The Journal of Clinical Endocrinology & Metabolism, 2023, 108, 2615–2625

https://doi.org/10.1210/clinem/dgad084
Advance access publication 27 March 2023
Clinical Research Article

Factors Predicting Long-term Outcome and the Need for

Surgery in Graves Orbitopathy: Extended Follow-up From
the CIRTED Trial
Peter Taylor,1 Rathie Rajendram,2 Stephanie Hanna,1 Victoria Wilson,3 Julie Pell,1 Chunhei Li,1
Anne Cook,4 Rao Gattamaneni,5 Nicholas Plowman,6 Sue Jackson,7 Robert Hills,8 Robert French,1

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Jimmy M. Uddin,2 Richard W. J. Lee,2,3,9,* and Colin M. Dayan;1,* on behalf of the CIRTED
investigators
1
Thyroid Research Group, Cardiff University School of Medicine, Cardiff CF14 4XN, UK
2
Moorfields Eye Hospital NHS Foundation Trust, London, EC1V 2PD, UK
3
Bristol Medical School, University of Bristol, UK
4
Manchester Royal Eye Hospital, Manchester, M13 9WH, UK
5
Christie Hospital, Manchester, M20 4BX, UK
6
Bristol Haematology and Oncology Centre, Bristol, BS2 8ED, UK
7
Centre for Appearance Research, University of the West of England, Bristol, BS16 2JP, UK
8
Nuffield Department of Population Health, University of Oxford, Oxford OX3 7LF, UK
9
Institute of Ophthalmology, Faculty of Brain Sciences, University College London, London EC1V 9EL, UK
Correspondence: Peter Taylor, PhD, Thyroid Research Group, C2 Link Corridor, UHW, Heath Park, Cardiff CF14 4XN, UK. Email: [email protected].
*Joint senior authors.

Abstract
Graves orbitopathy is both disabling and disfiguring. Medical therapies to reduce inflammation are widely used, but there is limited trial data
beyond 18 months of follow-up.
Methods: Three-year follow-up of a subset of the CIRTED trial (N = 68), which randomized patients to receive high-dose oral steroid with
azathioprine/placebo and radiotherapy/sham radiotherapy.
Results: Data were available at 3 years from 68 of 126 randomized subjects (54%). No additional benefit was seen at 3 years for patients
randomized to azathioprine or radiotherapy with regard to a binary clinical composite outcome measure (BCCOM), modified European Group
on Graves’ Orbitopathy score, or Ophthalmopathy Index.
Clinical Activity Score (CAS), Ophthalmopathy Index, and Total Eye Score improved over 3 years (P < .001). However, quality of life at 3 years
remained poor. Of 64 individuals with available surgical outcome data, 24 of 64 (37.5%) required surgical intervention. Disease duration of greater
than 6 months before treatment was associated with increased need for surgery [odds ratio (OR) 16.8; 95% CI 2.95, 95.0; P = .001]. Higher
baseline levels of CAS, Ophthalmopathy Index, and Total Eye Score but not early improvement in CAS were associated with increased
requirement for surgery.
Conclusion: In this long-term follow-up from a clinical trial, 3-year outcomes remained suboptimal with ongoing poor quality of life and high
numbers requiring surgery. Importantly, reduction in CAS in the first year, a commonly used surrogate outcome measure, was not associated
with improved long-term outcomes.
Key Words: graves orbitopathy, CAS, GOQOL, azathioprine, radiotherapy, CIRTED

Graves’ disease is common throughout the world (1). Active reducing tissue remodeling in the extraocular muscles, orbital
moderate to severe Graves orbitopathy (GO; also known as fat, and other periocular soft tissues (4, 7). Immunosuppressive
thyroid eye disease or thyroid-associated orbitopathy) occurs therapies, particularly corticosteroids, are the mainstay of treat­
in approximately 5% to 10% of cases of Graves’ disease (2, ment (3, 8, 9). More recently, randomized clinical trials have in­
3). It can be both visually disabling and cosmetically disfiguring dicated substantial potential benefit from novel/additional
and substantially impairs quality of life (2, 4-6). Middle-aged agents such as azathioprine (10), teprotumumab (11), mycophe­
women are predominantly affected, and the negative impact nolate, (12) and tociluzimab (13). However, these clinical trials
of GO can be prolonged. have largely been of modest duration, and follow-up has been
Current medical therapeutic strategies have been directed at most 18 months and often less. One exception is the study
toward suppression of orbital inflammation in the hope of of Leo et al that followed patients up for 4 to 10 years. This

Received: 20 June 2022. Editorial Decision: 8 February 2023. Corrected and Typeset: 19 April 2023
© The Author(s) 2023. Published by Oxford University Press on behalf of the Endocrine Society.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (https://creativecommons.
org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered
or transformed in any way, and that the work is properly cited. For commercial re-use, please contact [email protected]
2616 The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10

showed continued improvement and a requirement for surgery fractions over 2 to 3 weeks. Those randomized to receive aza­
in 27% of patients, but limited quality of life data was reported thioprine received this treatment for 48 weeks with doses ad­
(14). Alongside these advances, surgery continues to be required justed according to weight and safety blood monitoring. The
to reduce the degree of diplopia or proptosis (3) and in some majority of patients were recruited from 2 centers in the UK
cases to preserve vision, as well as to improve quality of life (Bristol and Moorfields; n = 105) (10). The CIRTED trial
(15). Frequently, surgical intervention is performed a year or identified potential benefit at 48 weeks for azathioprine, but
more after disease onset and so is rarely included in trial out­ no clear benefit was observed for radiotherapy (10).
comes. At present there is limited data on what proportion of in­ Following week 48 (completion of the original CIRTED
dividuals with moderate to severe GO require surgery despite study) follow-up was optional and patients were managed
extensive medical intervention. Taken together, there is a press­ by their local teams.
ing need for longer term data to establish which of the character­
istics of GO correlate with poor clinical outcomes once the
inflammatory phase of the disease has resolved. These may be Outcomes
key in assessing who will potentially benefit the most from In the original study, the coprimary outcomes were a BCCOM

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

new and costly interventions and provide insight into disease tra­ (Box 1) and OI. Secondary outcome measures included Total
jectory while awaiting for trials with longer duration of Eye Score (TES) as an additional assessment of disease severity
follow-up. and the patient-reported GOQOL score (6, 23, 24), which has
There are several well-established tools for assessing GO. 2 components: Visual Function (GOQOL-VF) and
Typically disease activity is measured using Clinical Activity Appearance (GOQOL-AP). CAS was also measured through­
Score (CAS) or by the inflammatory component of the vision, out. In this 3-year report we additionally calculated the
inflammation, strabismus, and appearance scale (3, 16, 17). EUGOGO score from the individual components of the trial
Other scores objectively assess the severity of GO including dataset (18, 19). As this did not include a measurement of
European Group on Graves’ Orbitopathy (EUGOGO) (18, lid retraction, we inferred the presence of lid retraction if there
19), Ophthalmopathy Index (OI) (20), Thyroid Eye Score, was a >2 mm difference in palpebral aperture between an in­
NOSPECS (21, 22), Graves Ophthalmopathy Quality of dividual patient’s eyes. On this basis, moderate to severe GO
Life (GOQOL) (23, 24), and the vision, inflammation, was classified as having 2 or more of ≥ 2 mm difference in pal­
and appearance component of the vision, inflammation, stra­ pebral aperture, moderate or severe soft tissue involvement,
bismus, and appearance scale (17). In addition to these meas­ exophthalmos ≥ 3 mm above normal for race and gender,
ures, a greater understanding of disease trajectory is essential and inconstant or constant diplopia. Hence, the indices we
for identifying which patients should be followed more used to assess longitudinal features and stability of disease
closely in clinic. Potential variables that might predict the over 3 years or follow-up were the OI, TES, CAS,
disease trajectory and likely need for additional interven­ EUGOGO, and GOQOL scores. In addition, whether oph­
tions such as surgery include patient characteristics such as thalmic surgery was required, which was entirely at the discre­
baseline GO severity scores as well as age, sex, and duration tion of the patients and their local ophthalmologists, was also
of disease as well as potent risk factors such as smoking recorded.
status.
The CIRTED trial was a multicenter, factorial design,
double-masked, randomized controlled trial based in the
UK. Results from the primary outcome at 1 year have been re­
Box 1 Calculation of the BCCOM
ported (10), but participants were also invited to attend an op­
tional follow-up assessment at 3 years. All patients in the BCCOM Major Criteria
study received high-dose oral corticosteroids and were • An improvement of ≥1 grade in diplopia score
randomized to receive either azathioprine or placebo and or­ • An improvement of >8 degrees of eye movement
bital radiotherapy or sham radiotherapy. These planned med­ in any direction
ical interventions were completed within 6 months of study • A reduction of ≥2 mm in proptosis
entry. The 3-year data in CIRTED enables us to identify im­ BCCOM Minor Criteria
portant predictors of disease outcome and treatment response, • A reduction of ≥2 mm in lid aperture
especially with regard to quality of life and the need for • An improvement of ≥1 grade in soft tissue
surgery. involvement
• An improvement in best-corrected visual acuity of
Methods ≥1 line on the Snellen chart
• Subjective improvement
Study Details and Procedures All items refer to the worst eye.
The CIRTED trial was a randomized multifactorial trial and Response to treatment is calculated as follows:
has been described in detail previously (10, 25). Baseline char­ Improved = improvement in ≥1 major criteria or ≥2
acteristics were obtained on age at enrollment, disease dur­ minor criteria
ation, ethnicity, sex thyroid status, study center, CAS, and No Change = improvement or deterioration in ≤1 mi­
Thyroid Eye Score. In brief, all patients received high-dose nor criterion
oral prednisolone in a tapering regime for 24 weeks (80 mg Worse = deterioration in ≥1 major or ≥2 minor criteria
per day, reduced to 20 mg per day by 6 weeks, 10 mg per (even if other criteria improve) or requiring rescue
day by 15 weeks, and 5 mg per day by 21 weeks). Those therapy (radiotherapy, intravenous steroid, or orbital
randomized to radiotherapy received 20 Gy of radiation ad­ decompression) at any point
ministered to the retrobulbar compartment in 10 to 12
The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10 2617

Statistical Analysis aside from ethnicity, where non-Caucasians were less likely
The impact of being randomized to azathioprine and radio­ to provide 3-year data (OR = 0.30; 95% CI .11, .79;
therapy against the original primary outcome measures of dis­ P = .02). Baseline characteristics of the 3-year study group
ease severity BCCOM and OI were studied first. CAS was a vs the remainder of the CIRTED cohort are shown in
coprimary outcome, although we anticipated all participants Supplementary Table 1 (27).
would have a significant improvement in CAS well before 3
years in accordance with the natural history of the disease Analysis of the Effect of Azathioprine and
(26) and the EUGOGO score was utilized as an additional Radiotherapy on Primary Outcomes at 3 Years
outcome. Analysis was undertaken adjusting for the same At the 3-year assessment, we observed no clear benefit in terms
confounders as in the initial CIRTED analysis; these included of improvement in BCCOM between individuals randomized
smoking status at the time of GO diagnosis, thyroid status on to azathioprine vs placebo. The adjusted OR for improvement
enrollment, previous corticosteroid use, sex, disease severity, in BCCOM in individuals randomized to azathioprine was
disease duration, age greater than 60 years, and disease activ­ OR(adj) = 0.51 (95% CI .08, 3.13; P = .47) (Table 2). There
ity (baseline CAS). Additional analyses were then undertaken

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

was also no clear benefit of improvement with orbital radio­
to explore the effect of being randomized to azathioprine or therapy vs sham radiotherapy in terms of improvement in
radiotherapy on TES, quality of life scores, and need for oph­ BCCOM. The adjusted OR for improvement was OR(adj) =
thalmological surgery. 2.17 (95% CI .43, 10.9; P = .35) (Table 2). There was no evi­
Multivariable linear and logistic regression models were dence of interaction between azathioprine and radiotherapy
used, adjusting for minimization variables, the factorial de­ (pint = 0.20).
sign, and the value of the outcome variable at baseline. No additional benefits above oral steroid were seen with ei­
Where appropriate, to allow easy comparison variables were ther azathioprine or orbital radiotherapy on OI. Individuals
standardized; these analyses are therefore presented as per randomized to azathioprine had an adjusted beta [(B)(adj)] of
standard deviation. Comparison between nonnormally dis­ 1.33 (95% CI −.68, 3.34; P = .19) on OI, and for those
tributed outcomes between groups was undertaken using the randomized to orbital radiotherapy, B(adj) was 0.11 (95% CI
Wilcoxon rank sum test. The original trial was a factorial de­ −1.93, 2.13; P = .92) on OI (Table 2). There was also no evi­
sign, and, as patients who followed up at 3 years were broadly dence of an interaction between azathioprine and radiother­
representative of the whole trial population, with no striking apy in their effect on OI (pint = 0.85).
benefit observed for either azathioprine or radiotherapy and No differences in change in CAS were observed between in­
no evidence of interaction between them, we combined treat­ dividuals who received treatment with azathioprine vs pla­
ment groups when exploring baseline ophthalmic and clinical cebo (B(adj) = 0.58; 95% CI −.20, 1.36; P = .14) or those
assessments and their relationship to subsequent clinical out­ who received radiotherapy vs sham radiotherapy (B(adj) =
comes to increase power. In these analyses we explored the re­ 0.20; 95% CI −.58, .98; P = .61) (Table 2). There was no evi­
lationship between the baseline clinical assessments of CAS, dence of interaction between azathioprine and radiotherapy in
OI, TES, GOQOL-VF, and GOQOL-AP and their respective their effect on CAS (pint = 0.76).
scores at 3 years. We also explored the association between We also observed no differences in terms of improvement in
baseline CAS (as a continuous measure), OI (in quartiles) EUGOGO score between individuals randomized to azathio­
TES, and EUGOGO scores as well as baseline characteristics prine vs placebo or radiotherapy vs sham radiotherapy. The ad­
on the need for ophthalmological surgery over the 3 years of justed OR for having odds of moderate to severe GO as opposed
follow-up. to mild GO in individuals randomized to azathioprine was
Statistical significance was defined in advance as a P value of OR(adj) = 2.48 (95% CI .14, 44.7; P = .54) and for radiotherapy
<.05. All statistical analyses were undertaken using STATA was OR(adj) = 0.58 (95% CI .03, 10.97; P = .72) (Table 1).
version 16 (STATACORP, College Station, TX, USA). Analysis of key individual components of primary outcomes
showed no clear benefits for azathioprine or radiotherapy on in­
Results dividual components apart from radiotherapy being associated
with reduced incidence of diplopia at 3 years (B = −0.56; 95%
Data were available after 3 years of follow-up from 68 of 126
CI −1.08, −.06; P = .03) (Supplementary Table 2) (27).
(54%) individuals initially randomized into the CIRTED
study. Of these 68 individuals, 64 provided data on their re­
quirement for ophthalmological surgery, 58 on CAS score, Analysis of the Effect of Azathioprine and
52 on BCCOM, 52 on GOQOL, and 49 on OI. The Radiotherapy on Additional Outcomes at 3 Years
EUGOGO score could be calculated from the individual meas­ Individuals randomized to azathioprine had a higher TES
urements recorded in our dataset (as set out previously) for 47 after adjustment than those who received placebo [TES
trial subjects. The 68 individuals with 3-year data were fairly (B)(adj) of 3.11; 95% CI .04, 6.19; P = .05), and in those
evenly distributed across the intervention groups: 17 were in randomized to orbital radiotherapy the B(adj) was −0.61
the placebo/sham radiotherapy group, 16 were in the (95% CI −3.87, 2·65; P = .71) (Table 2). There was also no
placebo/radiotherapy group, 16 were in the azathioprine/ evidence of an interaction between azathioprine and radio­
sham radiotherapy group, and 19 were in the azathioprine/ therapy in their effect on OI (pint = .31).
radiotherapy group. The derivation of study participants is Individuals who were randomized to receive azathioprine did
shown in Figure 1. not have higher GOQOL-VF than those who did not
Baseline characteristics of key risk factors for GO, potential (P = .32), nor did those randomized to receive radiotherapy (P
confounders, and disease levels at baseline (CAS/TES) for trial = .36). A potential late benefit of radiotherapy was seen with re­
subjects who provided 3-year data, compared to those who gard to GOQOL-AP [median 81.3 (interquartile range [IQR]
did not, did not show any substantial differences (Table 1) 62.5-93.8) vs median 68.9 (IQR 31.2-87.5); P = .03]. In
2618 The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Figure 1. Derivation of study participants.

contrast, no late benefit was observed for GOQOL-AP in those P = .62; r2 = 0.005). However, baseline CAS was associated
who received azathioprine [median 68.9 (IQR 31.3-87.5) vs me­ with CAS at 12 weeks (B(std) = 0.37; 95% CI .18, .55;
dian 84.4 (IQR 68.8-93.8) for placebo; P = .06]. P < .001; r2 = 0.14) and to a lesser extent CAS at 48 weeks
No clear benefit on reducing the odds of requiring any sur­ (B(std) = 0.24; 95% CI .04, .44; P = .02; r2 = 0.05). Not
gery was observed for either azathioprine (OR = 0.39; 95% even CAS at week 48 was associated with CAS at 3 years
CI .09, 1.69; P = .21) or radiotherapy (OR = .34; 95% CI (B(std) = 0.09; 95% CI −.18, .37; P = .49; r2 = 0.01).
.08, 1.52; P = .16) (Table 3). EUGOGO score fell progressively over the study from
Individuals randomized to receive azathioprine did not see a 70.8% of patients having moderate to severe GO at baseline
reduction in the need for lid surgery (OR = 0.60; 95% CI .12, to 38.9% having moderate to severe GO at 12 weeks, falling
3.04; P = .54); however, a borderline significant result was ob­ further to 32.1% at 48 weeks and 6.4% at 3 years. Baseline
served for radiotherapy (OR = 0.17; 95% CI .03, 1.03; EUGOGO score was associated with 12-week EUGOGO
P = .05) with the 95% CI crossing equality (Supplementary score (B = 0.53; 95% CI .26, .79; P < .001; r2 = 0.24) but
Table 3) (27). No benefit was seen for either azathioprine not 48-week score (B = 0.26; 95% CI −.02, .54; P = .07;
(OR = 0.39; 95% CI .06, 2.68) or radiotherapy (OR = 0.71; r2 = 0.05) or 3-year score (P = .64).
95% CI .09. 5.74; P = .75) in reducing the need for orbital de­ GOQOL-VF improved progressively over the study period
compression (Supplementary Table 4) (27). from a baseline median of 68.7 (IQR 43.3-87.5) to 93.8
(IQR 76.7-100) at 3 years (P < .001) (Fig. 2D). GOQOL-VF
Changes in Clinical Assessments of GO Over the at baseline was associated with GOQOL-VF at 3 years
Study Period (B(std) = 0.62; 95% CI .39, .85; P < .001; r2 = 0.35). GOQOL-
AP also rose over the study period, although to a lesser degree
OI fell over the study period from a mean at 12 weeks 9.45
than visual function, rising from a baseline median of 56.2
(SD 3.95) to 6.02 (SD 2.99) (P = <.001 at 3 years) (Fig. 2A).
(IQR 37.5 −75) to 75 (IQR 56.3-93.8; P = .002) (Fig. 2E).
OI at 12 weeks correlated with OI at 3 years (B(std) = 0.59;
GOQOL-AP at baseline was associated with GOQOL-AP at
95% CI .34, .84; P < .001; r2 = 0.35). OI at 48 weeks was
3 years (B(std) = 0.61; 95% CI .38, .83; P < .001; r2 = 0.35).
also correlated with OI at 3 years (B(std) = 0.42; 95% CI .21,
.63; P < .001; r2 = 0.28).
TES fell over the study period from a baseline mean of 14.9 Outcomes: Factors Predicting the Need for Surgery
(SD 6.27) to 6.33 (SD 4.65) (P < .001 at 3 years) (Fig. 2B). Over the Study Period (3 Years).
Baseline TES was not associated with TES at 3 years (B(std) =
−0.13; 95% CI −0.42, .16; P = .38; r2 = 0.01) but was associated Requirement for any surgical intervention
with TES at week 48 (B(std) = 0.50; 95% CI .32, .68; P < .001; Twenty-four of the 64 individuals (37.5%) who provided sur­
r2 = 0.25). Week 48 TES was not associated with TES at 3 years gical outcome data required surgery over 3 years of follow-up.
(B(std) = −0.05; 95% CI −.33, .23; P = .71; r2 = 0.002). The decision to perform surgery was left to the local clinical
CAS score also fell progressively over the study period from team and was not protocolized. The majority of interventions
a median baseline of 5 (IQR 4-5) to a 3-year median of 1 (IQR involved lid surgery (n = 16) or orbital decompression (n =
0-1; P = <.001) (Fig. 2C). By week 12, 17.8% of individuals 12), with strabismus correction (squint surgery) being less com­
had a CAS of 0 or 1, rising to 23.8% by 24 weeks 40.8% by mon (n = 2). Five individuals required both lid surgery and de­
week 48 and 58.6% by 3 years. Baseline CAS was not associ­ compression surgery. Eight of 49 respondents (16.33%) had
ated with CAS at 3 years (B(std) = 0.07; 95% CI −.20, .33; undergone thyroidectomy over the 3-year period.
The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10 2619

Table 1. Odds of providing 3 year data

Reason OR 95% CI P ORa,b 95% CIa Pa

Age
<60 years 1 1
≥60 years 2.25 (0.80, 6.28) .12 1.22 (0.39, 3.83) .74
Gender
Male 1 1
Female 1.35 (0.61, 2.99) .46 1.56 (0.63, 3.83) .33
Ethnicity
Caucasian 1 1
Non-Caucasian 0.39 (0.17, 0.86) .02 0.30 (0.11, 0.79) .02

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Thyroid state
Hypothyroid/euthyroid 1 1
Hyperthyroid 1.01 (0.42, 2.39) .99 0.77 (0.29, 2.02) .59
Recent use of steroids
No recent use 1 1
Use in last 6 months 0.83 (0.29, 2.38) .73 0.78 (0.29, 2.07) 0.61
Disease duration
<6 months 1 1
≥6 months 1.26 (0.58, 2.71) .56 1.40 (0.58, 3.41) .46
Smoking status
Nonsmoker 1 1
Smoker 0.57 (0.28, 1.17) .12 0.46 (0.20, 1.07) .07
Study center
Other 1 1
Bristol/Moorfields 0.68 (0.26, 1.77) .43 0.85 (0.27, 2.74) .79
Baseline CAS
2-3 1 1
4-5 0.49 (0.18, 1.38) .18 0.41 (0.13, 1.31) .13
6-7 0.64
Baseline TES 0.96 (0.28, 3.27) .94 (0.15, 2.77) .55
<22 1 1
≥22 0.58 (0.23, 1.51) .27 0.49 (0.16, 1.48) .21
Received Aza 1.07 (0.53, 2.16) .85 0.96 (0.43, 2.15) .92
Received RT 1.47 (0.73, 2.97) .28 1.50 (0.68, 3.29) .32

Sixty-eight individuals provided data; 58 did not.

P = P value against the null hypothesis of no association.
Abbreviations: Aza, azathioprine; CAS, Clinical Activity Score; OR, odds ratio; RT, radiotherapy; TES, Total Eye Score.
a
Adjusted for age group, ethnicity, smoking status, gender, thyroid state, disease duration, study center, recent steroid use, baseline CAS, baseline TES.
b
Four individuals with incomplete baseline CAS excluded from analysis.

Disease duration of greater than 6 months at baseline was as­ Using quartiles of OI we found no evidence of its associ­
sociated with the need for surgery, although with a wide confi­ ation with needing surgery (OR = 1.12; 95% CI .71, 1.77;
dence interval (OR = 16.8; 95% CI 2.95, 95.0; P = .001) P = .63); we also saw no clear evidence with TES (OR =
(Table 3). CAS at baseline assessed as a continuous measure 1.35; 95% CI .87, 2.09; P = .19) (Table 4). Components of
was associated with the need for surgery (B = 0.11; 95% CI OI including presence of diplopia at baseline was also not as­
.02, .20; P = .02) (Table 4). No other baseline variables including sociated with need for surgery (OR = 1.32; 95% CI .78, 2.44;
age, gender, ethnicity, thyroid state, recent use of steroids, smok­ P = .26), nor was proptosis (OR = 1.07; 95% CI .69, 1.65;
ing status, or baseline TES and CAS (split high vs low) predicted P = .76) or palpebral aperture (OR = 1.34; 95% CI .85,
the need for surgery (Table 3) (Supplementary Figs. 1 and 2) (27). 2.12; P = .20). EUGOGO score at baseline did not predict
Although CAS at baseline predicted the need for surgery, the need for surgery overall (Table 4).
prompt reduction in CAS to 0 or 1 by 12 weeks was not asso­
ciated with a reduction in the need for surgery overall (OR =
0.47; 95% CI .09, 2.55; P = .38) or lid surgery (OR = 0.90; Requirement for lid or decompression surgery
95% CI .16, 5.02; P = .90) or orbital decompression (OR = Of the 64 individuals who provided data, 16 needed lid sur­
0.54; 95% CI .06, 4.94; P = .59). gery during the 3 years. The needs for lid and decompression
2620 The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10

Table 2. Effect of azathioprine and radiotherapy on the binary composite clinical outcome, Ophthalmopathy Index, Clinical Activity Score, and
Total Eye Score at 3 years follow-up

Reason OR/B 95% CI P OR/Ba,b 95% CIa Pa

Binary clinical composite outcome improvement (N = 52)

Azathioprine 0.83 (0.21, 3.29) .70 0.51 (0.08, 3.13) .47
Radiotherapy 2.82 (0.64, 12.4) .17 2.17 (0.43, 10.9) .35
Ophthalmopathy Index (N = 49)
Azathioprine 1.36 (−0.35 3.06) .12 1.33 (−0.68, 3.34) .19
Radiotherapy −0.53 (−2.26, 1.21)) .54 0.11 (−1.92, 2.14() .92
Clinical Activity Score (N = 58)
Azathioprine 0.24 (−0.48, 0.96) .51 0.58 (−0.20, 1.36) .14

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Radiotherapy 0.16 (−0.57, 0.88) .66 0.19 (−0.58, 0.98) .61
Total Eye Score (N = 46)
Azathioprine 2.97 (0.31, 5.62) .03 3.11 (0.04, 6.19) .05
Radiotherapy −1.20 (−3.97, 1.57) .39 −0.61 (−3.87, 2.65) .71
EUGOGO classificationa (N = 49)
Azathioprine 1.67 (0.14, 19.8) .69 2.48 (0.14, 44.7) .54
Radiotherapy 0.42 (0.04, 4.94) .49 0.58 (0.03, 10.97) .72

P = P value against the null hypothesis of no association.

Adjusted for age group, ethnicity, smoking status, gender, thyroid state, disease duration, study center, recent steroid use, and baseline Clinical Activity Score.
Abbreviations: B, beta coefficient; EUGOGO, European Group on Graves’ Orbitopathy; OR, odds ratio.
a
Odds of having moderate to severe vs mild on EUGOGO classification.

surgery were also separately assessed. CAS at baseline as a Factors Predicting Improvement in Quality of Life
continuous measure was associated with the need for lid sur­ (GOQOL)
gery (P = .02), as was duration of disease prior to enrollment There was improvement overall after 3 years of follow-up in
(OR = 7.18; 95% CI 1.24, 41.4; P = .03) (Supplementary patients’ perception of their visual function (GOQOL-VF;
Table 2) (27). OI, TES, and EUGOGO score at baseline P < .001) and visual appearance (GOQOL-AP; P = .002)
were not predictive of the need for lid surgery (Table 4). No (Fig. 2D and 2E); however, individual trajectories were vari­
other baseline factors were predictive of the need for lid sur­ able (Fig. 3A). Some individuals showed minimal improvement
gery including diplopia at baseline (OR = 0.94; 95% CI .51, or even deterioration in GOQOL-AP (Fig. 3A), especially
1.73; P = .84). among those who did not undergo surgery (Fig. 3B). Overall,
Twelve of 64 individuals underwent decompression surgery 35% of individuals failed to improve their GOQOL-VF and
within 3 years. Similar to lid surgery, disease duration prior 42.4% failed to improve GOQOL-AP at 3 years (Fig. 3A).
to enrollment of greater than 6 months was predictive of the These percentages were even higher in those who did not
need for orbital decompression (OR = 52.5; 95% CI 4.10, have surgery, with 42.1% failing to improve GOQOL-VF
672; P = .002), but there was no significant association with and 50% failing to improve GOQOL-AP (Figs. 3B and 3C).
baseline CAS score (P = .39) (Table 4). Baseline diplopia was After 3 years, 15 of the 60 people (25%) who provided
also associated with need for orbital decompression (OR = GOQOL-VF scores and 32 of the 60 people (54.2%) who pro­
6.00; 95% CI 1.78, 20.3; P = .004), as was palpebral aperture vided GOQOL-AP at 3 years had a score of 75 or less, indicat­
(OR = 3.18; 95% CI 1.54, 6.57; P = .002) and OI (OR = 2.70; ing ongoing substantial impairment of quality of life. No
95% CI 1.33, 5.47; P = .006), with a possible association with baseline characteristics, potential confounders, or disease lev­
TES, with 5% requiring surgery in the lowest quartile but 37% els at baseline (CAS/TES) were associated with having in­
requiring surgery in the highest quartile (P = .05), (Table 4), but creased odds of having a GOQOL score less than 75 at 3
not degree of proptosis (OR = 0.06; 95% CI .03, .15; P = .16) years (Supplementary Tables 5 and 6) (27).
(Supplementary Table 4) (27). No individuals with mild GO on
baseline EUGOGO score required orbital decompression over
3 years. No other baseline variables predicted the need for de­ Discussion
compression surgery including age, gender, ethnicity, thyroid There is limited outcome data for GO following medical inter­
state, recent use of steroids, smoking status, or baseline TES ventions beyond 2 years. This 3-year report on prospectively
(Supplementary Table 4) (27). acquired subjective and objective assessments in just over
Baseline GOQOL-VF was lower in those who required or­ half of the participants in the CIRTED trial, who were broadly
bital decompression over the 3-year follow-up [median 58.3 representative of the overall study population, therefore adds
(IQR 28.6-78.6) vs those who did not, median 75 (IQR substantially to the current evidence base for long-term treat­
63.4-100); P = .02]. However no clear association was seen ment outcomes in this important cause of visual disability, dis­
between GOQOL-VF and the need for any surgery or lid sur­ figurement, and impaired quality of life. In summary, we
gery. Baseline GOQOL-AP was not predictive of individuals observed no clear benefit beyond the effect of high-dose oral
undergoing any form of surgery (Supplementary Table 5) steroids in individuals randomized to receive azathioprine or
(27). radiotherapy on BCCOM, OI, CAS, and EUGOGO score at
The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10 2621

Table 3. Odds of requiring any surgery by 3 years

Reason OR 95% CI P ORa,b 95% CIa Pa

Age
<60 years 1 1
≥60 years 0.70 (0.26, 3.11) .88 0.64 (0.14, 2.93) .56
Gender
Male 1 1
Female 1.0 (0.31, 3.22) .99 1.92 (0.31, 11.7) .48
Ethnicity
Caucasian 1 1
Non-Caucasian 0.27 (0.05, 1.37) .12 0.08 (0.01, 1.05) .05

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Thyroid state
Hypothyroid/euthyroid 1
Hyperthyroid 10.81 (0.22, 2.90) .74 0.44 (0.08, 2.40) .34
Recent use of steroids
No recent use 1 1
Use in last 6 months 0.82 (0.14, 4.84) .83 0.45 (0.03, 6.61) .56
Disease duration
<6 months 1 1
≥6 months 4.71 (1.50, 14.8) .008 16.8 (2.95, 95.0) .001
Smoking status
Nonsmoker 1 1
Smoker 0.8 (0.31, 2.43) .80 0.76 (0.16, 3.57) .73
Study center
Other 1 1
Bristol/Moorfields 0.71 (0.17, 2.97) .64 0.74 (0.09, 6.02) .11
Baseline CAS
2-3 1 1
4-5 5.72 (0.65, 50.3) .12 6.48 (0.41, 103) .19
6-7 11.6 (1.17, 114) .04 13.6 (0.78, 237) .07
Baseline TES
<22 1 1
≥22 0.81 (0.19, 3.59) .78 1.40 (0.18, 11.0) .75
Received Aza 0.90 (0.33, 2.49) .85 0.39 (0.09, 1.69) .21
Received RT 0.52 (0.19, 1.47) .22 0.34 (0.08, 1.52) .16

Sixty-four individuals provided data—24 had surgery; 40 did not.

Sixteen individuals had lid surgery; 12 individuals required orbital decompression; 2 required squint surgery. Five individuals required both lid surgery and
decompression surgery.
P = P value against the null hypothesis of no association.
Abbreviations: Aza, azathioprine; CAS, Clinical Activity Score; OR, odds ratio; RT, radiotherapy; TES, Total Eye Score.
a
Adjusted for age group, ethnicity, smoking status, gender, thyroid state, disease duration, study center, recent steroid use, baseline CAS, baseline TES.
b
Two individuals with incomplete baseline CAS excluded from analysis.

3 years. Furthermore, for azathioprine we also observed no of outcome measures did suggest radiotherapy reduced diplo­
additional benefit with regard to need for subsequent surgical pia (Supplementary Table 2) (27). Taken together this does
intervention or quality of life measures. This is in contrast to raise the possibility of late benefits from radiotherapy; how­
our original findings at 1 year, which showed a modest poten­ ever, caution is needed over interpretation of these marginal
tial benefit from azathioprine (10). The results for radiother­ benefits given our incomplete patient follow-up and the num­
apy are less clear. In our original study we found no ber of statistical tests being performed, which raises the possi­
evidence of benefit with radiotherapy (10), but in this longer- bility of a type 1 error.
term follow-up we did observe that those randomized to re­ The most striking finding from this 3-year follow-up study
ceive radiotherapy had a higher GOQOL-AP score [median was that quality of life outcomes in GO remain poor, as 35%
81.3 (IQR 62.5-93.8) vs median 68.9 (IQR 31.2-87.5); of individuals failed to improve their GOQOL-VF and 42.4%
P = .03] than those who did not. We also observed a potential failed to improve their GOQOL-AP at 3 years (Fig. 3). This fig­
reduction in the need for lid surgery (OR = 0.17; 95% CI .03, ure was even higher in those patients who did not undergo sur­
1.03; P = .05), although it should be highlighted that the 95% gery. Overall, after 3 years, 25% of patients had a GOQOL-VF
CI does cross equality. Analysis of key individual components of 75 or lower and 54.2% had a GOQOL-AP of 75 or lower,
2622 The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Figure 2. (A) Ophthalmopathy Index at 12 weeks, 48 weeks, and 3 years. (B) TES at baseline, 12 weeks, 48 weeks, and 3 years. (C) CAS at baseline, 12
weeks, 48 weeks, and 3 years. (D) GOQOL Visual Function at baseline, 12 weeks, 48 weeks, and 3 years. (E) GOQOL Visual Appearance at baseline, 12
weeks, 48 weeks, and 3 years. Abbreviations: CAS, Clinical Activity Score; GOQOL, Graves Ophthalmopathy Quality of Life; TES, Total Eye Score.

Table 4. Need for surgery over 3-year follow-up by baseline CAS, OI, and TES

Any Surgery Lid Decompression

No Yes % P No Yes % P No Yes % P

CAS .02 .02 .39

2 5 0 0 5 0 0 5 0 0
3 4 1 20 4 1 20 4 1 20
4 8 7 46.6 11 4 26.7 12 3 20
5 14 7 33.3 18 3 14.3 16 5 23.8
6 6 4 45.5 7 4 36.4 10 1 9.1
7 1 4 80 1 4 80 3 2 40
OI .63 .66 .006
Q1 8 5 38.4 18 5 38.5 13 0 0
Q2 12 5 29.4 14 3 17.7 15 2 11.7
Q3 7 5 41.6 9 3 25.0 8 4 33.3
Q4 11 8 42.1 14 5 26.3 14 5 26.3
TES 0.19 0.46 0.05
Q1 14 6 30 15 5 25 19 1 5
Q2 11 5 31.3 14 2 12.5 12 4 25
Q3 7 5 41.7 8 4 33.3 11 1 8.3
Q4 8 8 50 11 5 31.2 10 6 37.5
EUGOGO classification
Mild 9 6 40 3 3 50 15 33 68.8
Moderate to severe 27 18 40 .99 10 4 28.6 0.36 0 12 100

Abbreviations: CAS, Clinical Activity Score; EUGOGO, European Group on Graves’ Orbitopathy; OI, Ophthalmopathy Index; Q1, quartile 1; Q2, quartile 2;
Q3, quartile 3; Q4, quartile 4; TES, Total Eye Score.

indicating ongoing prolonged substantial impairment in the improvement in physician-conducted GO assessments, in par­
quality of life of these patients despite aggressive medical man­ ticular measures of inflammation (CAS) and overall orbital ef­
agement. This appears to contrast with the substantial fects/deformity (eg, OI and TES) (Fig. 2). Taken together, this
The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10 2623

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

Figure 3. (A) GOQOL Visual Function and Appearance at baseline and 3 years—overall. (B) GOQOL Visual Function at baseline and 3 years by surgical
intervention. (C) GOQOL Visual Appearance at baseline and 3 years by surgical intervention. Abbreviations: GOQOL, Graves Ophthalmopathy Quality of
Life.

indicates there is a substantial unmet need for improving out­ CAS to 0 or 1 in the first 12 weeks was not associated with im­
comes in this visually disabling and cosmetically disfiguring con­ proved outcomes with regard to needing surgery or improving
dition. In keeping with this, we confirm that although CAS is a quality of life, raising the possibility that early control of inflam­
valuable baseline measure (3), change in CAS did not correlate mation—a key part of current treatment approaches for GO—
with key surgical and quality of life outcomes. may not have the impact on long-term orbital remodeling that
In addition, we have been able to quantify the need for sur­ have been predicted (7). This contrasts with the improvements
gery beyond the inflammatory phase of GO. Despite extensive in proptosis as well as quality of life seen with disease-modifying
medical therapy, including high-dose oral (but not intraven­ therapies such as teprotumumab, although this also improves
ous) steroids in all individuals, 37.5% of patients still required CAS substantially (28, 29) and emphazies the need for longer-
surgical intervention, predominantly lid surgery or decom­ term follow-up and assessments of quality of life and responses
pression. This is comparable but somewhat higher than the other than CAS, to ensure benefits translate into improved out­
surgery rates in the study of Leo et al, although it is noted comes for patients.
that some patients in their study declined surgery (14). The An additional finding from our study was the consistent ob­
high requirement for surgery also indirectly highlights the servation that the need for surgery over 3 years was higher in
poor efficacy of oral steroids and/or azathioprine and radio­ those who had thyroid eye disease for more than 6 months be­
therapy in treating thyroid eye disease. It is possible that fore enrolment in CIRTED (OR = 16.8; 95% CI 2.95, 95.0;
more of our patients would also benefit from surgical inter­ P = .001), in particular orbital decompression (OR = 52.5;
vention, especially since a recent meta-analysis has highlighted 95% CI 4.10, 672; P = .002). This emphazies the importance
the patient-reported benefits of surgery (15). of surveillance and early case finding in GO, as in the UK
Whereas baseline CAS was predictive of the need for surgery TEAMeD 5 approach (http://www.btf-thyroid.org/TEAMeD-
overall (particularly lid surgery but not orbital decompression), 5) (30), which supports patients and endocrinologists to detect
it was not associated with quality of life at 3 years. Of note, base­ GO early. More than 80% of cases of GO arise at the same time
line CAS was not associated with CAS later on in the study, and or after the diagnosis of thyrotoxicosis, a period in which pa­
early suppression of inflammation as shown by a reduction in tients are normally under endocrine follow-up (31).
2624 The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10

Our study also highlights that OI and TES are associated with views expressed are those of the authors and not necessarily
need for decompression surgery over 3 years (Table 4), and OI at those of the NIHR or the Department of Health and Social
baseline is still associated with OI 3 years later. This perhaps re­ Care.
flects the robustness of these assessments but also provides evi­
dence of incomplete treatment efficacy in our study. Whereas
baseline CAS, OI, and TES are associated with surgical out­ Disclosures
comes, they do not appear to be able to clearly identify those pa­ The authors have nothing to disclose.
tients who do not need surgery; they are better at identifying
those at highest risk of surgery and identifying individuals who
need the closest attention (Table 4). Although baseline OI and Role of the Funding Source
TES are predictive of identifying those who need surgery, they The sponsor and funders had no role in study design, data col­
are not associated with quality of life at 3 years. Perhaps unsur­ lection, data analysis, data interpretation, or writing of the re­
prisingly, quality of life at baseline is predictive of quality of life at port. P.N.T., R.L., and C.M.D. had full access to all the data
3 years: baseline GOQOL-VF was lower in those who required

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

in the study and had final responsibility for the decision to sub­
orbital decompression over the 3 years [follow-up median score mit for publication.
58.3 (IQR 28.6-78.6)] vs those who did not [median score 75
(IQR 63.4-100) P = .02]. A positive change in GOQOL-VF
was observed in those who received orbital decompression [me­ Data Availability
dian 28.1 (IQR 18.3-36.5) vs 7.13 (IQR 0—25); P = .05] Restrictions apply to the availability of some or all data gen­
(Supplementary Table 6) (27). This suggests that orbital decom­ erated or analyzed during this study to preserve patient confi­
pression in particular improves GOQOL-VF in people with thy­ dentiality or because they were used under license. The
roid eye disease. Taken together, to improve GOQOL outcomes corresponding author will on request detail the restrictions
in GO we need to take more account of the baseline GOQOL-VF and any conditions under which access to some data may be
and GOQOL-AP scores and use these in conjunction to perhaps provided.
better define who might benefit from surgery or consider novel
agents such as teprotumumab.
Although our study is one of the largest in thyroid eye dis­ References
ease with longer follow-up than other studies, there are several
1. Taylor PN, Albrecht D, Scholz A, et al. Global epidemiology of
limitations. The sample size is still modest, which means we hyperthyroidism and hypothyroidism. Nat Rev Endocrinol.
are potentially underpowered to detect important effects 2018;14(5):301-316. https://doi.org/10.1038/nrendo.2018.18
such as identifying clear benefits from surgery in terms of qual­ 2. Bartalena L, Pinchera A, Marcocci C. Management of Graves’ oph­
ity of life scores. The number of statistical tests performed also thalmopathy: reality and perspectives. Endocr Rev. 2000;21(2):
raises the possibility of a type 1 error. As the 3-year assessment 168-199. https://doi.org/10.1210/edrv.21.2.0393
was optional, almost half the study participants were also not 3. Taylor PN, Zhang L, Lee RWJ, et al. New insights into the patho­
evaluated at this time point, although a majority did provide genesis and nonsurgical management of Graves orbitopathy. Nat
data and are representative of the original cohort. Rev ndocrinol. 2020;16(2):104-116. https://doi.org/10.1038/s415
74-019-0305-4
In conclusion, we have identified no clear evidence of benefit
4. Wiersinga WM. Advances in treatment of active,
for the addition of azathioprine to high-dose oral steroid in indi­
moderate-to-severe Graves’ ophthalmopathy. Lancet Diabetes
viduals with moderate to severe GO at 3 years. The results for Endocrinol. 2017;5(2):134-142. https://doi.org/10.1016/S2213-
orbital radiotherapy were less clear, but there may have been pos­ 8587(16)30046-8
sible benefit in the visual appearance domain for quality of life 5. Terwee C, Wakelkamp I, Tan S, Dekker F, Prummel MF, Wiersinga
and a potential reduction in the need for lid surgery. W. Long-term effects of Graves’ ophthalmopathy on health-related
Furthermore, we have identified that 3-year outcomes in GO des­ quality of life. Euro J Endocrinol. 2002;146(6):751-757. https://
pite the medical interventions we used are heterogenous and doi.org/10.1530/eje.0.1460751
poor, particularly for GOQOL-AP. A large proportion of pa­ 6. Dietrich A, Taylor P, White P, et al. Establishing the usefulness of
tients still need surgery, and in those who receive surgery there the GO-QOL in a UK hospital-treated population with thyroid
eye disease in the CIRTED trial. Psychol Health Med. 2018;23­
are potential benefits, particularly for GOQOL-VF. Despite con­
(Suppl 1):1341-1355. https://doi.org/10.1080/13548506.2018.
siderable effectiveness of therapy in improving GO indices such
1503693
as CAS, this is not reflected in quality of life scores, and future 7. Salvi M, Campi I. Medical treatment of Graves’ orbitopathy. Horm
trials of treatment and treatment strategies should reflect this. Metab Res. 2015;47(10):779-788. https://doi.org/10.1055/s-0035-
Our results also strongly suggest that intervening in GO early 1554721
will likely reduce the need for future surgical intervention, al­ 8. Dosiou C, Kossler AL. Thyroid eye disease: navigating the new
though rapid reduction in CAS was not associated with extra treatment landscape. J Endocr Soc. 2021;5(5):bvab034. https://
benefit. Longer-term studies, particularly for promising new doi.org/10.1210/jendso/bvab034
therapies, are also needed, with an emphasis on quality of life 9. Bartalena L, Kahaly GJ, Baldeschi L, et al. The 2021 European
measures and not just objective clinical assessments of GO. Group on Graves’ Orbitopathy (EUGOGO) clinical practice guide­
lines for the medical management of Graves’ orbitopathy. Euro J
Endocrinol. 2021;185(4):G43-g67. https://doi.org/10.1530/EJE-
Acknowledgments 21-0479
10. Rajendram R, Taylor PN, Wilson VJ, et al. Combined immunosup­
V.W. and R.W.J.L. received salary support from the National pression and radiotherapy in thyroid eye disease (CIRTED): a mul­
Institute for Health and Care Research (NIHR) Biomedical ticentre, 2 × 2 factorial, double-blind, randomised controlled trial.
Research Centre at Moorfields Eye Hospital, NHS Lancet Diabetes Endocrinol. 2018;6(4):299-309. https://doi.org/
Foundation Trust, and UCL Institute of Ophthalmology. The 10.1016/S2213-8587(18)30021-4
The Journal of Clinical Endocrinology & Metabolism, 2023, Vol. 108, No. 10 2625

11. Smith TJ, Kahaly GJ, Ezra DG, et al. Teprotumumab for 21. Werner SC. Classification of the eye changes of Graves’ disease. Am
thyroid-associated ophthalmopathy. N Engl J Med. 2017;376­ J Ophthalmol. 1969;68(4):646-648. https://doi.org/10.1210/jcem-
(18):1748-1761. https://doi.org/10.1056/NEJMoa1614949 29-7-982
12. Kahaly GJ, Riedl M, Konig J, et al. Mycophenolate plus methyl­ 22. Werner SC. Modification of the classification of the eye changes of
prednisolone versus methylprednisolone alone in active, Graves’ disease: recommendations of the Ad Hoc Committee of the
moderate-to-severe Graves’ orbitopathy (MINGO): a randomised, American Thyroid Association. J Clin Endocrinol Metab. 1977;
observer-masked, multicentre trial. Lancet Diabetes Endocrinol. 44(1):203-204. https://doi.org/10.1210/jcem-44-1-203
2018;6(4):287-298. https://doi.org/10.1016/S2213-8587(18)3002 23. Terwee CB, Dekker FW, Mourits MP, et al. Interpretation and val­
0-2 idity of changes in scores on the Graves’ ophthalmopathy quality of
13. Perez-Moreiras JV, Gomez-Reino JJ, Maneiro JR, et al. Efficacy of to­ life questionnaire (GO-QOL) after different treatments. Clin
cilizumab in patients with moderate-to-severe corticosteroid-resistant Endocrinol (Oxf). 2001;54(3):391-398. https://doi.org/10.1046/j.
Graves orbitopathy: a randomized clinical trial. Am J Ophthalmol. 1365-2265.2001.01241.x
2018;195:181-190. https://doi.org/10.1016/j.ajo.2018.07.038 24. Terwee CB, Gerding MN, Dekker FW, Prummel MF, Wiersinga
14. Leo M, Marcocci C, Pinchera A, et al. Outcome of Graves’ orbitop­ WM. Development of a disease specific quality of life questionnaire
athy after total thyroid ablation and glucocorticoid treatment: for patients with Graves’ ophthalmopathy: the GO-QOL. Br J

Downloaded from https://academic.oup.com/jcem/article/108/10/2615/7086917 by guest on 31 October 2023

follow-up of a randomized clinical trial. J Clin Endocrinol Metab.
Ophthalmol. 1998;82(7):773-779. https://doi.org/10.1136/bjo.8
2012;97(1):E44-E48. https://doi.org/10.1210/jc.2011-2077
2.7.773
15. Woo T, Li C, Ganesananthan S, et al. The effect of ophthalmic sur­
25. Rajendram R, Lee RWJ, Potts MJ, et al. Protocol for the combined
gery for Graves’ orbitopathy on quality of life: a systematic review
immunosuppression & radiotherapy in thyroid eye disease
and meta-analysis. Thyroid. 2022;32(2):177-187. https://doi.org/
(CIRTED) trial: A multi-centre, double-masked, factorial rando­
10.1089/thy.2021.0411
mised controlled trial. Trials. 2008;9:6-6. https://doi.org/10.1186/
16. Mourits MP, Prummel MF, Wiersinga WM, Koornneef L. Clinical
1745-6215-9-6
activity score as a guide in the management of patients with Graves’
26. Rundle FF, Wilson CW. Development and course of exophthalmos
ophthalmopathy. Clin Endocrinol (Oxf). 1997;47(1):9-14. https://
doi.org/10.1046/j.1365-2265.1997.2331047.x and ophthalmoplegia in Graves’ disease with special reference to
17. Dolman PJ, Rootman J. VISA classification for Graves orbitopathy. the effect of thyroidectomy. Clin Sci. 1945;5(3-4):177-194.
Ophthalmic Plast Reconstr Surg. 2006;22(5):319-324. https://doi. https://doi.org/10.1136/bjo.82.7.773
org/10.1097/01.iop.0000235499.34867.85 27. Taylor PN. Data from: CIRTED 3 Year Data Supplementary
18. Bartalena L, Baldeschi L, Dickinson A, et al. Consensus statement Online Material. Deposited December 26, 2022. https://doi.org/
of the European Group on Graves’ Orbitopathy (EUGOGO) on 10.5281/zenodo.7483891.
management of GO. Euro J Endocrinol. 2008;158(3):273-285. 28. Kahaly GJ. Immunotherapies for thyroid eye disease. Curr Opin
https://doi.org/10.1530/EJE-07-0666 Endocrinol Diabetes Obes. 2019;26(5):250-255. https://doi.org/
19. Bartalena L, Baldeschi L, Dickinson AJ, et al. Consensus statement 10.1097/MED.0000000000000493
of the European Group on Graves’ Orbitopathy (EUGOGO) on 29. Douglas RS, Kahaly GJ, Patel A, et al. Teprotumumab for the treat­
management of Graves’ orbitopathy. Thyroid. 2008;18(3): ment of active thyroid eye disease. N Engl J Med. 2020;382(4):
333-346. https://doi.org/10.1089/thy.2007.0315 341-352. https://doi.org/10.1056/NEJMoa1910434
20. Perros P, Crombie AL, Matthews JN, Kendall-Taylor P. Age and 30. BTF. TEAMeD-5. 2018.
gender influence the severity of thyroid-associated ophthalmop­ 31. Marcocci C, Bartalena L, Bogazzi F, Panicucci M, Pinchera A.
athy: a study of 101 patients attending a combined thyroid-eye clin­ Studies on the occurrence of ophthalmopathy in Graves’ disease.
ic. Clin Endocrinol (Oxf). 1993;38(4):367-372. https://doi.org/10. Acta Endocrinol (Copenh). 1989;120(4):473-478. https://doi.org/
1111/j.1365-2265.1993.tb00516.x 10.1530/acta.0.1200473