Journal of Neuro-Ophthalmology, December 1996, Volume 16, Issue 4

Visual Function in Neurofibromatosis

OBJECTIVES: To evaluate the visual function in patients with neurofibromatosis (NF) and to study the etiology and incidence of visual dysfunction associated with NF. PATIENTS AND METHODS: A total of 75 patient with diagnostic criteria for NF were evaluated. Neuro-ophthalmological examination as well as electrophysiological and imaging studies were performed. Special attention was given to the presence of visual dysfunction and to its correlation with the ophthalmic changes that were found. RESULTS: Ocular findings were present in 42 (56%) patients. Visual dysfunction was identified in only 11 (14.7%) patients. Visual acuity decrease was the most prevalent change, being present in eight (72.7%) of patients with visual dysfunction. Nystagmus, strabismus, visual field defects, and color vision defects were also detected. Therapy is also reviewed. The prognosis of the 11 patients with visual dysfunction was unfavorable, and is discussed. CONCLUSION: The prevalence of NF and ophthalmological findings related to it make the problem of visual dysfunction in NF a serious one, deserving of the attention of all ophthalmologists. Clinical examination, associated with complementary diagnostic techniques (mainly imaging studies), allows NF's identification, definition, and therapy. The high prevalence of asymptomatic ocular findings in NF (73.8%) highlights the role of imaging techniques in its evaluation.

journal of Neuro- Ophthalmology 16( 4): 234- 240, 1996. © 1996 Lippincott- Raven Publishers, Philadelphia Visual Function in Neurofibromatosis A. Castanheira- Dinis, M. D., Manuela Novais, M. D., Ivone Cravo, M. D., Fatima Campos, M. D., L. Gouveia- Andrade, M. D., and J. Ribeiro- da- Silva, M. D. Objectives: To evaluate the visual function in patients with neurofibromatosis ( NF) and to study the etiology and incidence of visual dysfunction associated with NF. Patients and Methods: A total of 75 patient with diag­nostic criteria for NF were evaluated. Neuro- ophthalmo-logical examination as well as electrophysiological and imaging studies were performed. Special attention was given to the presence of visual dysfunction and to its correlation with the ophthalmic changes that were found. Results: Ocular findings were present in 42 ( 56%) pa­tients. Visual dysfunction was identified in only 11 ( 14.7%) patients. Visual acuity decrease was the most prevalent change, being present in eight ( 72.7%) of pa­tients with visual dysfunction. Nystagmus, strabismus, visual field defects, and color vision defects were also detected. Therapy is also reviewed. The prognosis of the 11 patients with visual dysfunction was unfavorable, and is discussed. Conclusion: The prevalence of NF and ophthalmological findings related to it make the problem of visual dys­function in NF a serious one, deserving of the attention of all ophthalmologists. Clinical examination, associated with complementary diagnostic techniques ( mainly im­aging studies), allows NF's identification, definition, and therapy. The high prevalence of asymptomatic oc­ular findings in NF ( 73.8%) highlights the role of imag­ing techniques in its evaluation. Key Words: Neurofibromatosis- Visual function- Gli­omas- MR. Manuscript received April 10, 1995; accepted March 8, 1996. From the Departments of Paediatric Ophthalmology and Neuro- Ophthalmology, Santa Maria Hospital Eye Clinic, Lis­bon Faculty of Medicine, Lisbon, Portugal. Address correspondence and reprint requests to Dr. A. Castanheira- Dinis, Eye Clinic, Hospital de Santa Maria, Avenida Professor Egas Moniz, 1699 Lisboa Codex, Portugal. Neurofibromatosis ( NF) is a systemic disease transmitted as an autosomal dominant trait ( 1- 3). It is considered the most frequent of the phakoma-toses, its gene being present in one in every 3,000- 4,000 individuals ( 1- 3). Two types are well known, NF1 and NF2, but other types may exist with dif­ferent clinical and genetic features. However, ex­istence of subtypes as autonomous entities is still under investigation ( 1- 5). The high incidence and polymorphism of NF explain the many scientific reports on this disease that deal with clinical find­ings, genetic patterns, diagnosis, and therapeutic approaches. As it is a disease of neural crest origin, ocular findings are frequent and age- related. In NF1, oc­ular changes are frequently the first to be noticed, and their identification is accepted as a fundamen­tal criterion for NF classification. Thus, ophthal­mologists play a major role in diagnosing and de­fining this disease; it is, therefore, natural to find a large number of reports concerning NF in the oph­thalmic literature. Although ophthalmologists always evaluate vi­sual function in NF patients and prescribe the most appropriate measures for correction of any anom­alies found, results of such work are not usually published or appear only as complementary data. The study of visual function in NF patients, most of whom are children, incidence of related sys­temic disturbances, incidence of visual dysfunc­tion, characterization of its etiology, efficacy of therapeutic measures, and visual prognosis consti­tute the basis of the present study. PATIENTS AND METHODS From January 1990 to December 1994 ( 5 years), 78 patients who were considered to have NF were evaluated retrospectively at the Neuro- Ophthal­mology and Pediatric Ophthalmology Depart- 234 VISUAL FUNCTION IN NEUROFIBROMATOSIS 235 ments of the Santa Maria Hospital Eye Clinic. Of these 78, three patients were excluded from the study, including a 53- year- old man with criteria for NF1 and bilateral diabetic maculopathy; a 33- year-old man with criteria for NF1, a left eye ( grade C) proliferative vitreoretinopathy after retinal detach­ment; and a 12- year- old boy with criteria for NF1 and a disfiguring, posttraumatic, left eye corneal opacity. Of the remaining 75 patients, 72 ( 96%) fulfilled the National Institutes of Health Consen­sus Development Conference ( 6) for NF1, two were probably NF2, and one was likely to have an NF5. Data concerning those patients are shown in Table I. Most ( 72%) were children and sex distri­bution did not show significant differences. Whenever possible, patients were submitted to a full neuro- ophthalmological examination, electro­physiological studies, and radioimaging. Neuro-ophthalmological examination was adapted to pa­tient age and, whenever possible, included visual acuity assessment, evaluation of eye movements and pupillary reflexes, anterior segment biomi­croscopy, posterior segment examination with di­rect ophthalmoscopy, indirect ophthalmoscopy or 90 D lens, color vision assessment with Farn-sworth 15 and/ or 100 Hue tests, and visual field evaluation by kinetic perimetry using a Goldmann perimeter. For adults and verbal children, visual acuity was evaluated using a Snellen chart at 6 m, but for preverbal children, preferential looking and Stycar were the selected examination techniques. Preferential looking was assessed using Teller acu­ity cards and standard procedures. STYCAR stands for sheridan tests for young children and retarded are used to assess visual competence during infancy and early childhood, and in handicapped children. STYCAR includes several tests ( single- letter cards, miniature toy test, graded balls tests, rolling ball test, mounted balls test, Snellen equivalents, and others), each one be­ing specially adapted to a certain age group and to different levels of handicap. They are used only when an appropriate result cannot be obtained with preferential looking tests. In our patients, STYCAR was used for only two children ( case 3, a 3- year- old girl, and case 9, a 7- year- old girl). TABLE 1. Population ( n = 75) Mean age ( yrs) 14.1 Male 33 ( 44.0%) Female 42 ( 56.0%) Adults (> 16 yr) 21( 28.0%) Mean age ( yrs) 33.7 Children (= s16 yr) 54 ( 72.0%) Mean age ( yr) 6.5 In case 3, the mounted balls test was used. In this test, six plastic balls of 6.3, 5.1, 3.8, 3.5, 1.9, and 1.3 cm diameter are used. The child faces a 71.3 x 60.96 cm screen placed at 3 m, and the examiner holds a ball completely still in a position a few inches beyond the edge of the screen for 2- 3 s. Then, the ball is rapidly removed from the child's sight behind the screen, being presented again for 2- 3 s at the same distance from the op­posite edge of the screen. In response, the child should swiftly move his eyes to fixate the ball in the new position. The balls are presented in order of decreasing size and results are registered. Visual acuity is determined as follows: ball of 6.3 cm, 6/ 48; ball of 5.1 cm, 6/ 36; ball of 3.8 cm, 6/ 24; ball of 3.5 cm, 6/ 18; ball of 1.9, 6/ 12; and ball of 1.3 cm, 6/ 9. Smaller balls may be used: ball of .95 cm, 6/ 6; ball of .62 cm, 6/ 4.5; ball of .47 cm, 6/ 3; and ball of .32 cm, 6/ 2.5). In case 9, single- letter cards were used. In this test, cards are presented at 3 m and the child tries to match the letters pointed by the examiner with those displayed in a key- card. Visual acuity is de­termined as follows: O, 6/ 60; X, 6/ 36; V, 6/ 24; AT, 6/ 18; UX, 6/ 12; HT, 6/ 9; OX, 6/ 6; TV, 6/ 4.5; and HO, 6/ 3. In each age group, visual acuity was evaluated considering preferential looking normal values as a reference: neonate, 20/ 800; 1 month, 20/ 400- 20/ 200; 3- 4 months, 20/ 200- 20/ 100; 9- 12 months, 20/ 100- 20/ 50; 18 months, 20/ 50- 20/ 25; 24 months, 20/ 25; and 36 months, 20/ 20. Although these guidelines are useful to define normal and decreased visual acuity, careful interpretation was attempted in ev­ery case. Visual evoked potentials ( VEPs) were performed with " Biosistemas Studio" equipment and an Ol­ivetti computer. Each record was made by summa­rizing responses to 128 stimuli, presented at a fre­quency of 1.6 Hz under monocular viewing condi­tions. The patient was seated 125 cm in front of a square grating and fixated a 1- cm diameter red tar­get. A pattern stimulus was generated by a video monitor with a spatial frequency of 30' arc, 100% contrast, and acquisition time of 512 ms. Not only was the general morphology of the recording an­alyzed, but amplitude and latency of the P100 wave were quantified. VEP recording was attempted on every patient. Pattern stimuli were used when visual acuity was > 1/ 10 and patient age was 3= 4 years old. Flash stimuli were preferred when these conditions were not present. Electroretinography ( ERG) was performed with " BioSistemas Studio" equipment and an Olivetti computer whenever there were retinal lesions. / Neuro- Ophthalmol, Vol. 16, No. 4, 1996 236 A. CASTANHEIRA- DINIS ET AL. Flash ERGs were obtained with bilateral mydriasis after 1% tropicamide. Photopic ERGs were re­corded with the retina adapted to an ambient light intensity of - 300 lux. Scotopic ERGs were re­corded in a completely dark room with an analysis time of 128 ms. A complete study of each recording was carried out with photopic and scotopic retinal adaptation. Oscillatory potentials, amplitude ( mv), and latency ( ms) of scotopic b- waves were all eval­uated. Imaging studies- computed tomography ( CT) and/ or magnetic resonance ( MR)- were also per­formed in all cases. When NF2 was diagnosed, an audiogram was requested. Special attention was paid in evaluation of visual acuity to identification of other forms of visual dys­function and correlation to ocular changes that were found. RESULTS Results are shown in Tables 2 and 3. NF associated ocular changes were present in 42 ( 56%) of the 75 patients ( Table 2). Although 33 pa­tients had only one type of abnormal ocular find­ing, two different ocular manifestations were found in nine patients. One must emphasize that in this group of 42 cases, 31 ( 73.8%) patients did not show any evi­dence of visual dysfunction. Of these 42 patients, 16 were adults and 26 were children; thus, 76.2% of the adults presented ocular findings while only 48.1% of the children had ocular abnormalities. Lisch nodules were the ocular finding most fre­quently found ( 42.3%). It seems of interest to men­tion their appearance in a 3- year- old child. Optic pathway gliomas were found in 10 ( 13.3%) cases and retinal hamartomas in two cases. Three chil­dren had lens opacities and two of these had di­agnostic criteria for NF1. Sphenoid dysplasia or hypoplasia was present in three patients. One of these patients had a pulsatile proptosis in associa­tion with an orbital encephalocele and absence of the greater wing of the sphenoid. TABLE 2. Ocular/ orbital findings ( 42 patients) Finding n (%) Lisch nodules 32 ( 42.3) Optic pathway gliomas 10( 13.3) Lens opacities 3 ( 4.0) Retinal hamartomas 2 ( 2.7) Sphenoid dysplasia 3 ( 4.0) Congenital glaucoma 1 ( 1.3) Eleven patients had NF ocular changes associ­ated with visual dysfunction ( Table 3). Of these 11 patients, seven had more than one recognizable form of visual dysfunction and only in two cases ( cases 4 and 10) was that dysfunction unilateral. Table 4 shows, in detail, the ophthalmological manifestations found in these 11 patients as well as results of the different diagnostic tests performed and the different therapeutic options considered. Among the various forms of visual dysfunction that were recognized, decreased visual acuity was the most prevalent [( 72.7%) eight patients] affecting both eyes in six cases. Of these eight patients, six had optic pathway gliomas as a direct cause of that disturbance. In case 5, an 8- month- old child, a pre­cise evaluation of visual acuity was not possible, but the presence of nystagmus, and optic nerve chiasmal and optic tract glioma supported the presence of a visual acuity impairment. In case 4, the association of a combined pigment epithelial and retinal hamartoma affecting the macular re­gion definitely contributed to the visual acuity de­crease. In the remaining cases, a retinal hamarto­ma and cataract ( case 8) and congenital glaucoma ( case 11) were the probable causes for this type of visual dysfunction. Nystagmus was found in three ( cases 1, 2, and 5) patients and strabismus was present in two ( cases 3 and 4) patients. The etiology of these findings is, probably, secondary to the visual acuity decreases that, in these five patients, was due to the pres­ence of optic pathway gliomas. Four patients ( cases 1, 2, 6, and 7) had visual field defects. All of these had chiasmal gliomas, and a precise correlation between tumor location and morphology of the visual field defects was found. Five patients ( cases 1, 2, 7, 8, and 10) also had color vision defects. Electrophysiological studies, CT, and MR were performed whenever possible, confirming and completing the clinical findings. Therapeutic approaches are shown in Table 4, and included periodic surveillance, refractive cor­rection, trabeculectomy for congenital glaucoma in case 11, ventriculo- peritoneal cerebrospinal fluid ( CSF) shunting in case 9 for hydrocephalus control and chiasmal glioma radiotherapy in cases 6 and 7. As expected, visual function recovery was not pos­sible in many cases [ eight ( 72.7%) patients] given the presence of optic nerve atrophy or severe mac­ular lesions. Table 5 shows the systemic findings in these 11 patients and includes relevant data concerning their families. / Neuro- Ophthalmol, Vol. 16, No. 4, 1996 VISUAL FUNCTION IN NEUROFIBROMATOSIS 237 TABLE 3. Visual dysfunction ( 11 patients) Total Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case 9 Case 10 Case 11 (%) Visual acuity decrease Nystagmus Strabismus Visual fields defects Color vision defects + + - + + + + - + + + - + - - + - + - - 8 ( 72.7) 3 ( 27.3) 2( 18.2) 4 ( 36.4) 5 ( 45.4) Cafe- au- lait spots were present in 63.6% of cases. Cutaneous neurofibromas occurred in 18.2% of patients. In case 4, which is a very interesting since the patient had no diagnostic criteria for either of the two types of NF, there were cutaneous plexiform schwannomas. Here, the diagnostic hypotheses of schwannomatosis, a new type of NF, or an incom­plete form of NF2 were all considered. This clinical case was presented as a poster during the Euro­pean Congress on Paediatric Ophthalmology and Orthoptics in Maastricht, October 1994. Association with other central nervous system ( CNS) changes occurred in 72.7% of patients. Al­though most of these were asymptomatic and needed no surgical treatment, in case 4, a large left hypoglossal nerve schwannoma, causing tongue hemiatrophy with fasciculations, was scheduled for surgical excision. In case 9, ataxia was the result of the presence of pontine and cerebellar hamarto­mas; obstruction of the foramen of Monro caused hydrocephalus requiring ventriculo- peritoneal CSF shunting. Bony abnormalities were present in two of the three cases of sphenoid dysplasia or hypoplasia already mentioned. Table 5 also shows the incidence of NF in these patients' relatives ( 45.4%). It was common to find more than one relative affected. The patients of cases 6 and 7 are sisters. A positive family history was found in 50 of the 75 ( 66.7%) patients reported here. DISCUSSION Ocular changes were found in 42 of 75 ( 56%) patients. Although most of these 75 patients were children ( 72%), the incidence of these ocular find­ings was greater among the adults ( 76.2%) than children ( 48.1%). This is not surprising consider­ing the age- related expression of NF. Although ocular changes in NF were frequent, visual dysfunction was present in only 14.7% of patients. Probably, the re- evaluation of those 14 patients who did not return for follow- up would lead to detection of more cases of visual dysfunc­tion. The absence of reports on this subject makes comparison with other studies impossible, but, in our opinion, we find this low incidence easy to explain, since, in most cases, Lisch nodules were the only ocular finding and they do not cause vi­sual dysfunction. We found Lisch nodules in 42.3% of patients while most authors report values of 75- 95% ( 1,2,7,8). Since Lisch nodules become more frequent with age, this low incidence proba­bly results from the young age of most of the chil­dren included in this group ( mean age of 6.5 years). Optic pathway gliomas, as opposed to Lisch nodules, were almost always associated with vi­sual dysfunction and were the main cause in the present study. Of the 10 patients with glioma, 75 cases, nine ( 90.0%) were symptomatic. This inci­dence of optic glioma in NF, 13.3%, is almost iden­tical to those reported by most authors, - 15% ( 1,2,9- 11). NF- associated optic pathway gliomas have a bet­ter prognosis than do isolated optic gliomas, so clinical surveillance was a common option in these 10 patients ( 1,9). Radiotherapy was performed when the chiasm was involved and tumor growth was confirmed by clinical evaluation and imaging. Results were good, with tumor mass regression and clinical improvement without significant com­plications after 1 year of clinical follow- up. Optic pathway glioma therapy is still controversial. Of­ten the best option is surveillance, but, when growth is evident or a clinical deterioration occurs, therapy is mandatory and may consist of surgical excision, radiotherapy, or chemotherapy. In chias­mal gliomas, the rate of complications associated with surgery is high, so that radiotherapy is pre­ferred or, according to some authors, chemother­apy. Age is a crucial factor in the therapeutic de­cision ( 9,12- 14). The presence of hypothalamic glioma in case 6 was also an important factor in choosing radiotherapy. In case 9, CSF ventriculo-peritoneal shunting was required for hydrocepha- / Neuro- Ophthalmol, Vol. 16, No. 4, 1996 238 A. CASTANHEIRA- DIN1S ET Ah. TABLE 4. Ocular findings, diagnosis and treatment Case 1 2 3 4 5 6 7 8 9 10 Sex F M F F M F F M F F Age 16 yr 18 yr 3yr 4 yr 8 mo 15 yr 17 yr 11 yr 7yr 50 yr Type 1 1 1? 2? 1 1 1 1? 2? 1 1 Ocular finding VOD = 20/ 70; VOS = 20/ 50 ( astigmatism) Pulsatile proptosis Horizontal nystagmus OPD pallor Chiasmal glioma VOD = 20/ 50; VOS = 20/ 200 ( high myopia) Horizontal nystagmus Lisch nodules Cortical lens opacities OPD pallor, myopic crescent OD ON and chiasmal glioma Stycar: OD = 20/ 30 OS = 20/ 30 Strasbismus OD ON glioma VOD = Fingers at 1 m; VOS = 20/ 20 Strabismus RPE hamartoma OD ON glioma Follows objects Nystagmus OD papilledema; OS OPD pallor ON, chiasma and optic tracts glioma VOD = 20/ 20; VOS = 20/ 25 ( myopia) OPD pallor, myopic retinal changes ON and chiasma glioma VOD = 20/ 20 VOS = 20/ 20 Lisch nodules OPD pallor Right hemichiasma and right optic tract glioma VOD = 20/ 40; VOS = 20/ 70 ( amblyopia OS, anisometropia) Lisch nodules OS lens opacities Macular lesion OS ( hamartoma?) Stycar: RE = 20/ 20? LE: 20/ 20 ( searching VF) ( Myopia) Lisch nodules OPD pallor ( + OD) OD intracranial ON and chiasma glioma VOD = 20/ 20 VOS = 20/ 20 POAG OD, OS OD ON glioma Complementary diagnostic procedures VEP p, abn conduction OD, OS VF, bitemporal hemianopia F100; R/ G VEP f, abn conduction OD VF, bitemporal hemianopia F100 OD, OS, R/ G; B/ Y VEP f, abn conduction OD VEP f, abn conduction OD ERG, subnormal Angiography; echography Lab, N; thorax Rx, N; family Obs., N; ENT, N VEP p, abn conduction OD; OS (> OS) VF, bitemporal hemianopia F100, N VEP p, abn conduction OD VF, left homonymous hemianopia F100, OD R/ G; B/ Y OS: R/ G VEP p, abn conduction OS VF, N F100: R/ G OS Not cooperative for AV ( Snellen), VEP, VF VF, N; AP: OD f blind spot, superior nasal step VEP p: abn conduction OD Imaging MR CT MR RM CT MR CT CT MR CT MR MR CT MR CT Treatment Refraction Follow- up Refraction Follow- up Follow- up Schwannoma Hypoglossal Excision Follow- up Follow- up Refraction Radiotherapy Radiotherapy Refraction VPS Refraction Follow- up 4yr VOD: 20/ 200; VOS: 20/ 40 ( astigmatism) OD OPD: abnormal cupping OD congenital glaucoma ( buphthalmos) F100: R/ G OD VEP f; abn. conduction OD; OS ERG: subnormal CT Trabeculectomy Refraction Abn, abnormal; AP, automated perimetry; B/ Y, blue/ yellow; ENT, ears, nose, and throat; F, female; M, male; N, normal; OD, optic disc; ON, optic nerve; POAG, primary open angle glaucoma; R/ G, red/ green; RPE, retinal pigment epithelium; VEP, visual evoked potentials ( p, pattern; f, flash); VF, visual fields; VOS, visual acuity left eye; VPS, ventriculo- peritoneal shunting; VOD, visual acuity right eye. lus control. Identical approaches have been recom­mended by other authors ( 14,15). Retinal hamartomas were present in the two cases of probable NF2, which is in accordance with the greater incidence of these lesions in that type of NF ( 16- 18). Lens opacities are described only in NF2, in which they may occur in > 75% of the cases ( 19- 21). Among these 75 patients, there was one child with lens opacities and poorly defined diagnostic criteria for NF. This child, reported as case 8, had Lisch nodules, a retinal hamartoma, cafe- au- lait spots, and deep brain gliomas. It is our experience that, in some cases, full expression of NF occurs later, making classification difficult or even impos­sible at any given moment. In case 2, diagnostic / Neuro- Ophthalmol, Vol. 16, No. 4, 1996 VISUAL FUNCTION IN NEUROFIBROMATOSIS 239 TABLE 5. Systemic findings and family history Case 1 2 3 4 5 6 7 8 9 10 11 Sex F M F F M F F M F F M Age 16 yr 18 yr 3yr 4 yr 8 mo 15 yr 17yr 11 yr 7 yr 50 yr 4 yr Type 1 1 1? 2? 1 1 1 1? 2? 1 1 1 Cutaneous findings Cafe- au- lait spots Neurofibromas Plexiform schwannomas Cafe- au- lait spots Cafe- au- lait spots Cafe- au- lait spots Cafe- au- lait spots Cafe- au- lait spots Facial cutaneous neurofibromas Cafe- au- lait spots CNS findings Temporal arachnoidal cyst Encephalic hamartomas Brain stem glioma infiltration Left hypoglossal schwannoma Right foramen magnus tumor Hypothalamic glioma Left deep temporal glioma Left rolandic hamartoma or glioma Right hemiparesis Hypothalamic glioma Monro foramen obstruction ( hydrocephalus and VPS) Cerebellum and pons hamartomas ( ataxia) Macrocephalus Bony findings Right sph greater wing Dysplasia Orbital deformation and facial asymmetry Right sph greater wing hypoplasia Family history Sister, mother, and grandmother ( maternal), NF1 Sister, mother, and aunt ( maternal), NF1 Grandmother ( paternal) with identical tumor Daughter, cafe- au- lait spots Mother, NF1 CNS, central nervous system; F, female; M, male; Sph, sphenoid; VPS, ventriculo peritoneal shunting. criteria for NF1 were present and anterior subcap­sular lens opacities were found. This clinical fea­ture, although rare, is described by some authors ( 2). The two patients with sphenoid dysplasia or hy­poplasia did not have visual dysfunction related to the finding. In case 1 a pulsatile proptosis was present. These bony defects are frequently de­scribed in association with NF1 ( 1,3,22). The third case of sphenoid dysplasia reported here, al­though not having visual dysfunction, deserves special mention since it represents the classical as­sociation, described by Walsh and Hoyt, of ab­sence of the greater wing, orbital encephalocele, and plexiform neurofibroma of the frontal region and superior lid ( 1). Visual field defects were found in four patients with chiasmal gliomas. In five patients, children = s5 years of age, visual field evaluation was not performed. Kinetic perimetry was performed in every patient aged 5= 5 years, although the patient of case 9 did not cooperate with this test. Color vision defects were present in five pa­tients. Color vision was evaluated in all patients 3= 4 years of age using the Farnsworth 100- Hue ( except case 9, who did not cooperate). Red/ green defects were, as expected, the most prevalent. In case 2, there was also a blue/ yellow defect. Retinal changes related to high myopia are the most prob­able explanation for that finding. As shown in Table 5, different ancillary exami­nation techniques were used in order to obtain a better definition of visual dysfunction. VEPs were performed in nine patients and correlated well with the presence of optic gliomas or, in case 11, of optic nerve glaucomatous damage. The patient re­ported in case 5 abandoned follow- up before a VEP could be done, and the patient in case 9 was not cooperative. CT and MR were essential for locating and quan- / Neuro- OphtMmol, Vol. 16, No. 4, 1996 240 A. CASTANHEIRA- DINIS ET AL. tifying optic and CNS lesions and in order to arrive at a therapeutic decision. CONCLUSIONS Considering the high prevalence of NF in the general population, visual dysfunction in NF has become a significant problem deserving greater at­tention. Optic pathway and retinal lesions are usually untreatable and associated with permanent dam­age. Thus, it is of primary importance that NF pa­tients and their relatives undergo a neuro-ophthalmological examination as soon as the diag­nosis of NF is established or suspected in order to widen the therapeutic options, exclude other le­sions, improve visual prognosis, or increase sur­vival time. Ocular changes are easier to detect in adults, in whom they occur more frequently and in whom cooperation is much better, the latter being the rea­son why ocular findings were more prevalent in the adult group. However, 11 patients had visual dysfunction and, of those, eight ( 72.7%) were chil­dren. A prompt diagnosis is, therefore, mandatory and only feasible through an efficient interdiscipli­nary approach. Pediatricians should refer all NF cases for a complete neuro- ophthalmological ex­amination. Ophthalmologists must consider NF a possible cause of visual dysfunction and must rec­ognize the ocular changes that can cause visual dysfunction in NF. One must also emphasize the high frequency ( 73.8%) of asymptomatic ocular findings and high­light the importance of always complementing clinical assessment with other examination tech­niques, particularly neuroimaging, so that all ab­normalities may be recognized. REFERENCES 1. The Phakomatoses. In: Miller NR, ed. Walsh and Hoyt's clin­ical neuro- ophthalmology. Baltimore: Williams and Wilkins, 1988: 1747- 65. 2. Ragge NK. Clinical and genetic patterns of neurofibroma­tosis 1 and 2. Br } Ophthalmol 1993; 77: 662- 72. 3. Abnormalities affecting the entire eye and the phakoma­toses. In: Catalano RA, Nelson LB, eds. Pediatric ophthal­mology. A text atlas. Norwalk: Appleton and Lange, 1994: 57- 74. 4. Sieb JP, Schultheiss R. Segmental neurofibromatosis of the sciatic nerve: case report. Neurosurgery 1992; 31: 1122- 5. 5. Purcell SM. Schwannomatosis. An unusual variant of neu­rofibromatosis or a distinct clinical entity? Arch Dermatol 1989; 125: 390- 3. 6. Mulvihill MJH, Parry DM, Sherman JI, Pikus A, Kaiser- Kupfer MI, Eldridge R. Neurofibromatosis 1 ( Recklinghausen disease) and neurofibromatosis 2 ( bilateral acoustic neurofibro­matosis). An update. ( NIH Conference). Ann Intern Med 1990; 113: 39- 52. 7. Zehavi C, Romano A, Goodman RM. Iris ( Lisch) nodules in neurofibromatosis. Clin Gen 1986; 29: 51- 5. 8. Lubs M- LE, Bauer MS, Formas ME, Djokic B. Lisch nodules in neurofibromatosis type 1. N Engl ] Med 1991; 324: 1264- 6. 9. Wilson WB. Optic nerve gliomas: Treatment differences for the benign and malignant varieties. In: Tusa RJ, Newman SA, eds. Neuro- ophthalmological disorders: Diagnostic work- up and management. New York: Marcel Dekker, 1995: 163- 72. 10. Lund AM, Skovby F. Optic gliomas in children with neu­rofibromatosis type 1. Eur J Pediatr 1991; 150: 835- 8. 11. Matsui I, Tanimura M, Kobayashi N, Sawada T, Nagahara N, Akatsuka J- I. Neurofibromatosis type 1 and childhood cancer. Cancer 1993; 72: 2746- 54. 12. Cohen BH, Kaplan AM, Packer RJ. Management of intra­cranial neoplasms in children with neurofibromatosis type 1 and 2. Pediatr Neurosurg 1990- 91; 16: 66- 72. 13. Pierce SM, Barnes PD, Loeffler JS, McGinn C, Tarbell NJ. Definitive radiation therapy in the management of symp­tomatic patients with optic glioma. Cancer 1990; 65: 45- 52. 14. Petronio J, Edwards MSB, Prados M et al. Management of chiasmal and hypothalamic gliomas of infancy and child­hood with chemotherapy. / Neurosurg 1991; 74: 701- 8. 15. Rodriguez LA, Edwards MSB, Levin VA. Management of hypothalamic gliomas in children: an analysis of 33 cases. Neurosurgery 1990; 26: 242- 7. 16. Destro M, D'Amico DJ, Gragoudas ES et al. Retinal mani­festations of neurofibromatosis. Arch Ophthalmol 1991; 109: 662- 6. 17. Landau K, Yasargil GM. Ocular fundus in neurofibromato­sis type 2. Br J Ophthalmol 1993; 77: 646- 9. 18. Tonsgard JH, Carolyn SO. The ophthalmologic presenta­tion of NF- 2 in childhood. / Pediatr Ophthalmol Strab 1993; 30: 327- 30. 19. Bouzas EA, Freidlin V, Parry DM, Eldridge R, Kaiser- Kupfer MI. Lens opacities in neurofibromatosis 2: further significant correlations. Br J Ophthalmol 1993; 77: 354- 7. 20. Bouzas EA, Parry DM, Eldridge R, Kaiser- Kupfer MI. Vi­sual impairment in patients with neurofibromatosis 2. Neu­rology 1993; 43: 622- 3. 21. Evans DGR, Huson SM, Donnai D et al. A clinical study of type 2 neurofibromatosis. Quart } Med 1992; 304: 603- 18. 22. Smith JL, Bowen BC. Ocular pulsation in neurofibromato­sis. / Clin Neuro- Ophthalmol 1993; 13: 163- 70. / Neuro- Ophthalmol, Vol. 16, No. 4, 1996