OCR Text |
Show Journal of Neuro- Ophthalmology 20( 2): 138- 144, 2000. © 2000 Lippincott Williams & Wilkins, Inc., Philadelphia Insights Into the Three- dimensional Structure of the Oculomotor Nuclear Complex and Fascicles T. Umapathi, MRCP, Siow Wee Koon, B Mech Eng ( Hons), Rinta Paul Mukkam, B Mech Eng ( Hons), Loong Si Chin, FRACP, Tan Chai Beng, M Med ( Int Med), Tjia Helen T. L., M Med ( Int Med), and Wieslaw L. Nowinski, DSc, PhD The authors report the case of a patient with an ischemic lesion in the left midbrain. The patient presented with paresis of left inferior rectus, pupil, right superior rectus, convergence and transiently, of the left medial rectus. A lesion in the left dorsal midbrain close to the oculomotor nuclear complex, selectively involving the fascicles innervating the above muscles, is proposed. Fine magnetic resonance sections showed a consistent lesion in the left paramedian dorsal midbrain. A detailed, three-dimensional, schematic computer model of the oculomotor nucleus and fascicles was constructed. Using this model, the authors topographically validate the putative site of the lesion. The medial rectus subnucleus is divided into three subgroups, A, B, and C. Subgroup C is thought to be the site of the majority of neurons controlling convergence. In the above model, the putative lesion is closer to subgroup A than to C; this suggests that subgroup A, rather than subgroup C, may have a higher concentration of neurons involved in convergence. Key Words: Convergence- Oculomotor fascicles- Oculomotor nucleus- Three- dimensional model. The study of the oculomotor nucleus has generated tremendous interest since it was first described by Stilling in 1846 ( 1). Contributions from various workers for the last 150 years have helped us better understand the functional anatomy of the nucleus and its fascicles. The third cranial nuclear complex is made of various groups of cells, some of which are paired. The nerve fibers fan out from these subnuclei through the ventral midbrain before exiting as a number of roots at the interpeduncular fossa. To understand the various patterns of clinical deficits that can occur, it is essential to appreciate the detailed functional topography of the oculomotor nuclear complex and fascicles. It is not possible to Manuscript received October 13, 1999; accepted March 23, 2000. From the National Neuroscience Institute ( TU, LSC, TCB, TH), Singapore; the Department of Mechanical and Production Engineering ( SWK, RPM), National University of Singapore; and the Biomedical Laboratory ( WLN), Kent Ridge Digital Labs, Singapore. Address correspondence to T. Umapathi, MRCP, Department of Neurology, Johns Hopkins University, 600 North Wolfe Street, Meyer 6- 113, Baltimore, Maryland 21287- 7613. visualize the detailed organization of the subnuclei and the fascicles, even with the use of high- resolution magnetic resonance images ( 2). Much of our knowledge comes from the study of clinical cases and from various types of retrograde nerve studies in animals. Our objectives were: 1. to understand the three- dimensional structure of the oculomotor nucleus and its fascicles; 2. to test if a localized unilateral lesion in the dorsal midbrain can cause the pattern of opthalmoplegia seen in the patient described; 3. to hypothesize the possible concentration of neurons involved in convergence. MATERIALS AND METHODS Incorporating information available on the oculomotor nuclear complex and its fascicles ( 3- 11), a three-dimensional model was constructed using Unigraphics ( Version 13; Unigraphics Solution Pte. Ltd., Singapore), a computer- aided design software normally used in the design and manufacture of precision machines and tools. The position, the proportional size, and the relation to the red nucleus and substantia nigra of the oculomotor nuclear complex in the coronal plane were obtained and applied to this model from the Electronic Talairach atlas ( Thieme [ Thieme, NY] and Kent Ridge Digital Labs [ Singapore]) ( 12). CASE REPORT A 54- year- old woman presented with an acute history of vertigo, vomiting, and diplopia, with vertical separation of images maximum on looking up or down. She had no weakness, incoordination, or numbness of limbs. She did not smoke, and she had been taking low- dose estrogen for 4 months to treat menopausal symptoms. She had no significant medical history. On clinical examination, abnormalities were confined to the eyes ( Fig. 1). At rest, the right eye was lower and extorted. Elevation was reduced on upgaze OD ( pursuit 738 STRUCTURE OF THE OCULOMOTOR NUCLEAR COMPLEX AND FASCICLES 139 FIG. 1. Ocular movement abnormalities. and saccade), and the deficit was maximum in the abducted position; depression was reduced OS ( pursuit and saccade), and it was more obvious in the abducted position. Vertical oculocephalic reflex ( Fig. 2) and the Bell phenomenon ( Fig. 3) showed a similar degree of eye-movement deficit. A mild deficit was detected in left- eye adduction during the first examination; this resolved within a few hours. However, convergence was markedly impaired ( Fig. 4). The adequacy of the stimulus for convergence was confirmed by the miotic reaction OD. The right pupil was 1.5 mm, and the left pupil was 2.5 mm. The left eye was not reactive to light or accommodation. The pupillary reaction OD was normal. Results of the rest of the neurologic examination were normal. There was no evidence of thyroid eye disease, myasthenia gravis, or Brown tendon sheath syndrome. Magnetic resonance imaging showed a discrete lesion in the rostral part of the left midbrain ( Figs. 5A, B). Functional diffusion- weighted magnetic resonance images did not show any other lesions in either side of the midbrain ( Fig. 5C). The anticardiolipin antibody was raised ( immunoglobulin M, 35 MPL units/ mL). No other autoimmune clinical features or markers were present, and a diagnosis of primary antiphospholipid syndrome was made. Hormone replacement therapy was stopped, and the patient was anticoagulated. Over the next few weeks, the patient's deficits resolved gradually; convergence resolved first, followed by extraocular movements, and then mydriasis. In summary, this patient had a left paramedian and dorsal midbrain ischemic lesion that gave rise to elevation deficit, which was maximum in the abducted position OD. This appeared to be nuclear or infranuclear. OS there was a mydriatic pupil that did not respond to light or accommodation, and depression deficit that was maximum in the abducted position, which appeared to be nuclear or infranuclear. DISCUSSION We propose a lesion in the left midbrain involving the left inferior rectus, left pupil, and right superior rectus fascicles of the oculomotor nerve to explain the findings in the patient. We believe the mild depression deficit OS in the adducted position is solely due to the weakness of the inferior rectus. We would have expected greater weakness of depression in the adducted position if there were a concomitant left superior oblique palsy. Also, a lesion that extends as far caudally as to involve the region around the aqueduct in the caudal part of the midbrain, where the fourth fascicles traverse, might be expected to involve other structures, such as the superior cerebellar peduncle. Other than the vertigo at presentation, the pa- • w* m FIG. 2. Vertical oculocephalic reflex showing depression OS and elevation OD deficits. FIG. 3. Bell phenomenon showing the elevation deficit OD. J Neuro- Ophthalmol, Vol. 20, No. 2, 2000 140 T. UMAPATH1 ETAL. FIG. 4. Attempted convergence. tient did not have any objective cerebellar signs. Unfortunately, the resolution provided by the sagittal magnetic resonance image does not clarify this issue. We also feel that the lesion could not be in the oculomotor nucleus because the internal ophthalmoplegia and superior rectus palsy were unilateral ( 13). Wiest et al. ( 14) described a similar crossed vertical gaze palsy in a patient with mono- ocular elevation paresis and contralateral downgaze paresis. The lesion was postulated to be supranuclear, at the rostral interstitial nucleus of the medial longitudinal fasciculus. The same lesion could not explain our patient's deficits because the loss of Bell and oculocephalic reflexes were to the same extent as the volitional paresis, and there was ipsilateral mydriasis and convergence paresis. To validate our hypothesis, we constructed a three-dimensional schematic model of the oculomotor nuclear complex and its fascicles from information available from previous case reports and experimental studies. Stilling ( 1) is attributed the honor of first clearly describing the oculomotor nucleus, in 1846. von Gudden ( 15) divided the somatic nuclei of each side into dorsal and ventral parts and also established the partial decussation of the fascicles in the midbrain. Edinger ( 16) and Westphal ( 17) described the subnuclei subsequently named after them. Perlia ( 18) included these recently described subnuclei into the nuclear complex. He also recognized a median group of motor nerve cells between von Gudden's ventral nuclei by dividing the dorsal and ventral nuclei into rostral and caudal moieties. Over the next 50 years, various investigators, using data from clinical cases and retrograde studies, attempted to delineate further the functional anatomy of the oculomotor nuclear complex. In 1953, Warwick ( 3) published his classic study using retrograde degenerative methods in rhesus monkeys. He described the oculomotor nucleus as a complex of functional cell masses. The somatic nuclei for each extraocular muscle were paired and placed laterally with ipsilateral control, while the superior rectus subnuclei were located medially on either side of the midline, each one innervating the contralateral superior rectus muscle. For the first time, he attributed control of the levators of both sides to the caudal central nucleus J Neuro- Ophthalmol, Vol. 20, No. 2, 2000 confined to the caudal third of the nuclear complex. Retrograde tracer studies using horseradish peroxidase, wheat germ agglutinin, and tetanus toxin by Buttner- Ennever et al. ( 4,5) and Porter et al. ( 6) further refined Warwick's model of the oculomotor nucleus. In particular, they were able to subdivide the medial rectus sub-nuclei into three separate subgroups of uncertain function. The visceral subnuclei, both the paired anterior median and the Edinger- Westphal nuclei, occur rostrally. More caudally, the latter divides into slender lateral columns ( 7). The subnuclei serving individual extraocular muscles give rise to bundles of fascicles, which course ventrally and spread laterally through the red nucleus. The fascicular arrangement in the ventral midbrain, from medial to lateral, has been suggested as follows: pupillary fibers, inferior rectus, lid/ medial rectus, superior rectus, and inferior oblique ( 8). The Edinger- Westphal nucleus and inferior rectus subnucleus are in close proximity from the rostral end of the nuclear complex to the middle third. Their fibers then occupy the medial aspect of the fascicles. The inferior oblique subnucleus, which is most lateral in the middle third of the nuclear complex, gives rise to fibers that remain in the most lateral part of the fascicles. The fibers from the medially located contralateral superior rectus subnucleus cross and travel laterally, near the inferior oblique fascicle. The medial rectus subnucleus, located primarily in the ventral oculomotor nuclear complex and cauda- central nucleus, transmits fibers in an intermediate position. Burde et al. ( 9)* and Zaks et al. ( 10) have proposed a three- dimensional model of the rostro- caudal oculomotor fascicle arrangement based on studies in nonhuman primates, as well as in patients. The pupillary and inferior rectus fibers are most rostral, followed by the medial rectus, lid, superior rectus, and inferior oblique. Burde et al. ( 9) and Ksiazek et al. ( 19) also developed elegant schematic diagrams of the oculomotor nucleus. Using all of this information, we constructed a computerized, schematic, three- dimensional model of the oculomotor nucleus and the fascicles in the midbrain ( Figs. 6A- C). With this model, we were able to demonstrate that the contra- lateral superior rectus fascicle passes close to the ipsilateral inferior rectus and pupillary fibers before taking up a position caudal and lateral to the contralateral medial rectus and lid fibers. Therefore, a lesion at the rostral aspect of the fascicle just ventral to the third nuclear complex ( Fig. 6D) can cause dysfunction of the contralateral superior rectus, ipsilateral inferior rectus, and pupil. The better visualization afforded by the model may help resolve some of the controversies in the literature on the fascicular arrangement of the oculomotor nerve. Castro et al. ( 8) felt that the higher incidence of the synkinesis between the levator palpebrae and inferior rectus muscles, compared with the medial rectus muscles ( 95% and 67%, respectively) in aberrant regeneration of * With permission from Mosby Year Book. STRUCTURE OF THE OCULOMOTOR NUCLEAR COMPLEX AND FASCICLES 141 Ik MM I V V A W y 1 FIG. 5. Magnetic resonance images of midbrain showing an infarction in the left dorsal paramedian midbrain. A: T2- weighted axial. B: T2- weighted parasagittal. C: Diffusion axial images. the third nerve, argued for an intermediate position of the lid fibers between the inferior rectus and medial rectus fibers. Ksiazek et al. ( 19) reported two cases of patients with mydriasis, marked inferior rectus weakness, and medial rectus muscle paresis with sparing of the lids from ipsilateral fascicular lesions. This suggested proximity of the pupillary and inferior rectus fibers. In addition, it appeared that the medial rectus and inferior rectus fibers were adjacent without the lid fibers intervening. The mild, transient, left adduction deficit without ptosis in our patient appears to support the latter theory. However, on appreciating the three- dimensional orientation of these fibers, one realizes that both these ideas need not be mutually exclusive and can be reconciled. Visualizing the relation of the superior rectus fibers to the lid fibers and the potential association of the visceral, inferior rectus, medial rectus, and inferior oblique fibers, it is not too difficult to conceive- at least from a topographic point of view- divisional third- nerve lesions from intraaxial lesions ( 20). Schwartz et al. ( 21) described a patient with a third fascicular lesion sparing the inferior rectus and pupil. In our model, to spare these two fascicles, the lesion has to involve the region rostro- ventral to the nucleus ( affecting the lid and superior rectus) and extend laterally and dor-sally, catching the medial rectus and inferior oblique fibers. The lesion shown in the axial magnetic resonance image of this patient has a similar slanted shape. Shuaib and Murphy ( 22) described a case of medial rectus, inferior rectus, and levator palpebrae weakness J Neuro- Ophthalmol, Vol. 20, No. 2, 2000 T. UMAPATHI ET Ah. FIG. 6. A: Axial view showing arrangement of oculomotor fascicles in relation to the red nucleus and substangtia nigra ( the latter two structures are shown only on the left). B: Coronal view of third nucleus and fascicles in the midbrain ( seen rostrally). C: Sagittal view of fascicles as they traverse through the ventral midbrain and red nucleus toward the interpeduncular fossa. D: Lesion ( sphere) involving the left pupillary, left inferior rectus, and contralateral superior rectus fibers. E: Parasagittal view of oculomotor nucleus and fascicles ( with the exclusion of medial rectus fascicles) to show the proximity of the lesion to subgroup A, rather than to subgroup C, of the medial rectus subnucleus. 1, oculomotor nuclear complex; 2, oculomotor fascicles; 3, red nucleus; 4, substantia nigra; 5 cauda- central subnucleus and fascicle; 6, inferior oblique subnucleus and fascicle; 7A, subgroup A of medial rectus subnucleus; 7B, subgroup B of medial rectus subnucleus; 7C, subgroup C of medial rectus subnucleus; 8, superior rectus subnucleus and fascicle; 9, inferior rectus sub-nucleus and fascicle; 10, visceral ( parasympathetic) subnuclei and fascicle; 11, putative lesion. STRUCTURE OF THE OCULOMOTOR NUCLEAR COMPLEX AND FASCICLES 143 from a midbrain hemorrhage, which is difficult to explain by a single lesion ( 10). The clinical signs of intra-parenchymal bleeds, because of associated edema, often do not follow a fixed vascular pattern like infarctions do. Conceivably, a lesion rostro- ventro- lateral to the oculomotor nucleus, just before the superior rectus traverses to a position in between the medial rectus and inferior oblique fascicles, could cause the above set of signs. We believe that these examples, to some extent, validate the accuracy of the model. However, the data in our model contradict one major work ( 23). In his studies on cats, Bienfang described the superior rectus fascicle crossing and traversing the substance of the contralateral superior rectus subnuclei. Our patient had contralateral superior rectus palsy from an ipsilateral fascicular lesion. We feel that the superior rectus fascicles travel very close to each other, before decussating at the rostro- ventral part of the nuclear complex. To reconcile Bienfang's data, the inferior rectus and pupillary fibers have to travel in the midline between the right and left oculomotor nuclear components in our model. This is possible from a biophysical perspective. However, one would anticipate a high incidence of bilateral inferior rectus palsies in nuclear lesions if this were true. Furthermore, Burde and Loewy ( 7) found that in monkeys, the visceral axons course rostrally and ventrally. The close association of the superior rectus fascicles at the midline explains the bilaterality of involvement in nuclear lesions, with the contralateral side being worse. We feel that if the superior rectus traverses through the substance of the contralateral subnucleus, the latter observation would not be seen. Can lesions in the brainstem cause such localized involvement of the various fascicles? Intraaxial midbrain lesions involving oculomotor fascicles causing isolated inferior oblique ( 8); pupil ( 24); inferior oblique, superior rectus, medial rectus, lid, inferior rectus ( 22); and inferior oblique, superior rectus, medial rectus, lid paresis ( 22,25) have been reported. Divisional oculomotor nerve paresis has also been reported in intraaxial lesions ( 19). Hopf and Gutmann ( 26), in a study of 11 consecutive adults with diabetes, suggested that isolated third nerve palsy, with or without pupillary sparing, is much more likely on the basis of a focal mesencephalic infarct than a peripheral nerve lesion. In a review of published series, Thomke ( 27) found isolated third, sixth, and eighth cranial nerve dysfunction from lesions of the intraaxial parts of the nerve in the brainstem to be more common than extraaxial lesions. In fact, up to three quarters of isolated third- nerve lesions could be in the brain stem. In a series of ten patients with isolated superior oblique palsies from brainstem lesions ( 28), more than half had normal results on magnetic resonance images. Many of these patients were diagnosed on the basis of electrophysiologic abnormalities that suggested brainstem lesions. With such small lesions and with the wide separation of the individual fascicles of the oculomotor nerve before they exit at the interpeduncular fossa, both medial- to- lateral and rostral- to- caudal, it is not surprising that selective paresis of individual eye muscles can occur. The vascular supply of the oculomotor nuclei and fascicles is via a median group of arteries that arise from terminal bifurcation of the basilar artery at the origin of the superior cerebellar and posterior cerebral arteries ( 29). These arteries enter the brainstem through the interpeduncular fossa and course dorsally, producing a dense bundle to supply the midline structures up to and along the floor of the cerebral aqueduct. The terminal arborizations of these penetrating paramedian arteries are very fine to allow selective involvement. This is exemplified by the two cases reported by Bryan and Hamed ( 30) of nuclear third- nerve lesions that were small enough to actually spare the lid and pupil. The presence of convergence failure with normal adduction OS prompts an interesting hypothesis. There is evidence that not all of the motor neurons of the medial rectus subnuclei participate equally in versional and ver-gence movements ( 31). The retrograde tracer studies done by Buttner- Ennever and Akert ( 4), Buttner- Ennever et al. ( 5), and Porter et al. ( 6) have shown three different subgroups of medial rectus motorneurons in monkeys: subgroup A, located ventral and rostral; subgroup B, located dorsal and caudal; and subgroup C, located dorso-medial and rostral. Subgroup C is composed of the smallest cell bodies and can be labelled independently of the other subgroups by selective injections of radioactive tracers into the outer ( orbital) layer of the medial rectus muscle. Because the outer layer of the ocular muscles contains smaller muscle fibers, which are more likely involved in generating slower eye movements, it is hypothesized that subgroup C may be involved in convergence ( 32). Our patient had transient adduction failure OS. Subsequently, she had normal adduction on conjugate movement, but markedly impaired convergence. The area of the medial rectus subnuclei close to the suspected site of the lesion in our patient may contain a higher concentration of cell bodies mediating convergence. The edema associated with an acute ischemic lesion may have initially caused dysfunction in these cells, as well as in those mediating adduction independently or together with the exiting medial rectus fascicles. With resolution of edema, only convergence cells were affected. In the three- dimensional model, the suspected site of lesion in this patient is not close to subgroup C ( Fig. 6E). In fact, it is the rostro- ventral end of subgroup A that is closest. We would like to speculate that, at least in humans, this might be the site where there is a higher concentration of motor neurons mediating convergence. It is of course conceivable that the lesion could have affected supranuclear pathways on their way to the relevant medial rectus subgroup. We feel that this does not necessarily detract from our hypothesis. The very fact that the supranuclear fibers controlling convergence could be affected by the putative lesion, which is close to subgroup A, suggests that these fibers are headed for subgroup A, and the latter has an important role in controlling convergence. J Neuro- Ophthalmol, Vol. 20, No. 2, 2000 144 T. UMAPATHI ET AL. In summary, we describe a patient with crossed vertical paresis, ipsilateral mydriasis, and convergence paresis from a localized lesion at the individual fascicles of the third nerve. We demonstrate the topographic feasibility of such a lesion by constructing a schematic three-dimensional model. This case prompts us to propose that subgroup A, rather than subgroup C, of the medial rectus subnucleus may contain a higher proportion of neurons involved in convergence. REFERENCES 1. Stilling B. Disquisitiones de structura et functionibus cerebri. Je-nae: sumt. f. Maukii 1846. 2. Miller MJ, Mark LP, Ho KC, Haughton VM. Anatomic relationship of the oculomotor nuclear complex and medial longitudinal fascicles in the midbrain. Am J Neuroradiol 1997; 18: 111- 3. 3. Warwick R. Representations of the extra- ocular muscles in the oculomotor nuclei of the monkey. / Comp Neurol 1953; 98: 449- 503. 4. Buttner- Ennever JA, Akert K. Medial rectus subgroups of the oculomotor nucleus and their abducens internuclear input in the monkey. J Comp Neurol 1981; 197: 17- 27. 5. Buttner- Ennever JA, Grob P, Akert K. A trans- synaptic autoradiographic study of the pathways controlling the extraocular eye-muscles, using [ 125I] B- IIb tetanus toxin fragment. In: Cohen B, ed. Vestibular and oculomotor physiology. New York: New York Academy of Sciences, 1980: 57- 70. 6. Porter JD, Barton LG, Sparks DL. Innervation of monkey extra-oculomotor muscles: location of sensory and motor neurons by retrograde transport of horseradish peroxidase. J Comp Neurol 1983; 218: 208- 19. 7. Burde RM, Loewy AD. Central origin of the oculomotor parasympathetic neurons in the monkey. Brain Res 1980; 198: 434- 39. 8. Castro X, Johnson LN, Mamourian AC. Isolated inferior oblique paresis from brain- stem infarction. Arch Neurol 1990; 47: 235- 7. 9. Burde RM, Sarvino Pj, Trobe JD. Clinical decisions in neuro-opthalmology. 2nd ed. St Louis: Mosby Year Book, 1992: 247- 9. 10. Zaks R, Slamovits T, Burde R. Oculomotor brainstem anatomy: nuclei to fascicles. J Neurooncol 1994; 18: 241- 8. 11. Gauntt CD, Kashii S, Nagata I. Mononuclear elevation paresis caused by an oculomotor fascicular impairment. J Neuroopthalmol 1995; 15: 11- 4. 12. Nowinski WL, Bryan RN, Raghavan R. The electronic clinical brain atlas. Multiplanar navigation of the human brain. New York: Thieme, 1997. 13. Daroff RB. Ocular motor manifestations of brainstem and cerebellar dysfunction. In: Smith JL, ed. Neuro- opthalmology: symposium of the University of Miami and Bascom Palmar Eye Institute. Vol. 5. Hallandale, FL: Huffman Publishing Co., 1970: 104- 18. 14. Wiest G, Baumgartner, Schnider P, Trattnig S, Deecke L, Mueller C. Monocular elevation paresis and contralateral downgaze paresis from unilateral mesodiencephalic infarction. J Neurol Neurosurg Psychiatry 1996; 60: 579- 81. 15. von Gudden BA. Ueber die Kerne der Augenbewegungsnerven. Zbl. Augenheilk 1881; 5: 528- 9. 16. Edinger L. Ueber den Verlauf der zentralen Hirnnervenbahnen. Arch f Psychiat V Nervenkr 1885; 16: 858- 9. ' 17. Westphal C. Uber einen Fall von chronischer progressiver Lah-mung der Augenmuskeln ( Opthalmoplegia externa) nebst Bes-chreibung von Ganglienzellengruppen im Bereiche des Oculomo-toriuskems. Arch f Psychiat 1887; 18: 846- 69. 18. Perlia R. Die Anatomie des Oculomotoriuscentrums beim Men-schen. V Graefe's Arch f Ophthal 1889; 35: 287- 308. 19. Ksiazek SM, Slamovits XL, Rosen CE, Burde RM, Parisi F. Fascicular arrangement in partial oculomotor paresis. Am J Opthalmol 1994; 118: 997- 103. 20. Ksiazek SM, Repka MX, Maguire A, et al. Divisional oculomotor nerve paresis caused by intrinsic brainstem disease. Ann Neurol 1989; 26: 714- 8. 21. Schwartz TH, Lycette CA, Yoon SS, Kargman DE. Clinicoradio-graphic evidence for oculomotor fascicular anatomy ( letter). J Neurol Neurosurg Psychiatry 1995; 59: 338. 22. Shuaib A, Murphy W. Mesencephalic hemorrhage and third nerve palsy. CT: J Comput Tomogr 1987; 11: 385- 8. 23. Bienfang DC. Crossing axons in the third nucleus. Invest Opthalmol 1975; 14: 927- 31. 24. Shuaib A, Israelian G, Lee MA. Mesencephalic haemorrhage and unilateral pupillary deficit. J Clin Neuroopthalmol 1989; 9: 47- 9. 25. Nadeau SE, Trobe JD. Pupil sparing in oculomotor palsy: a brief review. Ann Neurol 1983; 13: 143- 8. 26. Hopf HC, Gutmann L. Diabetic 3rd nerve palsy: evidence for a mesencephalic lesion. Neurology 1990; 40: 1041- 5. 27. Thomke F. Isolated cranial nerve palsies due to brainstem lesions. Muscle Nerve 1999; 22: 1168- 76. 28. Thomke F, Christian H. isolated superior oblique palsies with elec-trophysiologically documented brainstem lesions. Muscle Nerve 2000; 23: 267- 70. 29. Porter JD, Barker RS. Anatomy and embryology of the oculomotor system. In: Miller NR, Newman NJ, eds. Walsh and Hoyt's clinical neuro- ophthalmology. 5th ed. Baltimore: Williams and Wilkins, 1998: 1055- 6. 30. Bryan JS, Hamed LM. Levator- sparing nuclear oculomotor palsy. Clinical and magnetic resonance imaging findings. J Clin Neuroopthalmol 1992; 12: 26- 30. 31. Mays LE, Porter JD. Neural control of vergence eye movements: activity of abducens and oculomotor neurons. J Neurophysiol 1984; 52: 743- 61. 32. Leigh RJ, Zee DS. Vergence eye movements. The neurology of eye movements. 2nd ed. Philadelphia: FA Davis Co., 1991: 264- 90. J Neuro- Ophthalmol, Vol. 20, No. 2, 2000 |