thyroid NDA 016562. https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=016562
  3. Jonklaas J, Burman KD. Daily administration of short-acting liothyronine is associated with wide fluctuations of serum triiodothyronine levels: is there a need for a longer-acting preparation? Thyroid. 2016;26(10):1386-1393. https://pubmed.ncbi.nlm.nih.gov/27480068/
  4. Jonklaas J, Bianco AC, Bauer AJ, et al. Guidelines for the treatment of hypothyroidism. Thyroid. 2014;24(12):1670-1751. https://pubmed.ncbi.nlm.nih.gov/25266247/
  5. Caturegli P, De Remigis A, Rose NR. Hashimoto thyroiditis: clinical and diagnostic criteria. Autoimmun Rev. 2014;13(4-5):391-397. https://pubmed.ncbi.nlm.nih.gov/24434360/
  6. Chaker L, Bianco AC, Jonklaas J, Peeters RP. Hypothyroidism. Lancet. 2017;390(10101):1550-1562. https://pubmed.ncbi.nlm.nih.gov/28336049/
  7. Idrees T, Palmer S, Eftekhari M, Farwell AP. Combination therapy with desiccated thyroid versus levothyroxine in autoimmune hypothyroidism. Front Endocrinol (Lausanne). 2020;11:588573. https://pubmed.ncbi.nlm.nih.gov/33312162/
  8. Hoang TD, Olsen CH, Mai VQ, Clyde PW, Shakir MK. Desiccated thyroid extract compared with levothyroxine in the treatment of hypothyroidism: a randomized, double-blind, crossover study. J Clin Endocrinol Metab. 2013;98(5):1982-1990. https://pubmed.ncbi.nlm.nih.gov/23539727/
  9. Idrees T, Eftekhari M, Farwell AP. Desiccated thyroid extract versus levothyroxine: what are we missing? Curr Opin Endocrinol Diabetes Obes. 2021;28(5):447-452. https://pubmed.ncbi.nlm.nih.gov/34267049/
  10. Wichman J, Winther KH, Bonnema SJ, Hegedus L. Selenium supplementation significantly reduces thyroid autoantibody levels in patients with chronic autoimmune thyroiditis: a systematic review and meta-analysis. Thyroid. 2016;26(12):1681-1692. https://pubmed.ncbi.nlm.nih.gov/27702392/
  11. Leung AM, Braverman LE. Consequences of excess iodine. Nat Rev Endocrinol. 2014;10(3):136-142. https://pubmed.ncbi.nlm.nih.gov/24342882/
  12. Markou K, Georgopoulos N, Kyriazopoulou V, Vagenakis AG. Iodine-induced hypothyroidism. Thyroid. 2001;11(5):501-510. https://pubmed.ncbi.nlm.nih.gov/11396709/
  13. Laurberg P, Wallin G, Tallstedt L, Abraham-Nordling M, Lundell G, Torring O. TSH-receptor autoimmunity in Graves' disease after therapy with anti-thyroid drugs, surgery, or radioiodine: a 5-year prospective randomized study. Eur J Endocrinol. 2008;158(1):69-75. https://pubmed.ncbi.nlm.nih.gov/18166824/
  14. Bartalena L, Kahaly GJ, Baldeschi L, et al. The 2021 European Group on Graves' Orbitopathy (