Article Text

Download PDFPDF
Medical treatment in thyroid eye disease in 2020
  1. Jwu Jin Khong1,2,3,
  2. Alan McNab1,2
  1. 1 Department of Surgery, University of Melbourne, Centre for Eye Research Australia Ltd, East Melbourne, Victoria, Australia
  2. 2 Orbital Plastics and Lacrimal Unit, Royal Victorian Eye and Ear Hospital, East Melbourne, Victoria, Australia
  3. 3 Department of Ophthalmology, Austin Health, Heidelberg, VIC, Australia
  1. Correspondence to Dr Jwu Jin Khong, Department of Surgery, University of Melbourne, Centre for Eye Research Australia Ltd, East Melbourne, VIC 3002, Australia; jwujinkhong{at}


Thyroid eye disease (TED) affects 25% of patients with Graves’ hyperthyroidism, where 1 in 20 patients has active, moderate-to-severe disease that will require medical treatment for reducing TED activity and severity. Intravenous corticosteroid has been the mainstay of treatment for active moderate-to-severe TED. With improved understanding of the pathophysiology of TED, immunotherapy targeting different molecular pathways including T cells, B cells, cytokines and cell surface receptors have been investigated in randomised clinical trials. This review provides an overview of the current advances in medical treatment including teprotumumab, tocilizumab, rituximab and mycophenolate and the indications for their use in the management of active, moderate-to-severe TED.

  • orbit
  • treatment medical
  • drugs

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Thyroid eye disease (TED) affects 25% of patients with Graves’ hyperthyroidism.1 The orbital involvement typically occurs within 18 months of onset of autoimmune thyroid disease.2 Disease presentation is heterogenous, ranging from mild disease with soft tissue involvement and lid retraction, to moderate and severe disease with extraocular muscle enlargement and increased fat volume, resulting in exophthalmos, ocular motility restriction and in 3.5%–6% dysthyroid optic neuropathy (DON).3

Without treatment, natural history studies have shown an initial active phase with orbital inflammation escalating over months, followed by a static, inactive phase after 12–18 months, where proptosis and ocular motility restriction improve but may persist.4 In a group of untreated TED cases observed over 12 months, 22% improved substantially, 64.4% showed minor or no improvement and 13.5% deteriorated progressively.5

Intravenous corticosteroid as primary treatment for active, moderate-to-severe TED and DON has been recommended by the European Group on Graves' Orbitopathy consensus group.6 Intravenous corticosteroid is effective in rapidly reducing orbital inflammation in 50%–80% of cases.7 8 However, a proportion are non-responders, and there are significant side effects for those taking prolonged and high-dose courses, with an 11% relapse rate on cessation of treatment at 12 weeks.8 9 Therefore new treatment modalities are needed and ideally they should fulfil two objectives: (1) to be better than corticosteroid in reducing inflammation, and (2) be able to show disease-modifying effects better than natural history.

Multiple immune-modulators have been explored in clinical trials over the last 6–8 years based on increasing understanding of the molecular mechanisms underlying the pathogenesis of TED. TED is likely mediated by cross reaction of thyroid auto-antibodies towards thyrotropin receptor (TSHR) and insulin-like growth factor 1 receptor (IGF-1R) on orbital fibroblasts. The B cell-led event triggers auto-antigen recognition on orbital fibroblasts and leads to T cell activation, recruitment of T cells, macrophages and mast cells into the orbit, and increasing proinflammatory cytokine release, activation of orbital fibroblasts, which further enhances cytokine release, myofibroblast and adipocyte proliferation and secretion of hyaluronic acid.10 Strategies of targeted treatment based on this pathophysiology model would logically include inhibition of T cell activation and T cell depletion, for example, ciclosporin, otelixizumab and teplizumab against CD3; B cell depletion with rituximab; inhibition of cytokine, for example, tocilizumab; anti-TNF alpha monoclonal antibodies, for example, infliximab, etanercept and adalimumab; monoclonal antibody specific for IGF-1R, for example, teprotumumab; TSHR inhibitors like KI-70; CD40 monoclonal antibodies, for example, CFZ533; PI3K intracellular pathway inhibitor, all with the aim of reversing or inhibiting the underlying disease pathophysiology.10–15 Generalised reduction of immune cell cycling with antimetabolites such as azathioprine and mycophenolate has also been included in the treatment trials of TED with promising results.16 17

As of July 2018, 33 interventional randomised controlled studies involving 1846 patients with TED with 21 interventions were examined in a network meta-analysis, which provided an overview of the current state of TED trial literature.18 Although oral and intravenous glucocorticoids, no treatment, combination of oral glucocorticoids and orbital radiation represents the majority of interventions studied in TED trials,18 new treatment trials with immunotherapies, representing a small fraction of treatment trials in TED, were already making a significant impact on clinical management of TED especially in steroid-resistant cases. Clinical trials in TED do not uniformly stratify risk factors contributory to TED, which include history of radioactive iodine, high TSHR antibody titre, unstable thyroid function, smoking and older age.19–21 There needs to be more reasonably large, uniform randomised controlled studies, and reproducible results to examine each of the new interventions to confirm their efficacy and safety profile.

This review aims to summarise the clinical evidence for emerging immune-modulator use, including biologics, in the treatment of active TED, and to provide a recommended treatment guideline for TED based on current evidence.


Corticosteroid has been the primary treatment for active, moderate-to-severe TED since the 1950s.22 However, the mode of administration has changed, with a preference for intravenous methylprednisolone (IVMP) over oral prednisolone in the last 10–15 years, as multiple randomised control trials have consistently shown IVMP is far better in efficacy and has fewer side effects when compared with oral prednisolone.7 8 22–24 In a landmark randomised control study comparing a cumulative dose of 4.5 g IVMP to 100 mg per day tapering dose of oral prednisolone in the treatment of active, moderate-to-severe TED, primary treatment response (a composite score) was 77% in the IVMP arm, compared with 51% in the oral prednisolone arm (p<0.01) by 3 months.7 On selecting common parameters to visualise treatment effect, proptosis reduction by 2 mm was achieved in 60% of the IVMP group versus 40% of the oral group (p<0.02) at 3 months; and reduction of clinical activity score (CAS) by more than 3 points was achieved in 77% of IVMP group, as compared with 51% of oral corticosteroid group (p<0.01). There is no statistical difference in diplopia rate between the two groups.7 There is then the question of which cumulative doses of IVMP will provide the best treatment response safely. A total of 156 patients were randomised to low dose (2.25 g), intermediate dose (4.98 g) and high dose (7.47 g) IVMP over 12 weeks for the treatment of active, moderate-to-severe TED.9 Overall high-dose IVMP provided significantly greater improvement in CAS at 12 weeks and ocular motility at 6 and 12 weeks compared with low-dose IVMP, lower relapse rates at 8% in high dose versus 11% in moderate to low dose at 12 weeks, without significant advantages for disease severity by 24 weeks, at the expense of more frequent major adverse events. Therefore, the study favoured intermediate dose for most cases of active moderate-to-severe disease, with the high-dose regime reserved for the most severe cases.9

IVMP was better tolerated than oral prednisolone, but not without significant side effects. In one study, 51% adverse events were reported in the oral prednisolone group compared with 17% in the IVMP group (p<0.001).7 In another study, adverse events rate was reported at 85.4% for oral prednisolone, commonly with Cushingnoid appearance as compared with 56.1% for the IVMP group (p<0.01).8 Notably severe adverse reactions have been reported with IVMP, especially in high doses above 8 g, with fulminant liver failure, cardiovascular events, stroke and death.22 These reported major adverse events mean judicious use of high-dose IVMP is necessary with a clearly set out safety monitoring protocol including at least monthly liver enzymes, biochemistry, glucose, and blood pressure check (see table 1).

Table 1

Comparative pricing, dosing regime and adverse effect profile for immunotherapies in thyroid eye disease

Orbital corticosteroid injection for TED has been investigated as an alternative to avoid systemic corticosteroid side effects. Prospective, interventional case series on orbital corticosteroid injections using 20 mg triamcinolone and 4 mg dexamethasone showed significant reduction in CAS over 4 months, improvement in upper (100%) and lower (69%) lid retraction, resolution of strabismus in 1 in 5 patients in 17 patients with active TED resistant to or intolerant of systemic corticosteroid.25 Similarly, fornix subconjunctival injections of 20 mg triamcinolone showed significant reduction in CAS, extraocular muscle thickness and eyelid retraction over 6 months in seven patients with active TED.26 The major limitations of periorbital steroid injection studies were the small number of cases and uncontrolled nature of the studies, making recommendation of its routine use unjustified. It may be of benefit in selected cases.


Given TSHR antibodies are likely pathogenic in TED, depleting B cells and reducing antigen presentation may lead to remission. Based on this hypothesis, rituximab, a chimeric human and mouse monoclonal antibody against CD20 antigen on B cells has been trialled in the treatment of active TED.27 The earliest cases of TED treated by rituximab were reported in 2006 by both Salvi et al and El Fassi et al.27 28 The first case by Salvi et al had moderately severe, active TED that failed to respond to weekly IVMP. The index case responded to two 1 g rituximab infusions with improved CAS from 5 to 2 in 3 months, with minimal changes in proptosis and ocular motility over 8 months.27 El Fassi et al reported two active and severe TED cases refractory to steroid, orbital radiation and orbital decompression. Case 1 was reported to show dramatic response to rituximab with reduction of CAS from 6 to 2 within 8 months and proptosis reduction by 3 mm, and the case 2 showed reduction of CAS from 5 to 1; both patients showed improvement in soft tissue signs and ocular motility.28 These initial encouraging results led to further study on rituximab by Salvi et al in nine patients with mild (2) and moderate-to-severe (7) active TED.29 Mean CAS decrease from 4.7 to 1.8 by 30 weeks, more significant than with IVMP, with effective B cell depletion from peripheral blood. Proptosis reduction and lid signs significantly improved in the rituximab-treated cases, but not more than IVMP.29

Case series on steroid-refractory, and sight threatening TED treated by rituximab appeared equally promising.30 31 Nine cases from Mitchell et al resistant to IVMP, who had DON or persistent moderate-to-severe active TED were treated with two 1 g rituximab infusions. CAS improved in all individuals. The four DON cases had reduction of NO SPECS grading from 6 to 4.30 In another retrospective cohort of six patients refractory to IVMP with progressive disease and DON, a mean reduction of CAS from 5.3 to 1.3 was observed at 8 weeks post rituximab. Improvement of visual acuity within 4 weeks in all DON cases was observed.31

Not all TED cases were met with success with rituximab rescue, with some cases developing major adverse reactions including a few cases of new onset DON. One case refractory to IVMP developed DON after the first rituximab infusion and required three-wall orbital decompression.32 Two moderate-to-severe active TED cases from a randomised control trial developed DON following rituximab infusions, one case 3 weeks and the other case 24 weeks after treatment.33 In another two patients with moderate-to-severe active TED, rituximab infusion caused rapid onset of orbital oedema and vision reduction, which was controlled with intravenous hydrocortisone with resolution of orbital swelling within 3 hours, most likely due to cytokine release syndrome.34

The efficacy of rituximab in moderate-to-severe TED was formally explored in two relatively small double-blind randomised controlled trials. In the Salvi et al trial, rituximab (n=15) either as two 1 g infusions or one 500 mg infusion was compared with IVMP (n=16). Mean CAS reduction was on par at 12 weeks in both the rituximab and IVMP groups, followed by significantly greater CAS reduction in the rituximab group from 16 weeks onwards, with either rituximab protocol. At 24 weeks, TED inactivation was observed in 100% of the patients treated with rituximab, compared with 69% after IVMP (p<0.05). At 52 weeks, rituximab was superior, with no reactivation of TED in the rituximab treatment arm, as compared with 31% reactivation rate in the IVMP group. Secondary endpoint results were mixed, with no significant reduction of mean proptosis and Gorman diplopia score in both treatment arms, although the rituximab group showed greater motility and improvement in quality of life scores at 52 weeks.34

Conversely, Stan et al randomised 25 patients with TED to either two 1 g rituximab or placebo saline infusions. The trial did not find differences between the two groups in mean CAS reduction at 24 or 52 weeks. Similarly, there were no significant differences in proptosis reduction, lid fissure changes, total orbital volume or diplopia score or quality of life between the two groups at 24 or 52 weeks. There were more adverse reactions in the rituximab group (80%) compared with the placebo group (27%), five of the six moderate-to-severe adverse reactions occurred in the rituximab group including two DON, vasculitis, infections (bronchitis and conjunctivitis) and gastrointestinal side effects.33

Given the contradictory results from these randomised controlled trials, rituximab currently cannot replace IVMP as primary treatment. Rituximab may have a role in treating severe, corticosteroid-resistant cases based on level IV evidence (defined as case series and case studies without concurrent controls). Larger clinical trials are warranted to confirm the efficacy of rituximab in active TED treatment.


Tocilizumab is a humanised monoclonal antibody against interleukin-6 receptor. The first study on tocilizumab was a prospective interventional study from Spain with 18 patients with steroid-resistant active TED who had failed three doses of weekly IVMP treatment. The patients were pulsed with 8 mg/kg of tocilizumab every 4 weeks until CAS was 1 or less. An average of 5.38 sessions (range 4–9) were required to achieve this endpoint. Mean reduction of CAS of 5.89 was achieved at 9 months. There was an impressive proptosis reduction in 72% of cases, where mean proptosis reduction was 3.92 mm, including one with 7 mm reduction. Improved extraocular motility >5° was observed in 83%, with 53.9% achieving resolution of diplopia in primary gaze with a minimum follow-up of 9 months. One patient had resolution of DON with three doses of tocilizumab. The adverse reaction profile was good, with mostly minor events such as tiredness, neutropaenia, upper respiratory infection and reversible elevation of liver enzymes.35

The encouraging results led to a randomised controlled trial by the same group. Perez-Moreiras et al randomised 32 patients to either four doses of monthly 8 mg/kg tocilizumab or placebo infusions in active, patients with moderate-to-severe TED who had incomplete response to at least three doses of IVMP or recurrence of active disease after stopping IVMP, and they were followed for up to 40 weeks. A primary endpoint of CAS reduction of two or more was achieved significantly more in cases on tocilizumab (93%) compared with the placebo group (59%) at 16 weeks. Disease inactivation was also significantly higher in the tocilizumab group (86.7%) as compared with 35.2% in the placebo group. However, proptosis reduction in the tocilizumab group was modest (1.5 mm mean proptosis reduction) and there was no difference in quality of life score comparing the two groups at 40 weeks. Reactivation rate was not reported. Two reported major adverse reactions were moderately raised transaminases and acute pyelonephritis.36

The randomised trial suggested that tocilizumab is efficacious in reducing orbital inflammation in steroid-resistant cases, with a small disease-modifying effect.36 Currently tocilizumab has a role in treating active, severe TED cases refractory to steroid treatment. More clinical studies are needed in defining the efficacy of tocilizumab in treating DON, and to reproduce the efficacy shown for active, moderate-to-severe TED.


In a phase II clinical trial that examined the efficacy of teprotumumab in 88 patients with active TED (CAS≥4) of <9 months duration,37 participants were randomised to three weekly teprotumumab (first infusion 10 mg/kg, subsequent seven infusions 20 mg/kg) over 24 weeks, or placebo infusions. At the end of 24 weeks, 69% of patients receiving teprotumumab as compared with 20% of patients on placebo had proptosis reduction of ≥2 mm and ≥2 CAS reduction. Proptosis reduction was noted as early as 6 weeks, with mean reduction of 3 mm in the teprotumumab group by 24 weeks.37 At week 24, a high response (proptosis reduction by ≥3 mm) was achieved in 54.8% in the teprotumumab group, compared with 8.9% in the placebo group. About 0% in teprotumumab group had no response, compared with 48.9% in the placebo group.38 Collectively, significant proptosis reduction (≥2 mm) was achieved in 71.5% in the teprotumumab group compared with 20% in the placebo group.38 There was also significantly greater CAS reduction from baseline in the teprotumumab treatment arm compared with placebo at each visit, improvement in quality of life scoring (GO-QOL) for visual functioning and subjective diplopia in the teprotumumab group compared with placebo across all time points. By 24 weeks, post-hoc analysis showed disease inactivation rate of 69% with teprotumumab as compared with 21% given placebo. Reactivation rate was not formally reported at 28 weeks, except for an observation of no rebound phenomenon.37 In terms of safety, 5 (11.9%) patients had to discontinue teprotumumab treatment due to major adverse reactions.37

The phase III OPTIC trial confirmed the reproducibility of proptosis reduction over 24 weeks, with 83% (34/41) in the teprotumumab group achieving proptosis reduction ≥2 mm compared with 10% (4/42) in the placebo group. Mean reduction in proptosis was 2.82 mm in the teprotumumab group, compared with 0.54 mm in the placebo group at week 24. In six participants who had radiological imaging pre-teprotumumab and post-teprotumumab treatment, the reduction in proptosis was attributed to both extraocular muscle and orbital fat volume reduction. Inactivation of active disease was more frequent (59% vs 21%), reduction in diplopia and improvement in GO-QOL score was significantly higher in the teprotumumab group compared with placebo at week 24. Teprotumumab was better tolerated in the phase III trial, and most events were mild to moderate in severity.

Teprotumumab has shown remarkable results in modifying TED disease activity and severity, visibly as proptosis reduction. It has become the first immunomodulator to get Food and Drug Administration approval for the treatment for TED. The long-term outcomes beyond 48 weeks are lacking, but additional data collection is ongoing from both trials.


Mycophenolate mofetil (MMF) is a prodrug of mycophenolic acid, an inhibitor of inosine monophosphate dehydrogenase involved in de novo purine synthesis. Mycophenolate is potently cytostatic on T and B cells.39 It is a first-line immunosuppressant in solid organ transplantation and lupus nephritis, and is frequently used off-label for other autoimmune diseases.40–42 The popularity of using mycophenolate lies in its favourable safety profile, with mainly gastrointestinal and haematological side effects that are fully reversible.40 The enteric-coated mycophenolate sodium (MPS) was designed to reduce the high incidence of gastrointestinal side effects associated with MMF, by allowing slower absorption than MMF. For transplant recipients, the usual dosage for MPS is 720 mg twice daily, which is therapeutically equivalent to MMF 1000 mg twice a day. However, both MMF and MPS were found to be similar in the incidence and requirement for dose changes for gastrointestinal adverse events.40

Two randomised controlled trials have investigated the efficacy of mycophenolate as either monotherapy or in combination with IVMP in TED. In a single centre trial, 174 patients with active, moderate-to-severe TED were randomised to either 3 g IVMP followed by tapering oral prednisolone (60 mg/day) or 1 g daily MMF for 24 weeks.43 The overall response rate was significantly superior in the MMF group compared with the glucocorticoid group at both 12 weeks (78.8% vs 51.3%, p<0.05) and 24 weeks (91.3% vs 67.9%, p<0.05). The proportion of patients with CAS reduction ≥2 was significantly greater in the MMF group (92.5%) than the glucocorticoid group (70.5%) at 24 weeks. A reduction in proptosis of at least 2 mm was noted in 68.8% on MMF as compared with 39.7% on glucocorticoid (p<0.05), where mean proptosis reduction was 3.2 mm and 3.4 mm for MMF compared with 1.9 mm and 2.2 mm for glucocorticoid, for the right and left eye, respectively. Diplopia was significantly improved in 90.4% in the MMF group compared with 63.6% in the glucocorticoid group at 24 weeks. TED reactivation was observed in five patients treated with glucocorticoids (6.4%) but in none of the MMF group. Adverse events occurred in 5% of patients treated with MMF, and 28% of patients treated with glucocorticoids.43 This study concluded that mycophenolate was highly effective and safer than glucocorticoids in reducing the activity and severity in active, moderate-to-severe TED.

The efficacy findings were less dramatic and mixed for the MINGO trial which compared MPS (720 mg daily for 24 weeks) and IVMP combination therapy, with IVMP monotherapy (cumulative 4.5 g over 12 weeks) in 164 patients with active, moderate-to-severe TED.17 At 12 weeks, the overall response rate was not significantly different between the treatment arms at 49% for the monotherapy group and 63% for the combination group. On post-hoc analysis, a significantly greater response was seen in the combination group (71%) compared with the monotherapy group (53%) by 24 weeks (p=0.03), and a sustained response was achieved in 67% of the combination group compared with 46% in the monotherapy group at 36 weeks (p=p 0.01) which suggested an additive effect of mycophenolate. Relapse rates were not significantly different at 24 or 36 weeks between the two treatment arms (11% for IVMP monotherapy and 8% for the combination group at 24 weeks).17 The addition of MPS also did not reduce severity of proptosis. Adverse effects were similar between monotherapy (20%) and combination therapy (25%), both IVMP and MPS were generally well tolerated. On the systemic safety analysis of mycophenolate use in TED, most side effects were mild with a reported side effect rate of 8.8% for gastrointestinal symptoms, 7.1% infection and 1.2% liver dysfunction in all mycophenolate-treated patients.44

While the two randomised controlled trials on mycophenolate showed discrepancy in the magnitude of effectiveness, both have shown a favourable treatment response with the addition of mycophenolate. The evidence suggests that mycophenolate is a useful adjuvant to IVMP, probably in patients with incomplete response to pulsed IVMP and in patients with TED reactivation in active, moderate-to-severe TED.


Azathioprine is a purine analogue which interrupts synthesis of purine ribonucleotides. It is an antiproliferative drug against fast dividing cells and has been commonly used in rheumatoid arthritis, organ transplant patients and inflammatory bowel disease.45

In the CIRTED trial involving 126 patients receiving a tapering dose of oral prednisolone (80 mg/day) over 24 weeks, they were randomised to either azathioprine 100–200 mg/day or placebo, 20 Gray orbital radiation or sham irradiation. The binary clinical composite outcome and ophthalmopathy index assessed at 48 weeks did not differ significantly between the azathioprine and placebo groups.16 No benefit was seen with long-term GO-QOL visual function or appearance for either azathioprine or orbital radiotherapy. On post-hoc analysis, the study showed that the azathioprine group had a higher probability of improvement in clinical composite outcome at 48 weeks when compared with placebo.16 However, firm conclusions cannot be made on the potential benefit of azathioprine, including prevention of relapse, because of a high rate of participant withdrawal from each study arm. A substantial number were due to side effects including abnormal blood tests, and some were due to deteriorated clinical conditions and others were not specified. The most common adverse events from azathioprine were mild infections, with no serious adverse events.

There is not enough evidence to support the use of azathioprine in active TED unless more positive studies emerge in the future.


Methotrexate (MTX), an antimetabolite that inhibits folic acid synthesis, which in turn inhibits T and B cell proliferation, is a widely used, safe, first-line immune-modulator for the treatment of rheumatoid arthritis.46 As yet, there has been no prospective study or randomised clinical trial to support the routine use of MTX in TED.

There have been four uncontrolled, retrospective studies that used MTX for TED in different clinical settings, either alone or in combination with systemic corticosteroid, another immunotherapy, or with orbital radiation, which makes isolating the efficacy of MTX difficult. Overall MTX use was associated with steroid sparing effect in about 60% of cases, reduction in CAS, with no clear data on modifying disease severity.47–50 In a study of 36 patients who used low-dose oral MTX (7.5 to 10 mg) after discontinuing corticosteroid due to side effects, CAS and ocular motility improvement were observed at 3, 6 and 12 months without improvement in visual acuity, proptosis or eyelid position.48 In another study, 19 patients with severe active TED minimally responsive to three 1 g IVMP were commenced on weekly MTX (mean dose 17.8 mg) and followed for a mean duration of 40 months (1206 days).47 About 91% (31/34) of eyes achieved an inflammatory index score of <3/8 (inactive disease) within a mean of 6 months (189 days). Four (21%) patients needed additional IVMP or radiotherapy to achieve disease control. Two (10%) patients progressed to DON. Reactivation rate was 8% after withdrawal of MTX. Overall 12 (63%) of 19 patients patients ended MTX treatment as TED was stable and inactive.47

Two ocular inflammation clinics used MTX at either 15 mg/week orally or 20 mg/week subcutaneously in 14 patients for 3–12 months who were dependent on oral prednisolone (average 32 mg/day). Nine (64%) of the 14 were able to discontinue oral prednisolone after a mean duration of 7.5 months. In nine patients on MTX for at least 6 months, 2 (22%) had improvement in proptosis, and 3 (33%) had improvement in visual acuity.50 In a multidisciplinary thyroid clinic, 23 patients with active moderate-to-severe TED were concurrently commenced on MTX at 20–25 mg/week and three doses of 500 mg IVMP.49 MTX was used for an average of 13 months. Eleven (46%) patients required second-line and three patients third-line immune-modulators (ciclosporin, azathioprine, rituximab). The group reported significant reduction in mean inflammatory index score from 5.5/10 to 2.7 at 25 weeks and 1.3 at 48 weeks, and a mean cumulative dose of IVMP at 2.71 g.49 Four (17%) patients failed to respond and needed orbital decompression for DON. Reactivation rate on cessation of MTX, and effect of MTX on disease severity were not reported.49 At 15–25 mg treatment dose, side effects were mild with mainly gastrointestinal symptoms, infections and a few cases of reversible increased liver enzymes.47 49 50


Ciclosporin is one of the earliest immune-modulators trialled in TED because the mechanism of its action specifically inhibits T cell activation by inhibiting interleukin-2 transcription.51 However, its use has not been widely accepted in the treatment of TED due to dose-dependent toxicity, especially renal impairment, which requires drug level monitoring.

Two early, small randomised controlled trials investigated the efficacy of ciclosporin, and both showed positive responses.52 53 Kahaly’s study found the addition of ciclosporin improved disease severity and reduced active disease relapse over 12 months.52 The combination of ciclosporin (5–7.5 mg/kg) and oral glucocorticoid was associated with a steeper drop in activity score compared with prednisolone over 10 weeks. Reduction of mean proptosis (3 mm in combined group vs 1 mm in prednisolone monotherapy group) was significant at 10 weeks but not at 6 or 12 months. Over 12 months, 8 (40%) in 20 on oral prednisolone alone had relapse of active disease, compared with 1 (5%) in 20 on ciclosporin and oral prednisolone. The study found dose-dependent renal toxicity, liver enzyme derangement and hypertension that were reversible and pneumonia that required cessation of treatment.52

In Prummel’s study, 36 subjects were randomised to either oral glucocorticoid or ciclosporin (7.5 mg/kg/day) for 3 months in severe TED. The oral prednisolone group did better at 12 weeks with 61% response rate as compared with 22% in ciclosporin. Only the prednisolone group showed improvement in total eye score, visual acuity and eye muscle enlargement, that was not observed in ciclosporin. For the non-responders to either treatment, combination therapy with lower dose of oral prednisolone and ciclosporin was continued for another 12 weeks, and this led to a further treatment response rate of 59%, with reduction in total eye score, eye muscle enlargement, proptosis and improved visual acuity.53 This study suggested that ciclosporin was inferior to oral prednisolone in efficacy as monotherapy. Combination treatment with ciclosporin and glucocorticoids can be considered in non-responders to either treatment for additional effects.


While intravenous corticosteroid is still the first line of treatment for active, moderate and severe TED including DON, targeted immunotherapies afford additional benefits especially for partial and non-responders to corticosteroid treatment. Addition of immunomodulators has the potential to reduce disease severity by proptosis reduction, disease inactivation and improving treatment response in corticosteroid refractory cases; the benefit needs to be weighed against side effects and cost. Mycophenolate is probably the current first line as adjuvant, as it is well tolerated for active, moderate-to-severe TED cases not initially responding to corticosteroid and in reactivation cases. There is evidence for the use of tocilizumab in active, severe TED cases not responsive to corticosteroid. There is some evidence for using rituximab in DON cases in refractory cases, but inconclusive evidence for its use in moderate-to-severe TED. Studies on teprotumumab appeared remarkable in disease severity reduction, and further long-term outcome data may change the treatment paradigm for active moderate-to-severe TED treatment in the future.

Key learning points

  • While intravenous corticosteroid is the current first-line management in moderate-to-severe active thyroid eye disease (TED), and in dysthyroid optic neuropathy (DON), emerging data on immunotherapies suggest additional benefits especially in partial and non-responders to corticosteroid treatment, pending data on long-term outcomes.

  • Immunotherapies trialled in TED consisted of antimetabolites that generally suppresses the immune system, and monoclonal antibodies that target specific antigens, cytokines and cells.

  • Mycophenolate is a relatively safe, useful addition to intravenous methylprednisolone in moderate-to-severe active TED cases with incomplete response to pulsed corticosteroid and in reactivation cases as a steroid-sparing agent.

  • Tocilizumab has been shown to be efficacious especially in reducing orbital inflammation in corticosteroid-resistant, severe active TED.

  • Rituximab’s role in active, moderate-to-severe TED is controversial, but level IV evidence supports the use of rituximab in refractory DON cases.

  • Teprotumumab is promising as an immunotherapy of choice for rapidly reducing orbital inflammation and in modifying disease severity in moderate-to-severe active TED, pending long-term outcome studies.


We would like to acknowledge Ms Way Tze Saw, a retail pharmacist, and Ms Ivy Tan, a hospital pharmacist, for their assistance with medication price listing and calculation.



  • Twitter @jjkhong

  • Contributors JJK is involved in the conception of review, literature analysis and synthesis, preparation of manuscript, critically revising the manuscript and responding to reviewers. AM critically appraised and edited the manuscript for the entire content, and approved the final version.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.