Results of Müllerotomy and Levator Aponeurosis Transposition for the Correction of Upper Eyelid Retraction in Graves Disease

Results of Müllerotomy and Levator Aponeurosis Transposition for the Correction of Upper Eyelid Retraction in Graves Disease

Results of Miillerotomy and Levator Aponeurosis Transposition for the Correction of Upper Eyelid Retraction in Graves Disease Emily]. Ceisler, MD, 1 ]...

6MB Sizes 0 Downloads 74 Views

Results of Miillerotomy and Levator Aponeurosis Transposition for the Correction of Upper Eyelid Retraction in Graves Disease Emily]. Ceisler, MD, 1 ]urij R. Bilyk, MD, 1 Peter A D. Rubin, MD, 1 William R. Burks, MD, 2 John W. Shore, MD 1•3 Background: Upper eyelid retraction in Graves disease may cause functional mor­ bidity and aesthetic deformity. Surgery to correct thyroid-related upper eyelid retraction may result in temporal undercorrection with failure to eliminate lateral eyelid retraction, leading in turn to a poor eyelid contour postoperatively. Methods: In 1984, one of the authors developed a new procedure for correcting moderate to severe upper eyelid retraction associated with Graves disease. The surgical technique consists of a MOIIerotomy and recession of the levator aponeurosis combined with medial transposition of the lateral horn of the levator aponeurosis. The procedure was performed on 37 patients (72 eyelids). Muller's muscle was used as the spacer to set the eyelid height. Transposition of the levator aponeurosis allowed adjustment of eyelid contour. Results: Thirty patients (58 eyelids) had excellent results, six (13 eyelids) had good results, and one (1 eyelid) had a poor result. No patient required re-operation for asym­ metry, unacceptable contour, or malposition. Only one eyelid had significant overcor­ rection, and only one eyelid had significant undercorrection, requiring further surgery. The most frequent unwanted effects were high eyelid crease (24 eyelids) and residual temporal flare (6 eyelids); however, most of these were seen early in the series before the lateral levator transposition modification was added. Conclusion: This procedure allows successful and simultaneous correction of both eyelid position and contour in patients with moderate to severe thyroid-related upper eyelid retraction. Ophthalmology 1995;102:483-492

Originally received: June 8, 1994. Revision accepted: October 16, 1994. 1 Department of Ophthalmology, Eye Plastics and Orbit Service, Mas­ sachusetts Eye and Ear Infirmary, Boston. 2 Ophthalmology Consultants, Margate, Florida. 3 Ophthalmic Consultants of Boston, Boston. Presented as a poster at the American Academy of Ophthalmology An­ nual Meeting, Chicago, November 1993. Each author states that he/she has no proprietary interest in the devel­ opment or marketing of this technique. Reprint requests to John W. Shore, MD, Ophthalmic Consultants of Boston, 50 Staniford St, 6th Fl, Boston, MA 02114.

Upper eyelid retraction is a well-recognized manifestation of ophthalmic Graves disease. It causes both functional morbidity in the form of exposure keratopathyjconjunc­ tivitis and aesthetic deformity. Many techniques have been described to correct upper eyelid retraction ofGraves disease. In our experience, these procedures have two major shortcomings: unpredictability with inconsistent eyelid position (over- and undercorrec­ tion) and contour abnormalities manifesting primarily as temporal undercorrection of the upper eyelid recession. The unpredictable absorption rate of autogenous and banked tissue spacers and other associated complications

483

Ophthalmology

Volume 102, Number 3, March 1995

Figure 1. Incision through skin and orbicularis oculi muscle at the level of the upper eyelid crease.

limit their usefulness in upper eyelid recession surgery. One of us (JWS) developed a surgical technique for cor­ rection of thyroid-related upper eyelid retraction, con­ sisting of Mullerotomy and recession of the levator apo­ neurosis combined with medial transposition of the lateral horn of the levator aponeurosis. The incision through Muller's muscle, near its origin, partially denervates the muscle and augments the upper eyelid recession. Muller's muscle then acts as an autogenous spacer to set the eyelid height. Transposition of the aponeurosis allows for ad­ justment of eyelid contour.

the orbital rim and secured with 6-0 Vicryl sutures passed through the substance of gland and the arcus marginalis. The skin and orbicularis muscle are elevated from the levator aponeurosis and tarsus by sharp dissection in the plane of the suborbicular fascia. In this way, the preseptal and pretarsal portions of the orbicularis oculi muscle are elevated off the levator aponeurosis and tarsal plate (Fig 2A). The levator aponeurosis is divided by making a hor­ izontal incision at the upper border of the tarsal plate. Muller's muscle is visualized deep to the levator aponeu­ rosis and it is not incised at this time. The levator apo­ neurosis is separated from the underlying Muller's muscle (Fig 2B). The peripheral vascular arcade is identified and avoided. Sharp dissection is used at all times during this stage of surgery. Blunt dissection and "rubbing" of the tissues to separate the levator aponeurosis from Muller's muscle result in bleeding or tearing of fibers, which is not desired. Sharp dissection is facilitated by the scleral pro­ tective lens, which acts as a dissecting platform, and by posterior/inferior traction placed on Muller's muscle/ conjunctiva by the surgical assistant with a cotton-tipped applicator. Sharp-tipped Westcott scissors are used to cut

Surgical Technique The patient is prepared for surgery and draped in the su­ pine position. A short-acting intravenous anxiolytic and amnestic agent is administered (midazolam, propofol) to minimize discomfort and blur the memory of the local anesthetic injection. The upper eyelid crease is marked with a marking pen and compared with the contralateral side to ensure symmetry. A protective scleral shell is placed over the eye, and a subcutaneous injection of local an­ esthetic containing epinephrine in a concentration of 1: 200,000 and hyaluronidase is administered. A 4-0 silk suture is passed through the upper eyelid margin centered above the pupil, and gentle inferior traction is applied to the upper eyelid. An incision is made through skin and orbicularis oculi muscle along the mark using a scalpel (Fig 1). Hemostasis is achieved with bipolar cautery. The orbital septum is identified and incised with Westcott scissors. The septal incision is enlarged medially and lat­ erally. The preaponeurotic fat pad is sharply dissected off the underlying levator aponeurosis. The fat pad is trimmed in those patients who need debulking and sculpting of the upper eyelid. The medial fat pad of the upper eyelid is debulked, if necessary. The lacrimal gland is carefully avoided, unless it has prolapsed and is present in the wound anterior to superolateral orbital rim. In this case, the orbital lobe ofthe lacrimal gland is repositioned behind

484

Figure 2. A, the levator aponeurosis is incised along the upper tarsal border. B, the dis inserted levator aponeurosis is reflected to show the underlying Muller's muscle. The peripheral vascular arcade is avoided (arrows).

Ceisler et al · Upper Eyelid Retraction in Graves Disease and separate the adhesions between the levator aponeu­ rosis and Muller's muscle. As the dissection continues superiorly in the eyelid, the undersurface of the levator aponeurosis appears as a white glistening fibrous band. In contrast, Muller's muscle is dark red in color and the muscle fibers are vertically oriented (cephalad to caudad). At times, the muscle fibers appear to be infiltrated with fat (i.e., the muscle itself has the appearance of fat, but a vertical orientation remains). The fibers have the consis­ tency of fat, but bleed liberally when cut; conversely, fat normally does not bleed when incised. Fatty infiltration and degeneration of Muller's muscle are common findings in patients with Graves orbitopathy. Bleeding, when it occurs, is from the fibers of Muller's muscle or conjunc­ tival vessels deep to the muscle. The bleeding usually can be controlled with bipolar cautery. The dissection plane between the levator aponeurosis and Muller's muscle is well-defined laterally and less well-defined medially. The dissection ends at the level of the superior transverse lig­ ament of Whitnall. Buttonholes in the posterior lamella need not be closed. Muller's muscle is partially denervated by a horizontal incision placed high in Muller's muscle, just inferior to Whitnall's ligament. First, Muller's muscle is infiltrated with local anesthetic containing epinephrine (Fig 3). A horizontal incision is made in Muller's muscle just inferior to Whitnall's ligament (Fig 4). The incision is carried me­ dially and laterally in a curvilinear fashion to parallel the upper border of tarsus. Again, the orbital lobe of the lac­ rimal gland is avoided. The underlying conjunctiva is not incised. Muller's muscle now is dissected sharply from the underlying conjunctiva and reflected inferiorly. Bleeding from the cut muscle fibers and the conjunctiva is common and may be ignored or controlled with a light application of bipolar cautery. We have found this dis­ section is not facilitated by radio frequency or pinpoint cautery tips, which shrink conjunctiva and shorten the posterior lamella of the upper eyelid. Sharp dissection with

Figure 3. With the detached levator (L) dissected free of underlying Muller's muscle (M) and retracted superiorly, a plane is hydrodissected between Muller's muscle and conjunctiva with local anesthetic. The su· perior border of the tarsal plate (T) also is seen.

Figure 4. After Muller's muscle is detached superiorly, it is reflected inferiorly and sharply dissected free of underlying conjunctiva.

Westcott scissors is the preferred method to separate Muller's muscle from underlying palpebral conjunctiva. Blunt dissection is ineffective. The dissection continues toward the eyelid margin until Muller's muscle is hinged and attached only at the upper border of tarsus. This dis­ section is presumed to effect at least partial denervation of Muller's muscle. The attachment of Muller's muscle at the upper tarsal border allows it to be used as a spacer for further lengthening of the entire upper eyelid. Alter­ natively, Muller's muscle may be excised if a spacer is not needed to lengthen the upper eyelid to the desired level. The levator aponeurosis is grasped with forceps and placed on stretch. The open blades of the Westcott scissors are introduced on either side of the levator aponeurosis such that the lateral horn of the levator aponeurosis can be divided (Fig 5). An incision is made in the horn just lateral to the temporal limbus and medial to the lacrimal gland. The tips of the scissors are aimed at the frontozy­ gomatic suture. An angling, yet vertically oriented, inci­ sion is made through the lateral horn of the levator apo­ neurosis. The incision in the aponeurosis continues in the superotemporal direction to the level of Whitnall's liga­ ment. In severe cases, the lateral portion of the superior transverse ligament of Whitnall also is divided. Care is

485

Ophthalmology

Volume 102, Number 3, March 1995 upper eyelid. The more the flap is moved medially, the more relaxation occurs in the lateral retractor mechanism of the upper eyelid and the more relaxation of the tem­ poral one half of the upper eyelid occurs. The rotation of the cut levator aponeurosis allows adjustment of the tem­ poral contour of the upper eyelid. The cut lateral horn of the levator aponeurosis is allowed to retract into the wound.

Figure 5. The cut edge of the levator aponeurosis is grasped, and the lateral horn (slotted line) is incised medial to the lacrimal gland. Muller's muscle is reflected inferiorly (M).

taken to avoid cutting into the orbital lobe of the lacrimal gland. An incision in the orbital lobe of the lacrimal gland will not harm the gland, but will result in bleeding that is difficult to control. As the medial aspect of the incised levator aponeurosis releases from the lateral horn, the en­ tire retractor mechanism of the upper eyelid is believed to "give." The incised nasal edge of the divided levator aponeurosis is grasped, and inferomedial traction is ap­ plied. This stretches the levator aponeurosis in an infe­ ronasal direction. The Westcott scissors are placed at the apex of the vertical incision in the levator aponeurosis, and the tips are angled toward the midline of the forehead. A back cut is made in the levator aponeurosis (Fig 6). This incision is medial to the lacrimal gland (Fig 6C). Thus, the lacrimal ductules are protected from injury. Again, the levator aponeurosis is believed to give as the medial relaxing incision in the central levator aponeurosis is made with the scissors. The back cut is continued to the midline of the levator aponeurosis or beyond. The cut edge of the aponeurosis is mobilized and transposed medially as a rotational/transpositional flap (Fig 7A). As the divided edge of the levator aponeurosis is moved medially, the <,trc of rotation adds length to the

486

Figure 6. A and B, once the lateral horn of the levator is detached, a back cut is made medially just inferior to Whitnall's ligament. C, the levator aponeurosis is reflected to demonstrate the position of the inner blade of the Westcott scissors. Notice that the lacrimal gland (arrow) is lateral to the incision.

Ceisler et al · Upper Eyelid Retraction in Graves Disease In women, a suture pass to incorporate deeper tissue also is made medially, in the midline, and laterally, thus cre­ ating a well-defined upper eyelid crease (Fig 8). In severe cases of eyelid retraction, the 4-0 silk traction sutures are taped to the cheek for 72 hours postoperatively.

Case Reports Case 1. A 37-year-old woman had bilateral upper eyelid retraction, eyelid lag, and lagophthalmos 3 months after iodine 131 (iodine 13 1) ablation of the thyroid gland for hyperthyroidism (Fig 9A). A diagnosis of Graves orbitopathy was made. On replacement therapy of levothyroxine (0.15 mg daily), thyroid function studies were repeated, and the results were normal. The bellies of the extraocular muscles were enlarged, and the tendons were of normal caliber on computed tomography scans. The orbital apex was only modestly crowded. The orbital involvement was symmetrical. Exposure keratopathy developed, prompting referral for surgery. Visual acuity was 20/20 in each eye. The patient had upper eyelid retraction in both eyes with lagophthalmos, eyelid lag,' poor eyelid closure, and exposure keratopathy of both eyes. There was no clinical evidence of restrictive myopathy or compressive optic neuropathy. Hertel exophthalmometry showed measurements of 22 mm in the right eye and 22 mm in the left with a base of90 mm. The palpebral

Figure 7. A, the lateral horn is seen grasped by the forceps . Its original position before transposition is indicated by the arro~. B, the lateral horn of the levator aponeurosis is transposed medially and sutured to the superior edge of Muller's muscle. The parasagittal view clearly dem­ onstrates the reconfiguration of eyelid anatomic structures.

The transposed lateral edge of the central levator apo­ neurosis is now stitched temporally to the cut superior edge of Muller's muscle with interrupted 7-0 silk sutures (Fig 7B). The position and tightness of the sutures are governed by the effect on the curvature and position of the upper eyelid. The more the aponeurosis is transposed medially, the less elevating power remains in the lateral eyelid and the more flattening of the temporal eyelid con­ tour is achieved. The medial transposition of the lateral levator aponeurosis effectively reduces the temporal flare ofthyroid-related eyelid retraction. Medial overcorrection (ptosis) is avoided by leaving the medial horn ofthe levator aponeurosis intact. The knots are temporally tied with a bow knot. The procedure is repeated on the contralateral side, if necessary. The scleral shells are removed, and the patient is asked to open the eyelids. The effect is observed with the patient in the sitting position. Adjustments are made to the position and tension of the sutures to set the final eyelid contour and position. Once symmetry of height and contour is achieved, square knots are tied. Skin and orbicularis muscle are sutured closed in a running fashion.

Figure 8. In women, in whom a well-defined upper eyelid crease is de­ sirable, a suture is passed through the anterior lamella and the superior border of tarsus/Muller's muscle.

487

Ophthalmology

Volume 102, Number 3, March 1995 and orbital findings were stable, and reconstructive eyelid surgery was offered for persistent exposure keratopathy. On preoperative examination in 1985, the patient was euthyroid while receiving replacement levothyroxine (0.15 mg daily). Corrected visual acuity was recorded as 20/20 in both eyes. Exophthalmos mea­ sured by Hertel exophthalmometry was 22.5 mm in both eyes with base of 93 mm. There was I mm of lagophthalmos on gentle eyelid closure. Eyelid lag and upper eyelid retraction were present (Fig lOA). The upper eyelids rested 1.5 mm above the upper corneosclerallimbus. The lower eyelids rested against the globe I mm inferior to the inferior corneoscleral limbus. The palpebral fissures measured 14.5 mm in both eyes. Results of ocular motility examination showed generalized restriction; however, the patient was free of diplopia. Results of slit-lamp biomicroscopy showed bilateral inferior punctate epithelial ker­ atopathy due to nocturnal lagophthalmos and an incomplete blink. Result of funduscopic examination were normal, as were results of the color vision and static perimetry tests. In February 1985, the patient underwent bilateral upper eyelid recession by medial levator aponeurosis transposition and Miillerotomy as described. The patient made an uneventful recovery and was followed for 3 years with stable eyelid position and contour. On the most recent examination, the palpebral fissures measured

Figure 9. A, the preoperative appearance of the patient before denervating Miillerotomy and upper eyelid levator aponeurosis recession with medial transposition. Notice the temporal flare and the marked eyelid retraction. B, the 6-month postoperative photograph demonstrates a normal eyelid contour, eyelid position, and crease and fold.

fissures measured 14 mm in the right eye and !3.5 mm in the left. There were 4 mm of upper scleral show on the right and 3.5 mm on the left. The temporal flare of the upper eyelid contour, typical of patients with Graves orbitopathy, was present bilaterally. In 1984, the patient underwent bilateral upper eyelid recession, medial levator aponeurosis transposition, and a Miillerotomy as described above. Postoperatively, excellent eyelid levels, contours, and symmetrical eyelid creases and folds were achieved (Fig 98). Six months postoperatively, the palpebral fissures measured 8 mm in each eye. The lagophthalmos and exposure keratopathy had resolved. The eyelid lag had improved by 80%. The temporal flare noted preoperatively was no longer present. The results were unchanged at last follow-up, 24 months after surgery. Case 2. A 30-year-old woman had symmetrical eyelid re­ traction, restricted ocular motility, proptosis, lagophthalmos, and exposure keratopathy secondary to thyroid-related orbitopathy. She had noticed eyelid retraction as early as 198!, 3 months after a renal artery bypass for left renal artery stenosis and re­ novascular hypertension. Hyperthyroidism was diagnosed 6 months later. Thyroid suppression and then ablation with 1131 was accomplished in 198!. The orbit and eyelid changes pro­ gressed slowly over the next 18 months. By 1984, the ophthalmic

488

Figure 10. A, preoperative photograph demonstrates upper eyelid re­ traction. (The pupils were dilated with tropicamide only.) B, 6 months after surgery, the eyelid contour and level are nearly normal, with a 1­ mm residual superior scleral show on the right. Symmetry to within 1.5 mm has been established. The eyelid crease and fold are well defined.

Ceisler et al · Upper Eyelid Retraction in Graves Disease The patient was treated conservatively with artificial tear sup­ plements for 3 months and was followed closely for any pro­ gression of her thyroid-related orbitopathy. Despite using arti­ ficial tears, the patient continued to have symptoms and signs of exposure keratopathy. No proptosis, diplopia, or optic neu­ ropathy was noted. The patient underwent bilateral levator transposition and Miillerotomy, with Frost sutures placed for 1 week. Postoperatively, the patient's upper eyelid retraction re­ solved and has remained stable for 3 years, with 2 mm of la­ gophthalmos bilaterally. The patient subsequently underwent hard palate-mucosal grafting to correct lower eyelid retraction (Fig liB).

Results

Figure 11. A, preoperative appearance of the patient. Notice the signif­ icant degree of upper and lower eyelid retraction. B, postoperative ap­ pearance after upper and lower eyelid surgery.

10.5 mm in the right eye and 9.5 mm in the left. The right upper eyelid margin rested 0.5 mm above the limbus. The eyelid con­ tour was symmetric, and temporal flare was not observed. There was no change of the upper eyelid position with a challenge of topical 2.5% phenylephrine applied to each cornea (Fig lOB). The punctate epithelial keratopathy disappeared without apply­ ing topical ocular lubricants or nocturnal eyelid taping. There was no lagophthalmos. Eyelid lag persisted and was unchanged. The patient was satisfied with the result, which was graded as good, and did not desire further eyelid surgery. Case 3. A 34-year-old woman was first seen in October 1989 with bilateral ocular foreign body sensation. Hyperthy­ roidism had been diagnosed 6 months previously, and the patient had undergone treatment with 1131 and partial thyroidectomy for goiter. She was otherwise healthy, although she did admit to a history of heavy alcohol use and continued to smoke one pack of cigarettes daily. Results of ophthalmologic examination dem­ onstrated corrected visual acuity of 20/20 in each eye, with no evidence of optic neuropathy by pupillary examination and Is­ hihara color-plate testing. Extraocular motility was decreased slightly in upgaze, but the patient did not have diplopia. Hertel exophthalmometry showed no significant proptosis. Upper and lower eyelid retraction was noted bilaterally, greater on the right, with an associated von Graefe sign and lagophthalmos. Upper scleral show (2-3 mm) was present (Fig llA). Punctate epithelial keratopathy was noted bilaterally on slit-lamp biomicroscopy. Results ofdilated funduscopic examination were unremarkable.

Thirty-seven patients (72 eyelids) with varying degrees of upper eyelid retraction underwent surgery by this tech­ nique from 1984 to 1994 (Table 1). Thirty-one patients (62 eyelids) required orbital decompression. Two patients who had asymmetric eyelid retraction required only uni­ lateral internal Mi.illerectomy on the opposite eyelid to establish symmetry. Exposure keratopathy was the indi­ cation for surgery in 34 patients (67 eyelids). Three pa­ tients (5 eyelids) were operated on for symptoms of ex­ posure without evidence of keratopathy or for aesthetic correction. Twenty-nine patients (58 eyelids) had lago­ phthalmos of 2 mm or greater on gentle eyelid closure. All patients were euthyroid at the time of surgery and had stable eyelid positions for at least 6 months before surgery. None had undergone previous eyelid surgery. Final results were graded as excellent in 30 patients (58 eyelids}, good in 6 ( 13 eyelids}, and poor in 1 ( 1 eyelid) (Table 1). Thirty-two of 35 patients (with bilateral pro­ cedures) achieved symmetric results. Lid lag was improved postoperatively in 70 eyelids. Early in the course of the study, a 2.5% phenylephrine challenge was used postop­ eratively in five patients and failed to produce eyelid re­ traction, indicative of at least a partial denervation of Muller's muscle. Before recognizing the need for aggressive lateral levator transposition, temporal undercorrection occurred in six eyelids (Table 2). All subsequent procedures included lat­ eral levator transposition with no further incidence ofun­ dercorrection. Other residual thyroid-related eyelid find-

Table 1. General Results (37 patients, 72 eyelids) Results Symmetric (bilateral cases with end result within 1.5 mm) Excellent (eyelid margin above pupil and 1 mm or more below the superior limbus, and symmetry within 1.5 mm) Good (eyelid encroaching on pupil or less than 1 mm from the superior limbus) Poor (eyelid margin above the limbus) Lid lag improved postoperatively

No. of Patients (eyelids) 32

30 (58) 6 (13) 1 (1)

36 (70)

489

Ophthalmology

Volume 102, Number 3, March 1995

Table 2. Residual Thyroid-related Eyelid Findings Findings

No. of Patients (eyelids)

Residual temporal flare Lid lag Lagophthalmos of > 1 mm Residual punctate epithelial keratopathy

31 (64) 2 (3) 4 (6)

(6*)

• All occurred early in series.

ings were eyelid lag (31 patients, 64 eyelids) which were asymptomatic, lagophthalmos of greater than 1 mm with gentle eyelid closure (2 patients, 3 eyelids), and residual punctate epithelial keratopathy (4 patients, 6 eyes). Postoperatively, a high eyelid crease (> lO mm) was seen in 24 eyelids (Table 3); however, these cases were early in the series. More significant complications were rare. One eyelid was overcorrected (encroaching on the pupil). One eyelid was undercorrected (above the limbus) and underwent surgery again, which was successful. One eyelid showed late retraction at 6 months, leading to un­ dercorrection. There were no cases of asymmetry leading to early or late re-operation where the sole purpose was to establish symmetry. There were no cases of unaccept­ able contour or eyelid malposition requiring re-operation. No overcorrection resulted in visually obstructing ptosis. No patient had persistent edema or required subsequent tarsorrhaphy or orbital decompression. There were no in­ fectious complications.

Discussion The cause of upper eyelid retraction in Graves disease is not completely understood. The etiology appears to be multifactorial. One accepted etiologic factor is increased sympathetic tone, causing overaction of Muller's mus­ cle. I-4 The eyelid retraction may or may not improve with treatment of the hyperthyroidism, despite return of the serum T3/T4 to normal levels. A second component is inflammation, fibrosis, and contracture of the inferior rectus muscle, causing increased tone and overaction of the synergistic superior rectus and levator muscles. 2-4 Le­ vator muscle hypertrophy was reported by Small 5 as a cause of hyperfunction of the upper eyelid; however, this was disputed by Feldon and Levin 6 in a volumetric study ofthe levator-superior rectus complex evaluated by com­ puted tomography scans. The authors implicate abnormal adhesions between the levator muscle and surrounding fixed orbital tissues as an important cause of the upper lid retraction. 2·6 Grove 7 noted similar adhesions, as well as fatty infiltration of the levator and Muller's muscle, as important contributors to the problem. Other factors may include eyelid inflammation and levator muscle contrac­ ture, fibrosis, and atrophy.ns Proptosis also may be a factor by causing a pseudo-eyelid retraction; however,

490

eyelid retraction is seen in many patients without pro­ ptosis.2·9 Thyroid-related upper lid retraction (and its conse­ quences) can be treated with medication or by undergoing surgery. Of the medication options, sympatholytic agents such as guanethidine, thymoxamine, reserpine, propran­ olol, and bethanidine have offered the best results. 9· 10 Un­ fortunately, long-term use is limited by local ocular tox­ icity as well as transient and variable efficacy. 9 · 1° Corti­ costeroids also may be ofsome use, primarily in the early, more inflammatory stages of the disease, by decreasing pseudo-eyelid retraction from proptosis as well as by de­ creasing inflammation and contraction in the eyelid re­ tractors. Surgery is the treatment of choice. Most authors agree that, unless emergent surgery is required for medically uncontrollable exposure keratopathy, one should be sure that the eyelid position has been stable for 6 to 12 months before undergoing surgery. Many surgical techniques have been described in the literature. In 1934, Goldstein 11 performed levator apo­ neurosis recession (inadvertently including Muller's mus­ cle), anchoring it under the brow with mattress sutures, using either a skin or conjunctival approach. Moran 12 cut through the levator aponeurosis and Muller's muscle hor­ izontally, varying the length of the incision based on test­ ing the extent ofthe correction by having the patient open and close the eyelid. von Blaskovics and Kettesy 13 severed the levator aponeurosis (and presumably Muller's muscle) from the superior tarsal border and retracted it with su­ tures. Henderson 14 varied his procedure: for 2 to 3 mm of retraction, he severed Muller's muscle from tarsus; for greater degrees of retraction, he also severed the levator aponeurosis from tarsus. Baylis and colleagues 15 used transconjunctival Mullerectomy for small amounts of re-

Table 3. Unwanted Effects No. of Eyelids High eyelid crease (> 10 mm) Overcorrection at 6 mos (eyelid encroaching on pupil) Undercorrection at 6 mos (eyelid above the limbus) Late retractions at 6 mos leading to undercor­ rection Asymmetry requiring early or late re-operation Persistent edema Unacceptable contour requiring revision Overcorrection resulting in visually obstruct­ ing ptosis Re-operation to correct eyelid malposition Subsequent tarsorrhaphy Subsequent orbital decompression • All occurred early in series.

24*

1 None None None None None None None

Ceisler et al · Upper Eyelid Retraction in Graves Disease traction and transcutaneous levator aponeurosis recession for larger amounts of retraction. It is difficult to compare their results with other series, however, because many of these patients had eyelid retraction for reasons other than Graves disease. 15 Putterman and co-workers3•4 · 16 - 18 have published ex­ tensively on the subject of thyroid-related upper eyelid retraction. In 1972, Putterman and U rist 3 presented two new techniques designed to improve on the unsatisfactory results they had obtained using previously described pro­ cedures. The authors believed that the poor results were due to the inability to quantitate the amount of surgery required to correct variable amounts of eyelid retraction. They described an internal, graded excision of Muller's muscle with or without recession of the levator aponeu­ rosis depending on the amount of correction needed. Ad­ justments could be made intraoperatively by using a sen­ sory eyelid block. Patients were awake, and eyelid height could be monitored and manipulated at surgery by pe­ riodically having them sit. Good results with this tech­ nique were later reported in large numbers of patients. The initial procedure was further modified to include only the temporal two thirds of Muller's muscle in the myec­ tomy because several early patients had had nasal over­ correction (ptosis). Putterman and Urist also noted that eyelid retraction frequently developed in the contralateral eyelid after unilateral upper eyelid surgery due to Hering's law of bilateral innervation, an important finding to re­ member when operating on any patient with Graves dis­ ease.4.J6-t8 Grove 8· 19 described the marginal myotomy as a means to weaken the levator muscle. This procedure allowed adhesions between the anterior levator apo­ neurosis and the overlying subcutaneous tissues to be divided. He believed that this procedure better ad­ dressed the subcutaneous inflammatory changes con­ tributing to lid retraction in a way that enhanced post­ operative results. 8·19 Results of pathologic examination of tissue obtained at the time of surgery demonstrated collagen proliferation, striated muscle fiber atrophy, and fatty infiltration. 7 Other authors have offered further variations. Dixon 20 believed Mullerectomy was usually sufficient to correct thyroid-related upper lid retraction. Harvey and Anderson 9 described an anterior approach for levator aponeurosis recession and Muller's muscle extirpation. They recommend cutting the lateral horn of the levator aponeurosis to avoid temporal undercorrection and also performed intraoperative adjustment of eyelid height and contour with patient cooperation. Hurwitz and Rodgers21 preferred Mullerectomy for mole retraction, Mullerec­ tomy with levator aponeurosis weakening for moderate retraction, and either Mullerectomy and levator aponeu­ rectomy or scleral graft implantation for severe retraction. The authors also described procedures to prevent post­ operative lateral undercorrection. An interesting tech­ nique was presented by Kohn 22 in 1982 in which the levator aponeurosis was recessed beyond the tarsal bor­ der, maintaining an attachment to tarsus by two pedicle tarsal rotation flaps. However, the eyelid tends to creep

slightly upward over many years, regardless of the tech­ nique used. Several authors have advocated the use of spacers. Callahan 23 used pretarsal orbicularis or collagen film with levator recession. Preserved sclera has been used widely as a spacer with proponents describing predictable results and minimal complications. 24- 26 A more recent study found only fair results and significant complication and concluded that, for upper eyelid lengthening, scleral graft­ ing offers no distinct advantage over other techniques. 27 We have found persistent eyelid edema and variable ab­ sorption to be a common and frustrating complications ofscleral grafts. Hard palate mucosa is another graft source advocated for various types of eyelid reconstruction, pri­ marily for the lower eyelid. Only a few of the patients described in one series had thyroid-related upper eyelid retraction. 28 Mechanical corneal abrasion is a potential complication because the epithelium of the hard palate is keratinized, unlike conjunctiva. More recently, most authors have advocated some variation on the theme oflevator aponeurosis and Muller's muscle recession or extirpation. 2 ·29 - 32 Modifications of earlier techniques have been suggested, such as recessing only the lateral portion of the levator/Muller's muscle complex and/or cutting the lateral horn of the levator to avoid the frequently observed complication of temporal undercorrection. 3031 We have developed a procedure that combines a graded Mullerotomy and levator recession with medial trans­ position of the levator aponeurosis. In the surgical pro­ cedure described, eyelid malformation is approached an­ atomically, allowing for predictable, successful, and si­ multaneous correction of both eyelid position and contour. Muller's muscle is used as a vascularized spacer to allow adjustment of eyelid height. The partial dener­ vation of the muscle weakens further the upper eyelid retractors and has an additive effect on lowering the eyelid. The anatomic course of innervation to Muller's muscle is not fully defined and may follow several routes, in­ cluding the levator muscle, the superior rectus muscle, and the vascular arcades ofthe eyelids. 33 Anatomic studies on rhesus monkeys by Collin et al 33 suggest that the in­ nervation is redundant, because neither levator transection nor a Fasanel a-Servat procedure completely denervates Muller's muscle. However, this study also confirmed that once sympathetic fibers to the muscle were transected, they failed to regenerate across scar tissue. Thus, although denervation of Muller's muscle using our procedure may not be complete, because the vascular arcades are left in­ tact, it is in all likelihood significant because the superior attachment ofMuller's muscle is transected completely and the muscle is dissected free of underlying conjunctiva. This supposition is supported clinically by the failure to induce eyelid retraction with a phenylephrine challenge postoperatively. The absence of denervating hypersensi­ tivity to phenylephrine challenge is explained by post­ operative fibrosis and scarring of Muller's muscle. Lateral levator aponeurosis transposition is very suc­ cessful in vertically lengthening the eyelid and achieving adjustment of eyelid contour, particularly laterally. The

491

Ophthalmology

Volume 102, Number 3, March 1995

enhanced lateral eyelid retraction typically seen in thyroid­ related orbitopathy may be due to several factors, includ­ ing full-thickness lateral eyelid inflammation and fibro­ sis,34 or selective fibrosis of the lateral levator aponeurosis (unpublished data, Grove AS; presented at the 1988 AAO Annual Meeting) or Muller's muscle (Victor Elner, MD, personal communication). The relatively weaker attach­ ment of the medial retinaculum as compared with the lateral retinaculum may subscribe to this finding. 35 Fi­ nally, the slightly divergent vector force of the tarsoliga­ mentous band to the orbital axis and even more so to the visual axis also may contribute to the lateral eyelid flare. 35 Regardless of the specific etiologic reasons, weakening of the lateral portion ofthe retractors is essential for adequate correction of Graves eyelid malposition. In addition, be­ cause of the weakness of the medial horn of the levator aponeurosis and the preferential age-related ratification ofthe medial retinaculum, 36 the medial horn ofthe levator aponeurosis should be left undisturbed to prevent medial overcorrection (ptosis). Using our technique, the desired eyelid contour is achieved by severing the lateral hm:n of the levator, trans­ posing the distal edge of the aponeurosis as a flap, and attaching it to Muller's muscle medially. The results are predictable and allow correction of both eyelid position and contour in patients with moderate to severe thyroid related upper eyelid retraction. Our results demonstrate that the technique offers both functional and aesthetic success, is reproducible, and achieves excellent correction of both eyelid position and contour. Although it can be used in all case of Graves eyelid retraction, it is most ef­ ficacious in patients with moderate to severe involvement.

References I. Blodi FC. Ophthalmopathy of Graves' disease. In: Sym­

2. 3. 4. 5.

6. 7. 8. 9. 10.

492

posium on Surgery of the Orbit and Adnexa: Trans New Orleans Acad Ophthalmol, 1973. St. Louis: CV Mosby, 1974;101-11. Hedin A. Eyelid surgery in dysthyroid ophthalmopathy. Eye 1988;2:20 1-6. Putterman AM, Urist M. Surgical treatment of upper eyelid retraction. Arch Ophthalmol 1972;87:401-5. Chalfin J, Putterman AM. Muller's muscle excision and levator recession in retracted upper lid. Treatment of thy­ roid-related retraction. Arch Ophthalmol 1979;97: 1487-91. Small RG. Upper eyelid retraction in Graves' ophthalmo­ pathy: a new surgical technique and a study ofthe abnormal levator muscle. Trans Am Ophthalmol Soc 1988;86:725­ 93. Feldon SE, Levin L. Graves' ophthalmopathy: V. Aetiology of upper eyelid retraction in Graves' ophthalmopathy. Br J Ophthalmol 1990;74:484-5. Grove AS Jr. Upper eyelid retraction and Graves' disease. Ophthalmology 1981 ;88:499-506. Grove AS Jr. Eyelid retraction treated by levator marginal myotomy. Ophthalmology 1980;87: I 013-18. Harvey JT, Anderson RL. The aponeurotic approach to eyelid retraction. Ophthalmology 1981 ;88:513-24. Dixon RS, Anderson RL, Hatt MU. The use of thymox­ amine in eyelid retraction. Arch Ophthalmol 1979;97:2147­ 50.

II. Goldstein I. Recession ofthe levator muscle for lagophthal­ mos in exophthalmic goiter. Arch Ophthalmol 1934; II :389­ 93. 12. Moran RE. The correction of exophthalmos and levator spasm. Plast Reconstr Surg 1956; 18:411-26. 13. von Blaskovics L, Kettesy A. Eingriffe am Auge: Eine An­ leitung zur operativen Tiitigkeit des Augenarztes, 2nd ed. Stuttgart: Ferdinand Enke, 1945:37-9. 14. Henderson JW. Relief of eyelid retraction. A surgical pro­ cedure. Arch Ophthalmol 1965;74:205-16. 15. Baylis HI, Cies WA, Kamin OF. Correction of upper eyelid retraction. Am J Ophthalmol 1976;82:790-4. 16. Putterman AM. Surgical treatment ofthyroid-related upper eyelid retraction. Graded Muller's muscle excision and le­ vator recession. Ophthalmology 1981;88:507-12. 17. Putterman AM. Surgical treatment of dysthyroid eyelid re­ traction and orbital fat hernia. Otolaryngol Clin North Am 1980; 13:39-51. 18. Putterman AM, Fett DR. Muller's muscle in the treatment of upper eyelid retraction: a 12-year study. Ophthalmic Surg 1986;17:361-7. 19. Grove AS Jr. Levator lengthening by marginal myotomy. Arch Ophthalmol 1980;98: 1433-8. 20. Dixon R. The surgical management of thyroid-related upper eyelid retraction. Ophthalmology 1982;89:52-7. 21. Hurwitz JJ, Rodgers KJA. Prevention and management of postoperative lateral upper-lid retraction in Graves' disease. Can J Ophthalmol 1983; 18:329-32. 22. Kohn R. Treatment of eyelid retraction with two pedicle tarsal rotation flaps. Am J Ophthalmol 1983;95:539-44. 23. Callahan A. Levator recession: with reattachment to the tarsus with collagen film. Arch Ophthalmol 1965;73:800­ 2. 24. Doxanas MT, Dryden RM. The use ofsclera in the treatment of dysthyroid eyelid retraction. Ophthalmology 1981 ;88: 887-94. 25. Crawford JS, Easterbrook M. The use of bank sclera to cor­ rect lid retraction. Can J Ophthalmol 1976;11:309-22. 26. Dryden RM, Soli DB. The use of scleral transplantation in cicatricial entropion and eyelid retraction. Trans Am Acad Ophthalmol Otolaryngol 1977;83:0P669-78. 27. Mourits MPh, Koornneef L. Lid lengthening by sclera in­ terposition for eyelid retraction in Graves' ophthalmopathy. Br J Ophthalmol 1991;75:344-7. 28. Cohen MS, Shorr N. Eyelid reconstruction with hard palate mucosa grafts. Ophthalmic Plast Reconstr Surg 1992;8: 183­ 95. 29. Levine MR, Chu A. Surgical treatment of thyroid-related lid retraction: a new variation. Ophthalmic Surg 1991 ;22: 90-4. 30. Older JJ. Surgical treatment of eyelid retraction associated with thyroid eye disease. Ophthalmic Surg 1991 ;22:318-23. 31. Harvey JT, Corin S, Nixon D, Veloudios A. Modified levator aponeurosis recession for upper eyelid retraction in Graves' disease. Ophthalmic Surg 1991;22:313-17. 32. Thaller VT, Kaden K, Lane CM, Collin JRO. Thyroid lid surgery. Eye 1987;1:609-14. 33. Collin JRO, Beard C, Wood I. Terminal course of nerve supply to Muller's muscle in the rhesus monkey and its clinical significance. Am J Ophthalmol 1979;87:234-46. 34. Waller RR. Eyelid malpositions in Grave's ophthalmopathy. Trans Am Ophthalmol Soc 1982;80:855-930. 35. Lemke BN. Anatomic considerations in upper eyelid re­ traction. Ophthalmic Plast Reconstr Surg 1991 ;7: 158-66. 36. Shore JW, McCord CD Jr. Anatomic changes in involutional blepharoptosis. Am J Ophthalmol 1984;98:21-7.