Enterogenous Cyst of the Orbital Apex and Superior Orbital Fissure

Enterogenous Cyst of the Orbital Apex and Superior Orbital Fissure

Enterogenous Cyst of the Orbital Apex and Superior Orbital Fissure David B. Leventer, BA, 1 John C. Merriam, MD, 1•2 Richard Defendini, MD, 1•3 Myles ...

4MB Sizes 0 Downloads 75 Views

Enterogenous Cyst of the Orbital Apex and Superior Orbital Fissure David B. Leventer, BA, 1 John C. Merriam, MD, 1•2 Richard Defendini, MD, 1•3 Myles M. Behrens, MD, 1,2 Edgar M. Housepian, MD/· 4 Steve LeQuerica, MD,S Andrew Blitzer, DDS, MD1·6 Background: Enterogenous cysts of the central nervous system are rare congenital tumors with a single layer of mucin-secreting epithelial cells resembling gastrointestinal epithelium. The tumor is located most commonly at lower cervical and cervicothoracic spinal levels; only 22 intracranial cases have been reported. To the authors' knowledge, this entity has not been described in the orbit. Methods: A 23-year-old woman with painful loss of vision and ophthalmoplegia in the left eye was treated with oral and intravenous corticosteroids for presumed orbital inflammation. After a cystic lesion in the left orbital apex was demonstrated on computed tomographic scan and magnetic resonance imaging, various diagnoses, including optic nerve tumor, granulomatous inflammation, lymphoma, vascular anomaly, and pseudotumor, were considered until transcranial biopsy established the correct diagnosis. The tumor subsequently recurred twice. Results: More than 3 years after the last recurrence, the patient has no pain but has unilateral optic atrophy, significant visual field loss, limited motility, and an anesthetic cornea in the left eye. Conclusion: The diagnosis of enterogenous cyst is difficult without adequate biopsy because the radiologic and clinical presentation of this rare tumor may be confused with other lesions. Previous attempts to explain intracranially placed enterogenous cysts offer no explanation for an orbital occurrence nor do they adequately describe a mechanism for an intracranial location in general. An embryologically based explanation that takes into account the occurrence of this entity from the caudal to rostral extent of the neuraxis is described. This theory suggests that the orbit is the most rostral possible location for an enterogenous cyst. Ophthalmology 1994;101:1614-1621

Originally received: October 18, 1993. Revision accepted: March 3, 1994. 1 College of Physicians and Surgeons, Columbia University, New York. 2 Edward S. Harkness Eye Institute, Columbia-Presbyterian Medical Center, New York. 3 Division of Neuropathology, Columbia-Presbyterian Medical Center, New York. 4 Department of Neurological Surgery, Neurological Institute, ColumbiaPresbyterian Medical Center, New York. 5 Neurology Group of North Jersey, 50 Mt. Prospect Avenue, Clifton. 6 Department of Otolaryngology, Head and Neck Surgery, ColumbiaPresbyterian Medical Center, New York. Reprint requests to John C. Merriam, MD, Edward S. Harkness Eye Institute, 635 West I 65th St, New York, NY 10032.

1614

Kubie and Fulton 1 first described two cases of central nervous system cysts lined by cells resembling alimentary canal cells in 1928. They called these tumors "teratomatous cysts." Today, the term teratoma is restricted to complex congenital tumors containing all three germ layers. 2 The terminology used by Kubie and Fulton and later by Hoefnagel et al 3 in 1962 is considered archaic. Since then, other terms, including neurenteric cyst, 4 •5 intestinome, 6 gastrocystoma, 7 enteric cyst, 8 archenteric cyst, 9 and enterogenous cyst, 10 have been used to describe these rare cysts of the central nervous system. Neurenteric cyst and enterogenous cyst have become the predominant terms used in the literature.

Leventer et al · Enterogenous Cyst Enterogenous cysts are most common in the lower cervical and upper thoracic spinal cord. Rare intracranial cysts have been reported. Similar to most other congenital epithelial cysts of the central nervous system, they grow by secretion of cell products rather than by cell proliferation. Slow growth accounts for the long clinical latency and the marked neural distortion they may produce before symptoms develop. When they attain a critical mass, however, symptoms can evolve rapidly. To our knowledge, this benign, congenital tumor has not been described in the orbit. This article describes the presentation and management of an enterogenous cyst of the orbital apex and reconciles the location of this lesion with the accepted theory of the embryogenesis of enterogenous cysts in the central nervous system.

Case Report A healthy 23-year-old woman from Colombia, South America, had sudden, severe retrobulbar pain and rapid loss of vision in the left eye in February 1986. A neurologist diagnosed optic neuritis and prescribed oral prednisone. On March 5, 1986, her visual acuity was 20/20 in the right eye and counting fingers in

Figure 1. Top, axial computed tomographic scan taken on March 28, 1986, demonstrates a globular nonenhancing mass, which extends through the superior orbital fissure with erosion of the lesser wing of the sphenoid. Bottom, coronal computed tomographic scan shows that the mass fills the apex of the orbit.

Figure 2. Magnetic resonance imaging taken on November 28, 1988, shows a bilobed density, which is hyperintense peripherally and lucent centrally. The mass extends through the superior orbital fissure.

the left, with a relative afferent pupillary defect and mild temporal disc pallor in the left eye. Corneal sensation was intact in both eyes. Computed tomographic (CT) scan and magnetic resonance imaging (MRI) demonstrated a mass in the apex of the left orbit (Fig l). The differential diagnosis included optic nerve tumor, schwannoma, lymphoma, sarcoidosis, other granulomatous disease, or pseudotumor. Results of routine laboratory evaluations for sarcoidosis and vasculitis were negative. After the patient was hospitalized for intravenous corticosteroid therapy, the pain subsided. A needle biopsy was not diagnostic. Oral dexamethasone was tapered and discontinued. On May I, 1986, visual acuity in the left eye was 20/20 with normal color perception, but a relative afferent defect persisted. Over the next few weeks, pain recurred with movement of the left eye, and late in July visual acuity in the left eye decreased to 20/50 with no color perception. Prednisone (100 mg daily) was started. On August 4, the orbit was explored via a Lynch incision. No mass was found, suggesting the possibility of pseudotumor. Postoperatively, the left pupil remained dilated and minimally reactive, and the patient had weakness of adduction and upgaze, but visual acuity improved. Oral prednisone was discontinued by mid September. Over the ensuing 2 years, the patient required intermittent oral prednisone for retrobulbar pain, but was generally well and able to work. However, in late 1988, she experienced severe pain, photophobia, nausea, decreased vision, and limitation of movement in the left eye, which progressed from an abduction deficit to complete ophthalmoplegia. An arteriogram and lumbar puncture were normal. On MRI, a bilobed mass was seen in the region of the superior orbital fissure (Fig 2); the radiologist suggested a vascular anomaly. The patient was treated initially with intravenous corticosteroids and subsequently with 60 to 80 mg prednisone orally daily. The patient became Cushingoid, and orbital radiotherapy was considered. By April 1989, prednisone had been reduced to 40 mg daily. Only mild weakness of abduction remained, and corrected visual acuity was 20/20. Repeat CT scan showed a mass between the

1615

Ophthalmology

Volume 101, Number 9, September 1994

superior and inferior orbital fissures, and the possibility of a schwannoma was suggested. On June 20, 1989, the orbit was unroofed via a left frontal craniotomy at the New York Neurological Institute, and a cystic mass was exposed in the superior orbital fissure. Clear, thick fluid was aspirated from the cyst. The cyst was opened, and a biopsy was performed. The cyst could not be removed in toto. Postoperatively, the pain and ptosis resolved rapidly, but residual deficits of adduction, abduction, and elevation persisted. Microscopically, the cyst was lined by a single layer of columnar epithelium , resembling gastric mucosa (Fig 3, left). Uniform nuclei without pleomorphism or mitoses were placed basally. The epithelial lining rested on dense connective tissue that in deeper layers became more cellular, loose, and well vascularized. The epithelium was periodic acid-Schiff-positive and mucicarmine-negative, a staining pattern typical of the nonacidic mucin of gastric and gallbladder epithelium (Fig 3, right). Ultrastructurally, the cells contained abundant, closely packed mucin granules, and the luminal surface was covered with very short microvilli but no cilia (Fig 4). Subsequently, punctate staining and a filamentary keratitis of the inferior cornea developed in the left eye, which were thought to be due to incomplete closure of the left upper lid, a poor Bell phenomenon, and corneal anesthesia. The patient also had various side effects from prolonged use of corticosteroids. Oral prednisone was discontinued on September 25, 1989. In late October 1989, the patient had recurrent severe pain, ophthalmoplegia, nausea, and vomiting. A recurrent cyst was detected by MRI (Fig 5). Pain was relieved initially with oral prednisone (80 mg), but the patient subsequently required hospitalization for intravenous methylprednisolone (250 mg every 4 hours) and analgesics. On admission to the Harkness Eye Institute on November 20, 1989, the corrected visual acuity in the left eye was 20/50+. The left pupil was dilated and nonreactive with a relative afferent defect, and the left eye was nearly immobile (Fig 6). Tangent and Goldmann visual fields were depressed in the left eye (Fig 7); the right visual field was normal. Sensation in the ophthalmic division of the trigeminal nerve was decreased; the left cornea was anesthetic. By CT scan, the cyst between the medial rectus muscle and optic nerve appeared smaller after I month of intensive corticosteroid therapy.

On November 22, 1989, the original Lynch incision was used to approach the apex of the left orbit via the ethmoid and sphenoid sinuses. The periorbita was adherent to the medial rectus muscle, and the posterior portion of the medial rectus appeared blue and ischemic. A small amount of green-brown fluid was encountered between the medial rectus muscle and the optic nerve. The optic nerve and the nerves innervating the superior rectus, medial rectus, superior oblique, and inferior rectus muscles were intact. Intraoperatively, the pupil returned to normal size. No drain was left in the wound. The patient had immediate relief of pain and was discharged on the third postoperative day with prednisone (80 mg daily), cephalexin (500 mg orally 4 times daily), and oral analgesics as needed. Pathologic study of the capsule showed dense fibrous tissue with necrotic muscle; no glandular epithelium was identified (Fig 8). By December 21, 1989, oral prednisone had been reduced to 5 mg, alternating with 2.5 mg daily. Visual acuity in the left eye was 20/25-2. Movement of the left eye had improved but was limited in upgaze and adduction. Levator function was 13 mm, with minimal ptosis. Despite taping the lids at night, superficial infiltrates developed in the inferior cornea in the left eye, which cleared with topical antibiotics. On January 10, 1990, the patient complained of nausea, photophobia, and pain along the supraorbital rim. She was readmitted the following day after receiving intravenous corticosteroids and intramuscular medication for pain and nausea in the emergency room. On CT scan, recurrence of the tumor was confirmed (Fig 9). The patient's condition improved rapidly while taking prednisone (80-120 mg daily). On January 18, 1990, with hypotensive anesthesia, the orbit was explored using the same medial incision to enter the ethmoid and sphenoid sinuses and gain access to the orbital apex. A fluidfilled cyst was excised. On postoperative CT scan, no residual tumor was seen. The patient was discharged on the fourth postoperative day and was prescribed prednisone (80 mg daily). Pathology of the specimen was similar to the previous biopsy (Fig 8); no glandular epithelium was found in the fibrous cyst wall. Movement of the globe improved slowly. The corneal epithelium remained irregular, and visual acuity fluctuated from 20/50 to 20/200. On CT scans in April 1990 and January 1991,

Figure 3. Left, the cyst is lined by a single layer of tall columnar epithelium resting on fibrous connective tissue. The nuclei are basal, the cytoplasm is pale and lacy, and no microvilli are evident on the apical surfaces (hematoxylin-eosin; original magnification, X160). Right, the epithelial cells are intensely periodic acid-Schiff-reactive after predigestion of the section with diastase (periodic acid-Schiff; original magnification, X 160).

1616

Leventer et al · Enterogenous Cyst fossa, 16 - 19 medulla oblongata and pons, 13 ·20- 22 prepontine area (unpublished data; Small JM, 1962), 19·23 ·24 cerebellopontine angle, 19·25 cerebellum,26 and the fourth ventricle.16·27 There is only one report of multiple enterogenous cysts occurring in the posterior fossa and supratentorial region. 16 To the best of our knowledge, this is the first report of an intracranial enterogenous cyst in the superior orbital fissure and orbital apex. This intracranial enterogenous cyst and six others that were reported previously deviated from the midline. 13 · 19·20·25·28 Five of these cysts, as well as the one in this current study, occurred on the left: three in the left cerebellopontine angle, one in the left prepontine cistern, and one adherent to the third cranial nerve. The sixth cyst was reported to have extended to the right pontomedullary junction. 13 Wilkins and Odom 29 grouped intraspinal enterogenous cysts according to histology. The simplest cyst, designated

Figure 4. The epithelial cells are filled with mucin granules and have short microvilli on their apical surfaces (original magnification, X3150).

the ethmoid and sphenoid sinuses had repneumatized; no mass was detected. A Goldmann visual field in July 1990 (Fig 10) was improved with that of November 21, 1989. By the fall of 1990, prednisone had been reduced to 2.5 to 5.0 mg daily. The patient was not weaned completely from prednisone until June 1991 . By the fall of 1990, abduction of the left eye was normal and adduction was nearly normal . Downgaze was limited, and the patient could not elevate the left eye, which was 2 mm enophthalmic. Persistent, variable staining and occasional filaments of the left cornea were managed with topical lubricating drops and ointment, taping the lids at night or applying a moisture shield, and pressure patching. Schirmer values were variable but generally low, and the patient said that when she cried, the left eye did not tear. Despite closing the inferior punctum with cautery and placing a 0.8-g gold weight in the left upper lid, variable inferior corneal staining was a persistent problem. She returned to school to finish training as a cosmetologist. In June 1993, the patient reported that the left eye teared when she cried. The left cornea remained anesthetic but the punctate keratitis resolved. Corrected visual acuity was 20/50 in the left eye with a small central field (Fig II). The left eye had nearly normal adduction but was limited in downgaze and did not elevate past the midline (Fig 12). The left optic nerve was atrophic.

Discussion Enterogenous cysts are rare congenital tumors that occur most frequently in the lower cervical and cervicothoracic spinal canal; few enterogenous cysts have been reported in the high cervical canal. 11 - 15 Injury from such lesions is due to spinal cord and nerve root compression. Intracranial cysts are rarer and have been repotted in the posterior

Figure 5. Top, axial magnetic resonance imaging taken on October 26, 1989, shows tumor recurrence at the apex of the left orbit. The tumor lies berween the medial rectus and optic nerve and above the optic nerve. A dilated vein (arrow) lies on top of the cyst. Bottom, sagittal magnetic resonance imaging shows the cyst extending from below to above the optic nerve.

1617

Ophthalmology

Volume 101, Number 9, September 1994

Top left, Figure 6. Complete ophthalmoplegia performed on ovember 20, 1989. The left eye is down and in and does not move in any gaze position. Top right, Figure 7. Goldmann visual field in the left eye on 21, 1989.

ovember

Bottom, Figure 8. Bottom left, capsule of cyst excised ovember 22, 1989, is made of dense fibrous tissue (periodic acid-Schiff; original mag· nification, X 160). Bottom right, some sections of the capsule also have necrotic striated muscle (periodic acid·Schiff; original magnification, X 160).

group A, is lined by a single layer of columnar or cuboidal epithelial cells with or without cilia. Group B may include other components of the gastrointestinal tract or the tracheobronchial tree such as smooth muscle or mucous glands. Group C contains group B features with the addition of ependymal or glial tissue. Most intracranial enterogenous cysts, including the cyst in the current study, fit into group A. Only three intracranial cysts reported in the literature are included in group B,20 •24- 25 and no intracranial lesion fits group C. The age of onset for intracranial and craniospinal enterogenous cysts in males ranges from 7 to 48 years of age, with a prevalence in the third and fourth decades.

1618

The age of onset in females ranges from birth to 77 years of age, with a prevalence in the fourth and fifth decades (Table 1). Intraspinal enterogenous cysts have a similar age and sex distribution. 11 Pain is a frequent symptom in intraspinal cysts, 22 but pain was the predominant symptom in only one intracranial cyst,12 and only three patients had rapid onset of symptoms (Table 1). 14• 18 •23 Our patient's clinical presentation, however, was dominated by the sudden onset of deep retrobulbar pain and rapid loss of vision and motility. The diagnosis of enterogenous cyst is difficult without an adequate biopsy because the neuroradiologic and clinical features may be confused with other cystic lesions

Leventer et al · Enterogenous Cyst

Figure 9. Computed tomographic scan taken on January 11, 1990, shows recurrent tumor at the apex of the left orbit. The medial orbital wall is recessed from previous surgery. Residual edema and inflammation also are seen.

such as dermoid, epidermoid, or arachnoid cysts, or cystic schwannoma. 26 In our patient, the initial diagnosis of optic neuritis was based solely on her painful loss of vision and improvement with corticosteroids. After CT scan and MRI showed a mass in the orbital apex, the differential diagnosis included glioma, neuroma, schwannoma, lymphoma, vascular anomaly, and pseudotumor. However, the diagnosis was questionable until the craniotomy. Serial CT and MRI showed a reduction in tumor size after highdose corticosteroids, suggesting that pain and ophthalmoplegia were due to an enlarging mass rather than an irritating secretion.

Embryology and Pathogenesis The pathogenesis of enterogenous cysts is not well understood, but it is thought to be an early developmental

Figure 10. Goldmann visual field in the left eye examined on July 24, 1990.

Figure 11. Goldmann visual field in the left eye examined July 19, 1993.

derangement in the presomite stage of germ layer differentiation. The most unifying theory places these cysts at the benign end of a wide spectrum of defects of midline closure. Abnormal contacts and cell displacements between peripheral (cutaneous) ectoderm dorsally, neuroectoderm in the midline, and endoderm ventrally result in communicating fistulas, dead-end sinus tracts, and sequestered cysts lined by epithelium. Associated with the epithelial aberrations are midline incursions and separations of mesodermal elements, most commonly spina bifida. The malformations are usually segmental, but extensive dysplasias and organ dislocations and aplasias, including anencephaly, comprise the more severe end of the spectrum.Z9 By this theory, epidermoid cysts and enterogenous cysts are directional variants of the same basic process: invagination of future-skin ectoderm in one patient and evagination of future-gut endoderm in the other. However, epidermoid and dermoid cysts occur both within and superficial to neural parenchyma, whereas very few enterogenous cysts, if any, are intramedullary, partly because neuroectodermal defects occur in the dorsal tangent of the neural tube and admit invaginating ectoderm openly, whereas they do not provide ready access to endoderm on the ventral side.

Figure 12. Eye movements in June 1993. The gold weight in the left upper lid is visible.

1619

Ophthalmology

Volume 101, Number 9, September 1994

Table 1. Summary oflntracranial Enterogenous Cysts* Reference

Age (yrs)

Sex

Group 29t

A

29 30 58 30 37

M F M F F M M

B A

Posterior fossa Posterior fossa Left cerebellopontine angle Posterior fossa Left prepontine cistern Left cerebellopontine angle Posterior fossa (cisterna magna)

40 48

M M

A

Craniocervical junction Cerebellum; vermis

Zalatnai20 (1987) van der Wal and Troost24 (1988) Hirai et al23 (1981)

Infant 4

F F

B B

30

F

A

Walls et aP 6 (1986)

40

F

A

F

A

Matson (1969) Kak et aP 3 (1990)

7 21

M M

Chavda et al22 (1985) Uematsu et aF 7 (1990) Harris et aP 9 (1991)

26

M

29

Umezu et al25 (1991) Giombini et al 18 (1981) Koksel et aF 4 (1990) Mehta et al26 (1984)

77

Afshar and 48 Scholtz27 (1981) Fabinyi and 54 Adams 12 (1979) Malcolm et al28 57 (1991) Small (unpublished Not data) (1962) known

F F Not Not known known

Symptoms

Location

Pontomedullary junction Anterior and to the right, from pontomedullary junction to C2 Brain stem

21

Pontomedullary junction To the left of mesencephalon and pons, adherent to CNIII Prepontine

Left CNIII, IV palsy, right hemiparesis, dysphagia

Left hemifacial spasm, left CNVII palsy Sudden onset of headache, nausea, vomiting, cerebellar ataxia Sudden onset of quadriparesis Headache, bilateral optic atrophy, right pyramidal sign, unsteady gait Complex cerebral malformation Left CNIII palsy

Sudden onset of headache, left hearing deficit, paresis of right inferior oblique muscle Gaze nystagmus, spastic hemiparesis, Multiple cysts in 4th ventricle right cerebellar ataxia Gaze nystagmus, depressed right Posterior 4th ventricle to aqueduct corneal reflex, cerebellar ataxia Premedullary to pontomedullary junction Pain Ventral pontomedullary junction to left cerebellopontine angle Prepontine

Not known

• Some data are not available for every case. t See Discussion for group classifications.

An underlying mechanism common to all midline maldevelopments is proposed. At the end of the second week of gestation, Hensen node, a center of rapid cell proliferation, arises on the dorsal surface of the embryonic disc (i.e., in the floor of the amniotic cavity) as the rostral pole of the primitive streak. The primitive notochordal process grows out of Hensen node as a cord of cells directed rostrally in the midline. As it tunnels forward directly beneath the ectodermal plate, it burrows into the ventral plate of endoderm, the archenteron, and develops a central canal. Hiatuses between the cells forming the floor of the notochordal process create openings into the yolk sac that temporarily connect this cavity, via the neurenteric canal and its caudal opening at Hensen pit, with the amniotic cavity. Beginning at the rostral end and advancing caudally, the canal and its perforations begin to close toward the end of the third week, and the notochordal process separates from the archenteron to become

1620

the notochord. Hensen pit remains open, and the formation of the notochord progresses caudally by axial migration of the node. During the third week, primitive mesoderm is laid down as a lateral plate of cells generated from the margins of the primitive streak. The new germ layer also extends forward beyond Hensen node until it reaches the future occiput, where paraxial somite formation begins on day 20, progressing caudally to reach the future first lumbar level (25th somite) on day 28 and the coccyx approximately 3 days later. The notochord induces the overlying midline ectoderm to become the neural plate, which proliferates to form the paramedial folds and invaginates medially to form the prominent neural grooves, the walls of which then fuse dorsally to form the neural tube. From the cephalospinal junction, neurulation advances forward and back. At every stage, Hensen pit marks the caudal extent of the neurenteric canal. The process is completed

Leventer et al · Enterogenous Cyst on day 26 anteriorly and day 28 posteriorly with closure of the polar neuropore. 30 This early period of germ layer differentiation and rapid cell proliferation, when dorsal and ventral epithelium are closely apposed before mesoderm intervenes, sets the stage for the midline epithelial errors. Enterogenous cysts are most common at cervical and upper thoracic levels, which suggests that they develop before or shortly after Hensen node begins its caudal migration, when the neurenteric canal openings are beginning to close. The occurrence of intracranial cysts in the posterior fossa also suggests a relatively early event. As expected, the large majority of enterogenous cysts are situated on the ventral aspect of the nervous system. The explanation for the minority of occurrences off the midline is less certain, but the same question arises with epidermal sequestrations. Why epithelial cysts are seldom multiple also is not clear. Our patient's lesion in the superior orbital fissure is the most rostral enterogenous cyst reported to date. Because the notochordal process ends in the future clivus at the level of the posterior clinoids, it might appear that the developmental derangement described cannot explain the occurrence of an epithelial rest of primitive endoderm as far forward as the orbit. The ocular apparatus, however, evolves by outward migration of a focal differentiation of neuroepithelium in the fourth week on the ventrolateral aspect of the future diencephalon; embryologically, this places our patient's lesion at the caudal border of the telencephalic vesicle, destined to become the third ventricle. Theoretically, then, the eye region may be the most rostral location possible for an enterogenous cyst. Its rarity, on the other hand, could reflect the unlikelihood of an archenteric sequestration occurring at the very tip of the neurenteric canal.

References I. Kubie LS, Fulton JF. A clinical and pathological study of

2. 3.

4. 5. 6. 7.

two teratomatous cysts of the spinal cord, containing mucus and ciliated cells. Surg Gynecol Obstet 1928;47:297-311. Rootman J. Diseases of the Orbit. Philadelphia: JB Lippincott, 1988;495-6. Hoefnagel D, Benirschke K, Duarte J. Teratomatous cysts within the vertebral canal. Observations on the occurrence of sex chromatin. J Neurol Neurosurg Psychiatry 1962;25: 159-64. Holcomb GW Jr, Matson DD. Thoracic neurenteric cyst. Surgery 1954;35: 115-21. Zulch KJ. Histological Typing of Tumours of the Central Nervous System. Geneva: World Health Organization, 1979; 59. Puusepp M. Variete rare de teratome sous-dural de Ia region cervicale (intestinome). Quadriplegie. Extirpation. Guerison complete. Rev Neurol 1934;2:879-86. Knight G, Griffiths T, Williams I. Gastrocystoma of the spinal cord. Br J Surg 1955;42:635-8.

8. Kahn AP, Hirsch JF, Da Cage C, et al. Les kystes enteriques intrarachidiens. A propos de trois observations. Neurochirurgie 1971;17:33-44. 9. Neuhauser EBD, Harris GBC, Berrett A. Roentgenographic features of neurenteric cysts. AJR Am J Roentgenol1958;79: 235-40. 10. Harriman DGF. An intraspinal enterogenous cyst. J Pathol Bacteriol1958;75:413-19. II. Agnoli AL, Laun A, Schonmayr R. Enterogenous intraspinal cysts. J Neurosurg 1984;61 :834-40. 12. Fabinyi GCA, Adams JE. High cervical spinal cord compression by an enterogenous cyst. Case report. J Neurosurg 1979;51 :556-9. 13. Kak VK, Gupta RK, Sharma BS, Banetjee AK. Craniospinal enterogenous cyst: MR findings. J Comput Assist Tomogr 1990;14:470-2. 14. Koksel T, Revesz T, Crockard HA. Craniospinal neurenteric cyst. Br J Neurosurg 1990;4:425-8. 15. Rodacki MA, Teixeira WR, Boer VHT, et al. Intradural, extramedullary high cervical neurenteric cyst. Neuroradiology 1987;29:588. 16. Walls TJ, Purohit DP, Aji WS, et al. Multiple intracranial enterogenous cysts. J Neurol Neurosurg Psychiatry 1986;49: 438-41. 17. Uematsu Y, Rojas-Corona RR, Llena JF, Hirano A. Epithelial cysts in the central nervous system, characteristic expression of cytokeratins in an immunohistochemical study. Acta Neurochir 1990;107:93-101. 18. Giombini S, Lodrini S, Migliavacca F. Intracranial enterogenous cyst. Surg Neurol 1981;16:271-3. 19. Harris CP, Dias MS, Brockmeyer DL, eta!. Neurenteric cysts of the posterior fossa: recognition, management, and embryogenesis. Neurosurgery 1991 ;29:893-8. 20. Zalatnai A. Neurenteric cyst of the medulla oblongata-a curiosity. Neuropediatrics 1987;18:40-1. 21. Matson DD. Neurosurgery of Infancy and Childhood, 2nd ed. Springfield: Charles C. Thomas, 1969;113-18. 22. Chavda SV, Davies AM, Cassar-Pullicino VN. Enterogenous cysts of the central nervous system: a report of eight cases. Clin Radio! 1985;36:245-51. 23. Hirai 0, Kawamura J, Fukumitsu T. Prepontine epitheliumlined cyst. Case report. J Neurosurg 1981;55:312-17. 24. van der Wal AC, Troost D. Enterogenous cyst of the brainstem-a case report. Neuropediatrie 1988; 19:216-17. 25. Umezu H, Aiba T, Unakami M. Enterogenous cyst of the cerebellopontine angle cistern: case report. Neurosurgery 1991 ;28:462-6. 26. Mehta VS, Chowdhury C, Bhatia R. Neuroenteric cyst of the cerebellum. Postgrad Med J 1984;60:287-9. 27. Afshar F, Scholtz CL. Enterogenous cyst of the fourth ventricle: case report. J Neurosurg 1981 ;54:836-8. 28. Malcolm GP, Symon L, Kendall B, Pires M. Intracranial neurenteric cysts. Report oftwo cases. J Neurosurg 1991;75: 115-20. 29. Wilkins RH, Odom GL. Spinal intradural cysts. In: Vinken PJ, Bruyn GW, eds. Handbook ofClinical Neurology. Vol.2, Pt.2. Amsterdam: North-Holland, 1976;55-1 02. 30. Willis RA. The Borderland of Embryology and Pathology, 2nd ed. London: Butterworths, 1962; 17-52.

1621