Figure 1. High-resolution surface-coil magnetic resonance imaging (MRI). A, Sagittal T2-weighted high-resolution 3-T MRI using surface coils in an aged-matched healthy individual. B, Sagittal T2-weighted high-resolution 3-T MRI using surface coils in a patient with blepharophimosis- ptosis-epicanthus inversus. The following structures can be observed: (1) superior rectus muscle, (2) levator palpebrae superioris, (3) orbital septum, (4) levator aponeurosis, (5) Müller muscle, (6) Whitnalls ligament, (7) tarsal plate, (8) preaponeurotic fat pad, (9) inferior rectus, (10) superior orbital rim, and (11) transverse fascial sheath expansion.
Figure 2. Intraoperative anatomic examination. Intraoperative photograph of the levator muscle during supramaximal levator resection in a representative patient with blepharophimosis-ptosis-epicanthus inversus showing the anterior thin, disorganized, aponeurosis-like structure followed by a thick, more organized musclelike structure. The junction of the 2 parts was approximately 20 to 25 mm from the anterior insertion of aponeurosis on the tarsal plate (black arrows).
Figure 3. Histopathologic examination results. In all images, the larger arrows indicate collagenous structure; asterisks indicate striated muscle. A-C, Hematoxylin-eosin staining of levator palpebrae superioris (LPS) muscle from a representative patient with blepharophimosis-ptosis-epicanthus inversus (BPES). A, High-power view of the anterior part showing smooth muscle fibers (arrowhead) with collagenous structure; a few blood vessels can be seen. B, Low-power view of the posterior end in a representative patient with BPES showing a collagenous structure with striated muscle fibers (asterisk) at the end. Minimal areas of fatty degeneration can be noted (2 small arrows). C, High-power view showing normal-appearing muscle fibers with peripheral nuclei suggestive of striated muscle (asterisk). The fibers are round to oval, with minimal variation in size. D-F, Hematoxylin-eosin staining of LPS muscle from a patient with simple severe congenital ptosis. D, The anterior end of the resected part with smooth muscle fibers suggestive of Müller muscle (arrowhead), as well as a dense collagenous structure suggestive of aponeurosis. E, Low-power view of the posterior end of the control sample showing fatty degeneration (2 small arrows) along with disorganized collagenous structure. F, High-power view of the posterior end of the control sample showing extensive fatty degenerative changes (2 small arrows) with no striated muscle fibers. G-I, Hematoxylin-eosin staining of LPS muscle from a normal tissue control (cadaver). G, The anterior part of the resected muscle with smooth muscle fibers suggestive of Müller muscle (arrowhead), as well as a dense collagenous structure suggestive of aponeurosis. H, Low-power view of the posterior part of the LPS showing compact and dense striated muscle (asterisk) with little collagenous tissue and almost no fatty degeneration. I, High-power view of the posterior end of the LPS showing in detail the striated muscle fibers (asterisk). (Original magnification: A-F, ×100; G-I, ×200.)
Decock CE, De Baere EE, Bauters W, Shah AD, Delaey C, Forsyth R, Leroy BP, Kestelyn P, Claerhout I. Insights Into Levator Muscle Dysfunction in a Cohort of Patients With Molecularly Confirmed Blepharophimosis-Ptosis-Epicanthus Inversus Syndrome Using High-Resolution Imaging, Anatomic Examination, and Histopathologic Examination. Arch Ophthalmol. 2011;129(12):1564–1569. doi:10.1001/archophthalmol.2011.348
Author Affiliations: Departments of Ophthalmology (Drs Decock, De Baere, Delaey, Leroy, Kestelyn, and Claerhout), Radiology (Dr Bauters), and Pathology (Dr Forsyth), Center for Medical Genetics (Drs De Baere and Leroy), Ghent University Hospital, Ghent, Belgium; and Department of Orbit & Oculoplasty, Bombay City Eye Hospital, Mumbai, India (Dr Shah).
Objective To study the basis of defective levator palpebrae superioris (LPS) function in blepharophimosis-ptosis-epicanthus inversus syndrome (BPES), an autosomal dominant eyelid malformation sometimes associated with ovarian dysfunction.
Methods Eight patients with molecularly proved BPES underwent high-resolution surface-coil 3-T magnetic resonance imaging before surgical intervention. The features of LPS muscle and adjoining connective tissue were compared with an age-matched control subject. During LPS resection for ptosis repair, detailed anatomic examination of the LPS was performed. Histopathologic characteristics were compared with normal control samples from a cadaver and a patient with simple severe congenital ptosis.
Results The most striking feature shown on magnetic resonance imaging was the thin, long anterior part of the LPS. During the operation, this consisted of a disorganized, thin, long aponeurosis. However, in the posterior part of the LPS, there was an organized thick structure suggestive of a muscle belly. Histopathologic examination revealed posteriorly well-formed striated muscle fibers in all patients with BPES but not in the control sample from the patient with simple severe congenital ptosis. These striated muscle fibers were comparable to those of the normal control tissue but were more intermixed with collagenous tissue and little fatty degeneration.
Conclusions The presence of striated muscle fibers in LPS of patients with BPES contrasts with the fatty degeneration in patients with simple severe congenital ptosis. To our knowledge, this is the first study providing novel insights into the pathogenesis of the eyelid malformation in BPES through extensive imaging, anatomic study, and histopathologic testing in a unique cohort of patients with molecularly proved BPES.
The blepharophimosis-ptosis-epicanthus inversus syndrome (BPES) is a complex eyelid malformation characterized by 4 major abnormalities that are present at birth: blepharophimosis, ptosis, epicanthus inversus, and telecanthus.1 In addition, lateral displacement of the inferior punctum, another important anatomic hallmark of BPES, is present in all patients.2
The inheritance of this syndrome is most often autosomal dominant, and it is the result of mutations in the FOXL2 gene (OMIM *605597), which encodes a forkhead transcription factor.3- 7
The diagnosis of BPES is based primarily on clinical findings. However, molecular genetic testing revealing a FOXL2 mutation is important to confirm the clinical diagnosis.7
Management of BPES requires multidisciplinary care, including an oculoplastic operation. Surgical correction of the eyelid malformation in BPES is recommended not only for cosmetic reasons but also because of functional implications, including amblyopia, strabismus, and refractive errors.8 Surgical management traditionally involves a medial canthoplasty for correction of blepharophimosis, epicanthus inversus, and telecanthus in children aged 3 to 5 years, followed approximately 1 year later by ptosis correction.
Most patients with molecularly proven BPES have severe congenital ptosis with poor LPS function. The pathophysiologic factors underlying this defective LPS function in BPES remains largely unexplained. Until now, this was commonly attributed to dysplasia or even absence of the levator palpebrae superioris (LPS).9 A recent study10 demonstrated that supramaximal LPS resection resulted in a significant increase in LPS function, which is not the case when conventional ptosis repair using a frontalis suspension procedure is performed in this condition.
To provide an anatomic substrate for our earlier findings, we set out to study the basis of defective LPS function in BPES through extensive use of magnetic resonance imaging (MRI) with surface coils, anatomic examination, and histopathologic testing in a unique cohort of patients with molecularly proved BPES.
Eight consecutive patients with molecularly proved BPES6 were included in this study. The patients' parents or guardians provided informed consent, and the study was conducted in accordance with the tenets of the Declaration of Helsinki, with formal ethics committee approval. In general, presence of the 4 major criteria (blepharophimosis, ptosis, epicanthus inversus, and primarily telecanthus) was initially used for a clinical diagnosis of BPES. In addition, a fifth anatomic hallmark (ie, lower eyelid malpositioning) was present in all patients. Mutation screening of the FOXL2 gene was performed as previously described4- 6 to confirm the diagnosis.
Detailed history and clinical evaluation was carried out in all patients to rule out any contraindication for MRI. All patients underwent high-resolution 3-T MRI with surface coil before the operation to correct ptosis. The MRIs were acquired using a 3-T scanner (Trio; Siemens, Erlangen, Germany). The scanning was performed with a high-resolution surface coil with an inner diameter of 60 mm. The surface coil was placed on the eye to be scanned and was centered and secured in place by tape. The scanning protocol included a sagittal and coronal T2-weighted sequence and a sagittal T1-weighted sequence. Both sequences had a field of view of 100 mm and an in-plane resolution of 0.4 mm. The images were taken with both eyes closed (resting position in slight downgaze). The features of the LPS muscle and adjacent connective tissue were studied. These were compared with findings of a high-resolution 3-T MRI scan with surface coil in a healthy, age-comparable control individual. The MRI findings were correlated with intraoperative anatomic and postoperative histopathologic findings.
The ptosis operation was performed in all patients with BPES between the ages of 4 and 13 years. All patients had severe ptosis with poor LPS function (ranging from 0-4 mm). During the operation, using supramaximal LPS resection, detailed anatomic examination of the LPS was performed.
Histopathologic examination was performed on the resected tissue using hematoxylin-eosin, desmin, and smooth muscle actin staining. The histopathologic findings were compared with those from examination of a normal control sample taken from a cadaver as well as a sample from a patient with simple severe congenital ptosis, also having an LPS function less than 4 mm and undergoing the same type of operation.
Data on the FOXL2 mutation screening, each patient's age at the time of the ptosis operation, and LPS aponeurosis length measured on the sagittal T2-weighted high-resolution surface-coil MRI scan are listed in the Table.
Representative MRI findings are depicted in Figure 1A (healthy control) and 1B (patient with BPES). In a sagittal T2-weighted high-resolution surface-coil MRI of a healthy individual (Figure 1A), several distinct structures could be observed in an anterior-posterior direction, starting with the thin skin and subcutaneous tissue. The preseptal part of the orbicularis oculi has an intermediate to hypointense signal intensity in T2-weighted images. The preseptal fat pad appears as a hyperintense structure, and the orbital septum forms a hypointense bandlike structure descending from the superior orbital rim and fusing with the LPS aponeurosis. The preaponeurotic fat is seen as a hyperintense structure and is identified just posteriorly to the orbital septum. The LPS splits into Müller muscle (which runs more posteriorly) on one side and into the aponeurosis on the other side. The aponeurosis attaches to the upper part of the tarsal plate after fusing with the orbital septum. Posteriorly, the aponeurosis fuses with the superior rectus muscle and a few connective septa between these two are observable on the image. The anterior, bandlike component of the common sheath between the LPS and superior rectus muscle is called the transverse superior fascial expansion and is seen as a hyperintense structure below the aponeurosis. Whitnalls ligament is identified as a hypointense V-shaped structure at the highest point of the LPS. From that point, the LPS changes direction and descends apically.
Figure 1B shows a sagittal T2-weighted high-resolution surface-coil MRI in a representative patient with BPES. To facilitate comparison with the age-matched control depicted in Figure 1A, the same numbering has been applied to both figures.
In all patients with BPES examined, there were several notable differences found compared with the healthy control. The volume of the preaponeurotic fat pad was increased in all patients. Whitnalls ligament appeared to be stretched, and the transverse superior fascial expansion was thinned or absent. The most striking feature was the thin aponeurosis of the LPS muscle, which was seen in all patients. This thinned aponeurosis was also much longer than in the control participant. The mean LPS aponeurosis length measured in our 10 patients with BPES was 24.9 mm (range, 21-28 mm). Because of the loss of signal intensity as the probe is placed deeper into the orbit, a muscle belly could not be discerned in the patients compared with the control participant.
During the ptosis operation using the supramaximal LPS resection, the LPS aponeurosis was consistently seen as a thin, very long, disorganized structure. On dissecting further posteriorly, the aponeurosis was replaced by a much thicker and well-organized structure suggestive of a muscle belly (Figure 2). The junction between the disorganized aponeurosis and organized musclelike structure was located 20 to 25 mm from the tarsal plate insertion in all patients with BPES.
Tissue of the LPS obtained during the ptosis operation in patients with BPES was subjected to histopathologic examination. Hematoxylin-eosin staining of the anterior part of the LPS (insertion on the tarsal plate) revealed smooth muscle fibers suggestive of Müller muscle and dense disorganized fibrous tissue representing the aponeurosis (Figure 3A). The posterior part showed the junction of the fibrous tissue of the aponeurosis into an area of well-formed striated muscle fibers. The striated muscle fibers were round to oval, with minimal variation in diameter. These findings suggest the presence of a LPS muscle belly in BPES. Some scattered areas of fatty degeneration were found (Figure 3B and C).
For comparison, a sample from a patient with severe congenital ptosis with very poor LPS function, as well as a normal cadaver LPS specimen, were examined.
The congenital ptosis specimen showed a similar histopathologic anterior aspect of the LPS as seen in the BPES specimen, with Müller muscle and disorganized fibrous tissue representing the aponeurosis (Figure 3D). However, in the posterior part, striated muscle fibers were absent, contrary to the finding in the BPES specimen. There were also extensive areas of fatty degeneration and dense, disorganized fibrous tissue (Figure 3E and F). These findings indicate the absence of an LPS muscle belly in patients with simple severe congenital ptosis.
The normal control specimen showed, in the anterior part of the LPS, a very similar structure of collagenous tissue, representing the normal aponeurosis, and a smooth muscle, representing Müller muscle, compared with the BPES and congenital ptosis specimens (Figure 3G). The posterior part of the normal control sample showed striated muscle fibers comparable to those seen in BPES, with round to oval shape and little variation in diameter. However, the muscle belly in the normal control differed from that in the BPES specimen because the muscle belly was denser, with much less connective tissue and almost no fatty degeneration (Figure 3H and I).
One of the cardinal features of the complex eyelid malformation in BPES is severe ptosis with almost no LPS muscle function. However, the pathophysiologic factors underlying LPS dysfunction in BPES remain largely unexplained. It is commonly suggested9 that poor LPS function could be attributed to dysplasia or even absence of the LPS. However, in a recent study, our group10 was able to demonstrate an increase in LPS function after supramaximal resection during ptosis repair of BPES. We suggested an anatomic substrate for these findings and, in an attempt to provide more insight into the basis of LPS dysfunction in BPES, we performed high-resolution surface-coil 3-T MRI before the ptosis operation in 8 patients.
High-resolution MRI with surface coils is currently the optimal way to visualize the minute structures of the orbit and periorbita. Surface coils significantly improve the signal to noise ratio, allowing thinner sections and increased spatial resolution compared with conventional imaging.11- 13 The role of surface coils in delineating normal orbital structures has been demonstrated in various studies.11- 16
The MRI findings in patients with BPES have previously been discussed in only 2 reports. Dollfus et al17 performed MRI of the LPS muscle in 5 affected members of a family with molecularly proven BPES. Absence of the LPS muscle was reported in 4 of the patients, and a very thin LPS muscle was noted in the fifth patient. The authors, however, had not used surface coils in their scanning protocol; this might explain the nonvisualization of the LPS muscle. Tronina et al18 performed similar preoperative MRI without surface coils in patients with BPES. Contrary to Dollfus et al, Tronina et al noted the presence of an LPS muscle. However, BPES was not molecularly confirmed in the latter study, and no detailed description of the morphologic characteristics of the LPS muscle was given.
Our study, using 3-T MRI with surface coil, revealed detailed information on the anatomy of the LPS and its adjoining structures in normal controls and patients with BPES. By using surface coils in our scanning protocol, we could identify the LPS and follow its entire course in all patients, up to its fusion with the superior rectus. We were able to describe detailed anterior structures of the LPS (Müller muscle and aponeurosis) as well as its surrounding connective tissues (Whitnalls ligament and transverse superior fascial expansion). At a more posterior site, it was not possible to distinguish the LPS from the closely related superior rectus muscle. This was the result of a too close connection between the LPS and the superior rectus in its posterior part as well as diminished resolution as the probe moved further from the surface coil. The most striking finding was the very thinned and elongated aponeurosis in all patients with BPES, with a mean length of 24.9 mm, being much longer than the 9 mm in our control participant and reported19,20 measurements ranging between 8 and 14 mm. Because of loss of MRI signal activity, no muscle belly could be identified.
These MRI findings were completed by detailed anatomic examination during surgery and histopathologic analysis of the resected tissue. The anatomic examination confirmed the MRI findings by revealing a thin, very long, disorganized aponeurosis. In addition, anatomic examination of further maximal resection showed a rather well-organized muscle belly. The muscle belly was located 20 to 25 mm from the tarsal plate insertion in all patients with BPES, explaining why it could not be demonstrated on MRI.
Histopathologic examination of the resected part of the LPS further corroborated our MRI and anatomic findings. The anterior part of the resected sample from patients with BPES showed a normal appearance of Müller muscle. However, the aponeurosis was replaced by disorganized connective tissue. Even more notably, histopathologic analysis confirmed the presence of well-formed striated muscle fibers in the posterior part of the resected specimen. The anterior part of the LPS muscle in the BPES sample resembled the findings in the control sample of severe simple severe congenital ptosis. However, striated muscle fibers are absent in simple severe congenital ptosis, and present in BPES. In 1955, Berke and Wadsworth21 reported the presence of loose areolar tissue and complete absence of striated muscle fibers in patients with simple severe congenital ptosis. This was later confirmed by other groups.22 The absence of striated muscle fibers explains the poor LPS function in patients with simple severe congenital ptosis. The striated muscle found in patients with BPES contained more connective tissue and fatty degeneration than what was seen in a normal cadaveric control specimen.
Because we observed well-formed striated muscle fibers in patients with BPES, there must be a different rationale for the poor LPS function seen in these patients. Combining results from the MRI, anatomic, and histopathologic evaluations, we provide an anatomic substrate for the LPS dysfunction seen in BPES. There is a rather well-formed muscle belly, but the fact that it is located too deeply in the orbit, in combination with a poor connection to the tarsal plate through a very long and thin aponeurosis, might explain the poor LPS function. This anatomic substrate also explains why supermaximal resection in BPES results in increased LPS function, as we previously demonstrated.10
In conclusion, to our knowledge, this is the first study combining high-resolution surface-coil MRI, anatomic, and histopathologic examinations, allowing extensive evaluation of the LPS muscle in patients with molecularly confirmed BPES. Our study provides new insights into the pathophysiologic characteristics of its dysfunction and offers a rationale for using distinct ptosis repair techniques in BPES operations.
Correspondence: Christian E. Decock, MD, Department of Ophthalmology, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium (Christian.email@example.com).
Submitted for Publication: December 24, 2010; final revision received March 17, 2011; accepted April 14, 2011.
Financial Disclosure: None reported.