Purpose
To compare the risk of developing compressive optic neuropathy in patients with active thyroid eye disease (TED) treated with corticosteroids with or without orbital radiotherapy.
Design
Retrospective single-center case-control study.
Methods
The clinical charts of 351 patients with active TED who received corticosteroids with or without orbital radiotherapy between 1999 and 2010 were reviewed. Patients with compressive optic neuropathy at the time of presentation were excluded. Group 1 received corticosteroids only and Group 2 received corticosteroids as well as orbital radiotherapy. The primary outcome measure was the development of compressive optic neuropathy. Secondary outcome measures were changes in other parameters indicating the activity of TED, including soft tissue inflammation, diplopia, ocular motility restriction, and appearance.
Results
There were 144 cases in Group 1 and 105 in Group 2. Both groups were matched for age, sex, and stability of thyroid function. The 2 groups differed only in the modality of treatment for active TED. The main indication for treatment in both groups was soft tissue inflammation. Corticosteroids were initiated an average of 2.6 months following symptom onset in Group 1 and 2.5 months in Group 2. Group 2 received orbital radiotherapy on average 4.2 months following the initiation of corticosteroid therapy and 8% (9/105) were intolerant to corticosteroids. At an average of 3.2 years follow-up, compressive optic neuropathy had developed in 17% (25/144) of Group 1 and 0% of Group 2 ( P < .0001), on average 5.5 months following the initiation of corticosteroid therapy. Although both groups experienced a significant reduction in periocular inflammation, the radiotherapy-treated group demonstrated a significantly greater improvement in ocular motility.
Conclusion
The rate of compressive optic neuropathy was significantly lower and improvement in ocular motility greater in patients receiving orbital radiotherapy in addition to corticosteroids. Patients with active TED appear to have an effective and sustained response to orbital radiotherapy combined with corticosteroids that is protective against disease progression and the development of compressive optic neuropathy.
Thyroid eye disease (TED) is an autoimmune disorder whose active (progressive) phase is characterized by inflammation and expansion of the orbital fat and extraocular muscles. Treatment of active TED is directed at quelling the immune reaction with the hope of minimizing proptosis, exposure keratopathy, and, in more severe cases, progression to compressive optic neuropathy or extraocular motility impairment. Corticosteroids and adjunctive external beam orbital radiation are two immunomodulators commonly used in the treatment of progressive TED. The efficacy of corticosteroids is attributed to anti-inflammatory and immunosuppressive properties. However, a significant number of patients (20%-40%) respond poorly or partially to corticosteroids, and recurrences are not infrequent on withdrawal and dose reduction. Furthermore their significant side effects can limit the dose and duration of effective therapy.
External beam orbital radiotherapy has been used in the treatment of TED for almost a century. It is presumed to work by inducing terminal differentiation in progenitor fibroblasts, suppressing the downstream consequences of fibroblast activation, and by reducing the secretion of proinflammatory cytokines from activated lymphocytes. Current high-energy linear accelerators provide targeted delivery of radiation to the retrobulbar orbital tissues.
The degree of heterogeneity between published studies of orbital radiotherapy in TED makes it difficult to compare outcomes and perform meta-analyses. According to most studies, orbital radiotherapy is especially effective for soft tissue inflammatory changes and recent extraocular muscle involvement ; however, its effect on progression of TED and in particular the risk of development of compressive optic neuropathy has not been sufficiently investigated. Currently the use of orbital radiotherapy is restricted to patients older than 35 years of age and without systemic vascular disease.
In general, corticosteroids show effect within a day, but the benefit is short-lived. Orbital radiotherapy may not show benefit for several days to weeks, but its effects are longer-lasting. Given this more sustained effect, our a priori hypothesis was that treatment with orbital radiotherapy in addition to corticosteroids was associated with a lower risk of compressive optic neuropathy compared with corticosteroid therapy alone.
Methods
All patients with active TED who received corticosteroids with or without orbital radiotherapy under the care of 2 clinicians (P.J.D. or J.R.) were retrospectively identified from electronic databases at the Vancouver Hospital Eye Care Centre and the BC Cancer Agency between January 1, 1999 and January 1, 2010. The study protocol, to perform a retrospective chart review of all patients meeting the study criteria, was approved by the University of British Columbia Clinical Research Ethics Board and adhered to the tenets of the Declaration of Helsinki (study ID: H11-02430).
The inclusion criteria were all patients with active TED who received corticosteroids with or without orbital radiotherapy and posttreatment follow-up of 12 months or longer. Group 1, the control group, received only corticosteroids. Group 2, in addition to concurrent or previous treatment with corticosteroids, received orbital radiotherapy. Exclusion criteria were compressive optic neuropathy at the time of initial presentation, treatment with orbital radiotherapy alone, a past history of optic neuropathy, previous orbital decompression surgery, or incomplete clinical records.
The following parameters were compared to evaluate how closely the 2 groups were matched: (1) demographics (age, sex); (2) history of diabetes, hypertension, or smoking; (3) stability of thyroid function at the time of treatment; (4) indication(s) for treatment of TED; (5) pretreatment and posttreatment activity and severity of TED; (6) time from presentation of symptoms to receiving corticosteroids and orbital radiotherapy; and (7) corticosteroid treatment regimen (oral/intravenous [IV]/both). Differences in proportions and mean values for continuous, normally distributed data were carried out using the χ 2 test or paired t test and the Mann-Whitney rank-sum test was used for nonparametric data. Statistical tests were performed using Sigmaplot version 12.5 for Windows (Systat Software, Inc, San Jose, California, USA).
The primary outcome measures were: (1) development of compressive optic neuropathy after presentation, based on reduced best-corrected visual acuity of 2 lines or more AND reduced color vision of 2 plates or more on Harvey-Rand-Rittler test AND a relative afferent pupillary defect in asymmetric cases with or without evidence of optic disc swelling and visual field defect; and (2) time from initiation of corticosteroid treatment and/or orbital radiotherapy to development of compressive optic neuropathy.
Secondary outcome measures were: (1) complications related to the effects of treatment and (2) change in other disease-related endpoints, using the VISA Classification system. The VISA classification measures 4 parameters of disease: (1) Vision documents the presence or absence of optic neuropathy (+1 = present, 0 = absent); (2) Inflammation is graded from 0 to 8: orbital pain 0-2, chemosis 0-2, eyelid edema 0-2, conjunctival injection 0-1, eyelid injection 0-1; (3) Strabismus is graded in 2 parts from 0 to 3: S1, diplopia (0 = none, 1 = diplopia with gaze, 2 = intermittent diplopia, and 3 = constant diplopia); S2, degree of ocular restriction is graded from 0 to 3 (measured using the pupillary light reflex 0 = >45 degrees, 1 = 30-45 degrees, 2 = 15-30 degrees, 3 = <15 degrees) ; (4) Appearance is graded 0-3 (1 for mild proptosis or lid retraction, 2 for moderate changes, and 3 for severe exposure keratopathy or globe prolapse).
The definition of active disease was based on the inflammatory score and subjective or objective findings of interval change in the 4 parameters of the VISA system. If the VISA score was less than 4 out of 8, and there was no deterioration, the patient was managed conservatively with cool compresses, nocturnal head elevation, and nonsteroidal anti-inflammatory drugs. In general, if the inflammatory grade was greater than 4 or if there was subjective or objective evidence of progression in inflammation or motility restriction, more aggressive therapy was offered, including oral or intravenous corticosteroids, orbital radiotherapy, and, in refractory cases, immunosuppressives. Orbital decompression was performed only in the postinflammatory phase, except in cases of refractory compressive optic neuropathy.
All patients with a VISA inflammatory score of greater than 4 were immediately given oral corticosteroids (prednisone 50 mg for 2-3 days) to assess the response. If positive, and depending on the severity of inflammation and rapidity of progression, the patient was offered IV corticosteroids (250-500 mg methylprednisolone weekly). If there was an improvement in the inflammatory score, this dose was reduced to the lowest maintenance dose. We offered orbital radiotherapy combined with corticosteroids (using similar doses as for inflammatory indications) as soon as the patient noticed diplopia or we detected ocular motility restriction.
The main indications for offering orbital radiotherapy to patients were the development of restriction in ocular motility, high doses, poor response, or intolerance to corticosteroids. In addition to these indications, disease-modulating agents were given when orbital radiotherapy was relatively contraindicated in patients less than 35 years of age or in the presence of systemic vascular disease.
Results
A total of 351 clinical charts of patients with active TED were reviewed and 249 patients fulfilled the inclusion criteria. One hundred and two patients were excluded from the study (37 had compressive optic neuropathy at the time of initial presentation, 9 were treated with orbital radiotherapy only, 21 had less than 12 months total follow-up, and 35 had incomplete clinical records). Group 1 consisted of 144 patients who received corticosteroids only while Group 2 consisted of 105 patients who, in addition to corticosteroid treatment, received orbital radiotherapy. The average length of follow-up of both groups was 3.2 ± 2.2 years (range 1.1-11.8 years), 3 years in Group 1 and 3.4 years in Group 2. No patient had less than 1.1 years of follow-up, and all had complete clinical records. All orbital radiotherapy patients received a total of 20 Gy in 10 fractions over 2 weeks, with 74% (78/105) receiving a tapering course of oral corticosteroids for a 3-week period during and immediately following orbital radiotherapy to reduce acute exacerbation of inflammation from radiotherapy treatment.
The average age at presentation was 53 ± 12 years (range 21-86 years) and 68% (170/249) of patients were female. There was no statistically significant difference in the demographics of age, sex, stability of thyroid function, smoking habit, or medical comorbidities between Group 1 and Group 2 ( Table 1 ). Table 2 demonstrates the number of patients receiving IV and/or oral corticosteroids in each group. A significantly greater proportion of Group 1 received both IV and oral therapy ( P < .002). In addition, a greater proportion of Group 2 received IV steroids alone or oral steroids alone ( P < .001). However, the total proportion of cases treated with IV steroids in Group 1 (100% (144/144)) and Group 2 (86% (91/105)) was not significantly different ( P = .49). Oral corticosteroids were used in 95% (137/144) of Group 1 and 61.9% (65/105) of Group 2 ( P = .04).
| Group 1 (Corticosteroids Only) | Group 2 (Corticosteroids & Radiotherapy) | P Value | |
|---|---|---|---|
| Number of patients | 144 | 105 | – |
| Mean ± SD age at onset of presenting symptoms (y) | 53 ± 12 Range 24-86 | 53 ± 11 Range 21-80 | .9 |
| Female sex, % (n) | 68% (96) | 71% (74) | .86 |
| Stable thyroid function, % (n) | 46% (66) | 50% (52) | .82 |
| Smoking habit, % (n) | |||
| At presentation | 46% (66) | 40% (42) | .65 |
| In past | 16% (23) | 15% (16) | .97 |
| Diabetes, % (n) | 4% (6) | 2% (2) | .55 |
| Hypertension, % (n) | 18% (26) | 16% (17) | .88 |
| Group 1 (Corticosteroids Only) | Group 2 (Radiotherapy & Corticosteroids) | P Value | Compressive Optic Neuropathy Group | |
|---|---|---|---|---|
| Number of patients | 144 | 105 | – | 25 |
| Corticosteroids | ||||
| Total (%) | 144 (100%) | 105 (100%) | .93 | 25 (100%) |
| Intravenous (%) | 144 (100%) | 91 (87%) | .49 | 25 (100%) |
| Intravenous alone (%) | 7 (5%) | 40 (38%) | <.001 | 4 (16%) |
| Oral (%) | 137 (95%) | 65 (62%) | .04 | 21 (84%) |
| Oral alone (%) | 0 | 14 (13%) | <.001 | 0 |
| Intravenous and oral (%) | 137 (95%) | 51 (49%) | .002 | 21 (84%) |
| Average total (range) cumulative dose of intravenous corticosteroids (g) | 4.3 (2-12) | 4.5 (2-9) | .3 | 6.5 (4-12) P < .001 |
| Disease-modifying drugs (%) | 32 (22%) | 12 (11.4%) | .09 | 12 (48%) P < .001 |
| Orbital decompression (%) | 58 (40%) | 31 (29%) | .28 | 19 (76%) P < .001 |
| Urgent orbital decompression | 5 (8.6% (5/58)) | 0 | .009 | 5 (26%) P < .001 |
| Time from presentation to orbital decompression (mo) | 9.6 | 10.1 | .9 | 4.3 P = .07 |
| Strabismus surgery (%) | 46 (32%) | 41 (39%) | .5 | 11 (44%) P = .45 |
The indications for corticosteroid treatment were: (1) periocular soft tissue inflammation, 80% in Group 1 (115/144) and 90% in Group 2 (95/105), P = .57; (2) strabismus, 44% in Group 1 (63/144) and 88% in Group 2 (92/105), P = .001; (3) intolerance to corticosteroids, 0% in Group 1 and 8.6% in Group 2 (9/105), P = .002.
The main indications for offering orbital radiotherapy to patients in Group 2 were: (1) the development of significant restriction in ocular motility (88% (92/105)); (2) total cumulative dose of corticosteroids reaching unsafe levels over 8 g (17% (18/105)); (3) intolerance to corticosteroids (8.6% (9/105)); and (4) inadequate control of disease activity with corticosteroids (6.5% (7/105)). Indications for offering other disease-modulating agents were similar to orbital radiotherapy: (1) high total cumulative dose of corticosteroids (14% (35/249)); (2) inadequate control of disease activity with corticosteroids, especially if ocular motility was not affected (4% (10/249)); or (3) where orbital radiotherapy was relatively contraindicated in patients less than 35 years of age or in the presence of systemic vascular disease (3%).
Corticosteroids were commenced an average of 2.6 months (Group 1) and 2.5 months (Group 2) after initial presentation ( P = .98). Group 2 received orbital radiotherapy on average 4.2 months following the initiation of corticosteroid treatment or 7.4 months (range 2.8-24 months) after initial presentation. One hundred percent of Group 1 (144/144) and 86.7% of Group 2 (91/105) received IV corticosteroids. The mean cumulative dose of IV corticosteroids received by Group 1 was 4.3 g (range 2-12 g) and Group 2 was 4.5 g (range 2-9 g). Owing to side effects, 4% (6/144) of Group 1 and 8% (8/105) of Group 2 discontinued corticosteroids. A greater number of patients (7/249) in Group 1 (32/144) were treated with other disease-modulating agents compared to Group 2 (13/105) ( P = .09), including cyclosporine and methotrexate ( Table 2 ). In Group 1 there was a significantly greater average cumulative dose of IV corticosteroids among those receiving disease-modulating agents (5.2 g) compared to those that did not (3.4 g) ( P = .02).
After an average follow-up of 3.2 years, compressive optic neuropathy developed in 17% (25/144) of patients in Group 1 and 0% in Group 2 ( P < .001). Compressive optic neuropathy developed an average of 5.5 months following the initiation of corticosteroid treatment. All patients with compressive optic neuropathy were recorded as having reduced best-corrected visual acuity of 2 lines or more AND reduced color vision of 2 plates or more on the Harvey-Rand-Rittler test AND a relative afferent papillary defect in asymmetric cases, with or without evidence of optic disc swelling (36% (9/25)) and a visual field defect (performed in 56% (14/25)). Nine patients were excluded from the study on the basis that they had received orbital radiotherapy only (without corticosteroids); none of these cases developed compressive optic neuropathy. Orbital radiotherapy was given to Group 2 at an average of 6.7 months after presentation and compressive optic neuropathy developed in Group 1 at an average of 8.1 months after presentation.
All 25 patients with compressive optic neuropathy had received IV corticosteroids and 21 of these (84%) had also received oral steroids ( Table 2 ). Twelve of the 25 patients with compressive optic neuropathy (48%) were also receiving disease-modulating agents and none were intolerant to corticosteroids. The mean cumulative dose of IV corticosteroids received by patients who developed compressive optic neuropathy was 6.5 g (range 4-12 g). This dose was significantly greater than that received by patients who did not develop compressive optic neuropathy ( P < .001). There were no diabetic patients among those who developed compressive optic neuropathy; 4% (6/144) of Group 1 and 2% (2/105) of Group 2 were diabetic ( P = .55). Eighteen percent (26/144) of Group 1 and 16.2% (17/105) of Group 2 were hypertensive. There were significantly more hypertensive patients (36% (9/25)) among those who developed compressive optic neuropathy ( P = .08). The average age of patients who developed compressive optic neuropathy was slightly greater compared with the rest of the cohort (56 vs 53 years), but this was not statistically significant ( P = .36), and fewer patients with compressive optic neuropathy had stable thyroid function at the time of presentation compared with the rest of the cohort (36% (9/25) vs 48% (108/224)) ( P = .15).
There was no significant difference in the need for subsequent orbital decompression surgery between Group 1 (40% (57/144)) and Group 2 (29% (31/105)), P = .28 ( Table 2 ). Of the 58 patients who underwent orbital decompression in Group 1, urgent decompression (within 2-4 weeks of developing compressive optic neuropathy or vision-threatening exposure keratopathy) was performed in 5 cases (8.6% (5/58)). All 5 cases were among those who developed compressive optic neuropathy, making the rate of urgent orbital decompression in this group 26% (5/19) (range 3-28 days). There were no cases of urgent decompression in Group 2. All cases presenting with compressive optic neuropathy, which may have undergone urgent decompression, were excluded from the study. Strabismus surgery was performed in 31.9% (46/144) in Group 1 and 39% (41/105) in Group 2 ( P = .5).
No patients in Group 2 developed orbital radiotherapy-induced retinopathy, optic neuropathy, or orbital malignancy up to the time of final follow-up. The proportion of patients in each group with side effects associated with a full course of corticosteroid treatment was similar; overall, 56% (139/249) of both groups reported side effects including hyperglycemia (15% (37/249)), insulin resistance (10% (25/249)), diabetes mellitus (2% (5/249)), osteoporosis (8% (20/249)), anxiety (6% (15/249)), depression (3% (7/249)), cataract (11% (28/249)), hypertension (14% (35/249)), dyspepsia (10% (25/249)), colitis (1% (3/249)), arrhythmia (0.4% (1/249)), and amenorrhea (0.4% (1/249)).
Table 3 summarizes the disease activity and severity scores at presentation and 6 months following treatment. Both groups experienced a statistically significant reduction in periocular inflammation; however, this was not significantly different between the 2 groups ( P = .19). Group 2 demonstrated a significantly greater improvement in diplopia ( P = .003) and restriction in motility ( P = .07) as compared with Group 1. There was no significant change of appearance (proptosis/eyelid retraction) in either group ( P = .9).
