Thyroid Eye Disease: Navigating the New Treatment Landscape

Chrysoula Dosiou; Andrea Lora Kossler


J Endo Soc. 2021;5(5) 

In This Article

Therapeutic Approach

Treatment of TED has evolved over the years, from nonspecific immunosuppression to targeted biologic therapies. EUGOGO has published clinical guidelines for the evaluation and management of TED.[20] The 2016 guidelines were published prior to the release of teprotumumab and the use of other biologic agents. Recent randomized controlled trials (RCTs) offer alternatives to traditional treatment. We will review treatment options for the various categories of activity and severity, with a focus on therapies that have shown benefit in RCTs.

Mild Active Thyroid Eye Disease

Supportive treatment is the mainstay in mild disease. This involves the maintenance of euthyroidism, strict tobacco avoidance, eye lubrication with drops or ointments, and selenium. Selenium, an antioxidant, when given as 100 mcg twice a day for 6 months, significantly decreased inflammation and improved QoL in patients with mild active TED.[22] Patients with signs of mild TED, per the EUGOGO guidelines, should be referred to centers with both endocrinological and ophthalmological expertise, except for the mildest cases that improve with restoring euthyroidism and topical lubricants.[20]

Moderate-severe Active Thyroid Eye Disease

Current medical therapies target this active stage in efforts to decrease inflammation, minimize worsening of functional ocular sequelae, and in some cases improve ocular signs, including proptosis and double vision. The mainstay of treatment for years has been steroids and orbital radiation (ORT).[20] In recent years, a number of biologics have been assessed in RCTs: rituximab, tocilizumab, and teprotumumab (Table 1).

Steroids. Corticosteroids have been used to treat TED since the 1950s. In 2001 and 2005, RCTs showed the superiority of IV over oral steroids.[23,24] Kahaly et al randomly assigned patients to oral vs IV steroids (methylprednisolone 500 mg IV weekly × 6 doses followed by 250 mg IV × 6 doses) and showed a response rate of 51% vs 77% at 12 weeks' follow-up, defined by improvement in 3 parameters of a composite ophthalmic end point.[24] There was significant improvement in disease severity, activity, and QoL, minimal improvement in proptosis (median decrease of 2 vs 1 mm), and no significant improvement in diplopia with IV compared with oral steroids.[21] Bartalena et al tried 3 different doses of IV steroids and showed that the 4.5-g dose was associated with the least toxicity, but the 7.47-g dose resulted in greater improvement in a composite ophthalmic outcome, CAS, and diplopia in 12 weeks, but not proptosis or QoL.[25] Proptosis improved by only 2 mm or more in 20% to 32% of patients and diplopia improved in 21% to 46%, with the 3 different doses at 12 weeks.[25] A recent meta-analysis confirmed the efficacy of IV steroids over oral steroids (risk ratio 1.51; 95% CI, 1.25–1.83[26]) and a significantly better side effect profile,[26] with findings confirmed in a second meta-analysis.[27] Meta-analysis of the RCT subset showed that IV steroids resulted in an average decrease of CAS of 2.5, inactivation of TED in 59%, and improvement of diplopia in about a third of patients.[28] Notably, steroids were shown to have either no effect[27] or a minor effect on proptosis (1.14 mm in RCTs).[28]

There are limited data regarding the long-term efficacy of IV steroids. In the Kahaly study, no relapse rates are given, though at the 6-month follow-up there were fewer patients who had undergone orbital decompression (14% vs 32%) and strabismus surgery (20% vs 35%) in the IV steroid group compared with the oral steroid group.[24] In a post hoc analysis of the Bartalena study,[25] it was found that, among patients with improvement in the composite index at 6 weeks, 65% remained improved compared to baseline at 12 weeks (at the end of the steroid course) and 53% at 24 weeks.[29] About one-third of patients were classified as unchanged and 12% were in the "deteriorated" category, compared to baseline, at 24 weeks.[29] Of the patients who showed no change at 6 weeks, 35% had improvement whereas 13% deteriorated at 24 weeks.[29]

Side effects of IV steroids[25,27,28,30] can include liver failure and death, especially if cumulative doses exceed 6 to 8 g, with a mortality rate of 0.6%.[28] However, serious adverse events have typically occurred in studies using daily and/or alternate single doses of more than 500 mg IV methylprednisolone,[28] such as those used for treatment of CON. Liver enzyme elevation is dose-dependent, with little risk of hepatotoxicity with current TED dosing regimens and liver enzyme monitoring.[31] A recent study showed that with the most commonly used IV steroid regimen (cumulative dose of 4.5 g), almost 39% of patients experienced at least one adverse event, with 91% of adverse events graded as mild.[32] Contraindications to IV steroids include recent hepatitis, liver dysfunction (5× elevation of liver enzymes), cardiovascular morbidity, severe hypertension, inadequately managed diabetes, and severe steroid-responsive glaucoma.[28]

Orbital Radiation. ORT was described as treatment for TED in the 1970s, given as 20 Gy per orbit over 10 days.[33] ORT may induce lymphocyte apoptosis and the terminal differentiation of orbital fibroblasts, which work together to break the inflammatory cycle.[34] Correct patient selection is critical. Patients with early, active, progressing, moderate-severe disease have the highest response rates.[35] Overall, efficacy data are mixed and long-term RCT data are lacking. Initial RCTs in the 2000s established the benefit of ORT with a response rate of 50% to 60%,[36,37] as defined by improvement in a composite ophthalmic outcome at 24 weeks and 12 months, respectively, while one RCT showed no effect.[38] The main outcome improved was eye motility (odds ratio [OR] 4.88 in a recent meta-analysis[39]), with no significant effect on proptosis, CAS, or lid aperture.[36,37] A review of 5 observational studies and 9 RCTs concluded with level 1 evidence that proptosis, eyelid retraction, and soft-tissue changes do not improve with ORT.[40] Despite improving extraocular motility, ORT did not affect the need for additional treatment or rehabilitative surgery in the next 1.5 years in an RCT of patients with moderate/severe TED,[36] whereas it decreased the need for further surgery in an RCT of patients with mild TED (from 84% to 66%).[37]

Three RCTs have studied the combination of ORT with steroids. High-dose oral steroids for 5 to 6 months plus ORT were more effective than either therapy alone, as assessed by a drop in the ophthalmopathy index in 2 small RCTs (N = 24, 26).[35,41] Intravenous steroids plus ORT improved more frequently a composite ophthalmic endpoint at 1 year compared with oral steroids plus ORT (87.8 vs 63.4%),[23] resulted in a greater decrease in CAS (2.8 vs 2) and fewer surgical procedures in follow-up (7% vs 22%), and had fewer side effects than the oral steroid/ORT group (56.1% vs 85.4%).[23] Both treatments had similar effects on proptosis (mean decrease of 1.3–1.6 mm) and diplopia (improvement in about 50%).

ORT's risks include retinopathy[42] and transient exacerbation of inflammatory symptoms. There is no increased risk of cataracts when using a high-voltage linear accelerator in fractionated doses,[43] while the risk of retinopathy is none[44] or very low (1%)[43] with appropriate irradiation techniques and doses. Even though there is a theoretical concern for malignancy, no radiation-induced malignancy was seen in the 2 largest long-term follow-up studies, with up to 29 and 36 years of follow-up.[44,45] Absolute contraindications are severe hypertension and diabetic retinopathy, and ORT should be avoided in patients younger than 35.[21,46]

Intravenous Steroids Plus Mycophenolate. Mycophenolate (MMF), a prodrug of mycophenolic acid, inhibits proliferation of T and B lymphocytes, suppresses antibody production, and modulates chemotaxis of activated lymphocytes.[47] A recent RCT (N = 164) showed that the combination of IV steroids for 12 weeks with MMF 360 mg orally twice a day for 24 weeks had an equivalent outcome to IV steroid monotherapy at 12 weeks but superior outcomes (defined as improvement in composite ophthalmic index in the most affected eye) by 24 weeks, increasing to 71% from 53%,[47] with a sustained response at 36 weeks. Notably, neither group had a significant benefit in proptosis, both groups had similar improvement in QoL and similar effects on diplopia score, while addition of MMF did not change the rates of relapse nor did it affect development of CON.[47] Side effects occurred in similar percentages of patients in the 2 groups (20% vs 25%), with gastrointestinal (GI) disorders seen more commonly in the combination group, with no patient discontinuing because of toxicity.[47]

Rituximab. Rituximab is a monoclonal antibody against CD20 used in 2 RCTs in 2015 with conflicting results.[48,49] An Italian RCT showed superiority of rituximab (1000 mg IV weekly × 2) vs IV methylprednisolone in improving CAS at 24 weeks in 32 patients with active moderate/severe TED.[48] A US RCT showed no difference between rituximab vs placebo, in 25 patients with active moderate/severe TED, in CAS at 24 or 52 weeks.[49] Neither study showed a significant effect on diplopia, proptosis, or QoL.[48,49] The Italian study involved younger patients, more smokers, and a much smaller percentage of recipients of prior radioiodine therapy compared with the US study. Importantly, in the Italian study the average duration of TED was 4 to 5 months, vs 10 to 12 months in the US study, suggesting that early intervention with rituximab may be needed to halt the inflammatory response.

Long-term effects of rituximab in the 2 RCTs were assessed at a year. In the Italian study, 31.2% of patients receiving IV steroids had reactivation of TED compared to none in the rituximab group, and fewer surgical procedures were carried out after rituximab vs after IV steroids (5/15 vs 12/16).[48] In the US study, treatment failure, defined as a CAS decrease of fewer than 2 points or need for additional therapy, occurred in 50% of rituximab-treated patients, similar to placebo.[49] Side effects with rituximab affected 21 of 28 total patients, with infusion reaction occurring in 13 of 28 and less common reactions being myalgias, skin reactions, optic neuropathy, GI side effects, and transient loss of vision.[48,49]

Tocilizumab. Tocilizumab is a monoclonal antibody against the interleukin 6 receptor. Its effect was investigated in active, steroid-resistant, moderate/severe TED in a single small RCT from Spain that randomly assigned 32 patients to 4 monthly cycles of IV tocilizumab at a dose of 8 mg/kg vs placebo.[50] Tocilizumab resulted in an improvement in CAS by at least 2 points in 93% of treated patients at week 16 vs 59% in the placebo group. The effect on proptosis was minimal (1.5 mm). Importantly, these effects were not significant at the 40-week follow-up. There was no significant effect on diplopia or QoL. Longer-term follow-up has not been published. Side effects were common, with more than one adverse event occurring in 60% of treated patients vs 24% of placebo, with the most common ones being infections, headache, cytopenias, and cholesterol abnormalities.[50] Important limitations of the study include its small size, short duration of treatment, and long recruitment period, with very variable disease duration among participants and the question of whether some patients had already entered the stable phase by the time of treatment. The results need to be confirmed in a larger RCT, of longer duration, ideally comparing tocilizumab against IV steroids or teprotumumab.

Teprotumumab. Teprotumumab is a monoclonal antibody against the IGF-1R. It blocks autoantibodies from attacking orbital fibroblasts, inhibits the cytokine cascade, prevents muscle and fat tissue remodeling, and stops hyaluronan buildup in the orbit. After phase 2 and phase 3 studies[51,52] demonstrated its efficacy, it became the first and only US Food and Drug Administration (FDA)-approved drug for TED in 2020. In the phase 3 RCT of 83 treatment-naive patients with active moderate/severe TED, IV teprotumumab (10 mg/kg followed by 20 mg/kg) every 3 weeks × 8 doses resulted in an improvement in proptosis of 2 mm or more in 83% of patients compared with 10% in the placebo group.[52] The mean change in proptosis from baseline was 3.32 mm, similar to that achieved via a single-wall orbital decompression. Overall response (a reduction of ≥ 2 in CAS plus a reduction in proptosis of ≥ 2 mm) occurred in 78% of patients vs 7% with placebo. Diplopia improved by 1 or more grades in 68% vs in 29% with placebo. There was significant improvement in Graves ophthalmopathy-specific QoL at 24 weeks.[52] Preliminary long-term data from the phase 2 and 3 clinical trials, 72 weeks after starting treatment, demonstrated maintenance of proptosis response in 53% and 56% of proptosis responders and maintenance of 1 or more grades of improvement in diplopia in 69% and 58% of diplopia responders, respectively.[53,54]

Side effects were experienced by 85% of patients on teprotumumab vs 69% on placebo, mostly grade 1 or 2. The most common were muscle spasms (30%), alopecia (20%,) hyperglycemia (10%), diarrhea (10%), and hearing impairment (10%). Teprotumumab is strictly contraindicated around pregnancy because of inhibition of IGF-1 signaling. Given its long half-life (20 days), a 6-month waiting period after completion of therapy is recommended before conception. Safety and effectiveness have not been established in pediatric patients; however, owing to growth hormone pathway inhibition, it should be avoided in this patient population. Finally, teprotumumab should be used with caution in patients with inflammatory bowel disease or uncontrolled diabetes.

Limitations of the previously listed RCTs include the fact that teprotumumab was assessed only in the treatment of naive TED patients and was not compared to IV steroids, the prior standard of care. More information regarding the long-term outcomes and predictors of response is needed. Finally, whether teprotumumab results in meaningful reduction in rates of rehabilitative surgery is still unknown.

Considerations Around Choice of Therapy. The 2016 EUGOGO guidelines recommended that IV steroids be considered as first-line therapy for active moderate to severe TED. Since then, phase 2 and 3 clinical trials led to FDA approval of teprotumumab for TED and additional biologic agents have been assessed in RCTs. Comparisons of therapies are limited by study heterogeneity of inclusion criteria, outcome measures, and follow-up data. New guidelines are needed to establish a new treatment algorithm for active moderate to severe TED.

Until then, a number of factors should be considered. First, the main manifestations of TED in any given patient may guide therapeutic choice, as the various therapies have differential effects on disease parameters. Teprotumumab significantly improves proptosis, double vision, CAS, and QoL at 24 weeks' follow-up with 72-week data reporting durability in the majority of proptosis and diplopia responders. Steroids and tocilizumab are mainly effective in reducing soft-tissue inflammation and CAS score, with little effect on diplopia or proptosis, at 12- and 16-week follow-up, respectively. However, the anti-inflammatory effect of tocilizumab was not durable at 40 weeks and long-term durability of CAS improvement in RCTs is lacking for steroids. Rituximab may also decrease inflammation when used early in active disease, but has no meaningful impact on diplopia and proptosis. ORT primarily ameliorates diplopia at the 24-week follow-up, and is most effective when used early in active progressive disease. Therefore, while it is reasonable to consider all of the listed options for active TED, teprotumumab is reported to be the most effective for patients with significant proptosis and/or diplopia. When diplopia is the main manifestation, ORT may also be considered in patients with early progressive disease, with the recognition that the majority of patients will still need rehabilitative surgery. When soft-tissue inflammation is the main manifestation, teprotumumab or steroids should be considered, with ORT, tocilizumab, and rituximab also showing some efficacy in RCTs.

Second, comorbidities may affect the choice of therapy. The potential for cardiovascular disease or hepatotoxicity may limit steroid use. Susceptibility to infection would make steroids, tocilizumab, and rituximab less desirable. Preexisting uncontrolled diabetes would limit the use of steroids, teprotumumab, and ORT. A history of inflammatory bowel disease and diabetes would require close monitoring with teprotumumab. The timing of future pregnancy would be a factor to consider when prescribing teprotumumab. The timing from the onset of TED may affect the efficacy of treatment, such as with rituximab and ORT, showing improved efficacy when used early in the disease. The presence of hypertension or preexisting retinopathy would make ORT less attractive. Patient age would preclude the use of teprotumumab in those younger than 18 and ORT in those younger than 35.

Finally, cost and availability influence the choice of therapy. Cost-benefit analyses of targeted therapies are lacking; however, the high cost of biologics is an important consideration and limits access to patients without insurance. Furthermore, variability and restrictions to insurance coverage can influence their use. Additionally, access to biologic medications such as teprotumumab may be limited because of drug availability or provider expertise. Moreover, teprotumumab is currently available only in the United States; therefore, the future international role of teprotumumab remains to be seen. Taking this into context, IV steroids are the more cost-effective option for patients presenting primarily with soft-tissue inflammatory signs, without significant proptosis or diplopia, particularly for patients without insurance coverage or those living outside the United States.

In summary, treatment options continue to evolve. Of all the RCTs reviewed, teprotumumab is the first drug to demonstrate a significant improvement in all clinical parameters: proptosis, diplopia, CAS, and QoL. However, there are many considerations when selecting therapy, including the main manifestations of TED, patient comorbidities, and the cost and availability of therapies. RCTs comparing IV steroids to teprotumumab are warranted and several questions regarding teprotumumab's efficacy and duration in certain populations need answers. Until then, teprotumumab should be considered, along with IV steroids, as a first-line therapy for active moderate to severe TED in the United States. New guidelines are needed to determine a new TED treatment algorithm, incorporating targeted therapies.

Moderate-severe Inactive Thyroid Eye Disease

Surgical rehabilitation is carried out in the "inactive" phase of TED because fibrotic disease is thought to be resistant to medical therapy. There is also a risk of worsening orbital inflammation if surgery is performed in the active phase. Therefore, it is common practice to ensure stable disease is present for at least 6 months. Exceptions to this delay include surgical interventions for sight-threatening disease.

Surgical management of stable TED is customized to individual patient needs and their unique presentation. The sequence of surgery considers 4 components of TED: 1) proptosis, 2) restrictive strabismus, 3) eyelid abnormality (retraction), and 4) cosmetic concerns (fat bags, rhytids, etc). Therefore, rehabilitative surgery is staged in the following order, though not all patients require all stages.

Orbital Decompression

The principle of orbital decompression is to expand the orbital space by widening the bony orbit and/or removing excessive orbital fat to address proptosis. This also relieves symptoms of orbital congestion and mechanical pressure on the optic nerve. The medial, lateral, and inferior orbital walls as well as orbital fat are amenable to decompression. Decompression of the orbital roof is fraught with serious potential complications and is typically avoided. Surgical complications include diplopia (5%-25%, depending on the technique) and vision loss (< 0.5%),[55,56] though these risks have been minimized with advances in surgical techniques.

Strabismus Surgery

Strabismus surgery adjusts the extraocular muscles and maximizes the area of single binocular vision. Strabismus surgery is performed following orbital decompression, if needed, because decompression carries a risk of inducing diplopia. Surgical success is inversely related to the degree of fibrosis and scarring in the extraocular muscles.

Eyelid Surgery

Eyelid retraction, the most common TED manifestation, is corrected with eyelid retraction repair, with the exact approach varying depending on the severity. Nonsurgical treatment with botulinum toxin (i.e., Botox) to the levator muscle or hyaluronic acid gel filler can be effective as a temporizing measure. Aesthetic concerns can be addressed with a combination of lasers, fillers, botulinum toxin, and eyelid surgery, as the last step of surgical rehabilitation.

Sight-threatening Thyroid Eye Disease

Sight-threatening orbitopathy, though present only in 3% to 5% of patients, is the most devastating complication of TED.[30]

Exposure Keratopathy. During active disease, corneal protection can be achieved with frequent topical lubrication. A moisture chamber or goggles can be used at night-time for nocturnal lagophthalmos. In nonresponsive cases, a tarsorraphy may be necessary.

Compressive Optic Neuropathy. CON is a medical emergency. Management options for CON include corticosteroids, external beam radiation, and surgical decompression, or a combination of these interventions. The 2016 EUGOGO guidelines recommend IV steroids (500–1000 mg of methylprednisolone for 3 consecutive days) as first-line therapy.[20] This course can be repeated after a week and is effective in approximately 40% of patients.[57,58] If CON is refractory to high-dose IV glucocorticoids, or there is rapid deterioration in visual function, surgical decompression is typically recommended to mechanically relieve the optic nerve compression.[20,58] Recently, case reports demonstrated that tocilizumab[59] and teprotumumab[60,61] were effective in treating CON. Sears et al. measured objective improvements in visual acuity, relative APD, color vision, proptosis, and visual field testing as early as 4 weeks after initiating teprotumumab for steroid-resistant CON, with orbital magnetic resonance imaging showing improvement in extraocular muscle size and optic nerve compression at 8 weeks.[60] Slentz and colleagues measured an objective improvement in visual acuity, proptosis, optic nerve edema, and optical coherence tomography retinal nerve fiber thickening 2 weeks after initiating teprotumumab for steroid-naive CON.[61] While these reports suggest biologic therapies may be considered as treatment options for CON, additional studies and long-term data are needed to validate these findings.