Functional Anatomy of the Anophthalmic Socket: Insights from Magnetic Resonance Imaging

Abstract

purpose. Changes in anophthalmic socket anatomy can significantly compromise esthetics and motility after enucleation. This study evaluated such changes by using magnetic resonance imaging (MRI) of enucleated and fellow orbits.

methods. High-resolution, surface coil MRI was performed in five patients after enucleation for uveal melanoma. Images were analyzed quantitatively to determine extraocular muscle (EOM) volume, maximum diameter, and length; orbital, optic nerve (ON), and orbital fat tissue (OFT) volume; and implant position. The fellow orbit was used as the control. Patients evaluated their satisfaction with the surgical results and were clinically examined.

results. Rectus EOM volume was slightly but not significantly reduced on the surgical compared with the control side. Oblique EOM and OFT volumes were unchanged. Rectus EOM length was significantly reduced in the surgical side, but maximum EOM diameter in central gaze position was slightly but not significantly greater on the enucleated side. Rectus EOM paths were not qualitatively changed by enucleation and continued to exhibit the influence of the connective tissue pulleys, which retained motility, as appropriate to EOM contraction. Implants were located significantly posterior to the normal globe location.

conclusions. Enucleation does not significantly change EOM volume, but shortens EOM paths, a change that would be expected to alter their mechanical properties. EOM pulleys appear to retain their functional role in enucleated orbits.

Surgical techniques for enucleation of the globe have been influenced by classic concepts of orbital anatomy. ^{1 2} Volume replacement and restoration of movement and comfort have been considered important aspects of a successful surgical outcome. ^{1 2 3 4} Volume deficiency and soft tissue displacement can contribute to complications such as enophthalmos and superior sulcus defect that compromise final results. ^{2 4} Restoration of saccadic eye movements occurring during speech is desirable because this greatly contributes to a normal facial expression. ⁵  

Recent studies have clarified the anatomic relationships of EOMs to the orbital connective tissues. It is now recognized that the rectus and inferior oblique EOMs pass through soft pulleys, connective tissue rings that constrain EOM paths and serve as functional mechanical origins. ^{6 7 8 9 10 11} Rectus pulleys are located at the EOM penetrations through the posterior Tenon’s fascia, much more anterior in the orbit than the true anatomic origins of the rectus EOMs in the annulus of Zinn. The normal globe center is 8 mm anterior to the array of rectus pulleys. ^{8 12 13} However, after enucleating the eye, most surgeons place the implant posterior to posterior Tenon’s fascia, and thus posterior to the pulleys that serve as the functional rectus EOM origins. ^{1 14} This relationship would be expected to alter profoundly the pulling directions of the rectus EOMs after enucleation. Another recent anatomic insight is that the rectus EOMs are bilaminar: the global layer (GL) of each EOM is in continuity with the tendinous insertion on the sclera, and acts to rotate the globe. The orbital layer (OL) inserts on the corresponding pulley to translate it posteriorly during EOM contraction. ^{7 8 9 10 11} Consequently, rectus tenotomy at enucleation disinserts each rectus EOM’s GL from the sclera, but leaves intact the OL insertion on the pulley. 

Magnetic resonance imaging (MRI) has been a useful tool for elucidating the functional anatomy of orbital soft tissues, including the EOMs, connective tissues, orbital fat tissue (OFT), and optic nerve (ON). ^{15 16 17} In orbits with intact globes, the functional anatomy of the EOMs has been characterized extensively by multipositional, surface coil MRI. ^{6 7 8 10 12 18 19 20 21 22 23 24 25 26 27 28 29 30} In enucleated orbits, low-resolution MRI has been used to evaluate changes in the vascularity of porous implants, ^{31 32} and computed radiographic tomography (CT) has been used to evaluate postenucleation socket changes. ⁴ This study was undertaken to evaluate changes in orbital soft tissues after enucleation, by using high-resolution surface coil MRI, and to correlate these changes with clinical findings. 

Methods

Five adult patients, three men and two women, who had undergone monocular enucleation for choroidal melanoma at the Ocular Oncology Center of the Jules Stein Eye Institute, were studied. Excluded were patients with vascular clips, cardiac pacemakers, prosthetic devices, or retained foreign bodies, claustrophobia, and pregnancy, in whom MRI was contraindicated. ³³ Also excluded were patients with a history of previous orbital surgery, trauma, or ultrasonographic evidence of extraocular extension of an intraocular tumor. All participants gave written informed consent to a protocol approved by the Institutional Review Board in conformity with the tenets of the Declaration of Helsinki. 

Mean subject age (± SD) was 49 ± 9 years (range, 37–62). Details of implant size and type, wrapping, and postoperative interval are presented in Table 1 . Four patients had been fitted with long-term cosmetic prostheses. In one patient a conformer was used in anticipation of fitting of a long-term prosthesis. The surgical technique in all cases consisted of conjunctival peritomy, isolation and tenotomy of all rectus EOMs after securing each with 5-0 sutures, superior and inferior oblique muscle disinsertion, clamping and sectioning of the ON in the posterior orbit, removal of the globe, insertion of a gentamicin-soaked implant posterior to Tenon’s fascia, suturing of the four rectus EOMs anterior to the implant, and closure of posterior Tenon’s fascia, anterior Tenon’s fascia, and conjunctiva in three layers. The postoperative course was uncomplicated in all patients. 

As a part of the study, a medical and ophthalmic history was recorded in each patient. In addition, each patient completed a self-evaluation questionnaire concerning postoperative comfort and prosthesis appearance and motility (satisfactory facial expression during speech), measured on a scale of 0 to 10, with 10 being the best outcome. Clinical examination was also performed, including the examination of ocular and prosthesis motility, eyelid position, and evaluation for enophthalmos and superior sulcus configuration. Patients were photographed in diagnostic gaze positions. The prosthesis motility defect (examined in comparison with the fellow eye in abduction, adduction, supraduction, and infraduction) was evaluated by the same examiner on a scale of 0 to 4 (0, completely mobile prosthesis; 1, 2, or 3, mild, moderate, and severe motility impairment, respectively; and 4, completely immobile prosthesis). 

Each patient underwent high-resolution, T1-weighted MRI scanning with a 1.5-T scanner (Sigma; General Electric, Milwaukee, WI). Each subject’s head was carefully stabilized in a supine position with the nose aligned to the longitudinal and the pupil to the transverse reference light beams of the scanner. An array of surface coils was deployed in phased pairs, two over each orbit, in a masklike closure held strapped to the face. An adjustable array of monocular, afocal illuminated fixation targets at nine diagnostic positions of gaze was secured 2.5 cm in front of each orbit with the center target centered over each eye. Subjects were coached to avoid unnecessary movements during scanning. Blinking was reduced by maximizing precorneal humidity using a transparent face mask and instructing subjects to avoid unnecessary blinks. Head movement was minimized by secure stabilization to the surface coil face mask and judicious use of padded restraints. Axial MR images were obtained at 3-mm thickness using a 256 × 192 matrix over a 10-cm² field of view to localize placement of subsequent higher resolution quasicoronal images, perpendicular to the long axis of the orbit. Multiple contiguous images 3 mm in thickness were obtained in a coronal plane relative to the head, using a 256 × 256 matrix over a 10-cm² field of view, giving a pixel resolution of 390 μm. Multiple contiguous quasicoronal images (perpendicular to the long axis of each orbit separately) of 2 mm thickness were then obtained with a 256 × 256 matrix over an 8-cm² field of view, giving pixel resolution of 313 μm. Quasicoronal imaging was obtained in central and eccentric gaze positions. Sets of contiguous, 2 mm thickness, quasisagittal images (parallel to the long axis of each orbit separately) were then similarly obtained at 313 μm. Digital MR images were converted to 8-bit tagged image file format (TIFF) quantified and analyzed on computer (Scion Image for Windows, ver. Beta 4.02; based on NIH Image for Macintosh, developed by Wayne Rasband, National Institutes of Health, Bethesda, MD, and the efilm workstation; eFilm Medical, Inc., Toronto, Ontario, Canada). Only images free from degradation by motion or other artifacts were analyzed quantitatively. Statistical analysis of findings was performed on computer (SPSS, ver. 8.0; SPSS Sciences, Chicago, IL). 

The volume of the EOMs, including the medial rectus (MR), inferior rectus (IR), lateral rectus (LR), superior rectus-levator palpebrae superioris complex (SR-LP), and superior oblique (SO) and the volume of the ON (or ON stump), orbital cavity, and prosthesis in central gaze were measured by summing the cross-sectional areas of each structure in every quasicoronal section and multiplying by the image plane thickness, as described elsewhere. ^{16 17} The volume of inferior oblique (IO) was measured by summing its cross-sectional area in every quasisagittal image plane and multiplying by the plane thickness. Only sections in which the respective EOM was clearly visible were included, to rule out anatomic distortions due to scarring. The volume of the OFT was measured by subtracting the volume of all EOMs, implant (surgical side) or globe (fellow side), prosthesis (surgical side), and ON from the total orbital volume, as described elsewhere. ^{16 17} Therefore, the volume of OFT included the lacrimal gland, vessels, nerves, and connective tissue. The length of SR-LP, IR, MR, and LR (in central gaze position) was evaluated by multiplying the number of planes showing the respective EOM by the image plane thickness. The maximum cross-sectional area of the rectus EOMs (in quasicoronal sections) was measured in central and maximally contracting secondary gaze positions. Thus, the areas of SR-LP, IR, MR, and LR were evaluated in supraduction, infraduction, adduction, and abduction, respectively. Subsequently, the percentage increase of the maximum EOM cross-sectional area between the central and maximally contracting secondary gaze positions was calculated. 

The paths of EOMs in the anophthalmic and control sockets were evaluated by measuring the distance from muscle centroids to selected reference planes at the image level 2 mm anterior to the anterior edge of the ON stump (enucleated side) or 2 mm anterior to the globe–ON junction (fellow side), in quasicoronal sections. The plane of the interhemispheric fissure and the plane perpendicular to it at the level of the cribriform plate of the ethmoidal bone were used as references for the vertical and horizontal axes, respectively. The area centroid of a cross-section is equivalent to the center of gravity of a shape of uniform density and thickness. ^{13 17}  

Because the EOMs pass through their encircling pulleys, rectus EOM pulley locations in the coronal plane have in the past been inferred from EOM paths defined in sets of quasicoronal MRI scans. ^{22 24 26} In secondary and tertiary gaze positions, discrete path inflections produced by the pulleys have been identified from similar MR images to additionally define the anteroposterior locations of the pulleys. ^{8 12 29} The current data set was adequate to determine the coronal plane locations of pulleys by the customary means, but connective tissue distortions and lack of a normally placed globe in the anophthalmic orbits precluded the determination of EOM path inflections in the customary manner from coronal plane images. Nevertheless, even axial images illustrated large EOM path inflections due to pulleys, at the sites of identifiable insertions of the OLs on the pulley tissues. We thus used a modified procedure for determination of the anteroposterior location of the rectus pulleys, assuming each rectus pulley to be located at the point where the connective tissue suspension attaches to the orbital surface of the EOM in axial images for the horizontal rectus EOMs or the location of the connective tissue ring for the vertical rectus EOMs. 

Implant position was evaluated by measuring the distance from the implant centroid (in the plane closest to the implant center) to the anterior edge of the optic chiasm (axial section), interhemispheric fissure (quasicoronal section) and cribriform plate of ethmoid bone (quasicoronal section). Findings were compared with the respective distances from the globe centroid. 

Results

The distance from the upper lid margin to the prosthetic corneal reflex (marginal-reflex distance, MRD) in the four patients who had permanent prostheses was 2.25 ± 1.25 (mean ± SD; range, 0.5–3.5) mm. This compares with 4.25 ± 0.28 (4–4.5) mm MRD from the anatomic cornea on the fellow side. Hertel exophthalmometry measurements (anteriorly from skin surface overlying the lateral orbital rim to corneal surface) were 15 ± 2 (12–17) mm and 21 ± 1 (19–21) mm in the anophthalmic and fellow sides, respectively. The prosthesis motility defect was 3.2 ± 1.0 (2–4), 2.5 ± 1.3 (1–4), 2.8 ± 0.5 (2–3), and 3.2 ± 0.5 (3 – 4) for adduction, abduction, supraduction, and infraduction, respectively. 

The volume of the orbital cavity of the anophthalmic and fellow sides was 26,300 ± 3,900 (mean ± SD; range, 21,400–32,100) and 26,400 ± 4,200 (21,300–33,000) mm³, respectively. These did not differ significantly (paired t-test, P = 0.33). The volumes of the ON stump (anophthalmic side) and the ON (fellow side) were 313 ± 88 (range, 183–413) and 569 ± 175 (326–780) mm³, respectively. The maximum ON cross-sectional areas on the anophthalmic and fellow sides were 30.5 ± 7.6 (range, 20.6–41.7) and 42.6 ± 10.7 (range, 25.9–53.8) mm², respectively. Both differences for the ON were statistically significant (paired t-test, P < 0.05). Prosthesis volume for the four patients having permanent prostheses was 2589 ± 175 (range, 2040–3769) mm³. Globe volume on the fellow side was 7246 ± 1102 (5486–8374) mm³. The volume of OFT on the anophthalmic and fellow sides was 16,600 ± 2,700 (13,500–20,700) and 15,700 ± 2,800 (range, 13,400–20,400) mm³, respectively, not statistically significant by paired t-testing. Representative images used for determination of the volumes of the rectus EOMs, including the MR, IR, LR, and SR-LP, are illustrated in Figure 1 . Volumes of EOMs were reduced minimally on the surgical compared with the fellow sides. The volume reductions were 3.6%, 4.5%, 6.0%, and 9.6%, respectively. The volumes of the superior oblique (SO) muscles were also determined from quasicoronal images (Fig. 1) , but did not differ significantly between the surgical and fellow sides. The volumes of the inferior oblique (IO) muscles were determined from quasisagittal images (Fig. 2) , and also showed no statistically significant differences between the surgical and fellow sides. Volumes of the EOMs are presented in Table 2 . 

The maximum cross-sectional area of each rectus EOM in central gaze was higher on the anophthalmic than the fellow side, although the difference was statistically significant only for the MR (Table 3) . As an index of contractility, the percentage increase in the maximum cross-sectional area of the rectus EOMs between central and diagnostic secondary gaze positions was determined from quasicoronal images, as illustrated in Figure 3 . This measure of contractility was higher in the fellow than in the surgical side, although differences were not statistically significant (Table 4) . 

The length of each rectus EOM was significantly reduced (10%–20%) on the anophthalmic compared with the fellow side (Table 5) . Differences in EOM volume, maximum diameter and length, were examined for possible correlations with clinical parameters, including age, gender, postoperative interval, ptosis, exophthalmometry measurements, appearance, prosthesis motility defects, and implant type, size, and wrapping method. None of these correlations was statistically significant. 

EOM pulleys were identifiable in enucleated orbits and appeared to move posteriorly from central to diagnostic gaze directions. The posterior shift of the MR pulley can be observed directly in Figure 4 . In this case, the pulley produced a striking EOM path inflection and moved posteriorly in adduction along with the associated connective tissues. Area centroids of the rectus EOMs in the quasicoronal image plane 2 mm anterior to the junction of the globe and ON were used as indicators of rectus pulley locations in fellow eyes, and comparable centroids were determined in the enucleated side in the image plane 2 mm anterior to the anterior edge of the ON stump (Fig. 5) . The distance between rectus EOM centroids and the horizontal and vertical reference planes did not differ significantly between the enucleated and fellow orbits (Fig. 5) , indicating that the rectus pulleys were not significantly displaced in anophthalmic orbits. 

The mean lateral distances from the implant centroid to the interhemispheric fissure and the inferior distance to the cribriform plate were not significantly different from the respective distances from the globe centroid (Fig. 5 , Table 6 ). This implies that mean implant position was along the center of the rectus EOM pulley array laying along the orbital axis. However, the anteroposterior distance from the optic chiasm to the implant centroid was significantly shorter than the respective distance to the globe centroid, implying a roughly 8.5-mm more posterior location of the implant than the normal globe (Fig. 6 , Table 7 ). There was some variation in this distance, ranging from 7.9 mm to 13.2 mm. Of course, the more posterior location of the implant relative to the normal globe affects the lateral position of the implant in a craniocentric coordinate system exemplified by the axial images of both orbits in Figure 4 . In axial views, the implant appeared slightly closer to the midline than the normal globe. 

Discussion

A high-resolution MRI scan showed that EOMs in anophthalmic orbits substantially retain their volume and that rectus EOM pulleys can be identified in anophthalmic orbits. A uniform finding was that the position of the implant in the anophthalmic orbit averages 8.5 mm posterior to that of the normal globe. 

MRI is a reliable tool for measuring orbital structures, including the EOMs. ^{16 17} Advantages of MRI include lack of exposure to ionizing radiation, high quality of soft-tissue contrast, capability of multiplanar imaging and availability of techniques for examination of EOM contractility and tissue movement. ^{15 16 17} In the case of anophthalmic sockets, the fellow orbit has been used as a control in previous studies. ⁴ Although volumetric studies of orbital structures may produce different results depending on imaging orientation, ¹⁷ care was taken in the current study to standardize imaging orientation so that results were directly comparable. Scans were performed with the prostheses in place. Despite previous reports ³⁴ of MRI signal artifacts from prosthesis, the MR images obtained were not degraded by artifacts, probably because of the absence of electrically conductive metallic pigments from the prosthesis worn by the current subjects. Despite the capability of the current MRI technique to resolve them separately, the SR and LPS EOMs were evaluated as a unit (complex) because of their close anatomic relationship, as described elsewhere. ¹⁷  

Investigators in other studies have claimed that orbital fat atrophy can occur after enucleation, possibly related to surgical manipulations (cautery) ¹⁴ or circulatory changes attributed to decreased ocular metabolism ³⁵ and can predispose to complications such as superior sulcus defect and enophthalmos. ^{14 36} However, researchers in other studies have concluded that the main cause of volume loss in enucleation is the removal of the globe per se and the authors have therefore emphasized the importance of replacing lost volume by an adequate implant. ^{2 4 36} In the present study, the amount of enophthalmos and ptosis in the enucleated orbits are in accordance with previous reports on prosthetic enophthalmos ³ and postenucleation ptosis, ³⁷ implying adequate volume replacement. Absence of a difference in volumes of EOMs and the OFT between the anophthalmic and the fellow sides implies minimal postoperative orbital soft tissue loss. The maximum ON diameter was significantly reduced, and the significant change in ON volume between the surgical and fellow sides can largely be attributed to axonal atrophy. 

The significantly shorter distance from the implant centroid to the optic chiasm (compared with the respective distance from the globe centroid) implies a relatively posterior intraorbital implant position. Previous studies have described a posteroinferior change in implant position, often described as a postoperative implant migration. ^{1 14} However, the distance from the implant centroid to the optic chiasm did not correlate with the postoperative interval (in the present study beginning at 3 months), suggesting that implant displacement is not progressive or that any potential intraorbital migration is completed very early in the postoperative course. The posterior intraorbital position of the implant detected here corresponds with the significantly reduced lengths of the rectus EOMs in the anophthalmic orbit, because these were sutured anterior to the implant. The disparity between the anatomically known length of the rectus EOMs ³⁸ and the length measured in MRI scans could be explained by the fact that measurements were performed in quasicoronal scans, thus corresponding to the projection of each EOM to a quasisagittal plane. According to a previous CT scan study, the SR and IR muscles are retracted after enucleation, an effect thought to cause redistribution of orbital fat, displacement of conjunctival fornices, and prosthesis tilting. ⁴ The significantly reduced EOM length may have functional implications. It is known that EOM force lessens as EOM length shortens past the point of maximum myofilament overlap. ³⁹ Length reduction may result in less contractile force. ^{39 40} The lower ratio of maximum diameter of rectus EOMs from central to diagnostic secondary gaze positions suggests less contractility. Changes in contractile properties of EOMs may underlie previous findings of significant reduction in the saccadic amplitude of artificial eyes compared with fellow healthy eyes. ⁴¹  

The previously described active-pulley hypothesis (APH) proposes that a pulley serves as the functional origin of the rectus EOM and that pulleys make coordinated, gaze-related translations along the EOM axis in secondary gaze positions (they move posteriorly during EOM contraction and anteriorly during EOM relaxation), to remain at a fixed distance from the insertion relative to the globe. ⁸ In the present study, EOM pulleys were identifiable in enucleated orbits and displayed a posterior movement during secondary gaze directions, implying that pulleys retain their functional roles after enucleation. Physiologic rectus pulley movement might be expected to shift the soft tissues of the fornices, with which they have close anatomic relationships. ¹³ It has long been assumed that prosthesis motility after enucleation is largely determined by the degree to which the rectus EOMs rotate the implant, and the degree to which rotation of the implant is coupled to the prosthesis. ^{1 4 14} The recent orbital and EOM anatomic findings challenge this assumption and support a greater role of the conjunctival fornices in prosthesis movement. The GLs of the rectus EOMs would be expected to have poor mechanical advantage in rotating orbital prostheses, because these are posterior to the rectus pulleys. However, rectus OLs that are inserted on the rectus pulleys should remain capable of moving the conjunctival fornices and perhaps move the prosthesis by a mechanism independent of implant rotation. Images such as those in Figure 4 suggest that pulley movements may be coupled to the conjunctival fornices with which they have intimate anatomic relationships, challenging the traditional concept of prosthesis motility. Therefore, we alternatively propose that direct movement of the fornices is principally responsible for prosthesis motility in the anophthalmic orbit, independent of implant rotation. This concept better explains the motility of prostheses that have no direct mechanical coupling to the implant. 

Based on the findings of the present study, enucleation does not produce significant changes in the volume and paths of EOMs, but it does produce a significant reduction in EOM length, which could affect the mechanical properties of EOMs. Further research in this field should perhaps include MRI scans in patients with implant–prosthesis coupling by mechanical pegs, as well as sequential imaging of the same orbit at various postoperative intervals. Additional data could be useful in modifying techniques to improve surgical results and reduce the need for reoperations. 

 Supported by National Eye Institute Grant EY08313 (JLD) and the Straatsma Retina Research Fund. JLD received an unrestricted award from Research to Prevent Blindness and is the Laraine and David Gerber Professor of Ophthalmology.

 Submitted for publication February 18, 2003; revised May 24, 2003; accepted June 2, 2003.

 Disclosure: E.T. Detorakis, None; R.E. Engstrom, None; B.R. Straatsma, None; J.L. Demer, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Implant Characteristics

Patient	Gender	Age	Implant Type	Implant Diameter (mm)	Wrapping Material	Postoperative Interval (mo)
1	Male	53	Molteno M-sphere	20	None	28
2	Male	62	Hydroxyapatite	20	Sclera	108
3	Female	46	Hydroxyapatite	18	Vicryl	60
4	Male	37	Molteno M-sphere	20	None	30
5	Female	48	Hydroxyapatite	18	None	3

Figure 1.

 

Contiguous, 2-mm thickness, quasicoronal MR images of anophthalmic left orbit (patient 3), demonstrating cross-sectional areas of EOMs. Images in posterior to anterior sequence are arranged from left to right and top to bottom. Structures outlined digitally in central panel.

Figure 2.

 

Contiguous, 2-mm-thick, quasicoronal MR images of anophthalmic left orbit (patient 5) demonstrating digitally outlined area of IO muscle in multiple quasicoronal sections. Images in medial to lateral sequence are arranged from left to right and top to bottom.

Table 2.

 

Muscle Volume in Anophthalmic and Fellow Orbits

Muscle	Anophthalmic	Fellow	P
MR	627 ± 89	651 ± 110	0.42
LR	597 ± 101	635 ± 106	0.16
IR	516 ± 151	541 ± 142	0.08
SR-LP	542 ± 144	599 ± 175	0.31
SO	314 ± 61	305 ± 76	0.76
IO	218 ± 26	217 ± 28	0.80

Table 3.

 

Maximum Cross-sectional Area of Rectus Muscles in Anophthalmic and Fellow Orbits

Muscle	Gaze Direction	Anophthalmic	Fellow	P
MR	Central	35.9 ± 4.8	30.4 ± 4.1	0.02
	Adduction	40.2 ± 0.2	37.9 ± 2.1	0.33
LR	Central	33.3 ± 5.5	33.2 ± 6.2	0.96
	Abduction	38.6 ± 2.2	41.2 ± 0.2	0.32
IR	Central	29.4 ± 9.9	27.9 ± 9.1	0.37
	Infraduction	30.1 ± 2.1	32.6 ± 4.9	0.41
SR-LP	Central	33.5 ± 7.8	30.7 ± 9.0	0.44
	Supraduction	34.9 ± 3.4	32.2 ± 4.3	0.14

Figure 3.

 

Two-millimeter-thick, quasicoronal MR images of anophthalmic right orbit (patient 1), demonstrating increase in maximal quasicoronal cross-sectional area of MR muscle in adduction and central gaze. The clinical appearance in both gaze positions is presented below the MR images. Note poor adduction of the right eye despite contractile thickening of the MR and relaxational thinning of the lateral rectus.

Table 4.

 

Percent of Increase in Maximum Muscle Cross Section from Primary to Principle Rectus Secondary-Gaze Direction

Muscle	Anophthalmic	Fellow	P
MR	18.8 ± 6.2	23.8 ± 7.7	0.15
LR	18.9 ± 8.6	24.5 ± 8.3	0.72
IR	30.0 ± 10.6	30.7 ± 8.1	0.77
SR-LP	12.9 ± 10.9	18.2 ± 13.2	0.26

Table 5.

 

Length of Rectus Muscles in Anophthalmic and Fellow Orbits

Muscle	Anophthalmic	Fellow	P
MR	25.6 ± 1.7	31.2 ± 2.3	< 0.001
LR	28.0 ± 2.0	31.6 ± 1.7	< 0.001
IR	26.4 ± 2.6	33.2 ± 1.1	< 0.001
SR-LP	26.0 ± 1.4	30.4 ± 2.2	=0.01

Figure 4.

 

Two-millimeter-thick axial MR images in right (left column) and central (right column) gaze, in patient 4, who underwent enucleation of the left globe with implantation of an unwrapped implant (I). The top row of images is the plane immediately superior and contiguous with the bottom row. The prosthesis (P) appears as a dark signal. The MR muscle is shorter on the anophthalmic left than on the normal right side, but the insertion of the connective tissues on the orbital side of the MR corresponds to the location of the pulley bilaterally. Note the posterior movement of the normal right MR pulley from right to central gaze and the simultaneous anterior movement of the left MR pulley and associated connective tissue of the conjunctival fornix. The left ON terminates posterior to the prosthesis and is less readily distinguishable from the surrounding cerebrospinal fluid that the normal right ON.

Figure 5.

 

Two-millimeter-thick quasicoronal MRI scan (patient 2), showing measurement of distances from the EOM and implant centroids to the plane of the interhemispheric fissure (represented by the nearly vertical line) and the plane perpendicular to it at the level of the cribriform plate of ethmoid bone (represented by the nearly horizontal line). Horizontal and vertical distances are presented in parentheses. Inset: the level of MRI scan in axial section.

Table 6.

 

Distance from EOM Centroids to Vertical and Horizontal Reference Planes (Interhemispheric Fissure and Ethmoidal Plate of Ethmoidal bone, respectively)

Muscle	Reference Plane	Anophthalmic	Fellow	P
MR	Vertical	3.9 ± 1.2	4.2 ± 1.1	0.41
	Horizontal	15.6 ± 4.3	14.4 ± 2.4	0.28
LR	Vertical	7.6 ± 2.1	6.6 ± 2.2	0.33
	Horizontal	40.2 ± 4.0	39.2 ± 3.9	0.27
IR	Vertical	12.3 ± 3.3	13.8 ± 4.1	0.18
	Horizontal	22.9 ± 3.9	20.1 ± 4.9	0.22
SR-LP	Vertical	6.4 ± 1.2	5.9 ± 2.2	0.38
	Horizontal	29.1 ± 3.2	28.3 ± 3.2	0.17

Figure 6.

 

Two-millimeter-thick axial MRI scan (patient 5) showing measurement of distance from the globe and implant centroids to the anterior edge of the optic chiasm. Note the digitally outlined prosthesis in the anophthalmic left orbit.

Table 7.

 

Distance from Globe and Implant Centroids to Vertical, Horizontal, and Anteroposterior Reference Planes (Interhemispheric Fissure, Ethmoidal Plate of Ethmoidal Bone, and Anterior Edge of Optic Chiasm, Respectively)

Reference Plane	Anophthalmic	Fellow	P
Anteroposterior	56.2 ± 5.5	64.8 ± 7.7	0.01
Lateral	32.0 ± 0.22	32.6 ± 2.0	0.53
Vertical	7.6 ± 3.4	7.9 ± 1.9	0.71