Original Investigation

Factors Affecting Eyelid Crease Formation Before and After Silicone Frontalis Suspension for Adult-Onset Myogenic Ptosis Richard C. Allen, M.D., Ph.D., F.A.C.S.*†, Esther S. Hong, M.D.*, M. Bridget Zimmerman, Ph.D.‡, Leslie A. Morrison, M.D.§, Jeffrey A. Nerad, M.D., F.A.C.S.║, and Keith D. Carter, M.D., F.A.C.S.*† *Department of Ophthalmology and Visual Sciences; †Department of Otolaryngology – Head and Neck Surgery; and ‡Department of Biostatistics, University of Iowa Hospitals and Clinics, Iowa City, Iowa; §Department of Neurology, University of New Mexico School of Medicine, Albuquerque, New Mexico; and ║Cincinnati Eye Institute, Cincinnati, Ohio, U.S.A.

Purpose: To evaluate factors that affect eyelid crease formation before and after frontalis suspension. Design: Nonrandomized, comparative, interventional case series. Methods: Sixty-three patients (125 eyes) with myogenic ptosis were included. Data collected included age, gender, previous surgeries, follow up, as well as pre- and postoperative margin reflex distance, palpebral fissure height, and levator function. Intraoperative maneuvers of incorporation of the levator aponeurosis into the skin closure, conservative fat excision, and conservative skin excision were recorded. Preand postoperative eyelid creases were graded by 2 masked, independent observers as “good,” “fair,” or “poor.” Results: The weighted κ coefficient between the graders was 0.68 (95% CI, 0.58–0.79) preoperatively and 0.70 (95% CI, 0.61–0.79) postoperatively. Evaluating preoperative eyelid crease grades, there was no significant difference with regard to age or gender (p = 0.83 or 0.69, respectively). Eyelid crease grade correlated with margin reflex distance (p = 0.0004) and palpebral fissure height (p = 0.002). There was no significant correlation of eyelid crease with levator function (p = 0.104). After frontalis sling, intraoperative maneuvers of incorporation of the levator aponeurosis into the incision, skin preservation, and fat preservation correlated with postoperative eyelid crease (p = 0.0004, 0.059, and 0.033, respectively). Conclusions: Preoperative levator function in patients with adult onset myogenic ptosis may be an inaccurate measure of true levator palpebrae strength. Reliance on levator function alone in decision making for surgical intervention in these patients may be misguided. The inclusion of the intraoperative maneuvers of incorporation of the levator aponeurosis into the skin incision and preservation of fat and skin results in a stronger eyelid crease after frontalis sling surgery. (Ophthal Plast Reconstr Surg 2015;31:227–232)

Accepted for publication April 25, 2014. This manuscript was presented, in part, at the American Society of Ophthalmic Plastic and Reconstructive Surgery Fall Meeting on October 2010 in Chicago, IL, and also at the Association for Research in Vision and Ophthalmology meeting on May 2010 in Fort Lauderdale, FL. This work was supported, in part, by an unrestricted grant from Research to Prevent Blindness, Inc., New York, New York. The authors have no financial or conflicts of interest to disclose. Address correspondence and reprint requests to Richard C. Allen, M.D., Ph.D., F.A.C.S., Department of Ophthalmology and Visual Sciences, University of Iowa Hospitals and Clinics, 200 Hawkins Drive, Iowa City, IA 52242. E-mail: [email protected] DOI: 10.1097/IOP.0000000000000272

Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015

F

rontalis suspension is often looked on as a last resort in ptosis correction, usually reserved for cases in which levator function (LF) is poor and ptosis is significant.1–3 Levator-based procedures for myogenic ptosis are often favored over frontalis suspension for a number of reasons: there may be a cosmetic advantage, the surgery is considered less complex without the potential introduction of a foreign material, and raising the eyelid is not dependent on eyebrow elevation.4 There are situations in which there is not a clear advantage of a levator-based procedure or frontalis sling, for example significant unilateral congenital ptosis with poor to fair LF and progressive myogenic ptosis (chronic progressive external ophthalmoplegia, oculopharyngeal muscular dystrophy [OPMD], and myotonic dystrophy). In these instances, levator-based surgeries are often favored, even though it has been argued that patients may be better served with a frontalis sling.5–9 In examining cosmetic results of frontalis sling surgery, there has been an emphasis on formation of an upper eyelid crease, which is desirable.10–12 In addition, a strong eyelid crease decreases eyelash ptosis.13 In the unoperated eyelid, formation of a eyelid crease is dependent on 2 factors: extensions of connective tissue from the levator aponeurosis to the dermis and the ability of soft tissue anterior to the levator aponeurosis and superior to the eyelid crease (skin, orbicularis muscle, and preaponeurotic fat) to fold over the crease.14–17 The fold of preaponeurotic fat is created by the fusion of the septum to the levator aponeurosis; thus, it is difficult to have a crease without a fold. Long-term success and formation of an eyelid crease after frontalis suspension has been shown to be affected by many factors; in particular, surgeons have advocated open, eyelid crease incisions rather than stab incisions, direct fixation of the sling to the anterior surface of the tarsus, and retroseptal placement of the frontalis sling (septal pulley), among others.11,18–24 The purpose of this study is to determine the preoperative and intraoperative factors that influence eyelid crease formation in a relatively homogeneous population of patients with progressive myogenic ptosis, before and after undergoing silicone frontalis suspension.

METHODS Sixty-three patients (125 eyelids) were identified who underwent silicone frontalis suspension; 62 of the patients had a diagnosis of OPMD, using criteria described by Brais et al.25; the remaining patients had a diagnosis of chronic progressive external ophthalmoplegia confirmed by muscle biopsy. The study was approved by the Human Research Review Committee (Protocol No. 03-398) at the University of New Mexico School of Medicine and was in compliance with HIPAA and the Declaration of Helsinki. Medical records were reviewed and the

227

Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.

Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015

R. C. Allen et al.

following data were obtained: age, gender, previous eyelid surgeries, and pre- and postoperative (with follow up in months) eyelid measurements including margin reflex distance (MRD), palpebral fissure height (PFH), and LF. Levator function was not measured in those patients who had a previous frontalis sling. Frontalis sling surgery was performed as previously described5; however, the following intraoperative details were obtained from the medical records: skin preservation (defined as 18 mm or more of skin remaining postoperatively between the eyelashes and the inferior border of the eyebrow cilia), preaponeurotic fat preservation (defined as no preaponeurotic fat resection), and release of the levator aponeurosis from the anterior surface of the tarsus with incorporation into the closure of the skin incision (Fig. A,B). All patients had the frontalis sling fixated to the anterior surface of the tarsus (2 mm inferior to the superior border of the tarsus) with interrupted 5-0 Mersilene sutures, and all patients had the frontalis sling placed in a retroseptal fashion, as described by Patrinely and Anderson.18 Surgical correction was bilateral in all cases except 1. Preoperative and postoperative photographs were blindly and independently graded by 2 of the authors. A standardized grading system was followed that outlined criteria for each grade (good, fair, and poor) with an exemplary photograph of a patient in each grade for reference by graders. A “good” eyelid crease (numerical score 3) was defined as having a distinct eyelid crease with a fold. A “fair” eyelid crease (numerical score 2) was defined as a distinguishable eyelid crease only. A “poor” eyelid crease (numerical score 1) was defined as no distinguishable eyelid crease. The weighted κ coefficient was computed to assess agreement between the 2 graders. Statistical analyses for the unoperated eyelids included the Wilcoxon signed-rank test to compare eyelid crease grade between eyes, Wilcoxon rank sum test to compare eyelid crease grade between gender, and Spearman correlation coefficient to examine the correlation of eyelid crease grade with age. For the age and gender analysis, the maximum eyelid crease grade between the 2 eyes was used. For analysis that tested the association of eyelid crease grade with eyelid measurements (MRD, PFH, and LF), where there were 2 pairs of data values for each patient (1 for each eye), linear mixed model analysis was used to account for the correlation of the eyelid crease grade between eyes from the same patient and the correlation of measurements between eyes from the same patient. The generalized linear mixed model was used in the analysis of postoperative eyelid crease grade. This analysis was used to examine the relationship of duration of follow up and preoperative eyelid crease grade with postoperative grade. Generalized linear mixed model was also used to identify demographic and clinic variables, as well as eyelid measurements that were associated with good postoperative eyelid crease grade. For each variable being tested, a logistic model was fitted that included the variable of interest as the independent variable, with follow-up duration and preoperative eyelid crease grade as covariates. From these fitted models, the effect size of the variable tested on postoperative eyelid crease grade was expressed as the odds ratio of a good postoperative eyelid crease. All the statistical analyses were performed using Statistical Analysis Software Version 9.2 (SAS, Cary, NC, U.S.A.).

RESULTS Eyelid Crease Formation in Unoperated Eyelids of Patients With OPMD. Ninety-seven eyelids of 49 patients were identified that had no previous eyelid surgery. All patients had a diagnosis of OPMD. The mean age was 61 years with 61% women and 39% men. The mean MRD, PFH, and LF were –0.29 mm ± 1.02 mm, 4.96 mm ± 1.33 mm, and 11.1 mm ± 2.10 mm, respectively. The analysis of eyelid crease grades showed that 40% of eyelids were graded “good,” 32% were graded “fair,” and 28% were graded “poor.” The 2 masked, independent graders had exact agreement on 69 eyelids (71%). There was no instance in which an eyelid was graded

228

“good” by one grader and “poor” by the other. The weighted κ coefficient between the graders was 0.68 (95% CI, 0.58–0.79). An average of the 2 graders was used in the remainder of the analysis with a score from 1 to 1.5 as “poor”, 2 to 2.5 as “fair,” and 3 as “good”. The Wilcoxon signed-rank test showed no significant difference between the 2 eyelid crease grades within a single individual (p = 0.69), indicating that the overall eyelid crease grades were symmetrical. The statistical analysis that examined the association of eyelid crease grade with the demographic factors showed no significant correlation with age (Spearman correlation, 0.03; 95% CI, –0.31 to 0.26; p = 0.83). The mean age by eyelid crease grade group was 62 years for good, 61 years for fair, and 60 years for poor. There was also no significant difference in eyelid crease grade between males and females (Wilcoxon rank sum test, p = 0.69). Linear mixed model analysis was performed to test any associations between the eyelid crease grade and the eyelid measurements (MRD, PFH, and LF). This showed a significant correlation of MRD with eyelid crease grade (p = 0.0004), with the good eyelid crease having the smallest MRD and the poor eyelid crease the largest negative MRD. The mean MRD in the good, fair, and poor eyelid crease groups were 0.2 mm ± 0.6 mm, –0.2 mm ± 1.1 mm, and –0.8 mm ± 1.0 mm, respectively. Palpebral fissure height also demonstrated a statistically significant association with the eyelid crease grade, with the lower PFH values associated with poor eyelid crease. The mean PFH in the good, fair, and poor eyelid crease groups were 5.5 mm ± 1.3 mm, 5.0 mm ± 1.4 mm, and 4.3 mm ± 1.0 mm, respectively (p = 0.002). The analysis of LF and eyelid crease grades showed that the LF mean was highest in the good eyelid crease group, with similar LF means for the poor and fair groups, although this was not significant at the 0.05 significance level (p = 0.104). The good, fair, and poor eyelid crease grades had a mean LF of 11.5 mm ± 1.7 mm, 10.8 mm ± 2.0 mm, and 10.8 mm ± 2.1 mm, respectively. Determinants of Postoperative Eyelid Crease Formation in Patients With Progressive Myogenic Ptosis who Underwent Silicone Frontalis Suspension. Sixty-three patients (125 eyelids) were identified who underwent silicone frontalis suspension. The mean age was 64 years; there were 37 women and 26 men. Eleven eyelids had undergone previous blepharoplasty, 28 eyelids had undergone previous levator advancement, and 14 eyelids had undergone previous frontalis suspension. The mean preoperative MRD, PFH, and LF were –0.246 mm (SD ±0.712), 5.09 mm (SD ±1.39), and 10.5 mm (SD ±2.36 mm), respectively. The mean postoperative MRD and PFH were 2.35 (SD ±0.864) and 7.81 (SD ±1.26). Postoperative eyelid crease grades by the 2 examiners showed 40% of eyelids to be graded “good” (score 3), 33% to be graded “fair” (score 2), and 27% to be graded “poor” (score 1). The agreement between the raters as measured by the weighted κ coefficient was 0.70 (95% CI, 0.61–0.79). For the remaining analysis, the score of the 2 examiners was averaged to give a numerical value per eyelid. The analysis was performed to determine if there was an association between preop eyelid crease grade and postop eyelid crease grade (Table 1). The analysis showed that those with good (p = 0.001) or fair (p = 0.016) eyelid crease at preop were more likely to have a good postop eyelid crease grade compared with those with a poor eyelid crease at preop. Likewise, those with a poor eyelid crease at preop were more likely to have a poor postop eyelid crease grade than those with a good eyelid crease grade at preop (p = 0.054). The data were analyzed to determine if there was an association of postoperative eyelid crease grade with the follow-up time (Table 2). Follow up was divided into ≤3 months, >3 to 6 months, >6 to 12 months, and >12 months, with the proportion of eyes that worsened in eyelid crease by 1 grade or more compared with the preop grade over the follow-up intervals. A higher percentage of eyes had a worsened eyelid crease grade with longer length of follow up (p = 0.002). A similar analysis was performed for the outcome of improvement of eyelid crease.

© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.

Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.

Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015

Eyelid Crease Formation Before and After Frontalis Sling

TABLE 1.  Correlation of preoperative eyelid crease with postoperative eyelid crease Good eyelid crease postop Preop eyelid crease Good (3) Fair (2–2.5) Poor (1–1.5)

Poor eyelid crease postop

n (eyes)

%

95% CI

%

95% CI

44 32 41

44 34 16

(29, 59) (22, 49) (8, 30)

19 24 42

(9, 36) (11, 45) (25, 60)

Logistic regression models for good postop eyelid crease and for poor postop eyelid crease were fitted by the generalized estimating equation (GEE) method. Preop eyelid crease as the independent variable in the mode was used to test the association between preop and postop eyelid crease grade. Those patients with a good (p = 0.001) or fair (p = 0.016) eyelid crease at preop were more likely to have a good eyelid crease at postop compared with those with a poor eyelid crease at preop. Those with a poor eyelid crease (p = 0.054) at preop were more likely to have a poor eyelid crease at postop compared with those with a good eyelid crease at preop.

The results from this analysis showed no significant trend (increasing or decreasing) in percentage with improved eyelid crease grade with the follow-up time (p = 0.872). Generalized linear mixed model analysis was performed to identify demographic and preoperative variables that may be associated with postoperative eyelid crease grades. Age and gender showed no statistically significant association with a good postoperative eyelid crease (Table 3). The adjusted odds ratio for a good postoperative eyelid crease was 0.62 (95% CI, 0.33, 1.14; p = 0.120) for age and 0.47 (95%CI, 0.10, 2.12; p = 0.253) for gender. Previous eyelid surgeries did not demonstrate a statistically significant association with a good postoperative eyelid crease after frontalis suspension (Table 3). The adjusted odds ratio for achieving a good eyelid crease in patients with previous blepharoplasty, levator advancement, and frontalis sling was 1.09 (95% CI 0.08, 14.57; p = 0.988), 2.75 (95% CI 0.45, 16.72; p = 0.535), and 2.44 (95% CI 0.20, 30.63; p = 0.626), respectively. Among preoperative eyelid measurements, LF was the only variable that showed a statistically significant association with good postoperative eyelid crease grade with an adjusted odds ratio of 1.86 (95%CI, 1.26, 2.76; p = 0.009). The good eyelid crease group had a mean preoperative LF of 12 mm, while both the fair and poor eyelid crease groups each demonstrated a mean LF measurement of 10 mm. The pre- and postoperative MRD and PFH failed to demonstrate a statistically significant association with postoperative eyelid crease grade (Table 3). An analysis of intraoperative maneuvers to eyelid crease grades was performed (Table 4). Ninety (72%), 38 (30.4%), and 88 (70.4%) eyelids had the intraoperative maneuver of fat preservation,

TABLE 2.  Association of eyelid crease with follow-up time Follow up (months) ≤3 >3–6 >6–12 >12 Follow up (months) ≤3 >3–6 >6–12 >12

Worse eyelid crease

N (eyes)

19 19 10 17

%

95% CI

10 31 38 71

(3–31) (14–55) (15–69) (39–91)

N (eyes)

19 24 14 16

Improved eyelid crease %

95% CI

30 33 21 31

(12–59) (16–56) (9–45) (10–64)

Upper: Percentage of eyelids with a worse eyelid crease (decrease of at least 1 grade from preop) by length of follow up. Only those eyelids with a grade of 2 or higher are included. Lower: Percentage of eyelid with an improved eyelid crease (increase of at least 1 grade from preop) by length of follow up. Only those eyelids with a grade of 2 or lower are included.

levator aponeurosis incorporation into the skin closure, and skin preservation, respectively. The greatest correlation was seen in patients who underwent incorporation of their levator aponeurosis into the skin closure. Seventy-one percent of patients who underwent this maneuver attained a good postoperative eyelid crease grade compared with 16% of those who did not undergo this maneuver. Furthermore, no patient who underwent incorporation of the levator aponeurosis into the skin closure attained a poor eyelid crease grade. Forty-three percent and 41% of patients in the respective fair and poor eyelid crease grade groups did not have their levator incorporated into the skin closure. This positive association between a good eyelid crease grade and incorporation of the levator aponeurosis into the skin closure was statistically significant (p = 0.0004). Fat preservation also showed a significant association with a good postoperative eyelid crease; 42% of patients who underwent this maneuver attained a good grade while 6% of them who did not undergo this maneuver attained a good grade (p = 0.033). Skin preservation showed the least correlation to having a good postoperative eyelid crease. When patients from the good eyelid crease group were compared with those from the poor eyelid crease group, the adjusted odds ratio was 5.38 (95% CI, 0.93–31.03; p = 0.059)

DISCUSSION This study examines 2 questions: what determines eyelid crease formation in the unoperated eyelid with myogenic ptosis and what determines eyelid crease formation in the eyelid that has undergone frontalis sling surgery? Preoperative and postoperative eyelid crease formation has a tendency to affect surgical planning in ptosis repair. In the preoperative situation, a weak or indistinct eyelid crease is thought to reflect the strength of the levator muscle, the measure of which is commonly LF. In the situation of significant ptosis and a less distinct eyelid crease, one may lean toward a frontalis sling, especially in the situation in which it is difficult to measure LF as in a young child. In the postoperative situation, a distinct eyelid crease is favored, and the surgeon may choose a surgery that will consistently produce a well-formed eyelid crease in addition to an appropriate eyelid height and contour. In the unoperated eyelid, the authors found no significant correlation between eyelid crease grade and age or gender. They do find that more ptosis (through measurements of MRD and PFH) results in a poor eyelid crease. There are 2 possible explanations for this. On one hand, a greater amount of ptosis may obviate the formation of a fold over the crease. On the contrary, if one assumes that more ptosis is directly correlated with a weaker levator muscle, then the crease grade reflects levator health. Levator function had the poorest correlation of the eyelid measurements with eyelid crease grade. The authors expected that there would be a greater correlation of eyelid crease with LF, as LF is usually thought to be a good measurement of levator health. The population in this study did have a relatively high LF (11.1 ± 2.1 mm), and the authors may not have been able to

© 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc.

229

Copyright © 2014 The American Society of Ophthalmic Plastic and Reconstructive Surgery, Inc. Unauthorized reproduction of this article is prohibited.

Ophthal Plast Reconstr Surg, Vol. 31, No. 3, 2015

R. C. Allen et al.

TABLE 3.  Multivariate analysis of preoperative data with postoperative eyelid crease grades Postop eyelid crease frequency (%) Variable

Level N (eyes) Good

Age

Factors affecting eyelid crease formation before and after silicone frontalis suspension for adult-onset myogenic ptosis.

To evaluate factors that affect eyelid crease formation before and after frontalis suspension...
372KB Sizes 0 Downloads 4 Views