figure 1. Müllerectomy for upper eyelid paresis in 4 patients with facial nerve palsy. Eyelid position before (A, E, I, and M) and after (B, F, J, and N) left eye müllerectomy. Lagophthalmos before (C, G, K, and O) and after (D, H, L, and P) left eye müllerectomy. Postoperative effects of lateral (D) and medial (N) canthal tightening of the lower eyelid can be seen.
Figure 2. Effect of müllerectomy and lower eyelid elevation on average ± standard deviation upper eyelid height by group.
figure 3. Effect of müllerectomy and lower eyelid elevation on average ± standard deviation lagophthalmos by group.
Figure 4. Effect of müllerectomy and lower eyelid elevation on superficial punctate keratopathy (SPK) by measuring the average ± standard deviation corneal exposure index in each group. The corneal exposure index is calculated as follows: percentage of the corneal surface with keratopathy × SPK severity (from 0 to +4).
Hassan AS, Frueh BR, Elner VM. Müllerectomy for Upper Eyelid Retraction and Lagophthalmos Due to Facial Nerve Palsy. Arch Ophthalmol. 2005;123(9):1221-1225. doi:10.1001/archopht.123.9.1221
Facial nerve palsy often results in symptoms of ocular irritation due to inadequate eyelid closure. Weakened protractor function results in relative upper eyelid retraction and contributes to lagophthalmos.
To evaluate the role of müllerectomy in the comprehensive surgical treatment of ocular exposure due to facial nerve palsy.
Thirty-four patients with chronic facial nerve palsy underwent unilateral transconjunctival removal of Müller muscle and were followed up for an average of 20 months postoperatively. Other procedures were performed to treat lower eyelid retraction, as required. Preoperative and postoperative ocular exposure symptoms, upper eyelid position, lagophthalmos, and keratopathy were compared.
Of the 59 preoperative symptoms, 15 (25%) resolved and 39 (66%) improved. Upper eyelid position was lowered by an average of 1.35 mm (P<.001). Lagophthalmos (P = .002) and corneal exposure (P<.001) were significantly improved. Three patients required levator aponeurosis repair, 2 for preexisting dehiscence and 1 for inadvertent aponeurosis transection.
Müllerectomy is a rapid, safe, and reproducible surgical method for lowering the upper eyelid and reducing ocular exposure symptoms and signs due to chronic facial nerve palsy.
The chief ocular morbidity of facial nerve palsy is corneal exposure, with its consequent symptoms and signs.1,2 Orbicularis oculi muscle paresis affects upper and lower eyelid function,1 which may be improved by static and reanimating surgical techniques. In the lower eyelid, static methods include canthal ligament tightening,3 midfacial elevation,4 and tarsorrhaphy.2 In the upper eyelid, in addition to tarsorrhaphy, methods include reanimation with nerve grafts,5 gold weight implantation,6 and placement of palpebral springs.5,6
Gold weights are the most common treatment for paresis of upper eyelid closure. However, their use is complicated by cosmetic appearance, upper eyelid and eyeball tenderness, weight-induced lagophthalmos when supine, and inflammation, malposition, or extrusion.7- 9 Palpebral springs are much less commonly used because they require custom manufacture, are difficult to place,10,11 and may extrude, may displace, or may become infected.6,12 Tarsorrhaphy is cosmetically disfiguring and causes objectionable reduction of the visual field. Therefore, a static technique to overcome morbidities due to implantation of foreign materials or tarsorrhaphy may be a welcome addition to the treatment of ocular complications of facial nerve palsy.
Upper eyelid position is determined by the dynamic balance between protractor and retractor tone. Excision of Müller muscle is used for upper eyelid retraction due to Graves eye disease.13,14 Müllerectomy reduces minor degrees of Graves eye disease–related eyelid retraction13 and improves lagophthalmos. We postulated that removal of Müller muscle in patients with upper eyelid orbicularis oculi muscle paresis would improve the balance between protractor and retractor tone in the upper eyelid and thereby reduce symptoms and signs of ocular exposure.
Patients undergoing müllerectomy had ocular symptoms and signs due to facial nerve palsy, for which they were seen at the Eye Plastic and Orbital Surgery Service of the University of Michigan Kellogg Eye Center. They were operated on by 2 of us (B.R.F. and V.M.E.) for symptoms due to orbicularis oculi paresis without prospect of improvement. The presence and severity of discomfort, tearing, asymmetry, photophobia, and pain/burning were documented preoperatively and postoperatively. Also noted were patient age, duration of palsy, stability of paresis, laterality of involvement, and other surgery performed for ocular exposure. Informed consent for surgery was obtained for all patients. This retrospective study was approved by the University of Michigan Institutional Review Board.
The best-corrected visual acuity was obtained preoperatively and postoperatively for all patients. Preoperative and postoperative superficial punctate keratopathy (SPK) was graded from 0 to +4 as follows: 0 indicates none; 1, mild; 2, moderate; 3, severe; and 4, epithelial ulceration. The percentage of corneal surface with SPK multiplied by the SPK grade was used to derive a corneal exposure index.4,15 The preoperative and postoperative positions of the upper eyelid were measured from the midpupil to the upper eyelid margin with the eye in primary gaze and with the coronal plane of the patient’s head perpendicular to the floor. Lagophthalmos was measured during gentle eyelid closure.
All patients had persistent stable eyelid paresis for longer than 6 months or palsies without prospect for further improvement. The indication for surgery was symptomatic exposure keratopathy, usually with lagophthalmos and often with reflex tearing.
Anesthesia consisted of intravenous sedation and local infiltration of the upper eyelid subcutaneously and subconjunctivally with 0.5% bupivacaine hydrochloride mixed in equal parts with 1% lidocaine with 1:100 000 epinephrine or with epinephrine supplementation to 1:50 000 as previously described.16 The upper eyelid was everted, and the conjunctiva was incised along the superior tarsal margin. Dissection of the conjunctiva from Müller muscle was performed superiorly for 15 mm. Müller muscle was then cut from its superior tarsal insertion, dissected from the levator aponeurosis superiorly for 13 to 15 mm, transected near its origin from the inferior surface of the levator muscle, and removed. Care was taken to avoid damage to the lacrimal gland ducts by not disturbing the lateral extensions of the Müller muscle.17 The conjunctivae were then repositioned and either sutured with a continuous 6-0 chromic suture with knots buried or allowed to heal spontaneously. The unsedated patient’s eyelid height and contour were then evaluated while the patient was seated and in primary gaze to ensure intact levator function. The goal was to reduce the signs and symptoms of ocular exposure while preserving the full binocular visual field. When preexisting levator dehiscence resulted in flattening of the upper eyelid contour or frank intraoperative ptosis, one or two 6-0 polyglactin 910 mattress sutures were placed between the levator aponeurosis and tarsal plate to restore a desirable upper eyelid height and contour. Bipolar cautery was used for hemostasis. No temporary tarsorrhaphy or traction sutures were placed.
Additional procedures, as required, were accomplished during the same operation or during the follow-up period. Lateral canthal ligament tightening with 4-0 polypropylene or polyglactin 910 was performed using a modified lateral tarsal strip procedure.3 Medial eyelid laxity was addressed by medial tarsal suspension or medial canthal ligament tightening.18,19 In 2 patients, midfacial elevation was combined with lower eyelid tightening.4 Symptomatic paralytic lower eyelid retraction was addressed when necessary, with spacer grafts consisting of upper eyelid tarsus,20 hard palate,21- 23 or donor sclera.24
Data are expressed as average ± standard deviation. The statistical significance of differences between preoperative and postoperative groups was determined using a 2-tailed t test. Differences between groups were considered statistically significant at P<.05.
Müllerectomy was performed on 34 eyelids of 34 patients, with an average follow-up of 20 months (range, 2-66 months). Seven patients had less than 5 months of follow-up (3 had 3 months and 4 had 2 months). The average patient age was 50 years (range, 10-82 years); there were 19 female and 1. male patients. Palsy duration averaged 7.5 years (range, 0.2-12.0 years), and palsy stability averaged 5.1 years (range, 0.2-12.0 years). Six patients had undergone removal of gold weights complicated by discomfort or extrusion; 1 patient had previously undergone removal of a malpositioned palpebral spring. Müllerectomy alone was performed in 18 of the 34 patients. Sixteen patients underwent additional procedures to elevate the lower eyelid: lateral and medial tightening in 15 (94%), tarsorrhaphy in 2 (12%), and midfacial lift in 2 (12%) of the patients.
The most common ocular symptoms were discomfort and tearing due to exposure (table). Postoperatively, 25% of symptoms resolved and 66% improved. Only 8% remained unchanged; 0 worsened, and all occurred among the 18 patients undergoing müllerectomy alone. Of the 16 patients who underwent müllerectomy plus lower eyelid surgery, all experienced improved or resolved symptoms. Postoperative visual acuity was improved in 11 (32%) of the patients, while visual acuity was unchanged in 18 (53%). Of the 34 patients, 5 (15%) demonstrated reduced visual acuity, 3 due to intercurrent ophthalmic disease unrelated to facial nerve palsy.
The upper eyelid height was reduced by surgery in 32 of the 34 eyelids (Figure 1). The average preoperative and postoperative heights were 4.29 ± 1.33 and 2.93 ± 1.1. mm, respectively. The reduction of height was statistically significant (1.35 ± 1.27 mm, P < .001), regardless of whether patients underwent müllerectomy alone (1.44 ± 1.27 mm, P< .001) or müllerectomy with lower eyelid elevation (1.23 ± 1.29 mm, p = .002) (Figure 2). Procedures used to raise the lower eyelid did not affect the upper eyelid height reduction. Postoperative height reductions for patients followed up for 5 months or more (1.36 ± 1.21 mm) did not differ significantly (p = .91. from those with a follow-up of 2 to 5 months (1.29 ± 1.58 mm).
Preoperative lagophthalmos was present in 30 eyes, measuring 3.03 ± 1.98 mm. The average postoperative lagophthalmos was 2.10 ± 2.1. mm. The difference between preoperative and postoperative lagophthalmos was statistically significant for the 34 patients (p = .002) (Figure 3). The 18 eyes that underwent müllerectomy alone showed a statistically significant reduction in lagophthalmos (1.18 ± 1.33 mm, P = .002), but the 16 eyes that underwent lower eyelid elevation surgery did not (0.66 ± 1.96 mm, P = .20). Of the 18 patients undergoing müllerectomy alone, 8 (44%) exhibited 1 mm or less of lagophthalmos postoperatively; of the 16 patients who underwent additional lower eyelid surgery, 4 (25%. had 1 mm or less of lagophthalmos (p = .25).
Preoperative SPK was present in all 34 eyes. The postoperative reduction in the corneal exposure index was statistically significant for all patients (P<.001) (Figure 4). In addition, the corneal exposure index was significantly improved for patients undergoing müllerectomy alone or in combination with lower eyelid surgery (P<.001 for both). In the former group, the cornea showed no SPK or trace staining of less than 5% of the cornea postoperatively in 9 (50%) of 18 patients, while only 2 (12%) of the 16 patients in the latter group showed similar minimal SPK (p = .02).
Of the 34 patients undergoing upper eyelid müllerectomy, 11 (32%) exhibited preoperative asymmetry of greater than 1 mm, with the affected eyelid being higher in all. Postoperative asymmetry of greater than 1 mm was present in 12 (35%) of the 34 patients, with the operative eyelid being lower in all but 1.
In 2 patients, intraoperative aponeurotic ptosis became evident after Müller muscle excision, due to existing levator aponeurosis dehiscence. These dehiscences, together with a transection of the levator aponeurosis that occurred in one patient, were successfully repaired intraoperatively. No other complications occurred during the operations. No patient developed wound infection, corneal abrasion, or abnormal conjunctival healing.
Exposure keratopathy due to facial nerve paralysis causes reduced visual acuity, discomfort, photophobia, burning, and reactive tearing.1,4 Contributing to ocular exposure are upper eyelid retraction, reduced excursion of eyelid blink, and lagophthalmos.1 Surgical treatment has consisted principally of gold weight1 or spring implantation.5,6 Gold weights may be complicated by infection, extrusion, migration, and cosmetic deformity.7- 9 In addition, discomfort and chronic redness may occur because of the weight of the implant on the surface of the eye or because of gold-induced chronic inflammation surrounding the implant.10,25 These complications lead to removal of the weight in 10% to 15% of the patients.7- 10,25,26 Implanted palpebral springs are difficult to fixate surgically.10,11 Despite modifications,27,28 they exhibit significant rates of infection, extrusion, and chronic inflammation.6,12 Over time, springs can also weaken11 and erode.29 Successful nerve graft reanimation still requires the patient to volitionally contract the orbicularis oculi muscle to blink, decreasing blink frequency and creating facial asymmetry with the action of volitional blinking. The method we describe does not require implantation of alloplastic materials, avoiding the most common complications of gold weights and palpebral springs, and does not increase asymmetry.
Introduced in 1965,14 transconjunctival müllerectomy is a treatment for mild Graves eye disease–associated upper eyelid retraction.13 Its adaptation for treatment of upper eyelid protractor paresis has several theoretical bases. With protractor weakness, retractor tone predominates and the resting position of the upper eyelid is higher. This dynamic interaction provides for a surgical method that weakens retractor tone and resistance to protractor function. By restoring a more physiologic balance between these muscle antagonists, upper eyelid position and closure, despite persistent orbicularis oculi muscle weakness, are improved. By weakening the retractor complex, the preload opposing orbicularis oculi muscle action is reduced, permitting the weaker orbicularis oculi muscle to close the eyelid against reduced resistance. The lower starting point of the upper eyelid reduces the palpebral fissure and, hence, the amount of nonlinear orbicularis oculi force generation required for eyelid closure. Moreover, because Müller muscle contains elastic afferent receptors that increase levator muscle tone when the Müller muscle is stretched,30 its removal should decrease levator tone. Müllerectomy addresses these theoretical considerations in the surgical treatment of upper eyelid protractor paresis.
Our results indicate that müllerectomy improved final upper eyelid height (Figure 2) in 32 (94%) patients with facial nerve palsy, while not increasing cosmetic postoperative asymmetry. Preoperative asymmetry is due to the paretic eyelid being higher than the contralateral eyelid, while postoperative asymmetry occurs because the paretic eyelid is lowered relative to the opposite side, a neutral trade that is healthier for the cornea. We found that the average reduction in height was 1.35 mm, similar to the 1.5-mm average of upper eyelid ptosis that is induced by pharmacologic paresis of the Müller muscle.31 we also found that certain procedures—tarsorrhaphy and canthoplasty—that we used for lower eyelid retraction did not reduce upper eyelid height in our series (figure 2).
Exposure keratopathy was consistently improved by müllerectomy. However, virtual resolution of keratopathy occurred in 50% of patients undergoing müllerectomy alone, but in only 12% of patients who underwent additional lower eyelid surgery. Furthermore, patients who underwent only müllerectomy had statistically significant less postoperative lagophthalmos, while those patients who had additional lower eyelid elevation did not (Figure 3). We speculate that the patients who underwent lower eyelid surgery had a greater loss of facial nerve fibers with substantially less reserve, some of which may have deteriorated in the postoperative period, as previously reported.4 Only 2 of these patients underwent comprehensive midfacial elevation and lower eyelid surgery, which is effective for reduction of keratopathy and lagophthalmos in these patients.4 most patients (14/16) who underwent additional lower eyelid surgery in this series were treated with other less effective measures. We postulate that the most effective rehabilitation for ocular involvement in facial nerve palsy may be müllerectomy combined with comprehensive midfacial elevation and lower eyelid surgery.
Müllerectomy does not require implantation of foreign material or permanent sutures. Postoperatively, no taping or traction suture is necessary. In our study, postoperative ptosis was avoided in patients who exhibited intraoperative ptosis due to preexisting aponeurotic dehiscence by placing a dissolvable suture at the apex of the eyelid arch, as performed in 3 patients.
Because this study is retrospective, it has inherent limitations, including intercurrent surgical treatment, no direct comparison with control or other treatment groups, and lack of prospective data collection. Nevertheless, the clinical effectiveness of müllerectomy is reflected by the fact that 91% of preoperative symptoms were resolved or improved by the surgery (Table), while visual acuity improved in 32% of patients. Only 2 patients had reduced visual acuity from the effects of facial nerve palsy in the postoperative follow-up period. Thus, müllerectomy for upper eyelid paresis due to facial nerve palsy significantly and reproducibly improves the symptoms and signs of exposure keratopathy and lagophthalmos, while maintaining cosmetically acceptable symmetry, as shown by analysis of our pooled data of 2 surgeons working independently.
Correspondence: Victor M. Elner, MD, PhD, Department of Ophthalmology, University of Michigan, 1000 Wall St, Ann Arbor, MI 48105 (firstname.lastname@example.org).
Submitted for Publication: April 29, 2004; final revision received November 19, 2004; accepted December 2, 2004.
Financial Disclosure: None.