Orbital Decompression for Graves' Ophthalmopathy by lnferomedial, by Inferomedial Plus Lateral, and by Coronal Approach MAARTEN Ph. MOURITS, MD,l LEO KOORNNEEF, MD, PhD,l WILMAR M. WIERSINGA, MD, PhD,2 MARK F. PRUMMEL, MD,2 ARIE BERGH OUT, MD,2 RUTH van der GAAG, PhIY
Abstract: To test the efficacy and safety of orbital decompression for Graves' ophthalmopathy, the authors studied the records of 60 consecutive patients who were operated on for dysthyroid optic neuropathy or for rehabilitative purposes. Patients decompressed for neuropathy were older, had less proptosis, and a shorter duration of eye disease than patients operated on for disfigurement. The authors compared the results of three surgical procedures iricluding the inferomedial, the inferomedial plus lateral, and the coronal approach. Regarding improvement of visual function, no difference was found between the three techniques. Patients in whom vision failed to recover had a high prevalence of diabetes mellitus. Proptosis reduction varied from 1 to 9 mm, depending on the number of walls decompressed. There was no net change in the prevalence of diplopia. Persistent complications were seen in less than 5% of all decompressions. The authors conclude that orbital decompression for Graves' ophthalmopathy is safe and efficacious, regardless of surgical procedure. However, the coronal approach gives the best cosmetic results. Ophthalmology 1990; 97:636-641
Surgical decompression of the orbit in Graves' ophthalmopathy is indicated when vision is threatened due to optic nerve compression and after failure of con-
Originally received: July 20, 1989. Revision accepted: November 29, 1989. Orbital Center, University of Amsterdam, The Netherlands. Department of Endocrinology. University of Amsterdam, Amsterdam. 3 Department of Ophthaimo-immunology, Netherlands Ophthalmic Research Institute, Amsterdam. 1
2
Presented at the First International Meeting on Graves' Ophthalmopathy. Montreal. Canada, Sept. 27-0ct. 1, 1988. Reprint requests to Maarten Ph. Mourits, MD. Orbita Centrum, Academisch Medisch Centrum A-2 116, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
636
servative treatment.! More recently, a high proportion of patients with Graves' ophthalmopathy requests reduction of proptosis solely for rehabilitative reasons. 2 Since the first description of an orbital decompression in Graves' ophthalmopathy by Dollingeil in 1911, several alternative surgical techniques have been advocated, including one-, two-, three-, and four-wall decompressions. The most widely used technique today is antral-ethmoidal decompression by a transantral approach as described by Walsch and Ogura4 in 1957. The results of transantral surgery in a large series of patients were published recently by DeSanto. 5 The major disadvantage reported with this method is increased postoperative motility imbalance. As detailed evaluation of alternative techniques for orbital decompression has not been reported, we present the results over a 5-year period of orbital decompression
MOURITS et al
•
ORBITAL DECOMPRESSION
Table 1. Clinical Characteristics of 60 Patients Who Underwent Orbital Decompression for Graves' Ophthalmopathy Inferomedial
No. of patients Laterality Sex M:F Age (yrs) + SO Preoperative Hertel (mm) Mean + SO Graves' Ophthalmopathy duration (yrs) Mean Range
Inferomedial Plus Lateral
Coronal
A*
Bt
A*
Bt
A*
13 1 unilateral 12 bilateral
16 6 unilateral 10 bilateral
ounilateral
5
5 bilateral
1 1 unilateral obilateral
ounilateral 7 bilateral
18 bilateral
4:9 54.4 ± 12.3
10:6 46.5 ± 12.0
3:2 53.6 ± 8.9
1 :0 42.0 ± 0
1:6 54.6 ± 10.7
0:18 38.5 ± 12.4
21 .9 ± 2.8
25.2 ± 3.7
25.4 ± 4.4
29.0 ± -
21 .9 ± 4.2
23.7 ± 3.6
0.5 0.2-2
4.0 0.5-29
1.5 0.1-27
1.5
0.5 0.2-1.2
3.5 1-12
7
Bt 18
ounilateral
SO = standard deviation. * For optic compression . t For rehabilitative reasons .
in Graves' ophthalmopathy, in which both the coronal6 and inferomedial (plus lateral) approaches are applied. The outcomes of the various surgical techniques are compared.
PATIENTS AND METHODS The clinical characteristics of all patients, operated on between January 1983 and December 1987, are given in Table 1. Twenty-five patients (49 orbits) underwent orbital decompression because of optic nerve compression as shown by a decrease of visual acuity, visual field defects abnormal visual-evoked potentials (VEPs), and/or dis~ e~ema. All b';1t one patient had bilateral optic neuropathy. Eighteen patIents had received preoperative treatment of prednisone (n = 16), cyclosporin (n = 1) or orbital irradiation (n=3). The remaining seven pati~nts underwent decompr~ssion without a trial of steroids and/or other preoperative measures because of a very rapid onset of visual dete.rioration ..Thirteen patients were operated on using the mferomedlal approach only. In five patients, this approach was combined with a lateral wall decompression to create more space inside the fibrotic orbie and to gain more proptosis reduction. Seven of these 25 patients underwent a coronal decompression. Orbital decompressions for rehabilitative reasons were not performed until all signs of disease activity had disappeared for at least six months. In women with bilateral eye disease and not too high a hair line, we chose the coronal approach. In all others, an inferomedial or inferomedial plus lateral approach was used. To evaluate the proptosis reduction as a function of surgical technique we performed an analysis of variance. SURGICAL TECHNIQUE
In the inferomedial approach, a skin incision through the lower eyelid is made approximately 10 to 15 mm below
the lashes after the lid curve, but bending downward laterally in the skin crease. The orbicularis muscle is split bluntly until the periosteum is reached. The periosteum and periorbita are dissected free of the orbital floor and medial orbital wall. The floor is then perforated and removed medially and laterally from the infraorbital nerve. Attention is paid not to damage this neurovascular bundle. The removal of the floor is carried out as far as the posterior wall of the antral cavity. The nasal orbital wall is removed up to the posterior ethmoidal artery. Finally, the exposed periorbita is incised in postero-anterior direction in the posterior part of the orbit and perpendicular to these cuts in the anterior part, enabling the orbital fat to herniate freely into the sinuses. This technique may be preceded by a lateral wall decompression using Kronlein's incision. 8 The temporalis muscle is dissected from the zygomatic bone and the periosteum is dissected from the lateral orbital wall, which is then perforated and removed, sparing the orbital rim. Again, the periorbita is incised in postero-anterior direction allowing the orbital fat to herniate into the infratemporal fossa. In the coronal approach,9 after shaving a strip of hair, a skin-muscle incision is made through the scalp 10 mm ~ehind the hair border from ear to ear. The subgaleal flap IS turned down to the supraorbital rim and then the dissection is taken subperiosteal. Care is taken to avoid injury to the supraorbital nerves. The periorbita, including the trochlea, is dissected off the bony walls all around the globe. Again, the temporalis muscle is partially freed from its origin and the lateral wall is removed, allowing a fingertip to pass through and leaving the rim intact. Medially, a large part of the lacrimal and ethmoidal bones are removed as far back to the posterior ethmoidal artery. Finally, the floor can be perforated and removed medially and laterally, sparing the infraorbital nerve. The periorbita is incised in postero-anterior direction all around the globe in the posterior part of the orbit and perpendicular to the former in the front. Drains are left behind in the infra637
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
•
NUMBER 5
NLP LP HM CF
•
20/400
coronal
20/200 20/100 20/80
JiJ.6
inferomedial
20/70
r
postop visual
20/60
.0
20/50
inferomedial plus lateral
20/40 20/30 20/25
06
Diabetes Mellitus
20/20
Fig I . Preoperative visual acuity versus 3-month postoperative visual acuity in 25 patients with dysthyroid optic neuropathy. The diagonal indicates no change in visual acuity. Points below the diagonal indicate an improvement and points above the diagonal indicate a postoperative reduction. c.u. = cornea! ulcer; CF = counting fingers; HM = hand motions; LP = light perception; NLP = no light perception. The asterisk indicates that both eyes improved after a second decompression.
acuity 202020~~2O
20
2s 30
40
50
preop visual acuity
2020
60 70 80 •
2020 20 100 200 400 CF
HM LP NLP
Table 2. Mean Proptosis Reduction after Orbital Decompression For Optic Compression (mm)
For Disfigurement (mm)
Surgical Approach
No. of Orbits
Preoperative Hertel
Proptosis Reduction
Range
Preoperative Hertel
Proptosis Reduction
Inferomedial Inferomedial plus lateral Coronal
25 + 26 10 + 1 14 + 36
21.9 25.4 21.9
2.0 4.3 4.7
2-5 3-6 2-9
25.2 29.0 23.7
3.0 4.0 4.6
temporal fossae. The skin-muscle flap is brought backward and closed with skin staples, which are removed after 10 days.
RESULTS VISUAL ACUITY
Follow-up varied from 0.5 to 5.5 years. Changes of visual acuity evaluated 3 months after decompression for optic neuropathy are shown in Figure 1. The changes of visual fields and VEPs were consistent with the alterations of the visual acuity. Visual function improved in 19 (76%) of25 patients who underwent decompression. Four (67%) of the six nonresponders had diabetes mellitus, of which two had a nonproliferative diabetic retinopathy. Of the remaining two failures, one patient showed an improvement of visual acuity, visual fields, and visual-evoked re638
Range
1-8 1-8
sponse after a second decompression procedure. Nonresponders to orbital decompression did not respond to subsequent high-dose prednisone treatment or 2000 rad orbital radiotherapy. Of the seven patients who underwent surgery without any form of pretreatment, five improved, whereas visual function remained unchanged in two despite postoperative steroid and/or irradiation treatment. In the patients operated on for rehabilitative reasons, no change in visual acuity was found. PROPTOSIS REDUCTION
Proptosis reduction was found to improve up to at least 6 months after the decompression. The mean proptosis reduction is specified in Table 2. The postoperative difference in proptosis between the two eyes (ocular balance) was I mm or less in 75% of all patients and in 70% of patients operated on for rehabilitative reasons. After a
MOURITS
et al •
ORBITAL DECOMPRESSION
29 28 Fig 2. Change of proptosis after {iifferent surgical approaches for orbital decompression for Graves' ophthalmopathy. The slope of the inferomedial plus lateral and of the coronal functions differs significantly from the inferomedial function. The inferomedial plus lateral and the coronal slopes do not differ significantly, indicating that when more walls are removed greater proptosis reduction is achieved.
27 26
25 24 23 22 21
t
20 19
mm
18
proptosis preop Hertel
postop Hertel
Table 3. Ocular Motility Three Months after Orbital Decompression For Optic Compression
For Disfigurement
Surgical Approach
No. of Patients
Worse
Better
Inferomedial Inferomedial plus lateral Coronal
11 + 16 5+ 1 5 + 18
0 0 0
0 2 0
11 3 5
o
o 1
15
0
2
19
4
3
28
Total
three-wall decompression, the mean proptosis reduction was significantly better than after a two-wall decompression (Fig 2). Sex, laterality, and preoperative Hertel value did not influence the proptosis reduction, nor did the indication for decompression (P=O.12). In patients with compressive optic neuropathy, in whom the proptosis reduction must be considered a beneficial side-effect of the operation, the mean reduction after the inferomedial technique also was less than that after the inferomedial plus lateral or coronal approach. MOTILITY
The motility changes after orbital decompression in the various surgical treatment groups are given in Table 3. Motility change was defined as a shift in grade of class IV of the NOSPECS classification system. Twenty-seven patients had pre- and postoperative diplopia in primary and reading positions. Diplopia developed postoperatively
Unchanged Worse (according to NOSPECS-grade)
2 2
Better
Unchanged
2
12 1
in four patients; in two only at the extremes of gaze and in two also in the primary position. On the other hand, five patients no longer experienced diplopia in the primary and reading position after orbital decompression. COMPLICATIONS
Other complications of orbital decompression in the various treatment groups are given in Table 4. A postoperative wound infection was seen in three patients after the inferomedial approach, which promptly responded to antibiotics. A transient corneal ulcer was found in two patients as a result of postoperative exposure; a persistent corneal clouding was present in one patient (mycotic ulcer). Four patients complained of neuralgic pain in the area of the infraorbital nerve. In two patients, this pain disappeared completely after the nerve had been dissected free via a second procedure. Finally, one diabetic patient lost vision in one of his eyes because of a postoperative
639
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
Dysthyroid optic neuropathy, occurring in less than 5% of patients with Graves' eye disease,1O is a potentially blinding disease, requiring careful attention and immediate medical and/or surgical intervention. In this series, the diagnosis of dysthyroid optic neuropathy was based on the presence of grossly enlarged eye muscles (assessed with the computed tomographic [CT] scan) in combination with at least two of the following symptoms: decreased visual acuity, visual field defects, increased VEP latency, impairment of color vision, or disc edema. When the Hertel values were relatively low, we performed an inferomedial decompression. When there was marked proptosis, a three-wall decompression was performed, either by the inferomedial plus lateral approach (primarily in male patients) or by the coronal approach (primarily in women). Our aim was to improve impaired visual function and to restore facial disfigurement. For the latter, we found a maximal retrodisplacement of the globes of less importance than symmetry (ocular balance). This may explain, at least in part, why our mean globe recession in the inferomedial groups is less than previously reported. I I Furthermore, our globe recessions in the rehabilitative group were typically the result of surgery alone; other modes of treatment were not involved and the majority of patients had shown stable proptosis for more than 2 years before the operation. Direct comparison with previous reports of the changes in visual function is limited because various factors such as indication for surgery, duration of eye disease, pretreatment, posttreatment, and combined treatment, and follow-up period contribute to the therapeutic outcome. For example, Hallin and associates l2 found that 77% of their patients improved after decompression, whereas Warren et al 13 state that visual acuity nearly always improved or at least was preserved. Hurwitz and Birt l4 noted substantial improvement in all but one operated eye after several decompressions and McCord 15 found improvement to 20/40 or better in 6 of 11 patients after surgery and in 5 patients after additional medical treatment and supervoltage irradiation. In this study, we were able to assess the effect of surgery alone in 20 of the 25 patients with optic neuropathy, because either no additional treatment was given or a marked improvement was found before the effect of additional treatment was expected. Visual function improved in 61 % (14 of 23 eyes) after the inferomedial approach due to surgery alone, in 66% (4 of 6 eyes) after the inferomedial plus lateral approach, and in 70% (7 of 10 eyes) after the coronal approach. These data suggest that for the visual outcome, the surgical approach in orbital decompression is of minor consequence. At the time 640
NUMBER 5
Table 4. Complications after Orbital Decompression
intraorbital hemorrhage. Postoperative numbness of the skin was not routinely recorded. However, if noted, the numbness typically proved transient and disappeared altogether within 3 to 6 months.
DISCUSSION
•
Inferomedial (n = 51)* Impaired smell Cellulitis Corneal ulcer Neuralgic pain Blind eye
1 2 1
Total
6 (12%)
Inferomedial Plus Lateral (n = 11)*
Coronal
(n = 50)*
2 3
1 1
1 (10%)
5 (10%)
* Number of orbits.
when postoperative visual acuity was recorded as shown in Figure 1 (e.g., after 3 months), visual acuity of 20/20 was rarely observed. In some patients, a further improvement was found up to 1 year postoperatively. In nonresponders to surgery, a postoperative CT scan was performed to determine if the decompression was adequate. In one patient, this led to a second decompressive procedure with good result. Four of the remaining five patients, who did not improve after surgery, were suffering from diabetes mellitus. This suggests that a vascular component might have been involved. Further support for this hypothesis is provided by the fact that the mean age of patients with compressive optic neuropathy is relatively high as compared with the general population of patients with Graves' ophthalmopathy. Trobe et al l6 found a mean age of patients with dysthyroid neuropathy of 61 years. In our series, the mean age was 54.3 years. However, the quality of the optic nerve vascularity is not the only factor in the pathogenesis of compressive optic neuropathy. It is well known that patients with dysthyroid optic neuropathy often have relatively low Hertel readings, although they show grossly enlarged eye muscles on CT scans. 9 ,17 In our series, the mean preoperative proptosis in patients, operated on for rehabilitative reason, was 24.4 mm, whereas the mean Hertel readings of the patients with compressive neuropathy was only 22.6 mm. The mean diameter of the eye muscles on the CT scan, however, was much higher in the neuropathy group than in the rehabilitative group. In general, we found a strikingly short duration of eye disease in patients with optic neuropathy (8.6 months at the moment of operation, Table 1). These observations suggest that dysthyroid optic neuropathy is not the end stage of a longstanding, therapy-resistant form of Graves' eye disease, but rather a peculiar variant ofthe disease in which the increased intraorbital pressure, 18 due to the grossly enlarged extraocular muscles and a tight orbital septum in combination with an impaired optic nerve vascularity, leads to a rapid loss of visual function. Despite the differences in character and extent of the operation techniques used, the surgical complications in this study did not vary significantly in the different groups (Tables 3, 4). Many authors however, report an increase of motility imbalance after surgical decompression. Hurwitz and Birt l4 found a worsening of ocular motility in 16 of 19 orbits, decompressed via a transantral approach,
MOURITS et al
•
ORBITAL DECOMPRESSION
whereas in DeSanto's5 series, postoperative diplopia developed in 52% of the patients. A high incidence of esotropia after ethmoidal decompression also has been reported by Fells.!9 In a review of surgical techniques for orbital decompressions in the United States, McCord!l found less iatrogenic muscle imbalance after translid antral-ethmoidal decompression (5.7%) than after transantral ethmoidal decompression (41.0%). Because of these reported complications, all our patients were subjected to meticulous orthoptic examinations pre- and postoperatively. In most patients operated on because of visionthreatening optic neuropathy we found that there was no significant change in ocular motility. These results support surgical intervention for cosmetic purposes. In the rehabilitative group, no change in class 4 of the NOSPECS classification system 20 was found in 12 of 16 patients, operated on using the inferomedial approach, and in 15 of 18 patients, using the coronal approach (Table 3). In a previously reported series of 11 patients, decompressed by a four-wall technique, Stranc and West,21 incising the periorbita, using both longitudinal and radial incisions, also had very few patients with postoperative diplopia. McCord I 5suggested that postoperative muscle imbalance is significantly reduced by complete removal of the inferior periorbita, which precludes the formation of fibrotic bands with the extraocular muscles. CONCLUSION
In the treatment of Graves' ophthalmopathy, surgery plays an important role. This study shows that the coronal approach is as good or better than the classic inferomedial (plus lateral) approach in terms of effectiveness and safety. The coronal approach also has the advantage, primarily in female patients, of no visible scars postoperatively, and therefore is the operation of choice in patients who seek relief of their exophthalmus for rehabilitative purposes. The simpler inferomedial approach is recommended for patients with compressive optic neuropathy (and mild proptosis) to restore visual function and to avoid muscle imbalance.
REFERENCES 1. Linberg JV. Anderson RL. Transorbital decompression: indications and results. Arch Ophthalmol1981; 99:113-9.
2 . Harting F, Koornneef L, Peeters HJF, Gillissen JPA. Fourteen years of orbital decompression in Graves' disease. Orbit 1986; 5:123-9. 3. Dollinger J. Die Druckentlastung der Augenhtihle durch Entfernung der auf3eren Orbitalwand bei hochgradigen Exophthalmus (Morbus Basedowii) und konsekutiver Hornhauterkrankung. Dtsc h Med Wochenschr 1911; 37:1888-90. 4. Walsch TE, Ogura JH. Transantral orbital decompression for malignant exophthalmos. Laryngoscope 1957; 67 :544-68. 5. DeSanto LW. Transantral orbital decompression. In: Gorman CA, Waller RR, Dyer JA, eds. The Eye and Orbit in Thyroid Disease. New York: Raven Press 1984; 231-51. 6. Krastinova 0, Rodallec A. Orbitopathie Basedowienne. Ann Chir Plast Esthet 1985; 30:351 - 8. 7. Koornneef L. Orbital bony and soft tissue anatomy. In: Gorman CA, Waller RR, Dyer JA, eds. The Eye and Orbit in Thyroid Disease . New York: Raven Press 1984; 5-23. 8. Krtinlein RU. Zur Pathologie und operativen Behandlung der Dermoidcysten der Orb ita. Beitr Klin Chir 1888; 4:149-63. 9. Stewart WB , Levin PS, Toth BA. Orbital surgery: the technique of coronal scalp flap approach to the lateral orbitotomy. Arch Ophthalmol 1988; 106:1724-6. 10. Char DH. Thyroid Eye Disease. Baltimore: Williams & Wilkins, 1985; 49,116. 11. McCord CD Jr. Current trends in orbital decompression. Ophthalmology 1985; 92:21 - 33. 12. Hallin ES, Feldon SE, Luttrell J. Graves' ophthalmopathy: III. Effect of transantral orbital decompression on optic neuropathy. Br J Ophthalmol 1988; 72:683-7. 13. Warren JD, Spector JG, Burde R. Long-term follow-up and recent observations on 305 cases of orbital decompression for dysthyroid orbitopathy. Laryngoscope 1989; 99:35-40. 14. Hurwitz JJ, Birt D. An individualized approach to orbital decompression in Graves' orbitopathy. Arch Ophthalmol1985; 103:660-5. 15. McCord CD. Orbital decompression for Graves' disease: exposure through lateral canthal and inferior fomix incision. Ophthalmology 1981 ; 88:533-41. 16. Trobe JD, Glaser JS, Laflamme P. Dysthyroid optiC neuropathy: clinical profile and rationale for management. Arch Ophthalmol 1978; 96: 1199-209. 17. Anderson RL, Linberg JV. Transorbital approach to decompression in Graves' disease. Arch Ophthalmol1981; 99:120-4. 18. Koornneef L. Eyelid and orbital fascial attachments and their clinical significance. Eye 1988; 2:130-4. 19. Fells P. Orbital decompression for severe dysthyroid eye disease. Br J Ophthalmol1987; 71:107-11. 20. Werner SC. Modification of the classification of the eye changes of Graves' disease: recommendations of the Ad Hoc Committee of the American Thyroid Association [Letter]. J Clin Endocrinol Metab 1977; 44:203-4. 21 . Stranc M, West M . A four-wall orbital decompression for dysthyroid orbitopathy. J Neurosurg 1988; 68:671-7.
641
Abstract: To test the efficacy and safety of orbital decompression for Graves' ophthalmopathy, the authors studied the records of 60 consecutive patients who were operated on for dysthyroid optic neuropathy or for rehabilitative purposes. Patients decompressed for neuropathy were older, had less proptosis, and a shorter duration of eye disease than patients operated on for disfigurement. The authors compared the results of three surgical procedures iricluding the inferomedial, the inferomedial plus lateral, and the coronal approach. Regarding improvement of visual function, no difference was found between the three techniques. Patients in whom vision failed to recover had a high prevalence of diabetes mellitus. Proptosis reduction varied from 1 to 9 mm, depending on the number of walls decompressed. There was no net change in the prevalence of diplopia. Persistent complications were seen in less than 5% of all decompressions. The authors conclude that orbital decompression for Graves' ophthalmopathy is safe and efficacious, regardless of surgical procedure. However, the coronal approach gives the best cosmetic results. Ophthalmology 1990; 97:636-641
Surgical decompression of the orbit in Graves' ophthalmopathy is indicated when vision is threatened due to optic nerve compression and after failure of con-
Originally received: July 20, 1989. Revision accepted: November 29, 1989. Orbital Center, University of Amsterdam, The Netherlands. Department of Endocrinology. University of Amsterdam, Amsterdam. 3 Department of Ophthaimo-immunology, Netherlands Ophthalmic Research Institute, Amsterdam. 1
2
Presented at the First International Meeting on Graves' Ophthalmopathy. Montreal. Canada, Sept. 27-0ct. 1, 1988. Reprint requests to Maarten Ph. Mourits, MD. Orbita Centrum, Academisch Medisch Centrum A-2 116, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
636
servative treatment.! More recently, a high proportion of patients with Graves' ophthalmopathy requests reduction of proptosis solely for rehabilitative reasons. 2 Since the first description of an orbital decompression in Graves' ophthalmopathy by Dollingeil in 1911, several alternative surgical techniques have been advocated, including one-, two-, three-, and four-wall decompressions. The most widely used technique today is antral-ethmoidal decompression by a transantral approach as described by Walsch and Ogura4 in 1957. The results of transantral surgery in a large series of patients were published recently by DeSanto. 5 The major disadvantage reported with this method is increased postoperative motility imbalance. As detailed evaluation of alternative techniques for orbital decompression has not been reported, we present the results over a 5-year period of orbital decompression
MOURITS et al
•
ORBITAL DECOMPRESSION
Table 1. Clinical Characteristics of 60 Patients Who Underwent Orbital Decompression for Graves' Ophthalmopathy Inferomedial
No. of patients Laterality Sex M:F Age (yrs) + SO Preoperative Hertel (mm) Mean + SO Graves' Ophthalmopathy duration (yrs) Mean Range
Inferomedial Plus Lateral
Coronal
A*
Bt
A*
Bt
A*
13 1 unilateral 12 bilateral
16 6 unilateral 10 bilateral
ounilateral
5
5 bilateral
1 1 unilateral obilateral
ounilateral 7 bilateral
18 bilateral
4:9 54.4 ± 12.3
10:6 46.5 ± 12.0
3:2 53.6 ± 8.9
1 :0 42.0 ± 0
1:6 54.6 ± 10.7
0:18 38.5 ± 12.4
21 .9 ± 2.8
25.2 ± 3.7
25.4 ± 4.4
29.0 ± -
21 .9 ± 4.2
23.7 ± 3.6
0.5 0.2-2
4.0 0.5-29
1.5 0.1-27
1.5
0.5 0.2-1.2
3.5 1-12
7
Bt 18
ounilateral
SO = standard deviation. * For optic compression . t For rehabilitative reasons .
in Graves' ophthalmopathy, in which both the coronal6 and inferomedial (plus lateral) approaches are applied. The outcomes of the various surgical techniques are compared.
PATIENTS AND METHODS The clinical characteristics of all patients, operated on between January 1983 and December 1987, are given in Table 1. Twenty-five patients (49 orbits) underwent orbital decompression because of optic nerve compression as shown by a decrease of visual acuity, visual field defects abnormal visual-evoked potentials (VEPs), and/or dis~ e~ema. All b';1t one patient had bilateral optic neuropathy. Eighteen patIents had received preoperative treatment of prednisone (n = 16), cyclosporin (n = 1) or orbital irradiation (n=3). The remaining seven pati~nts underwent decompr~ssion without a trial of steroids and/or other preoperative measures because of a very rapid onset of visual dete.rioration ..Thirteen patients were operated on using the mferomedlal approach only. In five patients, this approach was combined with a lateral wall decompression to create more space inside the fibrotic orbie and to gain more proptosis reduction. Seven of these 25 patients underwent a coronal decompression. Orbital decompressions for rehabilitative reasons were not performed until all signs of disease activity had disappeared for at least six months. In women with bilateral eye disease and not too high a hair line, we chose the coronal approach. In all others, an inferomedial or inferomedial plus lateral approach was used. To evaluate the proptosis reduction as a function of surgical technique we performed an analysis of variance. SURGICAL TECHNIQUE
In the inferomedial approach, a skin incision through the lower eyelid is made approximately 10 to 15 mm below
the lashes after the lid curve, but bending downward laterally in the skin crease. The orbicularis muscle is split bluntly until the periosteum is reached. The periosteum and periorbita are dissected free of the orbital floor and medial orbital wall. The floor is then perforated and removed medially and laterally from the infraorbital nerve. Attention is paid not to damage this neurovascular bundle. The removal of the floor is carried out as far as the posterior wall of the antral cavity. The nasal orbital wall is removed up to the posterior ethmoidal artery. Finally, the exposed periorbita is incised in postero-anterior direction in the posterior part of the orbit and perpendicular to these cuts in the anterior part, enabling the orbital fat to herniate freely into the sinuses. This technique may be preceded by a lateral wall decompression using Kronlein's incision. 8 The temporalis muscle is dissected from the zygomatic bone and the periosteum is dissected from the lateral orbital wall, which is then perforated and removed, sparing the orbital rim. Again, the periorbita is incised in postero-anterior direction allowing the orbital fat to herniate into the infratemporal fossa. In the coronal approach,9 after shaving a strip of hair, a skin-muscle incision is made through the scalp 10 mm ~ehind the hair border from ear to ear. The subgaleal flap IS turned down to the supraorbital rim and then the dissection is taken subperiosteal. Care is taken to avoid injury to the supraorbital nerves. The periorbita, including the trochlea, is dissected off the bony walls all around the globe. Again, the temporalis muscle is partially freed from its origin and the lateral wall is removed, allowing a fingertip to pass through and leaving the rim intact. Medially, a large part of the lacrimal and ethmoidal bones are removed as far back to the posterior ethmoidal artery. Finally, the floor can be perforated and removed medially and laterally, sparing the infraorbital nerve. The periorbita is incised in postero-anterior direction all around the globe in the posterior part of the orbit and perpendicular to the former in the front. Drains are left behind in the infra637
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
•
NUMBER 5
NLP LP HM CF
•
20/400
coronal
20/200 20/100 20/80
JiJ.6
inferomedial
20/70
r
postop visual
20/60
.0
20/50
inferomedial plus lateral
20/40 20/30 20/25
06
Diabetes Mellitus
20/20
Fig I . Preoperative visual acuity versus 3-month postoperative visual acuity in 25 patients with dysthyroid optic neuropathy. The diagonal indicates no change in visual acuity. Points below the diagonal indicate an improvement and points above the diagonal indicate a postoperative reduction. c.u. = cornea! ulcer; CF = counting fingers; HM = hand motions; LP = light perception; NLP = no light perception. The asterisk indicates that both eyes improved after a second decompression.
acuity 202020~~2O
20
2s 30
40
50
preop visual acuity
2020
60 70 80 •
2020 20 100 200 400 CF
HM LP NLP
Table 2. Mean Proptosis Reduction after Orbital Decompression For Optic Compression (mm)
For Disfigurement (mm)
Surgical Approach
No. of Orbits
Preoperative Hertel
Proptosis Reduction
Range
Preoperative Hertel
Proptosis Reduction
Inferomedial Inferomedial plus lateral Coronal
25 + 26 10 + 1 14 + 36
21.9 25.4 21.9
2.0 4.3 4.7
2-5 3-6 2-9
25.2 29.0 23.7
3.0 4.0 4.6
temporal fossae. The skin-muscle flap is brought backward and closed with skin staples, which are removed after 10 days.
RESULTS VISUAL ACUITY
Follow-up varied from 0.5 to 5.5 years. Changes of visual acuity evaluated 3 months after decompression for optic neuropathy are shown in Figure 1. The changes of visual fields and VEPs were consistent with the alterations of the visual acuity. Visual function improved in 19 (76%) of25 patients who underwent decompression. Four (67%) of the six nonresponders had diabetes mellitus, of which two had a nonproliferative diabetic retinopathy. Of the remaining two failures, one patient showed an improvement of visual acuity, visual fields, and visual-evoked re638
Range
1-8 1-8
sponse after a second decompression procedure. Nonresponders to orbital decompression did not respond to subsequent high-dose prednisone treatment or 2000 rad orbital radiotherapy. Of the seven patients who underwent surgery without any form of pretreatment, five improved, whereas visual function remained unchanged in two despite postoperative steroid and/or irradiation treatment. In the patients operated on for rehabilitative reasons, no change in visual acuity was found. PROPTOSIS REDUCTION
Proptosis reduction was found to improve up to at least 6 months after the decompression. The mean proptosis reduction is specified in Table 2. The postoperative difference in proptosis between the two eyes (ocular balance) was I mm or less in 75% of all patients and in 70% of patients operated on for rehabilitative reasons. After a
MOURITS
et al •
ORBITAL DECOMPRESSION
29 28 Fig 2. Change of proptosis after {iifferent surgical approaches for orbital decompression for Graves' ophthalmopathy. The slope of the inferomedial plus lateral and of the coronal functions differs significantly from the inferomedial function. The inferomedial plus lateral and the coronal slopes do not differ significantly, indicating that when more walls are removed greater proptosis reduction is achieved.
27 26
25 24 23 22 21
t
20 19
mm
18
proptosis preop Hertel
postop Hertel
Table 3. Ocular Motility Three Months after Orbital Decompression For Optic Compression
For Disfigurement
Surgical Approach
No. of Patients
Worse
Better
Inferomedial Inferomedial plus lateral Coronal
11 + 16 5+ 1 5 + 18
0 0 0
0 2 0
11 3 5
o
o 1
15
0
2
19
4
3
28
Total
three-wall decompression, the mean proptosis reduction was significantly better than after a two-wall decompression (Fig 2). Sex, laterality, and preoperative Hertel value did not influence the proptosis reduction, nor did the indication for decompression (P=O.12). In patients with compressive optic neuropathy, in whom the proptosis reduction must be considered a beneficial side-effect of the operation, the mean reduction after the inferomedial technique also was less than that after the inferomedial plus lateral or coronal approach. MOTILITY
The motility changes after orbital decompression in the various surgical treatment groups are given in Table 3. Motility change was defined as a shift in grade of class IV of the NOSPECS classification system. Twenty-seven patients had pre- and postoperative diplopia in primary and reading positions. Diplopia developed postoperatively
Unchanged Worse (according to NOSPECS-grade)
2 2
Better
Unchanged
2
12 1
in four patients; in two only at the extremes of gaze and in two also in the primary position. On the other hand, five patients no longer experienced diplopia in the primary and reading position after orbital decompression. COMPLICATIONS
Other complications of orbital decompression in the various treatment groups are given in Table 4. A postoperative wound infection was seen in three patients after the inferomedial approach, which promptly responded to antibiotics. A transient corneal ulcer was found in two patients as a result of postoperative exposure; a persistent corneal clouding was present in one patient (mycotic ulcer). Four patients complained of neuralgic pain in the area of the infraorbital nerve. In two patients, this pain disappeared completely after the nerve had been dissected free via a second procedure. Finally, one diabetic patient lost vision in one of his eyes because of a postoperative
639
OPHTHALMOLOGY
•
MAY 1990
•
VOLUME 97
Dysthyroid optic neuropathy, occurring in less than 5% of patients with Graves' eye disease,1O is a potentially blinding disease, requiring careful attention and immediate medical and/or surgical intervention. In this series, the diagnosis of dysthyroid optic neuropathy was based on the presence of grossly enlarged eye muscles (assessed with the computed tomographic [CT] scan) in combination with at least two of the following symptoms: decreased visual acuity, visual field defects, increased VEP latency, impairment of color vision, or disc edema. When the Hertel values were relatively low, we performed an inferomedial decompression. When there was marked proptosis, a three-wall decompression was performed, either by the inferomedial plus lateral approach (primarily in male patients) or by the coronal approach (primarily in women). Our aim was to improve impaired visual function and to restore facial disfigurement. For the latter, we found a maximal retrodisplacement of the globes of less importance than symmetry (ocular balance). This may explain, at least in part, why our mean globe recession in the inferomedial groups is less than previously reported. I I Furthermore, our globe recessions in the rehabilitative group were typically the result of surgery alone; other modes of treatment were not involved and the majority of patients had shown stable proptosis for more than 2 years before the operation. Direct comparison with previous reports of the changes in visual function is limited because various factors such as indication for surgery, duration of eye disease, pretreatment, posttreatment, and combined treatment, and follow-up period contribute to the therapeutic outcome. For example, Hallin and associates l2 found that 77% of their patients improved after decompression, whereas Warren et al 13 state that visual acuity nearly always improved or at least was preserved. Hurwitz and Birt l4 noted substantial improvement in all but one operated eye after several decompressions and McCord 15 found improvement to 20/40 or better in 6 of 11 patients after surgery and in 5 patients after additional medical treatment and supervoltage irradiation. In this study, we were able to assess the effect of surgery alone in 20 of the 25 patients with optic neuropathy, because either no additional treatment was given or a marked improvement was found before the effect of additional treatment was expected. Visual function improved in 61 % (14 of 23 eyes) after the inferomedial approach due to surgery alone, in 66% (4 of 6 eyes) after the inferomedial plus lateral approach, and in 70% (7 of 10 eyes) after the coronal approach. These data suggest that for the visual outcome, the surgical approach in orbital decompression is of minor consequence. At the time 640
NUMBER 5
Table 4. Complications after Orbital Decompression
intraorbital hemorrhage. Postoperative numbness of the skin was not routinely recorded. However, if noted, the numbness typically proved transient and disappeared altogether within 3 to 6 months.
DISCUSSION
•
Inferomedial (n = 51)* Impaired smell Cellulitis Corneal ulcer Neuralgic pain Blind eye
1 2 1
Total
6 (12%)
Inferomedial Plus Lateral (n = 11)*
Coronal
(n = 50)*
2 3
1 1
1 (10%)
5 (10%)
* Number of orbits.
when postoperative visual acuity was recorded as shown in Figure 1 (e.g., after 3 months), visual acuity of 20/20 was rarely observed. In some patients, a further improvement was found up to 1 year postoperatively. In nonresponders to surgery, a postoperative CT scan was performed to determine if the decompression was adequate. In one patient, this led to a second decompressive procedure with good result. Four of the remaining five patients, who did not improve after surgery, were suffering from diabetes mellitus. This suggests that a vascular component might have been involved. Further support for this hypothesis is provided by the fact that the mean age of patients with compressive optic neuropathy is relatively high as compared with the general population of patients with Graves' ophthalmopathy. Trobe et al l6 found a mean age of patients with dysthyroid neuropathy of 61 years. In our series, the mean age was 54.3 years. However, the quality of the optic nerve vascularity is not the only factor in the pathogenesis of compressive optic neuropathy. It is well known that patients with dysthyroid optic neuropathy often have relatively low Hertel readings, although they show grossly enlarged eye muscles on CT scans. 9 ,17 In our series, the mean preoperative proptosis in patients, operated on for rehabilitative reason, was 24.4 mm, whereas the mean Hertel readings of the patients with compressive neuropathy was only 22.6 mm. The mean diameter of the eye muscles on the CT scan, however, was much higher in the neuropathy group than in the rehabilitative group. In general, we found a strikingly short duration of eye disease in patients with optic neuropathy (8.6 months at the moment of operation, Table 1). These observations suggest that dysthyroid optic neuropathy is not the end stage of a longstanding, therapy-resistant form of Graves' eye disease, but rather a peculiar variant ofthe disease in which the increased intraorbital pressure, 18 due to the grossly enlarged extraocular muscles and a tight orbital septum in combination with an impaired optic nerve vascularity, leads to a rapid loss of visual function. Despite the differences in character and extent of the operation techniques used, the surgical complications in this study did not vary significantly in the different groups (Tables 3, 4). Many authors however, report an increase of motility imbalance after surgical decompression. Hurwitz and Birt l4 found a worsening of ocular motility in 16 of 19 orbits, decompressed via a transantral approach,
MOURITS et al
•
ORBITAL DECOMPRESSION
whereas in DeSanto's5 series, postoperative diplopia developed in 52% of the patients. A high incidence of esotropia after ethmoidal decompression also has been reported by Fells.!9 In a review of surgical techniques for orbital decompressions in the United States, McCord!l found less iatrogenic muscle imbalance after translid antral-ethmoidal decompression (5.7%) than after transantral ethmoidal decompression (41.0%). Because of these reported complications, all our patients were subjected to meticulous orthoptic examinations pre- and postoperatively. In most patients operated on because of visionthreatening optic neuropathy we found that there was no significant change in ocular motility. These results support surgical intervention for cosmetic purposes. In the rehabilitative group, no change in class 4 of the NOSPECS classification system 20 was found in 12 of 16 patients, operated on using the inferomedial approach, and in 15 of 18 patients, using the coronal approach (Table 3). In a previously reported series of 11 patients, decompressed by a four-wall technique, Stranc and West,21 incising the periorbita, using both longitudinal and radial incisions, also had very few patients with postoperative diplopia. McCord I 5suggested that postoperative muscle imbalance is significantly reduced by complete removal of the inferior periorbita, which precludes the formation of fibrotic bands with the extraocular muscles. CONCLUSION
In the treatment of Graves' ophthalmopathy, surgery plays an important role. This study shows that the coronal approach is as good or better than the classic inferomedial (plus lateral) approach in terms of effectiveness and safety. The coronal approach also has the advantage, primarily in female patients, of no visible scars postoperatively, and therefore is the operation of choice in patients who seek relief of their exophthalmus for rehabilitative purposes. The simpler inferomedial approach is recommended for patients with compressive optic neuropathy (and mild proptosis) to restore visual function and to avoid muscle imbalance.
REFERENCES 1. Linberg JV. Anderson RL. Transorbital decompression: indications and results. Arch Ophthalmol1981; 99:113-9.
2 . Harting F, Koornneef L, Peeters HJF, Gillissen JPA. Fourteen years of orbital decompression in Graves' disease. Orbit 1986; 5:123-9. 3. Dollinger J. Die Druckentlastung der Augenhtihle durch Entfernung der auf3eren Orbitalwand bei hochgradigen Exophthalmus (Morbus Basedowii) und konsekutiver Hornhauterkrankung. Dtsc h Med Wochenschr 1911; 37:1888-90. 4. Walsch TE, Ogura JH. Transantral orbital decompression for malignant exophthalmos. Laryngoscope 1957; 67 :544-68. 5. DeSanto LW. Transantral orbital decompression. In: Gorman CA, Waller RR, Dyer JA, eds. The Eye and Orbit in Thyroid Disease. New York: Raven Press 1984; 231-51. 6. Krastinova 0, Rodallec A. Orbitopathie Basedowienne. Ann Chir Plast Esthet 1985; 30:351 - 8. 7. Koornneef L. Orbital bony and soft tissue anatomy. In: Gorman CA, Waller RR, Dyer JA, eds. The Eye and Orbit in Thyroid Disease . New York: Raven Press 1984; 5-23. 8. Krtinlein RU. Zur Pathologie und operativen Behandlung der Dermoidcysten der Orb ita. Beitr Klin Chir 1888; 4:149-63. 9. Stewart WB , Levin PS, Toth BA. Orbital surgery: the technique of coronal scalp flap approach to the lateral orbitotomy. Arch Ophthalmol 1988; 106:1724-6. 10. Char DH. Thyroid Eye Disease. Baltimore: Williams & Wilkins, 1985; 49,116. 11. McCord CD Jr. Current trends in orbital decompression. Ophthalmology 1985; 92:21 - 33. 12. Hallin ES, Feldon SE, Luttrell J. Graves' ophthalmopathy: III. Effect of transantral orbital decompression on optic neuropathy. Br J Ophthalmol 1988; 72:683-7. 13. Warren JD, Spector JG, Burde R. Long-term follow-up and recent observations on 305 cases of orbital decompression for dysthyroid orbitopathy. Laryngoscope 1989; 99:35-40. 14. Hurwitz JJ, Birt D. An individualized approach to orbital decompression in Graves' orbitopathy. Arch Ophthalmol1985; 103:660-5. 15. McCord CD. Orbital decompression for Graves' disease: exposure through lateral canthal and inferior fomix incision. Ophthalmology 1981 ; 88:533-41. 16. Trobe JD, Glaser JS, Laflamme P. Dysthyroid optiC neuropathy: clinical profile and rationale for management. Arch Ophthalmol 1978; 96: 1199-209. 17. Anderson RL, Linberg JV. Transorbital approach to decompression in Graves' disease. Arch Ophthalmol1981; 99:120-4. 18. Koornneef L. Eyelid and orbital fascial attachments and their clinical significance. Eye 1988; 2:130-4. 19. Fells P. Orbital decompression for severe dysthyroid eye disease. Br J Ophthalmol1987; 71:107-11. 20. Werner SC. Modification of the classification of the eye changes of Graves' disease: recommendations of the Ad Hoc Committee of the American Thyroid Association [Letter]. J Clin Endocrinol Metab 1977; 44:203-4. 21 . Stranc M, West M . A four-wall orbital decompression for dysthyroid orbitopathy. J Neurosurg 1988; 68:671-7.
641
