- Email: [email protected]
Surgical treatment of severe dysthyroid ophthalmopathy — long-term results
Surgical treatment of severe dysthyroid ophthalmopathy - long-term results R. Ron6evid, D. Ron6evid
Clinic for Burns, Plastic and Reconstructive Surgery (Head: Prof. M. Karapandzic, MD, DMD, PhD), University Clinical Centre, Zvecanska 9, 11000 Beograd, Yugoslavia
SUMMAR Y. Conservative therapy in cases of severe dysthyroid ophthalmopathy (DO) has not given satisfactory results. Since 1986, 51 patients (99 orbits) with marked DO have been treated by surgical decompression of the orbits. The pr0gedure begins within the upper eyelid. The excision of the excessive skin and subcutaneous tissue is performedrand eyelid fat is removed. In order to obtain the correction of the upper eyelid retraction, the levator aponeurosis is divided by 213 transverse incisions in its central part and Muller's muscle is sectioned at the level of its tarsal insertion. Through an incision in the lower eyelid, the posterior part of the orbital floor, the lateral orbit wall, as well as the periorbital and intraorbital fat are removed. Through an incision made over the medial margin of the orbit, the ethmoidal part of the medial orbital wall and the retrobulbar fat are removed. The periorbital periosteum should be incised at several sites. After operation all patients showed a significant reduction of exophthalmos (5-11 mm, 7.16 mm on average), significant reduction of intraocular pressure, marked improvement in ocular muscle function, as well as considerable reduction or disappearance of subjective symptoms. There was an improvement in vision in 68 % patients who had impaired vision before the operation. There were no cases of subsequent impairment of vision or ocular motility. Mild relapse was recorded in three cases only and only one patient required unilateral reoperation. Strabismus surgery had to be performed in five patients due to unsatisfactory correction of double vision. It can be concluded that this method of orbital decompression gives very good functional and aesthetical long-term results.
INTRODUCTION
and 'four wall decompression' (Kennerdal and Maroon 1982), in which removal of the lateral half of the orbital roof is added to 'three-wall decompression' produce a better effect of decompression in comparison with previous methods. Roncevic and Jackson (1989) described a modified three-wall decompression with removal of the periorbital, intraorbital and retrobulbar fat with the significant effect of decompression. Olivari (1991) reported transpalpebral decompression by removal of the intraorbital fat. Personal observations are presented in this paper and modifications of the surgical approach for correction of severe DO, as well as functional and aesthetic long-term results of the treatment.
Various terms have been used for dysthyroid endocrine Graves' ophthalmopathy, i.e. exophthalmos. The pathogenesis of this condition remains speculative. Conservative treatment of marked dysthyroid ophthalmopathy (DO), both medical management and radiation therapy, does not give satisfactory results. Because of this, surgical treatment, i.e. orbital decompression, is now being used with increasing frequency. Various types of surgical decompression have been performed over the years. Dollinger (1891) performed the first recorded orbital decompression by removing a portion of the lateral orbital wall. Hirsch and Urbaneck (1930) proposed the removal of the orbital floor. Naffziger (1931) introduced the concept of resection of the orbital roof via a neurosurgical approach. Sewell (1936) reported the resection of the medial orbital wall. Walsh and Ogura (1957) introduced antral-ethmoidal decompression, i.e. removal of the orbital floor and medial wall via the transantral approach. Tessier (1969) and Wolfe (1979) introduced the concept of orbital expansion. Their procedures basically changed the concept of treating DO. Anderson and Lindberg (1981) removed the orbital floor and medial wall via a translid approach. 'Threewall decompression' in which lateral wall decompression is added to antral-ethmoidal decompression
MATERIALS AND METHODS Since 1986, 51 patients (99 orbits) with marked DO (exophthalmos 24-33mm) have been treated by surgical decompression of the orbits. There were 41 female patients, aged 18-73 years, and 10 male patients, aged 22-51 years. The majority of the patients had the typical clinical symptoms and findings of headache, diplopia, disturbance of eye motility, manifestations of optic neuropathy of varying degrees, increased intraocular pressure, eyelid retraction, oedema and hyperaemia of the conjunctiva. Decompression of the orbit is performed after a 355
356 Journal of Cranio Maxillo-Facial Surgery
,a,
la
C
D
I:
F
Fig. 1 - (A) Preoperative view of the patient with exophthalmos of 30 mm. (B) The preoperative CT scan shows marked elongation of the retrobulbar cone. (C) Appearance of patient 2 years after operation. (D) The postoperative CT scan shows the retrobulbar cone to be greatly shortened. (E) Appearance of patient 4 years after operation. (F) Appearance of patient 9 years after operation.
complete endocrine and ophthalmic assessment and treatment. The procedure for the correction begins within the upper eyelid. As for a blepharoplasty, the excision of the excessive skin and subcutaneous tissue is performed, the eyelid fat is removed and sometimes even the part of the enlarged lacrimal gland which
protrudes beyond the orbit. In order to obtain the correction of the upper eyelid retraction, the levator aponeurosis is divided by 2-3 transverse incisions in its central part and Muller's muscle is sectioned at the level of its tarsal insertion. An incision is made in the lower eyelid, 3 mm below the margin, and through
Surgical treatment of severe dysthyroid ophthalmopathy
357
B
C
this, the floor and the lateral wall of the orbit are explored and periorbital and intraorbital fat is removed, as much as possible. Using a small chisel and haemostat, the posterior part of the orbital floor and the lateral orbital wall are removed; the orbital margin remains intact. The bony bridge between the floor and the lateral wall defect is removed to produce a large continuous orbitectomy. The size of the defect is related to the severity of the exophthalmos. An attempt should be made to remove the bone from the orbital floor as close to the orbital apex as possible; the anterior part of the orbital floor should remain intact. The infraorbital nerve is protected during the floor osteotomy. This is achieved by removing bone on either side and then freeing the nerve. If possible, the sinus mucosa should not be opened. Small perforations of the sinus mucosa are frequently unavoidable, but are of no importance. Incision of the periorbital periosteum should be made in several sites. Following this, through an incision made over the medial margin of the orbit, the medial orbital wall is explored and its ethmoidal part is removed. This medial wall defect should be directed toward the orbital floor, and continuity should be established with the defect created previously. Using this same approach, the retrobulbar space is explored, and
Fig. 2 - (A) Preoperative view of the patient with exophthalmos of 29 mm. The face is rounded and swollen as a result of long term steroid therapy. (B) Appearance of patient 1 month after surgery. (C) Appearance of patient 5 years after surgery.
again, as much fat as possible is removed with a haemostat. The periorbital periosteum in this region should be incised at several sites. After establishment of satisfactory haemostasis, a thin vacuum drain is inserted into the retrobulbar space, and the incisions are closed. At the beginning of the operation, temporary blepharorrhaphy should be performed using two single sutures. These sutures are removed 5-7 days after the operation. In all cases, the operation was performed on both orbits in a single operation. After the operation, in all patients the subjective symptoms are observed, as well as the changes of eyelids, conjunctiva and cornea, the degree of exophthalmos by Hertel, the function of extraocular muscles, i.e. diplopia by Hess-Lancaster test, intraocular pressure, the changes in the disc, i.e. the changes in vision and field of vision, as well as the thickening and the elongation of the extraocular muscles by CT scan and echography.
RESULTS Although the results are evident immediately after the operation, the definitive results were not obtained until 2-3 months thereafter. Postoperatively, in all
358 Journal of Cranio Maxillo-Facial Surgery
A
B
C
Fig. 3 - (A) Preoperative view of the patient with asymmetric exophthalmos measuring 29 mm on the fight and 25 mm on the left. (13)Appearance of patient 1 year after surgery. (C) Appearance of patient 8 years after surgery.
A
B
Fig. 4 - (A) Preoperative view of the patient with exophthalmos of 33 ram. (B) Appearance of patient 21 days after surgery.
patients subjective s y m p t o m s are considerably reduced or have disappeared. The swelling o f the eyelids disappeared or was significantly reduced. D u e to unsatisfactory correction o f the eyelid retraction it was necessary to p e r f o r m lengthening o f the upper eyelid in three patients and lengthening o f the lower
eyelid in one patient. Conjunctival hyperaemia and chemosis, as well as exposure keratitis resolved in all patients. All patients showed a significant reduction o f exophthalmos (5-11 mm, 7.16 m m on average) with m a r k e d i m p r o v e m e n t or normalisation o f ocular
Surgical treatment of severe dysthyroid ophthalmopathy 359
A
B
C
Fig. 5 - (A) A 18-year-oldpatient after thyroidectomywith exophthalmos of 27 mm. (B) Appearance of patient 14 days after surgery. (C) Appearance of patient two years after surgery.
A
B
Fig. 6 - (A) Preoperative view of the patient after inadequately performed antral-ethmoidal decompressionwith injury to extraocular muscles. (B) Appearance of patient after described decompression of the orbits and chondroplasty of the anterior part of the orbital floors.
muscle function (Figs 1-7; Tables 1, 2). Due to the unsatisfactory correction of double vision, strabismus surgery had to be performed in five patients. Of 99 DOs performed, there was impairment of vision in 58
eyes, in 33 due to optic neuropathy and in 25 due to corneal changes. After the operation, the majority of these patients (68 %) showed improvement of vision and improvement or elimination of the field defect
360 Journal of Cranio Maxillo-Facial Surgery
A
B
Fig. 7 - (A) Preoperative view of patient after Naffziger's decompression and radiation therapy with pulsatile exophthalmos of 30 mm and with corneal opacification on the left. (B) Appearance of patient four years after described orbital decompression and osteoplasty of the orbital roofs.
Table 1 Changes in exophthalmos 3 months after orbital decompression (n = 99)
Reduction in mm
Number of eyes
5 6 7 8 9 10 11
9 19 36 22 9 3 1
DISCUSSION
Table 2 - Diplopia before and three months after orbital decompression (n = 51)
Before surgery After surgery
Without diplopia
Diplopia on lateral gaze
Diplopia on straight gaze
13 28
26 18
12 5
Table 3 - Vision changes three months after orbital decompression (n = 58)
Improvement Unchanged Impairment
had transient hypoaesthesia or anaesthesia of the infraorbital region. Finally, significant aesthetic improvement has been achieved in all patients.
40 18 0
(Table 3). In the patients with partial atrophy of the disc and irreversible changes in the cornea, there was no improvement of vision. There were no cases of subsequent impairment of vision or ocular motility. The intraocular pressure, measured by KPa, before operation was 1.5-4.0, mostly 3.0-3.5 (average 2.8). After the operation, the pressure was 1.5-3.0, mostly 1.5-2.5 (average 1.9). Milder relapse was recorded in only 3 cases and in only one patient was unilateral reoperation required. Echography and CT scan showed great reduction of extraocular muscle thickening and shortening of the retrobulbar cone in all operated patients. All patients
Conservative treatment rarely gives satisfactory improvement in cases of severe DO. Surgical treatment, i.e. orbital decompression, is indicated in such cases. Morax and HurbIi (1987) have highlighted the indications for surgical treatment most concisely and realistically. It should be noted that there is no excuse for further application of conservative treatment which does not give satisfactory results. The irreversible changes of optic nerve and extraocular muscles, i.e. disc atrophy and extraocular muscles fibrosis, should be avoided. If disc atrophy has developed, no improvement in vision can be achieved by orbital decompression. If extraocular muscles fibrosis has occurred, neither exophthalmos nor diplopia can be corrected satisfactorily by orbital decompression. Significant symmetrical decompression, particularly in the narrow orbital apex where there is the highest pressure, is achieved by the surgical treatment described. The elimination of this pressure is necessary for normalisation of the function of all orbital structures, especially of the function of the optic nerves and extraocular muscles. In addition, if the eyelid swelling and retraction is corrected, all signs and symptoms of DO will disappear or will be significantly decreased. If performed with precision, this procedure is without high risk (Roncevic and Jackson, 1989). Of course, the main effect of decompression is achieved by the widening of the orbit, i.e. by the removal of the orbital walls, but removal of fat has an important effect, especially in cases with severe DO. The removal of fat only, without the removal of the orbital walls (Olivari, 1991), cannot produce sat-
Surgical treatment of severe dysthyroid ophthalmopathy
isfactory decompression in cases with marked DO. By applying this method, without an additional operation, eyelid retraction cannot be satisfactorily corrected. Four-wall decompression can provide some disadvantages. The surgical procedure lasts longer, a part of the brain can prolapse through the opening in the orbital roof which leads to a decrease in the orbital volume, causing a pulsatile eye or pulsatile exophthalmos (Fig. 7). Transfrontal approach (Bartalena et al., 1989) with the removal of the roof, the lateral wall and part of the orbital floor can produce the same complication. Besides, this approach has a poorer decompression effect compared with KennerdellMaroon's four-wall decompression. Though, we do not believe that the four-wall decompression has a significantly greater decompression effect in comparison with the three-wall decompression, considering that the dura, i.e. the brain, lies against the opening in the orbital roof. Also, we do not believe that this method can produce globe retrodisplacement ranging from 14 to 16 mm (Kennerdall and Maroon, 1982; McCord, 1985), i.e. from 10 to 17 mm (Shorr et al., 1982). Globe retrodisplacement depends not only on bony opening, i.e. increased orbital volume, but, also, on the amount of periostial opening and condition of the orbital tissues, especially of the condition of the extraocular muscles. In order to obtain the functional as well as the aesthetic effect of decompression, even in severe cases of DO, it is enough to achieve globe retrodisplacement ranging from 7 to 9 mm which can be done using the method described. Globe retrodisplacement ranging from 14 to 16 mm or 17 mm almost always leads to marked enophthalmos. Antral-ethmoidal decompression (Walsh and Ogura, 1957) and all other previously used methods, are archaic methods because of their poor visibility and inadequate results. In the majority of up-to-now applied methods of orbital decompression, the orbital floor is frequently removed in conjunction with medial or medial and lateral wall decompression. Leone et al. (1989) recommend medial and lateral wall decompression with the removal of the lateral orbital rim. They claim that the removal of the orbital floor causes postoperative complications' involving extraocular muscle imbalance, infraorbital anaesthesia, hypo-ophthalmia, and recurrent sinusitis'. Decompression of the lateral wall of the orbit produces a relatively small effect. Removal of the lateral rim does not lead to decompression of the extraocular muscles and optic nerve, but there is always a visible deformity of some degree. If only the posterior part of the orbital floor is removed, it is practically impossible to produce hypoophthalmia. If orbital decompression is performed with precision, residual diplopia is the consequence of the initial disease process and not of the operation. The same will occur by decompression of only the medial and lateral walls if the preoperative exophthalmos was marked with gross diplopia. In such cases, there will be more frequent residual diplopia because of insufficient decompression. If the surgery is
361
performed carefully it is quite possible to avoid serious injury to the infraorbital nerve. Nevertheless, there is always temporary hypoaesthesia or anaesthesia. After decompression of the orbital floor, if the sinus mucosa is injured, the sinus will be filled with blood. Blood clots in the sinus are eliminated spontaneously through the natural sinus ostium as happens after facial bone injuries and orthognathic surgery. None of the patients from the group presented has had a sinusitis so far. Thus there is no reason to avoid decompression of the orbital floor, i.e. removal of its posterior part. Without this procedure it is impossible to achieve adequate decompression of the orbital apex, and it is frequently impossible, with marked exophthalmos, to get good functional and aesthetic effects. Apart from decompression in the medial orbital wall, Pearl et al. (1991) perform osteotomy of part of the lateral and lower orbital wall in the anterior part and the whole bony segment is rotated laterally. It is true that by applying this procedure the orbit is widened in its anterior part and its volume is increased, but this procedure gives a poor decompression effect on the orbital apex where the main problem lies. Similarly, orbital expansion does not give satisfactory decompression in the orbital apex. It has been frequently stated that orbital decompression does not usually improve extraocular muscle function and that after the operation, there is often a deterioration of extraocular muscle function. Waller et al. (1987) claim that decompression does not, as a rule, improve muscle function and that muscle dysfunction and diplopia were more apparent after the operation in the majority of cases, and that nearly 70 % of their total group of patients required one or more strabismus operations after orbital decompression. Shorr et al. (1992) state that they had a 34 % incidence of worsened muscle balance after orbital decompression. Similar statements have been made by De Santo (1980) and Garrity et al. (1992). What can be the cause of non-improvement and even deterioration of extraocular muscle function after orbital decompression? The majority of those who emphasize the poor function of extraocular muscles after decompression of the orbit, performed this operation via a transantral approach. Applying the above-mentioned approach, it is possible to injure the extraocular muscles and cause a worsening of their function, due to poor visibility in the course of the operation. If the anterior part of the orbital floor is removed, by any approach, the globe loses its support and drops toward the sinus (Fig. 7), which results in worsening of the diplopia. Moreover, if the operation is performed in such a way, the pressure in the orbital apex remains, i.e. a sufficient decompression effect, which would enable improvement in extraocular muscle function, is not achieved. In general, if the pressure in the orbital apex is not eliminated, a considerable improvement in extraocular muscle function and optic nerve function cannot be expected. Also, removal of orbital fat contributes to the decrease in pressure in the orbit and to the improvement in optic nerve and extraocular muscle function. Though
362 Journal of Cranio Maxillo-Facial Surgery
lateral decompression gives the least good results, in all the cases of decompression through the floor and medial wall, decompression through the lateral wall should also be performed. Lateral decompression, in addition prevents superfluous medial globe displacement which can also cause worsening of the diplopia. The openings made in all orbital walls should be continuous, without bone bridges between the walls, as well as in the region of the infraorbital nerve. In this way an important symmetrical decompression is achieved which enables the recovery of the optic nerve and balanced decongestion, i.e. the recovery of the extraocular muscles. Unbalanced recovery of the extraocular muscles can worsen diplopia. Strabismus operations should not be performed before orbital decompression. If orbital decompression is not performed before the strabismus operation, the latter will only lead to temporary improvement in extraocular muscle function, because it does not solve the basic cause of the muscle changes. Concerning the correction of eyelid retraction, there are many different opinions. In the course of the same operation it is almost always possible to correct the retraction of the eyelid successfully, using the method described. It is also felt that the blepharorrhaphy, maintained for 5-7 days, helps to maintain the increased levator length and that it is unnecessary in most cases to insert an interpositional graft to achieve this effect. Temporarily sutured eyelids produce a slight pressure on the globe which has an influence on moving the globe into the newly created space (Roncevic, 1991). It is possible to test the vision through the central part of the eyelids, between the two sutures. The described method of orbital decompression is logical, based on an understanding of the pathology, is less hazardous, is relatively easy to perform, gives consistently good functional and aesthetic results and allows rapid recovery.
Acknowledgement This study was supported by a grant from the Ministry of Science of Serbia.
References Anderson, R. L., J. V. Linberg : Transantral approach to decompression in Graves' disease. Arch. Ophthalmol. 99 (1981) 120-124 Bartalena, L., C. Marcoeci, F. Bogazzi, M. Panicuccu, G. BrunoBossio, R. Cantini, A. Lepri, G. Tusini, A. Pinchera : Orbital decompression for severe Graves' ophthalmopathy. J. Neurosurg. Sci. 33 (1989) 223-227
De Santo, L. W.: The total rehabilitation of Graves' disease. Laryngoscope 90 (1980) 1652-1678 Dollinger, J. : Die Druckentlastung der Augenh6hle durch Entfernung der fiusseren Orbitawand bei hochgradigem Exophtalmus (Morbus Basedowi) und konsekutiver Hornhauterkrankung. Dtsch. Med. Wochenschr. 37 (1911) 1888-1890 Garrity, J. A., D. D. Saggau, C. A. Bartley, V. Fatoureehi, C. A. Gorman: Torsional diplopia after transantral orbital decompression and extraocular muscle surgery associated with Graves' orbitopathy. Am. J. Ophthalmol. 113 (1992) 363-373 Hirsh, V. 0., J. Urbaneck- Behandlung eines exzessiven Exophthalmus (Basedow) dutch Entfernung von Orbitalfett yon der Kieferh6hle aus. Monatsschr. Ohrheilkd. Laryngorhinol. 64 (1930) 212-213 Kennerdell, J. S., J. C. Maroon: An orbital decompression for severe dysthyroid exophthalmos. Ophthalmology 89 (1982) 467-472 Leone, C. R. Jr., K. L. Piest, R. J. Newman: Medial and lateral wall decompression for thyroid ophthalmopathy. Am. J. Ophthalmol. 108 (1989) 160-166 McCord, C. D. Jr. : Current trends in orbital decompression. Ophthalmology 89 (1985) 21-33 Morax, S., T. Hurbli: Choice of surgical treatment for Graves' disease. J. Craniomaxillofac. Surg. 15 (1987) 174-181 Naffziger, H. C. : Progressive exophthalmos following thyroidectomy: its pathology and treatment. Ann. Surg. 94 (582-586) 1931 Olivari, N.: Transpalpebral decompression of endocrine ophthalmopathy/Graves' disease/by removal of intraorbital fat. Plast. Reconstr. Surg. 87 (1991) 627-641 Pearl, R. M., L. Vistenes, S. Troxel: Treatment of exophthalmos. Plast. Reconstr. Surg. 87 (1991) 236-244 Ron?eci~, R. : Thyrotoxic exophthalmos. Plast. Reconstr. Surg. 88 (1991) 371-372 Ron~evid, R., L T. Jackson: Surgical treatment of thyrotoxic exophthalmos. Plast. Reconstr. Surg. 84 (1989) 754-760 Sewall, E. C. : Operative control of progressive exophthalmos. Arch. Otolaryngol. 24 (1936) 621-624 Shorr, N., R. W. Neuhaus, H. L Baylis : Ocular motility problems after orbital decompression for dysthyroid ophthalmopathy. Ophthalmology 89 (1982) 323-328. Tessier, P.." Les exophthalmies. Expansion chirurgicale de l'orbite. Ann. Chit. Plast. 14 (1969) 207-214 Thawley, S. E., J. Ogura, J. R. Jacobs: Transantral approach for orbital decompression, In: Smith B. C. : Ophthalmic Plastic and Reconstructive Surgery, Mosby, St. Louis 1987, 1393-1400 Waller, R. R., L. W. De Santo, R. L Anderson." Management of thyroid ophthalmopathy, In: Smith B. C. : Ophthalmic Plastic and Reconstructive Surgery, Mosby, St. Louis 1987, 136%1392 Walsh, T. E., J. H. Ogura: Transantral orbital decompression for malignant exophthalmos. Laryngoscope 64 (1957) 544-568 Wolfe, S. A. : Modified three-wall orbital expansion to correct persistent exophthalmos or exorbitism. Plast. Reconstr. Surg. 64 (1979) 448-455
Dr R. Ron~:evifi, MD, DMD, PhD Clinic for Burns, Plastic and Reconstructive Surgery University Clinical Centre Zvecanska 9 11000 Beograd Yugoslavia Paper received 20 January 1995 Accepted 21 September 1995
Clinic for Burns, Plastic and Reconstructive Surgery (Head: Prof. M. Karapandzic, MD, DMD, PhD), University Clinical Centre, Zvecanska 9, 11000 Beograd, Yugoslavia
SUMMAR Y. Conservative therapy in cases of severe dysthyroid ophthalmopathy (DO) has not given satisfactory results. Since 1986, 51 patients (99 orbits) with marked DO have been treated by surgical decompression of the orbits. The pr0gedure begins within the upper eyelid. The excision of the excessive skin and subcutaneous tissue is performedrand eyelid fat is removed. In order to obtain the correction of the upper eyelid retraction, the levator aponeurosis is divided by 213 transverse incisions in its central part and Muller's muscle is sectioned at the level of its tarsal insertion. Through an incision in the lower eyelid, the posterior part of the orbital floor, the lateral orbit wall, as well as the periorbital and intraorbital fat are removed. Through an incision made over the medial margin of the orbit, the ethmoidal part of the medial orbital wall and the retrobulbar fat are removed. The periorbital periosteum should be incised at several sites. After operation all patients showed a significant reduction of exophthalmos (5-11 mm, 7.16 mm on average), significant reduction of intraocular pressure, marked improvement in ocular muscle function, as well as considerable reduction or disappearance of subjective symptoms. There was an improvement in vision in 68 % patients who had impaired vision before the operation. There were no cases of subsequent impairment of vision or ocular motility. Mild relapse was recorded in three cases only and only one patient required unilateral reoperation. Strabismus surgery had to be performed in five patients due to unsatisfactory correction of double vision. It can be concluded that this method of orbital decompression gives very good functional and aesthetical long-term results.
INTRODUCTION
and 'four wall decompression' (Kennerdal and Maroon 1982), in which removal of the lateral half of the orbital roof is added to 'three-wall decompression' produce a better effect of decompression in comparison with previous methods. Roncevic and Jackson (1989) described a modified three-wall decompression with removal of the periorbital, intraorbital and retrobulbar fat with the significant effect of decompression. Olivari (1991) reported transpalpebral decompression by removal of the intraorbital fat. Personal observations are presented in this paper and modifications of the surgical approach for correction of severe DO, as well as functional and aesthetic long-term results of the treatment.
Various terms have been used for dysthyroid endocrine Graves' ophthalmopathy, i.e. exophthalmos. The pathogenesis of this condition remains speculative. Conservative treatment of marked dysthyroid ophthalmopathy (DO), both medical management and radiation therapy, does not give satisfactory results. Because of this, surgical treatment, i.e. orbital decompression, is now being used with increasing frequency. Various types of surgical decompression have been performed over the years. Dollinger (1891) performed the first recorded orbital decompression by removing a portion of the lateral orbital wall. Hirsch and Urbaneck (1930) proposed the removal of the orbital floor. Naffziger (1931) introduced the concept of resection of the orbital roof via a neurosurgical approach. Sewell (1936) reported the resection of the medial orbital wall. Walsh and Ogura (1957) introduced antral-ethmoidal decompression, i.e. removal of the orbital floor and medial wall via the transantral approach. Tessier (1969) and Wolfe (1979) introduced the concept of orbital expansion. Their procedures basically changed the concept of treating DO. Anderson and Lindberg (1981) removed the orbital floor and medial wall via a translid approach. 'Threewall decompression' in which lateral wall decompression is added to antral-ethmoidal decompression
MATERIALS AND METHODS Since 1986, 51 patients (99 orbits) with marked DO (exophthalmos 24-33mm) have been treated by surgical decompression of the orbits. There were 41 female patients, aged 18-73 years, and 10 male patients, aged 22-51 years. The majority of the patients had the typical clinical symptoms and findings of headache, diplopia, disturbance of eye motility, manifestations of optic neuropathy of varying degrees, increased intraocular pressure, eyelid retraction, oedema and hyperaemia of the conjunctiva. Decompression of the orbit is performed after a 355
356 Journal of Cranio Maxillo-Facial Surgery
,a,
la
C
D
I:
F
Fig. 1 - (A) Preoperative view of the patient with exophthalmos of 30 mm. (B) The preoperative CT scan shows marked elongation of the retrobulbar cone. (C) Appearance of patient 2 years after operation. (D) The postoperative CT scan shows the retrobulbar cone to be greatly shortened. (E) Appearance of patient 4 years after operation. (F) Appearance of patient 9 years after operation.
complete endocrine and ophthalmic assessment and treatment. The procedure for the correction begins within the upper eyelid. As for a blepharoplasty, the excision of the excessive skin and subcutaneous tissue is performed, the eyelid fat is removed and sometimes even the part of the enlarged lacrimal gland which
protrudes beyond the orbit. In order to obtain the correction of the upper eyelid retraction, the levator aponeurosis is divided by 2-3 transverse incisions in its central part and Muller's muscle is sectioned at the level of its tarsal insertion. An incision is made in the lower eyelid, 3 mm below the margin, and through
Surgical treatment of severe dysthyroid ophthalmopathy
357
B
C
this, the floor and the lateral wall of the orbit are explored and periorbital and intraorbital fat is removed, as much as possible. Using a small chisel and haemostat, the posterior part of the orbital floor and the lateral orbital wall are removed; the orbital margin remains intact. The bony bridge between the floor and the lateral wall defect is removed to produce a large continuous orbitectomy. The size of the defect is related to the severity of the exophthalmos. An attempt should be made to remove the bone from the orbital floor as close to the orbital apex as possible; the anterior part of the orbital floor should remain intact. The infraorbital nerve is protected during the floor osteotomy. This is achieved by removing bone on either side and then freeing the nerve. If possible, the sinus mucosa should not be opened. Small perforations of the sinus mucosa are frequently unavoidable, but are of no importance. Incision of the periorbital periosteum should be made in several sites. Following this, through an incision made over the medial margin of the orbit, the medial orbital wall is explored and its ethmoidal part is removed. This medial wall defect should be directed toward the orbital floor, and continuity should be established with the defect created previously. Using this same approach, the retrobulbar space is explored, and
Fig. 2 - (A) Preoperative view of the patient with exophthalmos of 29 mm. The face is rounded and swollen as a result of long term steroid therapy. (B) Appearance of patient 1 month after surgery. (C) Appearance of patient 5 years after surgery.
again, as much fat as possible is removed with a haemostat. The periorbital periosteum in this region should be incised at several sites. After establishment of satisfactory haemostasis, a thin vacuum drain is inserted into the retrobulbar space, and the incisions are closed. At the beginning of the operation, temporary blepharorrhaphy should be performed using two single sutures. These sutures are removed 5-7 days after the operation. In all cases, the operation was performed on both orbits in a single operation. After the operation, in all patients the subjective symptoms are observed, as well as the changes of eyelids, conjunctiva and cornea, the degree of exophthalmos by Hertel, the function of extraocular muscles, i.e. diplopia by Hess-Lancaster test, intraocular pressure, the changes in the disc, i.e. the changes in vision and field of vision, as well as the thickening and the elongation of the extraocular muscles by CT scan and echography.
RESULTS Although the results are evident immediately after the operation, the definitive results were not obtained until 2-3 months thereafter. Postoperatively, in all
358 Journal of Cranio Maxillo-Facial Surgery
A
B
C
Fig. 3 - (A) Preoperative view of the patient with asymmetric exophthalmos measuring 29 mm on the fight and 25 mm on the left. (13)Appearance of patient 1 year after surgery. (C) Appearance of patient 8 years after surgery.
A
B
Fig. 4 - (A) Preoperative view of the patient with exophthalmos of 33 ram. (B) Appearance of patient 21 days after surgery.
patients subjective s y m p t o m s are considerably reduced or have disappeared. The swelling o f the eyelids disappeared or was significantly reduced. D u e to unsatisfactory correction o f the eyelid retraction it was necessary to p e r f o r m lengthening o f the upper eyelid in three patients and lengthening o f the lower
eyelid in one patient. Conjunctival hyperaemia and chemosis, as well as exposure keratitis resolved in all patients. All patients showed a significant reduction o f exophthalmos (5-11 mm, 7.16 m m on average) with m a r k e d i m p r o v e m e n t or normalisation o f ocular
Surgical treatment of severe dysthyroid ophthalmopathy 359
A
B
C
Fig. 5 - (A) A 18-year-oldpatient after thyroidectomywith exophthalmos of 27 mm. (B) Appearance of patient 14 days after surgery. (C) Appearance of patient two years after surgery.
A
B
Fig. 6 - (A) Preoperative view of the patient after inadequately performed antral-ethmoidal decompressionwith injury to extraocular muscles. (B) Appearance of patient after described decompression of the orbits and chondroplasty of the anterior part of the orbital floors.
muscle function (Figs 1-7; Tables 1, 2). Due to the unsatisfactory correction of double vision, strabismus surgery had to be performed in five patients. Of 99 DOs performed, there was impairment of vision in 58
eyes, in 33 due to optic neuropathy and in 25 due to corneal changes. After the operation, the majority of these patients (68 %) showed improvement of vision and improvement or elimination of the field defect
360 Journal of Cranio Maxillo-Facial Surgery
A
B
Fig. 7 - (A) Preoperative view of patient after Naffziger's decompression and radiation therapy with pulsatile exophthalmos of 30 mm and with corneal opacification on the left. (B) Appearance of patient four years after described orbital decompression and osteoplasty of the orbital roofs.
Table 1 Changes in exophthalmos 3 months after orbital decompression (n = 99)
Reduction in mm
Number of eyes
5 6 7 8 9 10 11
9 19 36 22 9 3 1
DISCUSSION
Table 2 - Diplopia before and three months after orbital decompression (n = 51)
Before surgery After surgery
Without diplopia
Diplopia on lateral gaze
Diplopia on straight gaze
13 28
26 18
12 5
Table 3 - Vision changes three months after orbital decompression (n = 58)
Improvement Unchanged Impairment
had transient hypoaesthesia or anaesthesia of the infraorbital region. Finally, significant aesthetic improvement has been achieved in all patients.
40 18 0
(Table 3). In the patients with partial atrophy of the disc and irreversible changes in the cornea, there was no improvement of vision. There were no cases of subsequent impairment of vision or ocular motility. The intraocular pressure, measured by KPa, before operation was 1.5-4.0, mostly 3.0-3.5 (average 2.8). After the operation, the pressure was 1.5-3.0, mostly 1.5-2.5 (average 1.9). Milder relapse was recorded in only 3 cases and in only one patient was unilateral reoperation required. Echography and CT scan showed great reduction of extraocular muscle thickening and shortening of the retrobulbar cone in all operated patients. All patients
Conservative treatment rarely gives satisfactory improvement in cases of severe DO. Surgical treatment, i.e. orbital decompression, is indicated in such cases. Morax and HurbIi (1987) have highlighted the indications for surgical treatment most concisely and realistically. It should be noted that there is no excuse for further application of conservative treatment which does not give satisfactory results. The irreversible changes of optic nerve and extraocular muscles, i.e. disc atrophy and extraocular muscles fibrosis, should be avoided. If disc atrophy has developed, no improvement in vision can be achieved by orbital decompression. If extraocular muscles fibrosis has occurred, neither exophthalmos nor diplopia can be corrected satisfactorily by orbital decompression. Significant symmetrical decompression, particularly in the narrow orbital apex where there is the highest pressure, is achieved by the surgical treatment described. The elimination of this pressure is necessary for normalisation of the function of all orbital structures, especially of the function of the optic nerves and extraocular muscles. In addition, if the eyelid swelling and retraction is corrected, all signs and symptoms of DO will disappear or will be significantly decreased. If performed with precision, this procedure is without high risk (Roncevic and Jackson, 1989). Of course, the main effect of decompression is achieved by the widening of the orbit, i.e. by the removal of the orbital walls, but removal of fat has an important effect, especially in cases with severe DO. The removal of fat only, without the removal of the orbital walls (Olivari, 1991), cannot produce sat-
Surgical treatment of severe dysthyroid ophthalmopathy
isfactory decompression in cases with marked DO. By applying this method, without an additional operation, eyelid retraction cannot be satisfactorily corrected. Four-wall decompression can provide some disadvantages. The surgical procedure lasts longer, a part of the brain can prolapse through the opening in the orbital roof which leads to a decrease in the orbital volume, causing a pulsatile eye or pulsatile exophthalmos (Fig. 7). Transfrontal approach (Bartalena et al., 1989) with the removal of the roof, the lateral wall and part of the orbital floor can produce the same complication. Besides, this approach has a poorer decompression effect compared with KennerdellMaroon's four-wall decompression. Though, we do not believe that the four-wall decompression has a significantly greater decompression effect in comparison with the three-wall decompression, considering that the dura, i.e. the brain, lies against the opening in the orbital roof. Also, we do not believe that this method can produce globe retrodisplacement ranging from 14 to 16 mm (Kennerdall and Maroon, 1982; McCord, 1985), i.e. from 10 to 17 mm (Shorr et al., 1982). Globe retrodisplacement depends not only on bony opening, i.e. increased orbital volume, but, also, on the amount of periostial opening and condition of the orbital tissues, especially of the condition of the extraocular muscles. In order to obtain the functional as well as the aesthetic effect of decompression, even in severe cases of DO, it is enough to achieve globe retrodisplacement ranging from 7 to 9 mm which can be done using the method described. Globe retrodisplacement ranging from 14 to 16 mm or 17 mm almost always leads to marked enophthalmos. Antral-ethmoidal decompression (Walsh and Ogura, 1957) and all other previously used methods, are archaic methods because of their poor visibility and inadequate results. In the majority of up-to-now applied methods of orbital decompression, the orbital floor is frequently removed in conjunction with medial or medial and lateral wall decompression. Leone et al. (1989) recommend medial and lateral wall decompression with the removal of the lateral orbital rim. They claim that the removal of the orbital floor causes postoperative complications' involving extraocular muscle imbalance, infraorbital anaesthesia, hypo-ophthalmia, and recurrent sinusitis'. Decompression of the lateral wall of the orbit produces a relatively small effect. Removal of the lateral rim does not lead to decompression of the extraocular muscles and optic nerve, but there is always a visible deformity of some degree. If only the posterior part of the orbital floor is removed, it is practically impossible to produce hypoophthalmia. If orbital decompression is performed with precision, residual diplopia is the consequence of the initial disease process and not of the operation. The same will occur by decompression of only the medial and lateral walls if the preoperative exophthalmos was marked with gross diplopia. In such cases, there will be more frequent residual diplopia because of insufficient decompression. If the surgery is
361
performed carefully it is quite possible to avoid serious injury to the infraorbital nerve. Nevertheless, there is always temporary hypoaesthesia or anaesthesia. After decompression of the orbital floor, if the sinus mucosa is injured, the sinus will be filled with blood. Blood clots in the sinus are eliminated spontaneously through the natural sinus ostium as happens after facial bone injuries and orthognathic surgery. None of the patients from the group presented has had a sinusitis so far. Thus there is no reason to avoid decompression of the orbital floor, i.e. removal of its posterior part. Without this procedure it is impossible to achieve adequate decompression of the orbital apex, and it is frequently impossible, with marked exophthalmos, to get good functional and aesthetic effects. Apart from decompression in the medial orbital wall, Pearl et al. (1991) perform osteotomy of part of the lateral and lower orbital wall in the anterior part and the whole bony segment is rotated laterally. It is true that by applying this procedure the orbit is widened in its anterior part and its volume is increased, but this procedure gives a poor decompression effect on the orbital apex where the main problem lies. Similarly, orbital expansion does not give satisfactory decompression in the orbital apex. It has been frequently stated that orbital decompression does not usually improve extraocular muscle function and that after the operation, there is often a deterioration of extraocular muscle function. Waller et al. (1987) claim that decompression does not, as a rule, improve muscle function and that muscle dysfunction and diplopia were more apparent after the operation in the majority of cases, and that nearly 70 % of their total group of patients required one or more strabismus operations after orbital decompression. Shorr et al. (1992) state that they had a 34 % incidence of worsened muscle balance after orbital decompression. Similar statements have been made by De Santo (1980) and Garrity et al. (1992). What can be the cause of non-improvement and even deterioration of extraocular muscle function after orbital decompression? The majority of those who emphasize the poor function of extraocular muscles after decompression of the orbit, performed this operation via a transantral approach. Applying the above-mentioned approach, it is possible to injure the extraocular muscles and cause a worsening of their function, due to poor visibility in the course of the operation. If the anterior part of the orbital floor is removed, by any approach, the globe loses its support and drops toward the sinus (Fig. 7), which results in worsening of the diplopia. Moreover, if the operation is performed in such a way, the pressure in the orbital apex remains, i.e. a sufficient decompression effect, which would enable improvement in extraocular muscle function, is not achieved. In general, if the pressure in the orbital apex is not eliminated, a considerable improvement in extraocular muscle function and optic nerve function cannot be expected. Also, removal of orbital fat contributes to the decrease in pressure in the orbit and to the improvement in optic nerve and extraocular muscle function. Though
362 Journal of Cranio Maxillo-Facial Surgery
lateral decompression gives the least good results, in all the cases of decompression through the floor and medial wall, decompression through the lateral wall should also be performed. Lateral decompression, in addition prevents superfluous medial globe displacement which can also cause worsening of the diplopia. The openings made in all orbital walls should be continuous, without bone bridges between the walls, as well as in the region of the infraorbital nerve. In this way an important symmetrical decompression is achieved which enables the recovery of the optic nerve and balanced decongestion, i.e. the recovery of the extraocular muscles. Unbalanced recovery of the extraocular muscles can worsen diplopia. Strabismus operations should not be performed before orbital decompression. If orbital decompression is not performed before the strabismus operation, the latter will only lead to temporary improvement in extraocular muscle function, because it does not solve the basic cause of the muscle changes. Concerning the correction of eyelid retraction, there are many different opinions. In the course of the same operation it is almost always possible to correct the retraction of the eyelid successfully, using the method described. It is also felt that the blepharorrhaphy, maintained for 5-7 days, helps to maintain the increased levator length and that it is unnecessary in most cases to insert an interpositional graft to achieve this effect. Temporarily sutured eyelids produce a slight pressure on the globe which has an influence on moving the globe into the newly created space (Roncevic, 1991). It is possible to test the vision through the central part of the eyelids, between the two sutures. The described method of orbital decompression is logical, based on an understanding of the pathology, is less hazardous, is relatively easy to perform, gives consistently good functional and aesthetic results and allows rapid recovery.
Acknowledgement This study was supported by a grant from the Ministry of Science of Serbia.
References Anderson, R. L., J. V. Linberg : Transantral approach to decompression in Graves' disease. Arch. Ophthalmol. 99 (1981) 120-124 Bartalena, L., C. Marcoeci, F. Bogazzi, M. Panicuccu, G. BrunoBossio, R. Cantini, A. Lepri, G. Tusini, A. Pinchera : Orbital decompression for severe Graves' ophthalmopathy. J. Neurosurg. Sci. 33 (1989) 223-227
De Santo, L. W.: The total rehabilitation of Graves' disease. Laryngoscope 90 (1980) 1652-1678 Dollinger, J. : Die Druckentlastung der Augenh6hle durch Entfernung der fiusseren Orbitawand bei hochgradigem Exophtalmus (Morbus Basedowi) und konsekutiver Hornhauterkrankung. Dtsch. Med. Wochenschr. 37 (1911) 1888-1890 Garrity, J. A., D. D. Saggau, C. A. Bartley, V. Fatoureehi, C. A. Gorman: Torsional diplopia after transantral orbital decompression and extraocular muscle surgery associated with Graves' orbitopathy. Am. J. Ophthalmol. 113 (1992) 363-373 Hirsh, V. 0., J. Urbaneck- Behandlung eines exzessiven Exophthalmus (Basedow) dutch Entfernung von Orbitalfett yon der Kieferh6hle aus. Monatsschr. Ohrheilkd. Laryngorhinol. 64 (1930) 212-213 Kennerdell, J. S., J. C. Maroon: An orbital decompression for severe dysthyroid exophthalmos. Ophthalmology 89 (1982) 467-472 Leone, C. R. Jr., K. L. Piest, R. J. Newman: Medial and lateral wall decompression for thyroid ophthalmopathy. Am. J. Ophthalmol. 108 (1989) 160-166 McCord, C. D. Jr. : Current trends in orbital decompression. Ophthalmology 89 (1985) 21-33 Morax, S., T. Hurbli: Choice of surgical treatment for Graves' disease. J. Craniomaxillofac. Surg. 15 (1987) 174-181 Naffziger, H. C. : Progressive exophthalmos following thyroidectomy: its pathology and treatment. Ann. Surg. 94 (582-586) 1931 Olivari, N.: Transpalpebral decompression of endocrine ophthalmopathy/Graves' disease/by removal of intraorbital fat. Plast. Reconstr. Surg. 87 (1991) 627-641 Pearl, R. M., L. Vistenes, S. Troxel: Treatment of exophthalmos. Plast. Reconstr. Surg. 87 (1991) 236-244 Ron?eci~, R. : Thyrotoxic exophthalmos. Plast. Reconstr. Surg. 88 (1991) 371-372 Ron~evid, R., L T. Jackson: Surgical treatment of thyrotoxic exophthalmos. Plast. Reconstr. Surg. 84 (1989) 754-760 Sewall, E. C. : Operative control of progressive exophthalmos. Arch. Otolaryngol. 24 (1936) 621-624 Shorr, N., R. W. Neuhaus, H. L Baylis : Ocular motility problems after orbital decompression for dysthyroid ophthalmopathy. Ophthalmology 89 (1982) 323-328. Tessier, P.." Les exophthalmies. Expansion chirurgicale de l'orbite. Ann. Chit. Plast. 14 (1969) 207-214 Thawley, S. E., J. Ogura, J. R. Jacobs: Transantral approach for orbital decompression, In: Smith B. C. : Ophthalmic Plastic and Reconstructive Surgery, Mosby, St. Louis 1987, 1393-1400 Waller, R. R., L. W. De Santo, R. L Anderson." Management of thyroid ophthalmopathy, In: Smith B. C. : Ophthalmic Plastic and Reconstructive Surgery, Mosby, St. Louis 1987, 136%1392 Walsh, T. E., J. H. Ogura: Transantral orbital decompression for malignant exophthalmos. Laryngoscope 64 (1957) 544-568 Wolfe, S. A. : Modified three-wall orbital expansion to correct persistent exophthalmos or exorbitism. Plast. Reconstr. Surg. 64 (1979) 448-455
Dr R. Ron~:evifi, MD, DMD, PhD Clinic for Burns, Plastic and Reconstructive Surgery University Clinical Centre Zvecanska 9 11000 Beograd Yugoslavia Paper received 20 January 1995 Accepted 21 September 1995