Predictors of Response to Transantral Orbital Decompression in Severe Graves' Ophthalmopathy

mayo Predictors of Response to Transantral Orbital Decompression in Severe Graves' Ophthalmopathy VAHAB FATOURECHI, M.D., CHARLES W. BEATTY, M.D.,

ERIK J. BERGSTRALH, M.S.,

JAMES A. GARRITY, M.D.,

• Objective: To identify factors that may help predict the outcome after transantral orbital decompression in Graves' ophthalmopathy. • Design: A retrospective study was conducted of 428 patients who had undergone an initial transantral orbital decompression for severe Graves' ophthalmopathy at the Mayo Clinic between November 1969 and May 1989. • Material and Methods: With use of logistic regression analysis, we assessed the preoperative characteristics, the early postoperative results, and the followup questionnaire data (obtained a median of 9.5 years postoperatively) from 304 female and 124 male patients with Graves' ophthalmopathy who had undergone transantral orbital decompression at a median age of 53 years. • Results: On multi varia te stepwise regression analysis, young age, male sex, and long duration of eye symptoms were predictors of severe initial proptosis (P<0.001). The only independent predictors of greater postoperative recession of proptosis were severity of initial proptosis and longer interval between opera-

Standard treatments of severe Graves' ophthalmopathy include surgical orbital decompression, orbital radiotherapy, and immunosuppression.1 For selection of appropriate therapy for a particular patient, knowledge of the likely response to that treatment is necessary. Although some information is available about predictors of response to orbital irradiation and corticosteroid therapy in Graves' ophthalmopathy,2-3 the published data about predictors of outcome after surgical orbital decompression are meager. Of the numerous methods of orbital decompression, selecFrom the Division of Endocrinology/Metabolism and Internal Medicine (V.F., C.A.G.), Section of Biostatistics (E.J.B., K.P.O.), Department of Ophthalmology (J.A.G., G.B.B.), and Department of Otorhinolaryngology (C.W.B.), Mayo Clinic Rochester, Rochester, Minnesota. Address reprint requests to Dr. Vahab Fatourechi, Division of Endocrinology/Metabolism, Mayo Clinic Rochester, 200 First Street SW, Rochester, MN 55905. Mayo Clin Proc 1994;69:841-848

GEORGE B. BARTLEY,

M.D.,

KENNETH P. OFFORD, M.S., AND COLUM A. GORMAN, M . B . , B . C H .

tion and postoperative examination (P<0.001). Patients with the most reduction of proptosis had the greatest improvement in visual acuity but more chance for postoperative development of continuous diplopia. Failure of prior corticosteroid or orbital radiation therapy did not affect the degree of recession of proptosis or improvement in visual acuity. On multi variate analysis for predictors of long-term overall patient satisfaction, only young age of the patient was of borderline significance (P = 0.05), and the only significant predictor of satisfaction with the postoperative eye appearance was an operation done primarily for cosmetic purposes (P = 0.012). • Conclusion: Although various factors may influence the outcome of orbital decompression in patients with Graves' ophthalmopathy, this study showed that the more pronounced the initial proptosis, the greater the degree of recession postoperatively. A higher degree of reduction of proptosis is associated with better visual acuity but also a greater likelihood of development of continuous diplopia. {Mayo Clin Proc 1994; 69:841-848)

tion of a procedure can be tailored to the patient's individual needs. For ease of operation, elimination of external incisions, and overall efficacy, we use the transantral route most frequently. In this retrospective study based on our extensive experience with transantral orbital decompression for Graves' ophthalmopathy,4 we identified the preoperative attributes associated with the success or failure of transantral decompression. PATIENTS AND METHODS Study Subjects.—Between November 1969 and May 1989, 428 patients underwent an initial transantral orbital decompression for severe Graves' ophthalmopathy at the Mayo Clinic. The indications for decompression were optic neuropathy in 50.7%, orbital congestive changes in 27.1 %, reduction of proptosis before eye muscle operation in 8.4%, a disfiguring proptosis in 7.9%, exposure keratitis in 841

© 1994 Mayo Foundation for Medical Education and Research

842

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

Mayo Clin Proc, September 1994, Vol 69

Fig. 1. A and B, Diagrams of transantral approach to removal of floor and medial wall of orbit in Graves' ophthalmopathy. (From Gorman and associates.5 By permission of the New England Journal of Medicine.) 4.7%, and severe corticosteroid-related side effects5 in 1.2%. The group was composed of 304 female and 124 male patients; their median age at orbital decompression was 53 years (range, 16 to 79). Therapies used before decompression included corticosteroids in 45% and orbital irradiation in 4%. Dermopathy was present in 12% of patients, and 13% had never had hyperthyroidism preoperatively. Most patients (88%) with hyperthyroidism had received radioactive iodine treatment, although some had received antithyroid therapy (37%) or had undergone thyroidectomy (13%) before therapy with radioactive iodine. Procedures.—Preoperative examination included determination of visual acuity, ocular motility, and eyelid position, slit-lamp biomicroscopy, funduscopy, and Krahn exophthalmometry. The method of transantral orbital decompression has been described previously.56 Briefly, the floor and the medial wall of the orbit were approached through a Caldwell-Luc antrostomy and were removed to allow expansion of the orbital contents into the opened sinuses (Fig. 1). Postoperatively, the eyes were examined within 6 months in 89% of the patients. Examination more than 6 months after operation was performed in 70% of the patients. Long-term follow-up by questionnaire was accomplished in 84% of the patients at a median of 9.5 years postoperatively. The baseline characteristics of the patients, operative results, complications of operation, short- and long-term results, and results of the patients' assessment of

the long-term benefits after total rehabilitation have been previously described.4 Statistical Analysis.—Statistical independence between observations was achieved by analyzing data collectively for each patient rather than for each individual eye. For proptosis, this approach involved the use of the mean Krahn reading for the two eyes. For visual acuity, a scoring system from 0 through 5 was arbitrarily devised, as follows: 0 = both eyes 20/20; 1 = only one eye 20/20; 2 = both eyes 20/25 to 20/40; 3 = one eye 20/25 to 20/40 and the other eye worse; 4 = one eye 20/50 to 20/100 and the other eye the same or worse; and 5 = both eyes 20/200 or worse. Predictors of improvement in papilledema, exposure keratitis, and visual field defects were not analyzed because more than 90% of patients who had these conditions before orbital decompression had complete resolution or substantially decreased severity after decompression.4 Univariate and multivariate stepwise linear regression was used to select a subset of independent predictors of extent of initial proptosis and reduction in proptosis after decompression. Predictors of initial visual acuity score, improvement in acuity after decompression (no or yes), worsening of acuity (no or yes), development of diplopia uncorrectable by prisms (no or yes), long-term patient satisfaction with eye appearance (5-point scale), and long-term overall satisfaction (no or yes) were analyzed with use of univariate and multivariate ordinal logistic regression. A stepwise backward algorithm was used. All variables were entered into the model at the

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

843

Table 1.—Overview of Factors Examined for Associations With Various End Points in 428 Patients Who Underwent Transantral Orbital Decompression for Graves' Ophthalmopathy* Factor Male sex Age at time of TAD (yr) Interval from eye symptoms to TAD (yr) Prior treatment Corticosteroids Retrobulbar injection of corticosteroids Orbital irradiation History of hyperthyroidism before TAD Dermopathy at time of TAD Thyroid function status at time of TAD Hypothyroid Euthyroid Hyperthyroid Primary indication for TAD ON No ON; preliminary to muscle or eyelid operation No ON; cosmetic reason No ON; inflammatory changes or corticosteroid-related side effects Findings on pre-TAD examination Krahn reading (mm) Keratitis Scotoma Diplopia score 0 = none 1 = only when tired 2 = on deviant gaze only 3 = correctable with prisms 4 = continuous, not correctable with prisms Acuity score 0 = both eyes 20/20 1 = only one eye 20/20 2 = both eyes 20/25 to 20/40 3 = one eye 20/25 to 20/40, other eye worse 4 = one eye 20/50 to 20/100, other eye same or worse 5 = both eyes 20/200 or worse Interval from TAD to early post-TAD eye examination (days)t Interval from TAD to questionnaire

(yr)t

No. with data

% with factor

428 428

29.0

Mean

SD

51.2

13.2

16-79

3.5

4.7

0.04-33

24.8

3.1

16.5-34.0

380

60.8

47.1

0-182

360

9.4

5.5

0.5-20.9

428 428

44.9

428 428

3.7 4.2

428 428

86.7 12.2

428

428

Range

4.2 88.6 7.2 50.7 13.1 7.9 28.3

423 423 425 424

31.7 44.5 33.3 0.9 24.1 4.7 37.0

426

27.7 24.2 18.3 16.4 9.6 3.8

*ON = optic neuropathy; TAD = transantral orbital decompression. flncluded only for the end points of change in Krahn exophthalmometry reading and change in acuity. ^Included only for the end point of patient satisfaction, as measured on the follow-up questionnaire. first step. The most nonsignificant variable was then removed, and the model was refitted. Again, the most nonsignificant factor was removed, and the process was repeated

until all remaining variables were significant. The variables examined as predictive factors are listed in Table 1. All statistical tests were two-sided. Because of the numerous

844

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

predictors and end points examined in this retrospective study, only P values of 0.01 or less were considered statistically significant. RESULTS Proptosis and Recession ofProptosis.—Extent (in millimeters) of proptosis before orbital decompression was univariately correlated (/"<0.001) with various factors (Table 2). Multivariate analysis showed only male sex, young age, and long duration of eye symptoms to be independent predictors of initial proptosis. The mean recession of proptosis, as measured within 6 months postoperatively, was 4.7 ± 2.15 mm (range, -2 to 11). Young age, absence of optic neuropathy, increased initial Krahn reading, and longer interval to postoperative eye examination were all univariately associated with more recession of proptosis (Table 3). On multivariate analysis, only severity of initial proptosis and later date of eye examination remained as independent predictors of greater recession of proptosis postoperatively. Prior failure of corticosteroid therapy, retrobulbar injection of corticosteroids, and orbital irradiation were not significantly associated with the degree of recession of proptosis. Visual Acuity.—The initial visual acuity score (mean ± SD) was 1.67 ± 1.46 (range, 0 to 5). The acuity score was 0 (20/20 in both eyes) for 28% of the patients, 1 (20/20 in only one eye) for 24%, 2 (20/25 to 20/40 in both eyes) for 18%, 3 or 4 for 26%, and 5 (20/200 or worse in both eyes) for 4%. The univariate predictors of poor initial visual acuity scores are shown in Table 4. On multivariate analysis, older age, absence of a history of hyperthyroidism, optic neuropathy, and scotoma remained as independent predictors of poor initial visual acuity. The change in visual acuity after orbital decompression (for the 362 patients with both preoperative and postoperative evaluations within 6 months) was analyzed separately for the 181 with 20/20 vision in at least one eye initially (acuity scores of 0 or 1) and for the other 181 patients with poorer initial vision (scores of 2, 3, 4, or 5). For those with good initial vision, the outcome was considered a failure relative to acuity if they had a "worse" (higher) postoperative score. This result occurred in 35 of the 181 patients (19%) with initial scores of 0 or 1. In 29 of these patients, the score worsened from 0 to 1 or from 1 to 2. The other six patients had larger decreases in acuity: from 0 to 2 in three patients, from 0 to 4 in one patient, and from 0 to 3 and 0 to 4 in one patient each. No significant predictors of a worse postoperative visual acuity score were noted among the 181 patients with initial scores of 0 or 1. Among the 181 patients with initial visual acuity scores of 2, 3,4, or 5, 108 (60%) had a "better" (lower) postoperative score, 65 (36%) had a score that remained the same, and 8

Table 2.—Predictors of Severity of Initial Preoperative Proptosis in 428 Patients Who Underwent Transantral Orbital Decompression for Graves' Ophthalmopathy Factor Sext Male Female Age (yr)t <50 50-59 >60 Duration of eye symptoms (yr)t <1.0 1.0-1.9 >2.0 Optic neuropathy No Yes Scotoma No Yes Acuity score 0,1* 2, 3,4, 5

No. of patients*

Krahn (mm, mean + SD)

122 301

25.8 + 2.87 24.5 ± 3.06

•cO.001

170 145 108

25.4 ± 2.89 "] 24.9 + 3.18 > 24.0 + 3.03 J

<0.001

133 98 188

23.8 ± 2.96 "I 24.6 + 2.99 > 25.8 ± 2.90 J

<0.001

211 212

25.5 ± 2.57 24.2 + 3.36

<0.001

236 186

25.5 + 2.69 24.113.34

<0.001

221 201

25.4 + 2.97 24.3 + 3.09

<0.001

P

*In some study patients, data for certain factors were not recorded. fSelected (on the basis of stepwise regression) as independent predictor of severity of initial proptosis. 120/20 in both eyes or one eye, respectively. (4%) had a worse score. The only significant predictor of an improved score after orbital decompression in these patients was a worse initial visual acuity score (P = 0.005). To determine whether the amount of recession of proptosis was related to improvement in acuity, we added the variable of change in Krahn exophthalmometry reading to the model, including the initial acuity score, and found a nearly significant (P = 0.013) effect. Among the 57 patients with proptosis recession of less than 4 mm, 44% had a better visual acuity score after decompression, whereas 71% of the 105 with proptosis recession of 4 mm or more had a better score (19 patients did not have postoperative Krahn readings). To determine whether failure to respond to other previous therapies had any effect on improvement in visual acuity in the subgroup of 217 patients whose primary indication for orbital decompression was optic neuropathy, we analyzed the improvement in or worsening of visual acuity after orbital decompression. In this group, 47% (101 patients) had received only corticosteroids before transantral orbital decompression, 5% (10) received prior orbital irradiation and corticosteroid therapy, and 46% (99) had no prior irradiation or corticosteroid therapy. In these three groups of patients, no statistically significant difference in improvement of vi-

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

845

Table 3.—Predictors of Recession of Proptosis After Transantral Orbital Decompression in 428 Patients With Graves' Ophthalmopathy

Factor Age (yr) <50 50-59 >60 Optic neuropathy No Yes Initial Krahn reading (mm)t <24 24-26 >27 Time to postoperative eye examination (days)t <90 90-182

No. of patients*

Krahn reading Change (mm, Decrease of >4 mm (%) mean ± SD)t

131 124 86

5.1 ±2.03 4.5 ±2.16 4.2 ±2.21

75

165 176

5.0 ±2.08 4.4 ±2.19

72 61

112 126 103

4.1 ±2.22 4.6 ±1.90 5.3 ±2.22

56] 68 >

240 101

4.3 ±1.97 5.5 ±2.36

60 80

61

1f

60 J

75 J

P

<0.001

0.014

<0.001

<0.001

*In some study patients, data for certain factors were not recorded. •[Calculated as preoperative reading minus early (within 6 months) postoperative reading. iSelected (on the basis of stepwise regression) as independent predictor of recession of proptosis. sion was found for various categories of initial visual acuity (Fig. 2). The other seven patients received retrobulbar injections of corticosteroids before operation. Postoperative Diplopia.^Data for preoperative and postoperative (within 6 months) evaluation of diplopia were available for 370 patients. Preoperatively, 116 patients had no diplopia, 93 had transient diplopia or diplopia on deviant gaze, 18 had diplopia correctable with prisms, and 143 had continuous diplopia that was uncorrectable with prisms. Transantral orbital decompression did not correct diplopia. In. fact, among the 227 patients with no diplopia, transient diplopia, or correctable diplopia preoperatively, 120 (53%) had continuous diplopia postoperatively. The only significant predictor for the development of continuous diplopia was the initial level of diplopia. Continuous diplopia developed postoperatively in 41% of the 116 patients with no preoperative diplopia, 63% of the 93 with transient preoperative diplopia or diplopia on deviant gaze only, and 72% of the 18 with correctable diplopia. The extent of reduction of proptosis was significantly (P - 0.001) associated with the development of continuous diplopia, even when adjustments were made for the initial level of diplopia. Continuous diplopia developed in 37% of patients with recession of proptosis of less than 4 mm and in 59% of those with recession of 4 mm or more. Long-Term Satisfaction of Patients.—On a follow-up questionnaire, patients were asked whether they were satisfied (no or yes) with their eye status. Of the 288 patients for

whom data were available for this question, 76% indicated that they were satisfied. Items on the questionnaire that were associated with overall satisfaction were inquiries about comfortable eyes, absence of diplopia, and good vision in at least one eye. When the factors listed in Table 1 were used as possible predictors (along with reduction of proptosis), only age at the time of orbital decompression had a weak (P = 0.05) association with overall satisfaction. The satisfaction rate was 81% for patients younger than 50 years of age at the time of orbital decompression and 71% for those 50 years old or older. The questionnaire also asked patients to classify the "appearance" of their eyes on a 5-point scale, in which 1 = very satisfied, 3 = acceptable, and 5 = very dissatisfied. Data for this question were available for 357 patients; 53% were very satisfied or satisfied, 36% indicated that their appearance was acceptable, and 11% were dissatisfied or very dissatisfied. Analysis of the factors listed in Table 1 also showed that the only significant predictor of satisfaction with eye appearance was an operation done for purely cosmetic purposes (P - 0.012). Among the 34 patients with a cosmetic indication for orbital decompression, 69% were satisfied or very satisfied, 31 % found their eye appearance acceptable, and none was dissatisfied. DISCUSSION Various factors may influence the outcome of Graves' ophthalmopathy, including the severity of the autoimmune

846

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

Table 4.—Predictors of Initial Preoperative Acuity Score in 428 Patients Who Underwent Transantral Orbital Decompression for Graves' Ophthalmopathy Factor Age (yr)t <50 50-59 >60 Duration of eye symptoms (yr) <1.0 1.0-1.9 >2.0 Prior systemic administration of corticosteroids No Yes Prior orbital irradiation No Yes History of hyperthyroidismf No Yes Indication for decompression! Optic neuropathy No optic neuropathy Inflammatory changes^ Preliminary to muscle or eyelid operation Disfiguring proptosis Scotomat No Yes Initial Krahn reading (mm) <24 24-26 _>27

No. of patients*

121

Acuity score >2 (%)

P

171 145 109

35") 49 >

<0.001

136 101 189

641 52

<0.001

236 190

44 53

0.002

408 18

47 67

0.011

57 369

68 45

<0.001

215 211

68 27 34

<0.001

27 6

<0.001§ <0.001§

235 189

30 70

<0.001

143 156 123

42

56 34

69 J

34 J

<0.001§

61

1f

40 J

<0.001

*In some study patients, data for certain factors were not recorded. tSelected (on the basis of stepwise regression) as independent predictor of initial acuity score. Ilncluded 5 patients without optic neuropathy whose primary indication was corticosteroid-related side effects (none of whom had an acuity score of >2). §In comparison with other indications. process, types of therapy for hyperthyroidism, and therapeutic interventions for the ophthalmopathy itself.7'22 Of the various treatment modalities available, orbital decompression—and particularly transantral orbital decompression—is one of the apparently highly effective surgical procedures.4·5 Selection of patients for transantral orbital decompression should include consideration of the primary value of the procedure—that is, the expansion of space in the bony orbit and the consequent improvement in corneal exposure and reduction in pressure on the optic nerve. Other effects are secondary and may vary, depending on factors that cannot be modified by decompression. For example, the presence of

optic nerve pallor indicates that improvement may not occur, regardless of the recession in proptosis. In our current study, we found that (1) the more severe the initial proptosis, the greater the degree of recession to be expected, (2) within 6 months after operation, the longer the interval to postoperative examination, the better the results of recession of proptosis, and (3) a higher degree of reduction of proptosis, although associated with better visual acuity, predicts a greater chance of development of new and continuous diplopia. The better results of recession of proptosis after a longer interval to postoperative examination suggest that some additional reduction of proptosis may

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

847

100

Q. O +-» CO

o

Û-

20/20

20/25

20/30 20/40 20/50-400 20/>400

Initial visual acuity Fig. 2. Changes in visual acuity within 6 months after transantral orbital decompression in 210 patients with optic neuropathy of Graves' disease. Bars indicate percentages of eyes that had improvement or worsening in vision after decompression. For each visual acuity category, percentages of eyes that had improvement in visual acuity are shown above zero line, and percentages that had a further decrease are shown below zero line, n = total number of eyes for each initial acuity group. (Only eyes with both preoperative and postoperative visual acuity measurements were included.) occur beyond 6 months. In our prior report,4 we suggested that a mean recession of only approximately 0.9 mm occurs beyond 6 months postoperatively. iWe also failed to find a difference in the outcome of surgical decompression between patients who had hyperthyroidism at the time of orbital decompression and patients who had either hypothyroidism after receiving radioactive iodine and were being treated or patients who had untreated hypothyroidism. Some investigators have suggested that thyroid dysfunction and persistence of hyperthyroidism indicate a worse prognosis for patients with ophthalmopathy.23 Euthyroid Graves' disease has been reported to be a predictor of poor prognosis in patients who receive orbital irradiation.2 In our study, the duration of eye symptoms was also not a predictor of response to orbital decompression. Similar findings have been reported in other studies of patients treated with corticosteroids.3 On univariate analysis, patients with optic neuropathy had less severe proptosis than did those without such neuropathy. The most likely explanation for this observation is that proptosis is a type of decompression. Nevertheless in

some patients, an anatomically narrow outlet of the orbit and possibly increased rigidity of orbital tissues may prevent forward displacement of the globe and result in compressive optic neuropathy and congestive signs and symptoms. Orbital decompression has been used not only as a primary mode of therapy but also as treatment after failure of corticosteroid or orbital radiation therapy.5·24"27 In our current study, the groups with and without previous corticosteroid or orbital radiation therapy had similar reduction in proptosis and improvement in visual acuity after transantral decompression. Patient satisfaction with eye appearance and general eye status was also unaffected by the prior use of systemic corticosteroids or orbital irradiation. Because of the small number of patients, however, the data related to irradiation should be interpreted with caution. In general, younger patients had better results for recession of proptosis and long-term satisfaction. Patients who underwent operation primarily for cosmetic purposes had higher degrees of satisfaction with the appearance of the eyes than patients who had other indications for transantral orbital decompression.

848

Mayo Clin Proc, September 1994, Vol 69

ORBITAL DECOMPRESSION IN GRAVES' OPHTHALMOPATHY

REFERENCES 1. Bahn RS, Gorman CA. Choice of therapy and criteria for assessing treatment outcome in thyroid-associated ophthalmopathy. Endocrinol Metab Clin North Am 1987; 16:391-407 2. Petersen IA, Kriss JP, McDougall IR, Donaldson SS. Prognostic factors in the radiotherapy of Graves' ophthalmopathy. Int J Radiât Oncol Biol Phys 1990; 19:259-264 3. Wiersinga WM, Smit T, Schuster-Uittenhoeve ALJ, van der Gaag R, Koornneef L. Therapeutic outcome of prednisone medication and of orbital irradiation in patients with Graves' ophthalmopathy. Ophthalmologica 1988; 197:75-84 4. Garrity JA, Fatourechi V, Bergstralh EJ, Bartley GB, Beatty CW, DeSanto LW, et al. Results of transantral orbital decompression in 428 patients with severe Graves' ophthalmopathy. Am J Ophthalmol 1993;116:533-547 5. Gorman CA, DeSanto LW, MacCarty CS, Riley FC. Optic neuropathy of Graves's disease: treatment by transantral or transfrontal orbital decompression. N Engl J Med 1974; 290:70-75 6. Garrity JA, McCaffrey TV, Gorman CA. Compression and decompression of orbital contents in Graves' ophthalmopathy. Acta Endocrinol (Copenh) 1989; 121(Suppl2):160-168 7. Shine B, Fells P, Edwards OM, Weetman AP. Association between Graves' ophthalmopathy and smoking. Lancet 1990;335:1261-1263 8. Kahaly G, Moncayo R, Stover C, Beyer J. Relationship of eye muscle antibodies with HLA phenotypes and thyroidstimulating immunoglobulins in endocrine orbitopathy. Res ExpMed 1991; 191:137-144 9. Atta HR, McCreath G, McKillop JH, Forrester JV, Thomson JA, Wilson R, et al. Ophthalmopathy in early thyrotoxicosis—relationship to thyroid receptor antibodies and effects of treatment. Scot Med J 1990;35:41-44 10. van der Gaag R, Wiersinga WM, Koornneef L, Mourits MP, Prummel MF, Berghout A, et al. HLA-DR4 associated response to corticosteroids in Graves' ophthalmopathy patients. J Endocrinol Invest 1990;13:489-492 11. Weetman AP, Zhang L, Webb S, Shine B. Analysis of HLADQB and HLA-DPB alleles in Graves' disease by oligonucleotide probing of enzymatically amplified DNA. Clin Endocrinol 1990;33:65-71 12. Werner SC, Feind CR, Aida M. Graves's disease and total thyroidectomy: progression of severe eye changes and decrease in serum long acting thyroid stimulator after operation. N Engl J Med 1967;276:132-138 13. Pequegnat EP, Mayberry WE, McConahey WM, Wyse EP. Large doses of radioiodide in Graves' disease: effect on ophthalmopathy and long-acting thyroid stimulator. Mayo Clin Proc 1967;42:802-811

14.

Solem JH, Segaard E, Ytteborg J. The course of endocrine ophthalmopathy during antithyroid therapy in a prospective study. Acta Med Scand 1979;205:111-114 15. Gwinup G, Elias AN, Ascher MS. Effect on exophthalmos of various methods of treatment of Graves' disease. JAMA 1982;247:2135-2138 16. Levitt MD, Edis AJ, Agnello R, McCormick CC. The effect of subtotal thyroidectomy on Graves' ophthalmopathy. World JSurg 1988;12:593-597 17. Frilling A, Goretzki PE, Grussendorf M, Erbslöh M, Roher H-D. The influence of surgery on endocrine ophthalmopathy. World JSurg 1990;14:442-446 18. Sridama V, DeGroot LJ. Treatment of Graves' disease and the course of ophthalmopathy. Am J Med 1989; 87:7073 19. Bartalena L, Marcocci C, Bogazzi F, Panicucci M, Lepri A, Pinchera A. Use of corticosteroids to prevent progression of Graves' ophthalmopathy after radioiodine therapy for hyperthyroidism. N Engl J Med 1989; 321:13491352 20. Tallstedt L, Lundell G, T0rring O, Wallin G, Ljunggren J-G, Blomgren H, et al. Occurrence of ophthalmopathy after treatment for Graves'hyperthyroidism. N Engl J Med 1992; 326:1733-1738 21. Karlsson FA, Dahlberg PA, Jansson R, Westermark K, Enoksson P. Importance of TSH receptor activation in the development of severe endocrine ophthalmopathy. Acta Endocrinol (Copenh) 1989; 121(Suppl 2):132-141 22. Bromberg N, Romaldini JH, Werner RS, Sgarbi JA, Werner MC. The evolution of Graves' ophthalmopathy during treatment with antithyroid drug alone and combined with triiodothyronine. J Endocrinol Invest 1992;15:191-195 23. Prummel MF, Wiersinga WM, Mourits MP, Koornneef L, Berghout A, van der Gaag R. Effect of abnormal thyroid function on the severity of Graves' ophthalmopathy. Arch Intern Med 1990;150:1098-1101 24. Mourits MP, Koornneef L, Wiersinga WM, Prummel MF, Berghout A, van der Gaag R. Orbital decompression for Graves' ophthalmopathy by inferomedial, by inferomedial plus lateral, and by coronal approach. Ophthalmology 1990; 97:636-641 25. Lamberg B-A, Granne B, Tommila V, Pelkonen R, Rinne J, Liesmaa M, et al. Orbital decompression in endocrine exophthalmos of Graves' disease. Acta Endocrinol 1985; 109:335-340 26. McCord CD Jr. Current trends in orbital decompression. Ophthalmology 1985; 92:21-33 27. Bartalena L, Marcocci C, Bogazzi F, Panicucci M, BrunoBossio G, Cantini R, et al. Orbital decompression for severe Graves' ophthalmopathy: results of a three-wall operative technique. J Neurosurg Sci 1989;33:323-327