Microanatomy of the Orbital Apex Computed Tomography and Microcryoplaning of Soft and Hard Tissue Robert Alan Goldberg, MD/ Kambiz Hannani, BS/ Arthur W. Toga, PhD2 Purpose: The anatomy of the orbital apex is characterized by a complex interplay between critical bony and neural structures. Traditional methods used to study this region include dissection, static sections, and computed tomography (CT). Tomographic techniques are very useful in understanding these complex relationShips, but the resolution of conventional CT and magnetic resonance imaging (MRI) is not sufficient to recognize the intricate details of the optic nerve canal and associated structures. The purpose of this study is to determine the value of microcryoplaning and computer reconstruction in visualizing the orbital apex in detail not previously possible, at any orientation in three-dimensional space . . Methods: Microcryotomy of the orbital apex area was performed on cadaver specimens, and images of each cryosection were digitized. Subsequently, the lesser wing of sphenoid bone and optic nerve were outlined to allow for spatial manipulation and three-dimensional visualization of the orbital apex. Results: The authors present reconstructed computer images of the orbital apex in coronal and axial planes with CT correlation. Clinically important anatomic points and landmarks as well as potential pitfalls are demonstrated. Conclusion: Microcryoplaning and computer reconstruction are useful techniques in viewing the detailed anatomy of the orbital apex. Although microcryoplaning has the limitation of poor ~oft tissue detail, the resolution of captured images is much greater than those obtained from CT or MRI scans; the improved resolution allows for accurate CT correlations. The technique has utility in education, surgical planning, and quantitative analysis of orbital apical anatomy. Ophthalmology 1992;99:1447-1452
The orbital apex is a critical anatomic region. Through this compact bony region pass the nerves that provide afferent and efferent function for the visual system. For
Originally received: November 25, 1991. Revision accepted: February 25 , 1992. 1 Jules Stein Eye Institute, UCLA School of Medicine, Los Angeles. 2 Laboratory of Neuro Imaging, UCLA Department of Neurology, Los Angeles. Supported in part by the Karl Kirschgessner Foundation, Ophthalmology Endowment Fund (Dr. Goldberg) and by the National Science Foundation (D1R #89-08174) (Dr. Toga) and the National Institutes of Health (ROI-RR05956) (Dr. Toga). Reprint requests to Robert Alan Goldberg, MD, 100 Stein Plaza, Los Angeles, CA 90024-7006.
the surgeon operating in this area, the carotid artery, as it winds its way through the cavernous sinus, represents a constant danger. I Detailed knowledge of the intricate anatomic relationships in the orbital apex is critical for accurate diagnosis and management of disorders affecting this region. 2 Previous publications have dealt with the apical microanatomy.3-6 None, however, have described the use of cryomicrotomy in combination with computer-aided reconstruction in their analysis. Microcryoplaning, which was developed by Rauschning, has been used for anatomical correlative comparisons with magnetic resonance imaging (MRI) and computed tomography (CT) scans in the past. 7- 9 This technique prevents deformation, loss of color, or fluid drainage and, therefore, allows for the visualization of the undisturbed anatomy of soft and hard tissue.
1447
Ophthalmology
Volume 99, Number 9, September 1992
In this report, we present an analysis of orbital apical microanatomy, which has been obtained by combining the techniques of microcryoplaning and computer modeling. This technology allows us to examine the threedimensional microanatomy of the orbital apex at any arbitrary orientation with a degree of accuracy not before attainable. Thus, we can generate precise reconstruction of images in any plane for correlation with CT and MR images.
Methods Microcryotomy was coronally performed on four halfhead, different male specimens. The heads were originally bisected by a band saw and were kept frozen at -70° C before cryosectioning. Using the band saw, the orbital area was removed and subsequently embedded within a cryostat mounting frame. The cryosectioning was performed at the speed of 4 cm/sec; the temperature was set to - 20° C, and the slice thickness varied between 20 and 50 J,Lm. Images of each cryosection were collected every 100 J,Lm using the Digistat high-resolution color, digitally controlled camera, and the data were stored on optical disks using the VMS operating system on V AX mainframes. The digital resolution was 10242 by 24 bits (8 for red, green, and blue, respectively). Subsequently, the lesser wing of the sphenoid bone and the optic nerve were outlined for one specimen to form contours. Each successive contour in the xy plane was assigned a z value and was reconstructed by triangulation to create a three-dimensional surface model of the optic nerve and canal. 10 Furthermore, the cryosections were stacked to reformulate a volume that could be resampled at any angle for multiplanar viewing of the cryosections.
Results The regions of particular interest to the ophthalmologist have been labeled on the sphenoid bone in Figure I. The optic canal, which lies completely within the lesser wing of the sphenoid bone, courses medially and superiorly with a 30° angle relative to the horizontal plane. The canal indents into the sphenoid sinus as a bony impression, which is often visible above the carotid impression on the lateral wall of the sphenoid sinus. The roof of the canal is separated from the anterior cranial fossa by a thin portion of the sphenoid bone, which thickens anteriorly. The canal is separated from the superior orbital fissure by the optic strut and the anterior clinoid process, which form the anterior and posterior portions of the lateral wall of the canal, respectively. Images obtained by computer reconstruction of the orbit are compared to a CT scan in Figure 2. The threedimensional resampled sphenoid bone (which appears in green) demonstrates the thickness of the roof of the canal compared with the thin medial wall separating the canal from the sphenoid and ethmoidal sinuses. In the area of the lateral optic canal strut, the canal is completely embedded within the sphenoid bone (Fig 3).
1448
Figure 1. Bony orbital anatomy. A, frontal view ot right orbital apex. The entrance to the optic nerve canal (black arrow) is the most superior and medial structure in the apex. The superior orbital fissure is lateral and inferior to the optic nerve canal (white arrowhead) and is separated from the canal by a bony strut (open arrow), which posteriorly blends with the anterior clinoid. The anterior and posterior ethmoidal arteries (letters "A" and "P") form a surgically useful imaginary line which points directly to the medial canal entrance. B, isolated anterior view of the sphenoid bone. The relationship between the optic nerve canal (artow) and the sphenoid sinus (curved arrow) can be appreciated. In the superolateral wall of the contralateral sphenoid sinus, the bony indentation of the optic nerve can be seen (arrowhead).
The lateral wall of the canal is formed by the anterior clinoid process, which is substantially more sturdy than the medial wall of the canal. Note that at this level the canal is coursing superiorly and may, therefore, be located superiorly as well as laterally to the sphenoid sinus. The coronal sections obtained by the cryoplaning were resampled to create axial images (Fig 4). The nerve (closed arrow) is reconstructed and is seen coursing between the lateral and medial bony neighbors, the anterior clinoid process and sphenoid sinus. Note that the diameter of the nerve increases proximally. Figure 5 represents coronal and axial CT scans of the orbitaLapex taken at different angles. In perfect coronal
Goldberg et al . Microanatomy of the Orbital Apex
Figure 2. Coronal cut through orbital apex. A, photograph of coronal cryosection. The optic nerve (arrow), muscle cone, superior orbital fissure (arrowhead), inferior orbital fissure (letter "I"), and sphenoid sinus (curved arrow) are seen. B, corresponding coronal CT scan. C, computer reconstruction of optic nerve (yellow) and lesser wing of the sphenoid (green) superimposed on the same coronal cryosection, demonstrating relationship of the optic nerve canal and the sphenoid sinus which share a thin bony wall. The thicker roof of the optic nerve canal, which borders the anterior cranial fossa, is also appreciated.
scans (Fig 5A), the pterygoid processes can be seen; furthermore, the turbinates and teeth are only present in more angled coronal scans (Fig 5B). The knowledge of the approximate level of the axial CT scan (Fig 5C) is crucial if correct structural correlations such as that of the superior orbital fissure versus the optic nerve canal are to be made.
Discussion The orbital apex is a complex region with many nerves and arteries that transverse the superior orbital fissure and the optic canal. This region is susceptible to various diseases and injuries. 2 The intracanalicular portion of the optic nerve may be damaged within the canal by sphenoid
fractures, edema, hemorrhage, stretching, or tension resulting in axonal lesions or shock waves from trauma. I ,3 Visual deficits also may arise from bone abnormalities such as callus formation, osteopetrosis, and fibrous dysplasia, and from tumors such as gliomas and meningioma, which transverse the optic canal. Surgical intervention may reverse the visual deficit caused by these lesions and injuries. I ,II-14 The diagnosis and surgical treatment of these lesions requires an intimate knowledge of the orbital apex microanatomy. We have created a series of computer-generated images to provide, for instructional purposes, direct comparisons to CT scans. Before comparing cryosections and CT scans, however, some factors may need to be considered. A CT scan represents an average of tissue densities within the width of the slice; many cryosections may be required to
1449
Ophthalmology
Volume 99, Number 9, September 1992 compared with scans and are, therefore, acceptable for CT anatomic references. Despite the enhanced detail of the soft and hard tissue structures seen in cryosections compared with CT scans, the technology of cryosectioning has some limitations. Although hard and soft tissue demarcation is clear (Figs 2 and 3), images captured in monochrome and digitized into binary format lack the soft tissue details seen in direct photography of sections. Furthermore, the texture of soft tissue may resemble that of nearby hard structures as seen when comparing the optic nerve with surrounding tissue in Figure 4A. Although color-captured images minimize the above problem, increased shading gradients need to be incorporated into software to improve monochrome images. Regardless of this limitation, cryoplaning is a useful tool in detailing hard and soft tissue relationships and correlative CT scan/MRI studies. Depending on the angle of the scan, CT scans may be misleading. Coronal sections are rarely perfect in the vertical plane; often these sections are rotated counterclockwise in the x-axis, which affects the size of nerves and muscles observed. In determining the size of the optic nerve or canal for diagnostic purposes, one must realize that oblique cuts result in different dimensions of the canal. Furthermore, the lateral rectus muscle will always be cut obliquely in a coronal scan. Therefore, before analyz-
Figure 3. Coronal cut through right orbital optic nerve canal entrance. A. the canal becomes a closed bony space as the lateral strut (L) bridges from the sphenoid body to the anterior clinoid process (letter "C"). separating the optic nerve (black arrow) from the superior orbital fissure and cavernous sinus (letter "S"). The intimate relationship of the optic nerve canal to the sphenoid sinus (curved arrow) can be appreciated; the thin bone that separates the canal from the sinus is removed in extracranial optic nerve canal decompressions. B, corresponding coronal CT scan. As is often the case. a lucent region representing pneumatization of the clinoid is noted in the clinoid process (letter "C"). which might be confused with the optic nerve canal (black arrow) by an unsophisticated observer.
include all structures averaged in one CT scan. 8 Cryosections can have a resolution of I Mm/pixel, whereas CT scan resolution does not exceed 0.5 mm/pixel. Furthermore, the CT resolution is greatly reduced after multiplanar reformatting 15; reformatted images obtained from volume-resampling of cryosections, however, retain their high resolution depending on the thickness of each slice. Hence, images rendered ' from microcryoplaning and computer remodeling have superior resolution when
1450
Figure 4. Axial cut through the optic nerve canal. A. resampled computer image. The medial relationship of the sphenoid sinus (curved arrow) and the lateral relationship of the clinoid process (letter "C") and lateral strut (open arrow) to the optic nerve canal (closed arrow) are well visualized. B, computed tomographic image at the same level. Note that the optic nerve canal (black arrow) and superior orbital fissure (white arrowhead) are both"well visualized.
Goldberg et al . Microanatomy of the Orbital Apex
Figure 5. Coronal and axial CT scans. A, B, coronal CT scans taken at different planes. Figure 5A represents a perfect coronal scan in which the pterygoid plates (letter "0") can be seen. Because of patient positioning limitations, coronal scans are typically in a more axial plane so that the turbinates (letter "T") as well as the mandible and teeth (letter "M") can be visualized (Fig 5B). C, axial CT scan. This axial cut is approximately 5 mm inferior to the optic canal. It is not uncommon for the superior orbital fissure (white arrowhead) to be confused with the optic nerve canal (see Fig 4B for comparison). The optic nerve canal should always be demarcated by the anterior clinoid and lateral bony strut.
ing a CT scan, the exact level and angle of cut must be understood. We have often observed the superior orbital fissure to be confused with the optic canal in axial scans. One landmark that can help in the canal's identification is the anterior clinoid process, which separates the superior orbital fissure from the canal. Since the canal is the most superior structure underneath the anterior cranial fossa, scans containing the optic canal should immediately follow those of the cranial fossa in sequential CT scans. Furthermore, the visualization of the three-dimensional structure of the sphenoid bone may be challenging to the beginner. Traditional coronal and axial images of the lesser wing of the sphenoid bone in the orbital apex fail to illustrate the orientation of the wing in space. Computed reconstructed images (Fig 2C), however, allow the observer to appreciate the angle and relationship of the sphenoid lesser wing as it thickens anteriorly and envelops the optic nerve. The optic canal also may be mistaken for pneumatized regions of the sphenoid bone. Pneumatization of the base of the pterygoid plates and the greater wing of the sphenoid bone occur regularly; extensions into the anterior clinoid process are less frequent. 6 These extensions, however, may be mistaken for the optic canal (Fig 3B). By recognizing
that the canal touches the sphenoid sinus and passes medial to the anterior clinoid process, the canal can be identified. Intimate knowledge of the orbital apical microanatomy is not only crucial for diagnostic purposes but also in the success of surgical procedures. Transnasal operations must avoid the carotid artery on the immediate lateral wall of the sinus. 6 • 11 The optic nerve is separated from the sphenoid sinus by only thin sinus mucosa without a bony covering in 4% of the population making it especially vulnerable during surgical procedures involving the sphenoid sinus. 5 Also note that the location of the canal may vary within the sinus; since the optic canal courses superiomedially through the sphenoid bone, the canal may be located near the roof of the sphenoid sinus (Fig 5B). The microanatomy of the region can further dictate the choice of surgical procedures. Because of the canal's location and to avoid the carotid artery, we perform our bony optic canal decompressions through the superiorlateral wall of the sphenoid sinus. II To ophthalmologists, study of the fascinating anatomic relationships in the critical and concentrated territory of the orbital apex should provide a satisfying challenge. The investment of the time studying this anatomy will pay rich dividends in terms of increased ability to make clinical
1451
Ophthalmology
Volume 99, Number 9, September 1992
and radiographic diagnosis of lesions in this region. For surgeons venturing into this region, of course, the knowledge of the detailed microanatomy of the orbital apex is of paramount importance in minimizing the potentially disastrous complications that result from misdirected surgery in this tight region where critical structures are separated by millimeters. Microcryoplaning and computer reconstruction allow us to examine this anatomy in three dimensions with precision not before attainable. In addition to its use as a teaching tool, it is an elegant technology with which to compare anatomic variations. Further work in our laboratory will quantify variations of normal anatomy, which adds an additional level of intricacy to the intellectual study and clinical management of the orbital apex.
References I. Osguthorpe JD, Sofferman RA. Optic nerve decompression. Otolaryngol C1in North Am 1988;21:155-69. 2. Niho S, Niho M, Niho K. Decompression of the optic canal by the transethmoidal route and decompression of the superior orbital fissure. Can J Ophthalmol 1970;5:22-40. 3. Ghobrial W, Amstutz S, Mathog RH. Fractures of the sphenoid bone. Head Neck Surg 1986;8:447-55. 4. Habal MB, Maniscalco JE, Rhoton AL Jr. Microsurgical anatomy of the optic canal: correlates to optic nerve ex- , posure. J Surg Res 1977;22:527-33.
1452
5. Fujii K, Chambers SM, Rhoton AL Jr. Neurovascular relationships of the sphenoid sinus. A microsurgical study. J Neurosurg 1979;50:31-9. 6. Bansberg SF, Harner SG, Forbes G. Relationship of the optic nerve to the paranasal sinuses as shown by computed tomography. Otolaryngol Head Neck Surg 1987;96:331-5. 7. Lufkin R, Rauschning W, Seeger L, et a1. Anatomic correlation of cadaver cryomicrotomy with magnetic resonance imaging. Surg Radiol Anat 1987;9:299-302. 8. Rauschning W, Bergstrom K, Pech P. Correlative craniospinal anatomy studies by computed tomography and cryomicrotomy. J Comput Assist Tomogr 1983;7:9-13. 9. Rauschning W. Computed tomography and cryomicrotomy of lumbar spinal specimens. A new technique for multiplanar anatomic correlation. Spine 1983;8: 170-80. 10. Toga AW. Three-dimensional Neuroimaging. New York: Raven Press, 1990;189-210. 11. Sofferman RA. Sphenoethmoid approach to the optic nerve. Laryngoscope 1981 ;91: 184-96. 12. Blinkov SM, Gabibov GA, Tcherekayev VA. Transcranial surgical approaches to the orbital part of the optic nerve: an anatomical study. J Neurosurg 1986;65:44-7. 13. Habal MB. Clinical observations on the isolated optic nerve injury. Ann Plast Surg 1978;1:603-7. 14. Takahashi M, Itoh M, Kaneko M, et al. Microscopic intranasal decompression of the optic nerve. Arch OtorhinolaryngoI1989;246:113-6. 15. Zonneveld FW, Koornneef L, Hillen B, de Slegte RG. Normal direct multiplanar CT anatomy of the orbit with correlative anatomic cryosections. Radiol C1in North Am 1987;25:381-407.
The orbital apex is a critical anatomic region. Through this compact bony region pass the nerves that provide afferent and efferent function for the visual system. For
Originally received: November 25, 1991. Revision accepted: February 25 , 1992. 1 Jules Stein Eye Institute, UCLA School of Medicine, Los Angeles. 2 Laboratory of Neuro Imaging, UCLA Department of Neurology, Los Angeles. Supported in part by the Karl Kirschgessner Foundation, Ophthalmology Endowment Fund (Dr. Goldberg) and by the National Science Foundation (D1R #89-08174) (Dr. Toga) and the National Institutes of Health (ROI-RR05956) (Dr. Toga). Reprint requests to Robert Alan Goldberg, MD, 100 Stein Plaza, Los Angeles, CA 90024-7006.
the surgeon operating in this area, the carotid artery, as it winds its way through the cavernous sinus, represents a constant danger. I Detailed knowledge of the intricate anatomic relationships in the orbital apex is critical for accurate diagnosis and management of disorders affecting this region. 2 Previous publications have dealt with the apical microanatomy.3-6 None, however, have described the use of cryomicrotomy in combination with computer-aided reconstruction in their analysis. Microcryoplaning, which was developed by Rauschning, has been used for anatomical correlative comparisons with magnetic resonance imaging (MRI) and computed tomography (CT) scans in the past. 7- 9 This technique prevents deformation, loss of color, or fluid drainage and, therefore, allows for the visualization of the undisturbed anatomy of soft and hard tissue.
1447
Ophthalmology
Volume 99, Number 9, September 1992
In this report, we present an analysis of orbital apical microanatomy, which has been obtained by combining the techniques of microcryoplaning and computer modeling. This technology allows us to examine the threedimensional microanatomy of the orbital apex at any arbitrary orientation with a degree of accuracy not before attainable. Thus, we can generate precise reconstruction of images in any plane for correlation with CT and MR images.
Methods Microcryotomy was coronally performed on four halfhead, different male specimens. The heads were originally bisected by a band saw and were kept frozen at -70° C before cryosectioning. Using the band saw, the orbital area was removed and subsequently embedded within a cryostat mounting frame. The cryosectioning was performed at the speed of 4 cm/sec; the temperature was set to - 20° C, and the slice thickness varied between 20 and 50 J,Lm. Images of each cryosection were collected every 100 J,Lm using the Digistat high-resolution color, digitally controlled camera, and the data were stored on optical disks using the VMS operating system on V AX mainframes. The digital resolution was 10242 by 24 bits (8 for red, green, and blue, respectively). Subsequently, the lesser wing of the sphenoid bone and the optic nerve were outlined for one specimen to form contours. Each successive contour in the xy plane was assigned a z value and was reconstructed by triangulation to create a three-dimensional surface model of the optic nerve and canal. 10 Furthermore, the cryosections were stacked to reformulate a volume that could be resampled at any angle for multiplanar viewing of the cryosections.
Results The regions of particular interest to the ophthalmologist have been labeled on the sphenoid bone in Figure I. The optic canal, which lies completely within the lesser wing of the sphenoid bone, courses medially and superiorly with a 30° angle relative to the horizontal plane. The canal indents into the sphenoid sinus as a bony impression, which is often visible above the carotid impression on the lateral wall of the sphenoid sinus. The roof of the canal is separated from the anterior cranial fossa by a thin portion of the sphenoid bone, which thickens anteriorly. The canal is separated from the superior orbital fissure by the optic strut and the anterior clinoid process, which form the anterior and posterior portions of the lateral wall of the canal, respectively. Images obtained by computer reconstruction of the orbit are compared to a CT scan in Figure 2. The threedimensional resampled sphenoid bone (which appears in green) demonstrates the thickness of the roof of the canal compared with the thin medial wall separating the canal from the sphenoid and ethmoidal sinuses. In the area of the lateral optic canal strut, the canal is completely embedded within the sphenoid bone (Fig 3).
1448
Figure 1. Bony orbital anatomy. A, frontal view ot right orbital apex. The entrance to the optic nerve canal (black arrow) is the most superior and medial structure in the apex. The superior orbital fissure is lateral and inferior to the optic nerve canal (white arrowhead) and is separated from the canal by a bony strut (open arrow), which posteriorly blends with the anterior clinoid. The anterior and posterior ethmoidal arteries (letters "A" and "P") form a surgically useful imaginary line which points directly to the medial canal entrance. B, isolated anterior view of the sphenoid bone. The relationship between the optic nerve canal (artow) and the sphenoid sinus (curved arrow) can be appreciated. In the superolateral wall of the contralateral sphenoid sinus, the bony indentation of the optic nerve can be seen (arrowhead).
The lateral wall of the canal is formed by the anterior clinoid process, which is substantially more sturdy than the medial wall of the canal. Note that at this level the canal is coursing superiorly and may, therefore, be located superiorly as well as laterally to the sphenoid sinus. The coronal sections obtained by the cryoplaning were resampled to create axial images (Fig 4). The nerve (closed arrow) is reconstructed and is seen coursing between the lateral and medial bony neighbors, the anterior clinoid process and sphenoid sinus. Note that the diameter of the nerve increases proximally. Figure 5 represents coronal and axial CT scans of the orbitaLapex taken at different angles. In perfect coronal
Goldberg et al . Microanatomy of the Orbital Apex
Figure 2. Coronal cut through orbital apex. A, photograph of coronal cryosection. The optic nerve (arrow), muscle cone, superior orbital fissure (arrowhead), inferior orbital fissure (letter "I"), and sphenoid sinus (curved arrow) are seen. B, corresponding coronal CT scan. C, computer reconstruction of optic nerve (yellow) and lesser wing of the sphenoid (green) superimposed on the same coronal cryosection, demonstrating relationship of the optic nerve canal and the sphenoid sinus which share a thin bony wall. The thicker roof of the optic nerve canal, which borders the anterior cranial fossa, is also appreciated.
scans (Fig 5A), the pterygoid processes can be seen; furthermore, the turbinates and teeth are only present in more angled coronal scans (Fig 5B). The knowledge of the approximate level of the axial CT scan (Fig 5C) is crucial if correct structural correlations such as that of the superior orbital fissure versus the optic nerve canal are to be made.
Discussion The orbital apex is a complex region with many nerves and arteries that transverse the superior orbital fissure and the optic canal. This region is susceptible to various diseases and injuries. 2 The intracanalicular portion of the optic nerve may be damaged within the canal by sphenoid
fractures, edema, hemorrhage, stretching, or tension resulting in axonal lesions or shock waves from trauma. I ,3 Visual deficits also may arise from bone abnormalities such as callus formation, osteopetrosis, and fibrous dysplasia, and from tumors such as gliomas and meningioma, which transverse the optic canal. Surgical intervention may reverse the visual deficit caused by these lesions and injuries. I ,II-14 The diagnosis and surgical treatment of these lesions requires an intimate knowledge of the orbital apex microanatomy. We have created a series of computer-generated images to provide, for instructional purposes, direct comparisons to CT scans. Before comparing cryosections and CT scans, however, some factors may need to be considered. A CT scan represents an average of tissue densities within the width of the slice; many cryosections may be required to
1449
Ophthalmology
Volume 99, Number 9, September 1992 compared with scans and are, therefore, acceptable for CT anatomic references. Despite the enhanced detail of the soft and hard tissue structures seen in cryosections compared with CT scans, the technology of cryosectioning has some limitations. Although hard and soft tissue demarcation is clear (Figs 2 and 3), images captured in monochrome and digitized into binary format lack the soft tissue details seen in direct photography of sections. Furthermore, the texture of soft tissue may resemble that of nearby hard structures as seen when comparing the optic nerve with surrounding tissue in Figure 4A. Although color-captured images minimize the above problem, increased shading gradients need to be incorporated into software to improve monochrome images. Regardless of this limitation, cryoplaning is a useful tool in detailing hard and soft tissue relationships and correlative CT scan/MRI studies. Depending on the angle of the scan, CT scans may be misleading. Coronal sections are rarely perfect in the vertical plane; often these sections are rotated counterclockwise in the x-axis, which affects the size of nerves and muscles observed. In determining the size of the optic nerve or canal for diagnostic purposes, one must realize that oblique cuts result in different dimensions of the canal. Furthermore, the lateral rectus muscle will always be cut obliquely in a coronal scan. Therefore, before analyz-
Figure 3. Coronal cut through right orbital optic nerve canal entrance. A. the canal becomes a closed bony space as the lateral strut (L) bridges from the sphenoid body to the anterior clinoid process (letter "C"). separating the optic nerve (black arrow) from the superior orbital fissure and cavernous sinus (letter "S"). The intimate relationship of the optic nerve canal to the sphenoid sinus (curved arrow) can be appreciated; the thin bone that separates the canal from the sinus is removed in extracranial optic nerve canal decompressions. B, corresponding coronal CT scan. As is often the case. a lucent region representing pneumatization of the clinoid is noted in the clinoid process (letter "C"). which might be confused with the optic nerve canal (black arrow) by an unsophisticated observer.
include all structures averaged in one CT scan. 8 Cryosections can have a resolution of I Mm/pixel, whereas CT scan resolution does not exceed 0.5 mm/pixel. Furthermore, the CT resolution is greatly reduced after multiplanar reformatting 15; reformatted images obtained from volume-resampling of cryosections, however, retain their high resolution depending on the thickness of each slice. Hence, images rendered ' from microcryoplaning and computer remodeling have superior resolution when
1450
Figure 4. Axial cut through the optic nerve canal. A. resampled computer image. The medial relationship of the sphenoid sinus (curved arrow) and the lateral relationship of the clinoid process (letter "C") and lateral strut (open arrow) to the optic nerve canal (closed arrow) are well visualized. B, computed tomographic image at the same level. Note that the optic nerve canal (black arrow) and superior orbital fissure (white arrowhead) are both"well visualized.
Goldberg et al . Microanatomy of the Orbital Apex
Figure 5. Coronal and axial CT scans. A, B, coronal CT scans taken at different planes. Figure 5A represents a perfect coronal scan in which the pterygoid plates (letter "0") can be seen. Because of patient positioning limitations, coronal scans are typically in a more axial plane so that the turbinates (letter "T") as well as the mandible and teeth (letter "M") can be visualized (Fig 5B). C, axial CT scan. This axial cut is approximately 5 mm inferior to the optic canal. It is not uncommon for the superior orbital fissure (white arrowhead) to be confused with the optic nerve canal (see Fig 4B for comparison). The optic nerve canal should always be demarcated by the anterior clinoid and lateral bony strut.
ing a CT scan, the exact level and angle of cut must be understood. We have often observed the superior orbital fissure to be confused with the optic canal in axial scans. One landmark that can help in the canal's identification is the anterior clinoid process, which separates the superior orbital fissure from the canal. Since the canal is the most superior structure underneath the anterior cranial fossa, scans containing the optic canal should immediately follow those of the cranial fossa in sequential CT scans. Furthermore, the visualization of the three-dimensional structure of the sphenoid bone may be challenging to the beginner. Traditional coronal and axial images of the lesser wing of the sphenoid bone in the orbital apex fail to illustrate the orientation of the wing in space. Computed reconstructed images (Fig 2C), however, allow the observer to appreciate the angle and relationship of the sphenoid lesser wing as it thickens anteriorly and envelops the optic nerve. The optic canal also may be mistaken for pneumatized regions of the sphenoid bone. Pneumatization of the base of the pterygoid plates and the greater wing of the sphenoid bone occur regularly; extensions into the anterior clinoid process are less frequent. 6 These extensions, however, may be mistaken for the optic canal (Fig 3B). By recognizing
that the canal touches the sphenoid sinus and passes medial to the anterior clinoid process, the canal can be identified. Intimate knowledge of the orbital apical microanatomy is not only crucial for diagnostic purposes but also in the success of surgical procedures. Transnasal operations must avoid the carotid artery on the immediate lateral wall of the sinus. 6 • 11 The optic nerve is separated from the sphenoid sinus by only thin sinus mucosa without a bony covering in 4% of the population making it especially vulnerable during surgical procedures involving the sphenoid sinus. 5 Also note that the location of the canal may vary within the sinus; since the optic canal courses superiomedially through the sphenoid bone, the canal may be located near the roof of the sphenoid sinus (Fig 5B). The microanatomy of the region can further dictate the choice of surgical procedures. Because of the canal's location and to avoid the carotid artery, we perform our bony optic canal decompressions through the superiorlateral wall of the sphenoid sinus. II To ophthalmologists, study of the fascinating anatomic relationships in the critical and concentrated territory of the orbital apex should provide a satisfying challenge. The investment of the time studying this anatomy will pay rich dividends in terms of increased ability to make clinical
1451
Ophthalmology
Volume 99, Number 9, September 1992
and radiographic diagnosis of lesions in this region. For surgeons venturing into this region, of course, the knowledge of the detailed microanatomy of the orbital apex is of paramount importance in minimizing the potentially disastrous complications that result from misdirected surgery in this tight region where critical structures are separated by millimeters. Microcryoplaning and computer reconstruction allow us to examine this anatomy in three dimensions with precision not before attainable. In addition to its use as a teaching tool, it is an elegant technology with which to compare anatomic variations. Further work in our laboratory will quantify variations of normal anatomy, which adds an additional level of intricacy to the intellectual study and clinical management of the orbital apex.
References I. Osguthorpe JD, Sofferman RA. Optic nerve decompression. Otolaryngol C1in North Am 1988;21:155-69. 2. Niho S, Niho M, Niho K. Decompression of the optic canal by the transethmoidal route and decompression of the superior orbital fissure. Can J Ophthalmol 1970;5:22-40. 3. Ghobrial W, Amstutz S, Mathog RH. Fractures of the sphenoid bone. Head Neck Surg 1986;8:447-55. 4. Habal MB, Maniscalco JE, Rhoton AL Jr. Microsurgical anatomy of the optic canal: correlates to optic nerve ex- , posure. J Surg Res 1977;22:527-33.
1452
5. Fujii K, Chambers SM, Rhoton AL Jr. Neurovascular relationships of the sphenoid sinus. A microsurgical study. J Neurosurg 1979;50:31-9. 6. Bansberg SF, Harner SG, Forbes G. Relationship of the optic nerve to the paranasal sinuses as shown by computed tomography. Otolaryngol Head Neck Surg 1987;96:331-5. 7. Lufkin R, Rauschning W, Seeger L, et a1. Anatomic correlation of cadaver cryomicrotomy with magnetic resonance imaging. Surg Radiol Anat 1987;9:299-302. 8. Rauschning W, Bergstrom K, Pech P. Correlative craniospinal anatomy studies by computed tomography and cryomicrotomy. J Comput Assist Tomogr 1983;7:9-13. 9. Rauschning W. Computed tomography and cryomicrotomy of lumbar spinal specimens. A new technique for multiplanar anatomic correlation. Spine 1983;8: 170-80. 10. Toga AW. Three-dimensional Neuroimaging. New York: Raven Press, 1990;189-210. 11. Sofferman RA. Sphenoethmoid approach to the optic nerve. Laryngoscope 1981 ;91: 184-96. 12. Blinkov SM, Gabibov GA, Tcherekayev VA. Transcranial surgical approaches to the orbital part of the optic nerve: an anatomical study. J Neurosurg 1986;65:44-7. 13. Habal MB. Clinical observations on the isolated optic nerve injury. Ann Plast Surg 1978;1:603-7. 14. Takahashi M, Itoh M, Kaneko M, et al. Microscopic intranasal decompression of the optic nerve. Arch OtorhinolaryngoI1989;246:113-6. 15. Zonneveld FW, Koornneef L, Hillen B, de Slegte RG. Normal direct multiplanar CT anatomy of the orbit with correlative anatomic cryosections. Radiol C1in North Am 1987;25:381-407.
