Orbit, 2015; 34(2): 92–98 ! Informa Healthcare USA, Inc. ISSN: 0167-6830 print / 1744-5108 online DOI: 10.3109/01676830.2014.999096
RESEARCH REPORT
The Role of Tissue Resection Length in the Determination of Post-Operative Eyelid Position for Muller’s Muscle-Conjunctival Resection Surgery Dan B. Rootman1, Justin Karlin2, Grant Moore1, and Robert Goldberg1 1
Division of Orbital and Ophthalmic Plastic Surgery, Jules Stein Eye Institute, University of California, Los Angeles, California, USA and 2Sackler School of Medicine, Tel Aviv University, New York, USA
ABSTRACT Purpose: To investigate the relationships between pre-operative marginal reflex distance (MRD), tissue resection length, phenylephrine response, and change in MRD with surgery for a cohort of individuals undergoing Muller’s muscle conjunctival resection (MMCR) surgery. Methods: All cases of MMCR surgery performed over a 13-year period at a single institution were screened for entry. Individuals with adequate photographic documentation and follow up were included. Patients with previous or concurrent upper eyelid, orbital or eyebrow disease of surgery were excluded. Marginal reflex distance (MRD) was calculated based on photographs utilizing public domain software. Data was plotted for inspection and appropriate statistical tests were performed. Results: During the study period 198 eyes fit criteria for analysis. A loose association between tissue resection length and change in MRD with surgery was found (r = 0.176, p50.05); this relationship was not significant in ANOVA analysis (p = 0.367). There was a strong association between MRD change with surgery and preoperative MRD (r = 0.498, p50.01). Approximately 28% of the sample responded to 2.5% phenylephrine drop instillation with a greater than 2 mm increase in MRD. The response to phenylephrine was strongly associated with pre-operative MRD (r = 0.441, p50.01). A regression on change in MRD with surgery with tissue resection, phenylephrine response 42 mm and pre-operative MRD as variables revealed a model with preoperative MRD as the only significant predictor (p50.01). Conclusion: Tissue resection length and phenylephrine response play small roles relative to pre-operative MRD in the determination of change in MRD with MMCR surgery. Keywords: Eyelids, Muller’s muscle-conjunctival resection, ptosis
INTRODUCTION
over time. Various strategies involving the addition or subtraction of Muller’s muscle resection length and/ or tarsus based on pre-operative ptosis or phenylephrine response have been proposed to improve predictability of the final eyelid position.1,5,7,8,10,11,13 However, such refinements in these published series have not been conclusively shown to increase overall predictability. It is yet unclear how varying technique and tissue resection length affect final outcome in MMCR surgery. The purpose of this study was to directly
Since originally described in 1975,1 the Muller’s muscle conjunctival resection (MMCR) procedure has gained significant popularity.2 This procedure has been utilized in a wide variety of ptosis cases, by a number of authors, producing mostly good results.1,3–12 Classically, a resection length to eyelid elevation ratio of 1:4 mm7 has been applied, however a number of refinements to this nomogram have been offered
Received 6 July 2014; Accepted 12 December 2014; Published online 13 March 2015 Correspondence: Dan B. Rootman, Division of Orbital and Oculoplastic Surgery, Jules Stein Eye Institute, 100 Stein Plaza, University of California, Los Angeles, CA 90095, USA. Tel: (310) 206 8250. Fax: (310) 825 9263. E-mail: [email protected]
92
Tissue Resection in MMCR investigate this relationship and to define preoperative factors that may predict post-operative eyelid position in MMCR surgery.
MATERIALS AND METHODS All cases of MMCR performed at the Jules Stein Eye Institute between January 1, 2000, and May 1, 2013, were screened for study entry. The dates were chosen as electronic clinical photographs were available for each patient in this time frame. This study was approved by the Institutional Review Board at the University of California-Los Angeles and investigations were performed according to the Declarations of Helsinki. Eyes were included if they had MMCR surgery in isolation from any other upper eyelid or brow procedures and if pre-operative standard and post phenylephrine photos, as well as at least one postoperative photo between 1 and 12 months after surgery were available. Eyes were excluded for criteria related to imaging if it not sufficiently clear or if they were not looking directly forward in primary position, in line with the camera. Patients with a history of previous ptosis or brow surgery or a systemic disease that may affect lid position such as thyroid eye disease, myesthnia gravis or chronic progressive external ophthalmoplegia were also excluded. Cases in which levator function was below 12 mm were also excluded, as were eyes with surgery involving resection of tarsus. Clinical histories were reviewed for study eligibility. Tissue resection amounts were extracted from preoperative plans, confirmed by operative reports. Cases with disagreement between the plan and report, without written clarification in the chart, were eliminated. All photographs were reviewed in a standard manner. All measurements were made using ImageJ (National Institutes of Health, Bethesda, MD) public domain software. The marginal reflex distance-1 (MRD) was measured from standardized photographs as the distance from the center of the pupil to the lowermost point of the upper eyelid in the midpupillary line. This measurement was used as a proxy for the marginal reflex distance, and avoids variability in the angular position of the flash in photography. Average corneal diameter for females (11.64 mm) and males (11.77 mm), as measured by Rufer et al.,14 was utilized as a reference in setting the measurement scale. MRD response to phenylephrine was measured as a change in MRD, 5 minutes after the instillation of 2.5% phenylephrine solution to the inferior fornix. Data were plotted raw and assessed for outliers and inconsistencies. Outliers were remeasured by a second reviewer. Paired t-tests were used to compare pre- and post-operative eyelid position. For ANOVA !
2015 Informa Healthcare USA, Inc.
93
analyses, continuous variables were binned in to 0.5 mm or 1 mm increments. For linear regression and correlation analyses, continuous data was utilized untransformed. Regression analyses were designed to test specific a priori hypotheses regarding the relative contribution of tissue resection length, eyelid position at presentation and phenylephrine response on postoperative change and are limited to such. All analyses were performed with IBM SPSS Statistics for Mac version 21.0 (SPSS Inc., an IBM Company, Somers, NY). The Bonferroni correction was applied where appropriate.
RESULTS Sample After applying study criteria, 198 eyes in 130 people were included in the final sample. Unilateral cases accounted for 30.8% of the surgeries and bilateral surgery was performed in 69.2% of eyes. Mean (SD) follow up was 3.78 (2.14) months.
Surgical Technique Under monitored anesthesia care, 2% lidocaine with 1:100,000 epinephrine was injected in to the upper eyelid. The eyelid was everted over a Desmarres retractor and the conjunctiva was dried. Resection length was marked on the conjunctival surface with a caliper at the 1/3 junction of the tarsus medially and laterally. Fine forceps were then used to grasp and elevate the conjunctiva and Muller’s muscle. An MMCR clamp was applied, and the Desmarres removed. A chromic gut suture was then passed under the clamp running in both directions with the ends externalized to skin. Resection was completed with a #15 Bard Parker blade. The eyelid was everted and gentle pressure applied for hemostasis.
Overall Effect of Surgery The overall mean effect of MMCR surgery was a 1.5 mm increase in the MRD from pre-operative to post-operative measurements. This was highly statistically significant (p50.01). Tissue resected and the response to surgery Resection length was on average (SD) 6.5 mm (1.5 mm), although the modal resection length was 8 mm representing 23% of total resections. A scatter plot of MRD change with surgery against the amount of tissue resected did not demonstrate a clear trend, with high variability (Figure 1). Pearson’s correlation analysis confirms the loose relationship (r = 0.176,
94 D. B. Rootman et al.
FIGURE 1. Scatter plot of change in marginal reflex distance (MRD) with surgery against tissue resected.
FIGURE 2. Mean change in marginal reflex distance (MRD) with surgery for each amount of tissue resected (df = 6, F = 1.1, p = .367).
p50.05). Regression analysis produces a significant (p50.05) model, explaining 3.1% of the total system variability. Dividing the mean change in MRD with surgery (1.5 mm), by the mean resection (6.5 mm) reveals a ratio of 0.23 mm MRD change per mm of resection, or 1 mm change in MRD for each 4 mm resected. The mean MRD change with surgery did not significantly differ for each mm of tissue resected (p = 0.367; Figure 2). When comparing eyes with 4 mm to eyes with 8-mm tissue resection lengths, the change in MRD with surgery did not significantly differ (t = 1.5, p = 0.133). However, the post-operative MRD did significantly differ between these groups, with the 4 mm resection demonstrating 0.7 mm higher postoperative MRD than the 8-mm resection group (p50.05). The pre-operative MRD for 4 mm resection
FIGURE 3. Scatter plot of change in marginal reflex distance (MRD) with surgery against pre-operative MRD.
FIGURE 4. Mean change in marginal reflex distance (MRD) with surgery for each category of pre-op MRD (df = 8, F = 11.1, p50.01).
patients was 1.3 mm higher than 8 mm resection patients (p50.01).
Pre-operative MRD and the response to surgery A similar scatter plot of MRD change with surgery against the pre-operative MRD demonstrated a more convincing trend and lower variability (Figure 3). Pearson’s correlation analysis confirms the relationship (r = 0.498, p50.01). Regression analysis produces a significant (p50.01) model, explaining 25.0% of the total system variability. This model suggests that for each increase of 1 mm in pre-operative MRD, the change in post-operative MRD will decrease by 0.54 mm. Restated, as the pre-operative MRD increases, the change with surgery decreases (see Figure 4). Orbit
Tissue Resection in MMCR
95
FIGURE 5. Scatter plot of pre-operative marginal reflex distance (MRD) against tissue resection length (mm).
FIGURE 6. Scatter plot of change in marginal reflex distance (MRD) with phenylephrine against pre-operative MRD.
Phenylephrine response and response to surgery A greater than 2-mm increase in MRD was noted after 2.5% phenylephrine drop instillation in 28.3% (n = 39) of the 138 eyes in which phenylephrine was applied. A greater than 1-mm increase was noted in 65.9% and a greater than 0.5-mm increase was noted in 83.3% of the same sample. Individuals who respond to phenylephrine with a 42-mm change in MRD had a 0.9-mm lower preoperative MRD than non-responders (p50.01). Additionally, these individuals had a greater change in MRD with surgery by 0.8 mm (p50.01). They did not have a significant difference in the final postoperative MRD (p40.05). For the group of individuals who demonstrated a42 mm increase in MRD with phenylephrine installation, the association between tissue resection length and change in MRD with surgery was stronger than in the full sample (r = 0.420, p50.01). Additionally, the pre-operative MRD was more highly correlated with surgical change in MRD (r = 0.568, p50.01). Tissue resected and pre-operative MRD are also more highly associated with each other in this group (r = 0.516, p50.01). Linear regression on change in MRD with surgery with resection length and pre-operative MRD as predictors was performed for individuals with phenylephrine response 42 mm. This revealed a significant (p50.01) model, explaining approximately 30% of the variability with only pre-operative MRD as a significant coefficient (p50.01). Tissue resection length did not maintain statistical significance in this model.
(r = 0.382, p50.05). This relationship is stronger than the one between the amount of tissue resected and change with surgery (r = 0.176, p50.05). Additionally, there was a strong association (r = 0.441, p50.01) between pre-operative MRD and change in MRD with phenylephrine (Figure 6). This association was stronger than the one between change in MRD with phenylephrine and change with surgery (r = 0.348, p50.01). A regression model with tissue resection length, pre-operative MRD and change in MRD with phenylephrine as predictors of change in MRD with surgery was fit. This model was significant (p50.01), and explained approximately 26% of the variability in the data. The only significant predictor in the model was pre-operative MRD (Table 1).
Tissue resected, pre-operative MRD, phenylephrine response and surgery There is a strong correlation (Figure 5) between the amount of tissue resected and pre-operative MRD !
2015 Informa Healthcare USA, Inc.
Subgroup analysis All analyses were performed for unilateral cases alone and no differences were noted relative to the bilateral analysis. Only bilateral analyses are presented here.
DISCUSSION This series of analyses, performed on a large high quality sample of MMCR surgeries, has demonstrated that the most important factor in determining postoperative MRD change is the pre-operative MRD. The amount of tissue resected and phenylephrine response do not appear to play statistically significant roles in determining surgical change. The weak association between resection length and post-operative MRD may be a statistical byproduct of the moderate relationship between resection length and pre-operative MRD (lower had a greater amount of tissue resected and vice versa). This may be a case of correlation rather than causation.
96 D. B. Rootman et al. TABLE 1. Linear regression of predictive factors on marginal reflex distance (MRD) change with surgery. Variable
Beta
p Value
Tissue resected (mm) Pre op MRD (mm) MRD responder
0.055 0.468 0.399
0.384 50.01 0.060
There are a number of key limitations to consider in this discussion. Initially it should be noted that a resection length of between 4 mm and 9 mm of tissue was performed in 98% of cases, and no tarsus was involved. Thus, these results may not apply to resections outside of that range, or to Fasanella– Servat type procedures. The data is also photographic in nature and may not accurately represent the dynamic eyelid position, although variability should be randomly distributed. Further, we did not apply a strict rule of 4:1 in each case, as some patients with latent ptosis may have had larger resections than what might be calculated based on the degree of ptosis evident in photographs. Again this variability should be evenly distributed statistically. Additionally, our statistical models did not account for all variables potentially important in the determination of outcome and are not able to make comments regarding overall importance of pre-operative MRD in the system, rather only in relation to a priori selected variables. Finally, it would be ideal to perform a randomized study comparing standard tissue resection (for instance 8 mm) to variable (based on algorithm). Such a study is in development. With regard to the role for the amount of tissue resected. Our result that there does not appear to be a significant relationship between resection length and outcome may be less surprising than on initial glance when the literature is critically reviewed. In Putterman et al.’s original article,1 they suggested removing approximately 8 mm of tissue for each case, regardless of the pre-operative MRD. They did note making some adjustment based on the phenylephrine response, although they did not specify the details. A re-analysis of their published results using linear regression and correlation statistics reveals an overall regression coefficient for pre-operative MRD on postoperative change in MRD of 0.49, which is strikingly similar to that reported here ( 0.53). The correlation between these variables in Putterman’s article is even greater at r = 0.82. Although clearly there are substantial differences in methods and measurements between the original study and our results, it is interesting to note that without varying the resection length, Putterman described very close association between pre-operative and post-operative MRD. In a later article, Putterman suggested making a modification to the resection length based on
phenylephrine testing, resecting 1 mm less for phenylephrine over-responders and 1 mm more for phenylephrine under-responders although overand under-response were not clearly defined. This article11 described 232 eyelids (an undetermined proportion using this variable formula) and found that 75% of acquired ptosis cases achieved a post-op MRD of between 2.5 and 5. They did not report associations between pre-operative and post-operative MRD. Guyuron et al.7 used a similar 7 mm-8 mm-9 mm rule for resection of conjunctiva and Mu¨llers muscle based on the phenylephrine response (slight – perfect – exaggerated). They described the mean change in eyelid position after surgery in 43 patients to be 2.1 mm. Presuming (they did not report this value) a mean resection length of 8 mm, they concluded by simple division that 1 mm of resection is required for 0.25 mm of eyelid elevation. No inferential statistics examining the relationship between these variables was offered. Weinstein and Buerger made a similar calculation 3 years later,13 and this heuristic has since persisted in the literature.15 Dresner used this ratio to create a rule for varying resection based on pre-operative ptosis. He reported 5 a series of 146 eyelids in which the tissue resection was based on the pre-operative ptosis with a modified 1:4 ratio. This rule was applied primarily in patients with a 42-mm response to phenylephrine, and the small number of patients with a smaller response had slightly more tissue removed. He found that 84% of patients achieved MRD symmetry between the eyelids within 0.5 mm. Additionally, he demonstrated a strong linear relationship between resection and change in MRD with surgery (r = 0.6). Dresner’s results on first pass appear disparate from ours. However some differences in the sample may explain these variations. It is not completely clear how many individuals had a phenylephrine response 42 mm in his study, but it would appear as though this number would be above to 80%. Subgroup analysis of strong (42 mm) phenylephrine responders in our sample reveals similar results. For patients demonstrating an MRD change with phenylephrine 42 mm, the relationship between tissue resection and MRD change on our study was much closer to Dresner’s (r = 0.420). However the relationships between pre-operative MRD, tissue resection length change with phenylephrine and MRD change with surgery are all stronger for patients with 42 mm change in MRD with phenylephrine. In linear regression analysis, again only pre-operative MRD maintains significance for this group. Unfortunately in Dresner’s study, no relationships between pre-operative ptosis and resection were calculated, nor were multivariate models fit. With data presented in that article, it is not possible to determine whether the association between resection Orbit
Tissue Resection in MMCR and outcome is an effect of a strong relationship between pre-operative ptosis and resection length. Additionally, the strong relationship reported in Dresner’s article has not been replicated in the literature. Perry et al. noted astutely from these articles that a variety of resection lengths appear to produce similar results.10 They went on to reason that in order to improve predictability of the procedure, the phenylephrine response can be used as a literal guide to outcome. They presumed that 10% phenylephrine instillation will maximally stimulate the Muller’s muscle and thus exactly simulate a 9-mm (most of the Muller’s muscle) resection. Thus an ideal phenylephrine response should be recreated surgically with a 9-mm MMCR resection. Under-correction with phenylephrine could be managed by adding 1-mm tarsal resection for each 1-mm under-correction. With this guide, they found that 82% of eyes required a tarsectomy, and 87% of patients overall achieved symmetry of 50.5 mm difference in MRD between the eyelids. This is again similar to others.3,5,11 They did not however report pre-operative, post-operative or change in MRD and did not compare to a ‘‘standard’’ 1:4 nomogram; thus, we are not certain whether the algorithm improved outcome. Ben Simon et al.3 also noted that a variety of resection lengths have been proposed for similar preoperative MRD heights and questioned the linear relationship between these variables. Using a familiar 1:4 ratio for resection, they analyzed 131 cases of MMCR. Again this group achieved similar levels of symmetry as previously reported at 81%. However they did not find a strong association between the amount of tissue resected and the amount of ptosis correction (r = 0.2). Putterman’s8 group later also addressed this same variability by attempting to create statistical model to accurately predict post-operative change in MRD by tissue resection length (accounting for preoperative MRD and phenylephrine response). They found that an extremely complex set of coefficients and transformations was required to predict outcome based on the amount of tissue resected. Simple linear relationships did not accurately describe outcome. Additionally, a closer look at the data in this study (8) reveals remarkable similarities to those presented here. In Table 2 of that article, they describe correlation between the amount of resection and the postoperative change in MRD of r = 0.2, which is similar to both ours at r = 0.17 and that reported by Ben Simon at r = 0.2.3 As would be expected, pre-operative MRD and the amount of tissue resected were highly correlated (0.6). Pre-operative MRD was associated with post-operative MRD (r = 0.3); however, not as closely as our data suggests (r = 0.5). These findings tend to support the hypothesis that the relationship !
2015 Informa Healthcare USA, Inc.
97
between the amount of tissue resected and the final lid position is weak overall, and may be a secondary effect of pre-operative planning. The relationship between resection and outcome has also been assessed using histological data, which tends to further support our contention. Zauberman et al.16 examined histopathology specimens of MMCR procedures and found no association between the amount of Muller’s muscle resected and post-operative outcome. They additionally did not find a significant correlation between caliper measurements of resected tissue length intra-operatively and postoperative lid position. Other groups have used different methodology and also drawn the conclusion that the amount of Muller’s muscle resected is not associated with change in MRD.17 Jakobiec et al. in studying Fasanella-Servat type operations also suggest that amount of Muller’s muscle found in surgical specimens does not affect clinical outcome.18 Overall, published data suggest that the MMCR procedure can correct ptosis in a large variety of patients including congenital, acquired, Horner’s syndrome and complex ocular surface disease patients.1,5,9,19,20 The bulk of the literature present similar symmetry and MRD results with varying techniques, resection nomograms, and patient populations. These convergent results in divergent populations support the contention that technique and population do not appear to appreciably alter outcome. These observations are further supported by data presented here suggesting that the wide variability in surgical response to differential resection length (Figure 1) is dependent more on the pre-operative MRD than the extent of resection (Figure 3). Even when accounting for phenylephrine response and tissue resection, the most important variable appears to be pre-operative lid height. This understanding carries broad implications for the approach to MMCR surgery and our conceptualization of its mechanism. This article does not present data that could support a pathophysiologic theory for the efficacy of MMCR surgery, and thus further studies are currently underway attempting to understand this process. However, we would contend that maintaining eyelid position is a complex, dynamic and active pathologic process involving visual, tactile and proprioceptive systems and ptosis is a state in which these eyelid position systems are disregulated. Ptosis surgery could be thought of similarly to strabismus surgery, in that it places the eyelid in a new position and allows the system to recalibrate and maintain a more physiologic spatial arrangement. For MMCR surgery specifically, Putterman may have been right in the beginning, that 8 mm of resection is appropriate for everyone. More accurately stated, resecting 8 mm of Muller’s muscle and
98 D. B. Rootman et al. conjunctiva may produce the same outcome as resecting 6 mm or 10 mm, whether that outcome is positive or negative.
DECLARATION OF INTEREST The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES 1. Putterman AM, Urist MJ. Muller Muscle-Conjunctiva Resection - Technique For Treatment Of Blepharoptosis. Arch Ophthalmol 1975;93(8):619–623. 2. Aakalu VK, Setabutr P. Current ptosis management: a national survey of ASOPRS members. Ophthal Plast Reconstr Surg 2011;27(4):270–276. 3. Ben Simon GJ, Lee S, Schwarcz RM, et al. Muller’s muscleconjunctival resection for correction of upper eyelid ptosis: relationship between phenylephrine testing and the amount of tissue resected with final eyelid position. Arch Facial Plast Surg 2007;9(6):413–417. 4. Carruth BP, Meyer DR. Simplified Mu¨ller’s muscleconjunctival resection internal ptosis repair. Ophthal Plast Reconstr Surg 2013;29:11–14. 5. Dresner SC. Fruther modifications of the Muller Muscle Conjunctival Resection Procedure for Blepharoptosis. Ophthal Plast Reconstr Surg 1991;7(2):114–122. 6. Georgescu D, Epstein G, Fountain T, et al. Mu¨ller muscle conjunctival resection for blepharoptosis in patients with poor to fair levator function. Ophthalmic Surg Lasers Imaging 2008;40:597–599. 7. Guyuron B, Davies B. Experience with the Modified Putterman Procedure. Plast Reconstr Surg 1988;82:775–780. 8. Mercandetti M, Putterman AM, Cohen ME, et al. Internal Levator Advancement by Muller’s Muscle–Conjunctival Resection. Arch Facial Plast Surg 2001;3:104–110.
9. Michels KS, Vagefi MR, Steele E, et al. Mu¨ller muscleconjunctiva resection to correct ptosis in high-risk patients. Ophthal Plast Reconstr Surg 2007;23:363—366. 10. Perry JD, Kadakia A, Foster JA. A new algorithm for ptosis repair using conjunctival Mu¨llerectomy with or without tarsectomy. Ophthal Plast Reconstr Surg. 2002;18: 426–9. 11. Putterman AM, Fett DR. Muller’s muscle in the treatment of upper eyelid ptosis: a ten-year study. Ophthalmic Surg 1986;17(6):354–360. 12. Putterman AM, Urist MJ. Muller Muscle Conjunctival Resection Ptosis Procedure. Ophthalmic Surg Lasers 1978; 9(3):27–32. 13. Weinstein GS, Buerger GF. Modifications Of The Mullers Muscle-Conjunctival Resection Operation For Blepharoptosis. Am J Ophthalmol 1982; 93(5):647–651. 14. Rufer F, Schroder A, Erb C. White-to-white corneal diameter: normal values in healthy humans obtained with the Orbscan II topography system. Cornea. 2005; 24(3):259–61. 15. Ben Simon GJ, Lee S, Schwarcz RM, et al. External levator advancement vs Mu¨ller’s muscle-conjunctival resection for correction of upper eyelid involutional ptosis. Am J Ophthalmol 2005;140:426–432. 16. Zauberman NA, Koval T, Kinori M, et al. Mu¨ller’s muscleconjunctival resection for upper eyelid ptosis: correlation between amount of resected tissue and outcome. Br J Ophthalmol 2013;97:408–411. 17. Morris WR, Fleming JC. A histological analysis of the Mullerectomy: redefining its mechanism in ptosis repair. Plast Reconstr Surg 2011;127:2333–2341. 18. Buckman G, Jakobiec FA, Hyde K, et al. Success Of the fasanella-servat operation independent of Mullers smooth-muscle excision. Ophthalmology 1989; 96(4):413–418. 19. Glatt HJ, Putterman AM, Fett DR. Muller’s muscleconjunctival resection procedure in the treatment of ptosis in Horner’s syndrome. Ophthalmic Surg 1990; 21(2):93–96. 20. Mazow ML, Shulkin ZA. Mueller’s muscle-conjunctival resection in the treatment of congenital ptosis. Ophthal Plast Reconstr Surg 2011;27(5):311–312.
Orbit
Copyright of Orbit is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
RESEARCH REPORT
The Role of Tissue Resection Length in the Determination of Post-Operative Eyelid Position for Muller’s Muscle-Conjunctival Resection Surgery Dan B. Rootman1, Justin Karlin2, Grant Moore1, and Robert Goldberg1 1
Division of Orbital and Ophthalmic Plastic Surgery, Jules Stein Eye Institute, University of California, Los Angeles, California, USA and 2Sackler School of Medicine, Tel Aviv University, New York, USA
ABSTRACT Purpose: To investigate the relationships between pre-operative marginal reflex distance (MRD), tissue resection length, phenylephrine response, and change in MRD with surgery for a cohort of individuals undergoing Muller’s muscle conjunctival resection (MMCR) surgery. Methods: All cases of MMCR surgery performed over a 13-year period at a single institution were screened for entry. Individuals with adequate photographic documentation and follow up were included. Patients with previous or concurrent upper eyelid, orbital or eyebrow disease of surgery were excluded. Marginal reflex distance (MRD) was calculated based on photographs utilizing public domain software. Data was plotted for inspection and appropriate statistical tests were performed. Results: During the study period 198 eyes fit criteria for analysis. A loose association between tissue resection length and change in MRD with surgery was found (r = 0.176, p50.05); this relationship was not significant in ANOVA analysis (p = 0.367). There was a strong association between MRD change with surgery and preoperative MRD (r = 0.498, p50.01). Approximately 28% of the sample responded to 2.5% phenylephrine drop instillation with a greater than 2 mm increase in MRD. The response to phenylephrine was strongly associated with pre-operative MRD (r = 0.441, p50.01). A regression on change in MRD with surgery with tissue resection, phenylephrine response 42 mm and pre-operative MRD as variables revealed a model with preoperative MRD as the only significant predictor (p50.01). Conclusion: Tissue resection length and phenylephrine response play small roles relative to pre-operative MRD in the determination of change in MRD with MMCR surgery. Keywords: Eyelids, Muller’s muscle-conjunctival resection, ptosis
INTRODUCTION
over time. Various strategies involving the addition or subtraction of Muller’s muscle resection length and/ or tarsus based on pre-operative ptosis or phenylephrine response have been proposed to improve predictability of the final eyelid position.1,5,7,8,10,11,13 However, such refinements in these published series have not been conclusively shown to increase overall predictability. It is yet unclear how varying technique and tissue resection length affect final outcome in MMCR surgery. The purpose of this study was to directly
Since originally described in 1975,1 the Muller’s muscle conjunctival resection (MMCR) procedure has gained significant popularity.2 This procedure has been utilized in a wide variety of ptosis cases, by a number of authors, producing mostly good results.1,3–12 Classically, a resection length to eyelid elevation ratio of 1:4 mm7 has been applied, however a number of refinements to this nomogram have been offered
Received 6 July 2014; Accepted 12 December 2014; Published online 13 March 2015 Correspondence: Dan B. Rootman, Division of Orbital and Oculoplastic Surgery, Jules Stein Eye Institute, 100 Stein Plaza, University of California, Los Angeles, CA 90095, USA. Tel: (310) 206 8250. Fax: (310) 825 9263. E-mail: [email protected]
92
Tissue Resection in MMCR investigate this relationship and to define preoperative factors that may predict post-operative eyelid position in MMCR surgery.
MATERIALS AND METHODS All cases of MMCR performed at the Jules Stein Eye Institute between January 1, 2000, and May 1, 2013, were screened for study entry. The dates were chosen as electronic clinical photographs were available for each patient in this time frame. This study was approved by the Institutional Review Board at the University of California-Los Angeles and investigations were performed according to the Declarations of Helsinki. Eyes were included if they had MMCR surgery in isolation from any other upper eyelid or brow procedures and if pre-operative standard and post phenylephrine photos, as well as at least one postoperative photo between 1 and 12 months after surgery were available. Eyes were excluded for criteria related to imaging if it not sufficiently clear or if they were not looking directly forward in primary position, in line with the camera. Patients with a history of previous ptosis or brow surgery or a systemic disease that may affect lid position such as thyroid eye disease, myesthnia gravis or chronic progressive external ophthalmoplegia were also excluded. Cases in which levator function was below 12 mm were also excluded, as were eyes with surgery involving resection of tarsus. Clinical histories were reviewed for study eligibility. Tissue resection amounts were extracted from preoperative plans, confirmed by operative reports. Cases with disagreement between the plan and report, without written clarification in the chart, were eliminated. All photographs were reviewed in a standard manner. All measurements were made using ImageJ (National Institutes of Health, Bethesda, MD) public domain software. The marginal reflex distance-1 (MRD) was measured from standardized photographs as the distance from the center of the pupil to the lowermost point of the upper eyelid in the midpupillary line. This measurement was used as a proxy for the marginal reflex distance, and avoids variability in the angular position of the flash in photography. Average corneal diameter for females (11.64 mm) and males (11.77 mm), as measured by Rufer et al.,14 was utilized as a reference in setting the measurement scale. MRD response to phenylephrine was measured as a change in MRD, 5 minutes after the instillation of 2.5% phenylephrine solution to the inferior fornix. Data were plotted raw and assessed for outliers and inconsistencies. Outliers were remeasured by a second reviewer. Paired t-tests were used to compare pre- and post-operative eyelid position. For ANOVA !
2015 Informa Healthcare USA, Inc.
93
analyses, continuous variables were binned in to 0.5 mm or 1 mm increments. For linear regression and correlation analyses, continuous data was utilized untransformed. Regression analyses were designed to test specific a priori hypotheses regarding the relative contribution of tissue resection length, eyelid position at presentation and phenylephrine response on postoperative change and are limited to such. All analyses were performed with IBM SPSS Statistics for Mac version 21.0 (SPSS Inc., an IBM Company, Somers, NY). The Bonferroni correction was applied where appropriate.
RESULTS Sample After applying study criteria, 198 eyes in 130 people were included in the final sample. Unilateral cases accounted for 30.8% of the surgeries and bilateral surgery was performed in 69.2% of eyes. Mean (SD) follow up was 3.78 (2.14) months.
Surgical Technique Under monitored anesthesia care, 2% lidocaine with 1:100,000 epinephrine was injected in to the upper eyelid. The eyelid was everted over a Desmarres retractor and the conjunctiva was dried. Resection length was marked on the conjunctival surface with a caliper at the 1/3 junction of the tarsus medially and laterally. Fine forceps were then used to grasp and elevate the conjunctiva and Muller’s muscle. An MMCR clamp was applied, and the Desmarres removed. A chromic gut suture was then passed under the clamp running in both directions with the ends externalized to skin. Resection was completed with a #15 Bard Parker blade. The eyelid was everted and gentle pressure applied for hemostasis.
Overall Effect of Surgery The overall mean effect of MMCR surgery was a 1.5 mm increase in the MRD from pre-operative to post-operative measurements. This was highly statistically significant (p50.01). Tissue resected and the response to surgery Resection length was on average (SD) 6.5 mm (1.5 mm), although the modal resection length was 8 mm representing 23% of total resections. A scatter plot of MRD change with surgery against the amount of tissue resected did not demonstrate a clear trend, with high variability (Figure 1). Pearson’s correlation analysis confirms the loose relationship (r = 0.176,
94 D. B. Rootman et al.
FIGURE 1. Scatter plot of change in marginal reflex distance (MRD) with surgery against tissue resected.
FIGURE 2. Mean change in marginal reflex distance (MRD) with surgery for each amount of tissue resected (df = 6, F = 1.1, p = .367).
p50.05). Regression analysis produces a significant (p50.05) model, explaining 3.1% of the total system variability. Dividing the mean change in MRD with surgery (1.5 mm), by the mean resection (6.5 mm) reveals a ratio of 0.23 mm MRD change per mm of resection, or 1 mm change in MRD for each 4 mm resected. The mean MRD change with surgery did not significantly differ for each mm of tissue resected (p = 0.367; Figure 2). When comparing eyes with 4 mm to eyes with 8-mm tissue resection lengths, the change in MRD with surgery did not significantly differ (t = 1.5, p = 0.133). However, the post-operative MRD did significantly differ between these groups, with the 4 mm resection demonstrating 0.7 mm higher postoperative MRD than the 8-mm resection group (p50.05). The pre-operative MRD for 4 mm resection
FIGURE 3. Scatter plot of change in marginal reflex distance (MRD) with surgery against pre-operative MRD.
FIGURE 4. Mean change in marginal reflex distance (MRD) with surgery for each category of pre-op MRD (df = 8, F = 11.1, p50.01).
patients was 1.3 mm higher than 8 mm resection patients (p50.01).
Pre-operative MRD and the response to surgery A similar scatter plot of MRD change with surgery against the pre-operative MRD demonstrated a more convincing trend and lower variability (Figure 3). Pearson’s correlation analysis confirms the relationship (r = 0.498, p50.01). Regression analysis produces a significant (p50.01) model, explaining 25.0% of the total system variability. This model suggests that for each increase of 1 mm in pre-operative MRD, the change in post-operative MRD will decrease by 0.54 mm. Restated, as the pre-operative MRD increases, the change with surgery decreases (see Figure 4). Orbit
Tissue Resection in MMCR
95
FIGURE 5. Scatter plot of pre-operative marginal reflex distance (MRD) against tissue resection length (mm).
FIGURE 6. Scatter plot of change in marginal reflex distance (MRD) with phenylephrine against pre-operative MRD.
Phenylephrine response and response to surgery A greater than 2-mm increase in MRD was noted after 2.5% phenylephrine drop instillation in 28.3% (n = 39) of the 138 eyes in which phenylephrine was applied. A greater than 1-mm increase was noted in 65.9% and a greater than 0.5-mm increase was noted in 83.3% of the same sample. Individuals who respond to phenylephrine with a 42-mm change in MRD had a 0.9-mm lower preoperative MRD than non-responders (p50.01). Additionally, these individuals had a greater change in MRD with surgery by 0.8 mm (p50.01). They did not have a significant difference in the final postoperative MRD (p40.05). For the group of individuals who demonstrated a42 mm increase in MRD with phenylephrine installation, the association between tissue resection length and change in MRD with surgery was stronger than in the full sample (r = 0.420, p50.01). Additionally, the pre-operative MRD was more highly correlated with surgical change in MRD (r = 0.568, p50.01). Tissue resected and pre-operative MRD are also more highly associated with each other in this group (r = 0.516, p50.01). Linear regression on change in MRD with surgery with resection length and pre-operative MRD as predictors was performed for individuals with phenylephrine response 42 mm. This revealed a significant (p50.01) model, explaining approximately 30% of the variability with only pre-operative MRD as a significant coefficient (p50.01). Tissue resection length did not maintain statistical significance in this model.
(r = 0.382, p50.05). This relationship is stronger than the one between the amount of tissue resected and change with surgery (r = 0.176, p50.05). Additionally, there was a strong association (r = 0.441, p50.01) between pre-operative MRD and change in MRD with phenylephrine (Figure 6). This association was stronger than the one between change in MRD with phenylephrine and change with surgery (r = 0.348, p50.01). A regression model with tissue resection length, pre-operative MRD and change in MRD with phenylephrine as predictors of change in MRD with surgery was fit. This model was significant (p50.01), and explained approximately 26% of the variability in the data. The only significant predictor in the model was pre-operative MRD (Table 1).
Tissue resected, pre-operative MRD, phenylephrine response and surgery There is a strong correlation (Figure 5) between the amount of tissue resected and pre-operative MRD !
2015 Informa Healthcare USA, Inc.
Subgroup analysis All analyses were performed for unilateral cases alone and no differences were noted relative to the bilateral analysis. Only bilateral analyses are presented here.
DISCUSSION This series of analyses, performed on a large high quality sample of MMCR surgeries, has demonstrated that the most important factor in determining postoperative MRD change is the pre-operative MRD. The amount of tissue resected and phenylephrine response do not appear to play statistically significant roles in determining surgical change. The weak association between resection length and post-operative MRD may be a statistical byproduct of the moderate relationship between resection length and pre-operative MRD (lower had a greater amount of tissue resected and vice versa). This may be a case of correlation rather than causation.
96 D. B. Rootman et al. TABLE 1. Linear regression of predictive factors on marginal reflex distance (MRD) change with surgery. Variable
Beta
p Value
Tissue resected (mm) Pre op MRD (mm) MRD responder
0.055 0.468 0.399
0.384 50.01 0.060
There are a number of key limitations to consider in this discussion. Initially it should be noted that a resection length of between 4 mm and 9 mm of tissue was performed in 98% of cases, and no tarsus was involved. Thus, these results may not apply to resections outside of that range, or to Fasanella– Servat type procedures. The data is also photographic in nature and may not accurately represent the dynamic eyelid position, although variability should be randomly distributed. Further, we did not apply a strict rule of 4:1 in each case, as some patients with latent ptosis may have had larger resections than what might be calculated based on the degree of ptosis evident in photographs. Again this variability should be evenly distributed statistically. Additionally, our statistical models did not account for all variables potentially important in the determination of outcome and are not able to make comments regarding overall importance of pre-operative MRD in the system, rather only in relation to a priori selected variables. Finally, it would be ideal to perform a randomized study comparing standard tissue resection (for instance 8 mm) to variable (based on algorithm). Such a study is in development. With regard to the role for the amount of tissue resected. Our result that there does not appear to be a significant relationship between resection length and outcome may be less surprising than on initial glance when the literature is critically reviewed. In Putterman et al.’s original article,1 they suggested removing approximately 8 mm of tissue for each case, regardless of the pre-operative MRD. They did note making some adjustment based on the phenylephrine response, although they did not specify the details. A re-analysis of their published results using linear regression and correlation statistics reveals an overall regression coefficient for pre-operative MRD on postoperative change in MRD of 0.49, which is strikingly similar to that reported here ( 0.53). The correlation between these variables in Putterman’s article is even greater at r = 0.82. Although clearly there are substantial differences in methods and measurements between the original study and our results, it is interesting to note that without varying the resection length, Putterman described very close association between pre-operative and post-operative MRD. In a later article, Putterman suggested making a modification to the resection length based on
phenylephrine testing, resecting 1 mm less for phenylephrine over-responders and 1 mm more for phenylephrine under-responders although overand under-response were not clearly defined. This article11 described 232 eyelids (an undetermined proportion using this variable formula) and found that 75% of acquired ptosis cases achieved a post-op MRD of between 2.5 and 5. They did not report associations between pre-operative and post-operative MRD. Guyuron et al.7 used a similar 7 mm-8 mm-9 mm rule for resection of conjunctiva and Mu¨llers muscle based on the phenylephrine response (slight – perfect – exaggerated). They described the mean change in eyelid position after surgery in 43 patients to be 2.1 mm. Presuming (they did not report this value) a mean resection length of 8 mm, they concluded by simple division that 1 mm of resection is required for 0.25 mm of eyelid elevation. No inferential statistics examining the relationship between these variables was offered. Weinstein and Buerger made a similar calculation 3 years later,13 and this heuristic has since persisted in the literature.15 Dresner used this ratio to create a rule for varying resection based on pre-operative ptosis. He reported 5 a series of 146 eyelids in which the tissue resection was based on the pre-operative ptosis with a modified 1:4 ratio. This rule was applied primarily in patients with a 42-mm response to phenylephrine, and the small number of patients with a smaller response had slightly more tissue removed. He found that 84% of patients achieved MRD symmetry between the eyelids within 0.5 mm. Additionally, he demonstrated a strong linear relationship between resection and change in MRD with surgery (r = 0.6). Dresner’s results on first pass appear disparate from ours. However some differences in the sample may explain these variations. It is not completely clear how many individuals had a phenylephrine response 42 mm in his study, but it would appear as though this number would be above to 80%. Subgroup analysis of strong (42 mm) phenylephrine responders in our sample reveals similar results. For patients demonstrating an MRD change with phenylephrine 42 mm, the relationship between tissue resection and MRD change on our study was much closer to Dresner’s (r = 0.420). However the relationships between pre-operative MRD, tissue resection length change with phenylephrine and MRD change with surgery are all stronger for patients with 42 mm change in MRD with phenylephrine. In linear regression analysis, again only pre-operative MRD maintains significance for this group. Unfortunately in Dresner’s study, no relationships between pre-operative ptosis and resection were calculated, nor were multivariate models fit. With data presented in that article, it is not possible to determine whether the association between resection Orbit
Tissue Resection in MMCR and outcome is an effect of a strong relationship between pre-operative ptosis and resection length. Additionally, the strong relationship reported in Dresner’s article has not been replicated in the literature. Perry et al. noted astutely from these articles that a variety of resection lengths appear to produce similar results.10 They went on to reason that in order to improve predictability of the procedure, the phenylephrine response can be used as a literal guide to outcome. They presumed that 10% phenylephrine instillation will maximally stimulate the Muller’s muscle and thus exactly simulate a 9-mm (most of the Muller’s muscle) resection. Thus an ideal phenylephrine response should be recreated surgically with a 9-mm MMCR resection. Under-correction with phenylephrine could be managed by adding 1-mm tarsal resection for each 1-mm under-correction. With this guide, they found that 82% of eyes required a tarsectomy, and 87% of patients overall achieved symmetry of 50.5 mm difference in MRD between the eyelids. This is again similar to others.3,5,11 They did not however report pre-operative, post-operative or change in MRD and did not compare to a ‘‘standard’’ 1:4 nomogram; thus, we are not certain whether the algorithm improved outcome. Ben Simon et al.3 also noted that a variety of resection lengths have been proposed for similar preoperative MRD heights and questioned the linear relationship between these variables. Using a familiar 1:4 ratio for resection, they analyzed 131 cases of MMCR. Again this group achieved similar levels of symmetry as previously reported at 81%. However they did not find a strong association between the amount of tissue resected and the amount of ptosis correction (r = 0.2). Putterman’s8 group later also addressed this same variability by attempting to create statistical model to accurately predict post-operative change in MRD by tissue resection length (accounting for preoperative MRD and phenylephrine response). They found that an extremely complex set of coefficients and transformations was required to predict outcome based on the amount of tissue resected. Simple linear relationships did not accurately describe outcome. Additionally, a closer look at the data in this study (8) reveals remarkable similarities to those presented here. In Table 2 of that article, they describe correlation between the amount of resection and the postoperative change in MRD of r = 0.2, which is similar to both ours at r = 0.17 and that reported by Ben Simon at r = 0.2.3 As would be expected, pre-operative MRD and the amount of tissue resected were highly correlated (0.6). Pre-operative MRD was associated with post-operative MRD (r = 0.3); however, not as closely as our data suggests (r = 0.5). These findings tend to support the hypothesis that the relationship !
2015 Informa Healthcare USA, Inc.
97
between the amount of tissue resected and the final lid position is weak overall, and may be a secondary effect of pre-operative planning. The relationship between resection and outcome has also been assessed using histological data, which tends to further support our contention. Zauberman et al.16 examined histopathology specimens of MMCR procedures and found no association between the amount of Muller’s muscle resected and post-operative outcome. They additionally did not find a significant correlation between caliper measurements of resected tissue length intra-operatively and postoperative lid position. Other groups have used different methodology and also drawn the conclusion that the amount of Muller’s muscle resected is not associated with change in MRD.17 Jakobiec et al. in studying Fasanella-Servat type operations also suggest that amount of Muller’s muscle found in surgical specimens does not affect clinical outcome.18 Overall, published data suggest that the MMCR procedure can correct ptosis in a large variety of patients including congenital, acquired, Horner’s syndrome and complex ocular surface disease patients.1,5,9,19,20 The bulk of the literature present similar symmetry and MRD results with varying techniques, resection nomograms, and patient populations. These convergent results in divergent populations support the contention that technique and population do not appear to appreciably alter outcome. These observations are further supported by data presented here suggesting that the wide variability in surgical response to differential resection length (Figure 1) is dependent more on the pre-operative MRD than the extent of resection (Figure 3). Even when accounting for phenylephrine response and tissue resection, the most important variable appears to be pre-operative lid height. This understanding carries broad implications for the approach to MMCR surgery and our conceptualization of its mechanism. This article does not present data that could support a pathophysiologic theory for the efficacy of MMCR surgery, and thus further studies are currently underway attempting to understand this process. However, we would contend that maintaining eyelid position is a complex, dynamic and active pathologic process involving visual, tactile and proprioceptive systems and ptosis is a state in which these eyelid position systems are disregulated. Ptosis surgery could be thought of similarly to strabismus surgery, in that it places the eyelid in a new position and allows the system to recalibrate and maintain a more physiologic spatial arrangement. For MMCR surgery specifically, Putterman may have been right in the beginning, that 8 mm of resection is appropriate for everyone. More accurately stated, resecting 8 mm of Muller’s muscle and
98 D. B. Rootman et al. conjunctiva may produce the same outcome as resecting 6 mm or 10 mm, whether that outcome is positive or negative.
DECLARATION OF INTEREST The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES 1. Putterman AM, Urist MJ. Muller Muscle-Conjunctiva Resection - Technique For Treatment Of Blepharoptosis. Arch Ophthalmol 1975;93(8):619–623. 2. Aakalu VK, Setabutr P. Current ptosis management: a national survey of ASOPRS members. Ophthal Plast Reconstr Surg 2011;27(4):270–276. 3. Ben Simon GJ, Lee S, Schwarcz RM, et al. Muller’s muscleconjunctival resection for correction of upper eyelid ptosis: relationship between phenylephrine testing and the amount of tissue resected with final eyelid position. Arch Facial Plast Surg 2007;9(6):413–417. 4. Carruth BP, Meyer DR. Simplified Mu¨ller’s muscleconjunctival resection internal ptosis repair. Ophthal Plast Reconstr Surg 2013;29:11–14. 5. Dresner SC. Fruther modifications of the Muller Muscle Conjunctival Resection Procedure for Blepharoptosis. Ophthal Plast Reconstr Surg 1991;7(2):114–122. 6. Georgescu D, Epstein G, Fountain T, et al. Mu¨ller muscle conjunctival resection for blepharoptosis in patients with poor to fair levator function. Ophthalmic Surg Lasers Imaging 2008;40:597–599. 7. Guyuron B, Davies B. Experience with the Modified Putterman Procedure. Plast Reconstr Surg 1988;82:775–780. 8. Mercandetti M, Putterman AM, Cohen ME, et al. Internal Levator Advancement by Muller’s Muscle–Conjunctival Resection. Arch Facial Plast Surg 2001;3:104–110.
9. Michels KS, Vagefi MR, Steele E, et al. Mu¨ller muscleconjunctiva resection to correct ptosis in high-risk patients. Ophthal Plast Reconstr Surg 2007;23:363—366. 10. Perry JD, Kadakia A, Foster JA. A new algorithm for ptosis repair using conjunctival Mu¨llerectomy with or without tarsectomy. Ophthal Plast Reconstr Surg. 2002;18: 426–9. 11. Putterman AM, Fett DR. Muller’s muscle in the treatment of upper eyelid ptosis: a ten-year study. Ophthalmic Surg 1986;17(6):354–360. 12. Putterman AM, Urist MJ. Muller Muscle Conjunctival Resection Ptosis Procedure. Ophthalmic Surg Lasers 1978; 9(3):27–32. 13. Weinstein GS, Buerger GF. Modifications Of The Mullers Muscle-Conjunctival Resection Operation For Blepharoptosis. Am J Ophthalmol 1982; 93(5):647–651. 14. Rufer F, Schroder A, Erb C. White-to-white corneal diameter: normal values in healthy humans obtained with the Orbscan II topography system. Cornea. 2005; 24(3):259–61. 15. Ben Simon GJ, Lee S, Schwarcz RM, et al. External levator advancement vs Mu¨ller’s muscle-conjunctival resection for correction of upper eyelid involutional ptosis. Am J Ophthalmol 2005;140:426–432. 16. Zauberman NA, Koval T, Kinori M, et al. Mu¨ller’s muscleconjunctival resection for upper eyelid ptosis: correlation between amount of resected tissue and outcome. Br J Ophthalmol 2013;97:408–411. 17. Morris WR, Fleming JC. A histological analysis of the Mullerectomy: redefining its mechanism in ptosis repair. Plast Reconstr Surg 2011;127:2333–2341. 18. Buckman G, Jakobiec FA, Hyde K, et al. Success Of the fasanella-servat operation independent of Mullers smooth-muscle excision. Ophthalmology 1989; 96(4):413–418. 19. Glatt HJ, Putterman AM, Fett DR. Muller’s muscleconjunctival resection procedure in the treatment of ptosis in Horner’s syndrome. Ophthalmic Surg 1990; 21(2):93–96. 20. Mazow ML, Shulkin ZA. Mueller’s muscle-conjunctival resection in the treatment of congenital ptosis. Ophthal Plast Reconstr Surg 2011;27(5):311–312.
Orbit
Copyright of Orbit is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
