impairment - bmj

42ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:492-495

MRI in neurofibromatosis 1. The nature andevolution of increased intensity T2 weighted-lesions and their relationship to intellectualimpairment

R E Ferner, R Chaudhuri, J Bingham, T Cox, R A C Hughes

AbstractThirty eight patients with neuro-fibromatosis 1 (NFI) had neurologicalexaminations, intellectual assessmentsand MRI scans. Increased intensitylesions on T2 weighted images werefound in 13 patients. These abnormalitieswere more common in patients agedunder 18 years. The lesions occurredpredominantly in the basal ganglia,brainstem and cerebellum, and weremultiple in 11 patients. They did not pro-duce symptoms or neurological deficit inany patient and did not enhance withgadolinium-meglumine-triamine-pen-tacetic acid contrast medium (Gd-DTPA). In 2 patients, however, theabnormalities exerted mass effect dis-torting the brain and in 3 patients theyoccurred in conjunction with knowngliomas. The lesions remainedunchanged over a three year follow upperiod. The nature of the lesions isuncertain but the fact that they may pro-duce mass effect and occur in associationwith gliomas suggests that they havemalignant potential. There was no corre-lation between the presence of theseabnormalities and intellectual impair-ment.

(7 Neurol Neurosurg Psychiatry 1993;56:492-495)

UMDS, Guy'sHospital, London, UKDepartment ofNeurologyRE FemerR A C HughesDivision ofRadiological SciencesR ChaudhuriJ BinghamT CoxCorrespondence to:Dr Femer, Department ofNeurology, Guy's Hospital,St Thomas's Street, LondonSEI 9RT, UKReceived 19 February 1992

and in revised form23 June 1992.

Accepted 27 July 1992

Neurofibromatosis 1 is a common autosomaldominant disease with a minimum prevalenceof 1 in 5000.' Diagnostic criteria formulatedby the National Institutes of HealthConsensus Development Conference2 are setout in table 1. The complications of the dis-order are legion and neurological manifesta-tions range from gliomas to intellectualimpairment. The majority of patients have an

IQ in the low average range. Specific learningproblems such as attention deficit, visual-spa-tial problems and reading difficulties, occur inbetween 40-60% of suffers.34

Table 1 Diagnostic features of neurofibromatosis 12

1 Cafe au lait spots2 Two or more neurofibromas or one plexiform neurofibroma3 Axillary or groin freckling4 Lisch nodules5 Optic nerve glioma6 A first degree relative with NF 17 A distinctive osseous lesion such as sphenoid wing dysplasia

or thinning of the long bone cortex with or withoutpseudarthrosis

2 or more criteria are required for diagnosis.

The cause of the intellectual difficulties hasnot been established. However, Rosman andPearce5 showed in a post mortem study thatpatients with NF1 and intellectual impair-ment had heterotopias in the subcortical anddeep white matter and disordered cerebralarchitecture. Patients with NFI and averageintelligence had similar less severe changesbut these abnormalities were absent in con-trol subjects. They proposed that the intellec-tual problems were linked to migrationalabnormalities in the brain of the developingfetus.MRI is now the investigation of choice for

visualising the distinction between grey andwhite matter in the brain. Increased intensitylesions have been detected on T2 weightedMRI brain scans in children and young adultswith NFl. They occur predominantly in thebasal ganglia, brainstem and cerebellum.67-9They do not show mass effect and do notenhance with Gd-DTPA. The nature of theabnormalities is uncertain, but it has beensuggested that they may be hamartomas orslow growing gliomas. The lesions are distin-guishable from those in multiple sclerosis andvascular disease by their larger size and differ-ent location. Another possibility is that theselesions represent grey matter heterotopias.Mirowitz et al observed that 7 of 35 patientswith NFl had increased signal intensitylesions in the basal ganglia which were moreprominent on TI than on T2 weighted MRI.They suggested that the signal characteristicsand morphology were compatible with greymatter heterotopias containing Schwann cellsand/or myelin.'0 Sevick et al reported that thelesions increase in frequency in early child-hood but tend to resolve with increasing age.9Two recent studies have attempted to cor-

relate the presence of these lesions with learn-ing problems in children with NFl. Duffneret aft found increased intensity T2 weightedlesions in 29 of 47 children and Dunn andRoos" detected these abnormalities in 16 of31 children. There was, however, no correla-tion between the presence of increased inten-sity lesions on MRI and learning difficultiesin the children studied.As part of a larger study on intellectual

impairment in NFl, we performed brainMRI on 38 adults and children with NFI.We evaluated the scans for the occurrence ofgeneral neuroradiological abnormalities andincreased intensity T2 weighted lesions. Wedetermined whether the presence of these T2weighted abnormalities was age related andscans were performed at intervals to assess

492 on June 5, 2022 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.5.492 on 1 M

ay 1993. Dow

nloaded from

http://jnnp.bmj.com/

MRI in neurofibromatosis 1

the evolution of the lesions. We also consid-ered whether intellectual impairment is relat-ed to the presence of T2 weightedabnormalities in this patient group.

MethodsWe are reporting all MRI head scans carriedout on 38 patients with NF1 in our hospitalbetween 1987-91 are reported. These includ-ed 32 patients from a study of 100 patientsattending our neurofibromatosis clinic andtaking part in a study of intellectual impair-ment. The subjects included 16 children andtheir ages ranged from 3-63 years (mean 25years). Neurological and psychometric assess-ment.

All patients had general medical and neu-rological examinations. Twenty eight patientshad formal IQ testing with the Wechsler IQscales.12 13 Estimates of intellectual abilitywere based on school performance and occu-pational history in the remaining 10 subjects.

MRIMRI brain scans were carried out with a 1-5Tesla superconducting system (PhilipsGyroscan S15, Philips Medical Systems).Images were taken in the axial, coronal andsagittal planes, with a slice thickness of 5 mmto 8 mm and an interscan distance of 0-5 mmto 0-8 mm. Ti weighted Spin Echo imageswere obtained with a Repetition Time (TR)of 500 ms and Echo Time (TE) of 20 ms. Tiweighted scans were performed following theadministration of Gd-DTPA paramagneticcontrast medium in 16 cases. Initially TIInversion Recovery sequences were used onthe first 19 patients with a TR 1500 ms, TE30 ms and TI 300 ms to look for grey matterheterotopias. T2 weighted Spin Echo imageswere carried out using TR of 2000 ms to3000ms and TE 50/100 ms to 20/120 ms.Follow up scans were available in tenpatients. A Chi square test was used to assessthe relationship between the presence ofincreased intensity T2 weighted abnormalitieson the scans and intellectual impairment.

Post mortem studiesA neuropathological study of two brains wascarried out. The post mortem brains wereboth immersion fixed and sliced. The firstwas sliced in the coronal plane and the sec-ond case was sliced in the sagittal plane tocorrespond with images taken on MRI.

StatisticsDifferences between proportions were testedwith Chi squared or Fisher's exact tests andtwo-tailed probabilities are quoted.

ResultsAbnormalities on MRI head scans were pre-sent in 28 of 38 patients (table 2). No greymatter heterotopias were found in any of the19 patients on whom Tl Inversion Recoverysequences were used.

Increased intensity lesions on T2 weighted

Table 2 Abnormalities on MRI head scans in 38 patientswith NFl

Known associatioms Other associations

Increased intensity T2 13 Multiple sclerosis 2weighted lesionsOptic nerve glioma 6 Vascular disease 2Parenchymal glioma 3 Cerebral atrophy 1Dilated ventricles 4 Chromophobe adenoma 1Increased intensity Ti 1 Posterior fossa 1weighted lesions arachnoid cyst

Meningeal dysplasia 1Venous malformation 1

arising from corpusstriatum vessels

Total patients affected 24 Total patients affected 10

Total number of patients with abnormal scans = 28normal scans = 10

Note: Some patients had more than one abnormality

Table 3 Sites of increased intensity lesions on T2weighted scans

Sites Number ofpatients

Basal ganglia 11Cerebellum 4Brainstem 3Lateral ventricle wall 2Cerebral peduncle 2Corpus callosum 2Frontal lobe 1Parietal lobe 1Internal capsule 1Optic tract 1Insular cortex 1

images were noted in 13 patients. In 3 casesthe lesions were single and were found in thebasal ganglia, cerebellum and deep whitematter adjacent to the insular cortex. Tenpatients had multiple lesions which occurredpredominantly in the basal ganglia, particu-larly the globus pallidus (table 3, fig 1). Thelesions did not cause symptoms or neurologi-cal deficit in any patient and did not enhancewith Gd-DTPA. However, in 2 cases thelesions produced mass effect and distorted

Figure 1 Axial 72 weighted image (TRSE 2000, TE80) in an asymptomatic patient with NFl. There arebilateral increased intensity lesions in the globus pallidusmore marked on the right. They did not enhance afterintravenous injection ofGd-DTPA.



j.com/

J Neurol N


ay 1993. Dow

nloaded from


Ferner, Chaudhuri, Bingham, Cox, Hughes

Figure 2 Axial T2weighted image throughthe brainstem (TRSE2000, TE 80) in anasymptomatic 17year oldboy with NFl. There is anincreased intensity lesionon the left side of the ponsand cerebellum which isdistorting the fourthventricle and deviating itto the right. The lesion didnot enhance with Gd-DTPA and has remainedunchangedfor 3 years.

the brain. A 17 year old man had a lesion onthe left side of the pons which distorted thefourth ventricle (fig 2). An eight year old girlwas found to have a lesion in the brainstemwhich produced mass effect but did notevolve over a 2 year period (fig 3). Six otherpatients had interval scans over the same

period which did not show changes in thelesions. One patient had asymptomaticincreased intensity lesions in the basal gangliaand cerebellum in addition to a known brain-stem glioma. Two patients had asymptomaticlesions in the cerebellum, basal ganglia, andinternal capsule in association with an opticnerve glioma. Although follow up scansshowed an increase in the size of the gliomas

Figure 3 Axial T2weighted image throughthe brainstem (TRSE2000, TE 80) in anasymptomatic eightyearold girl with NFl. There isan extensive area ofincreased signal in the ponsand middle cerebellarpeduncles especially theleft. The lesion is causingswelling and distortion ofthe fourth ventricle anddisplacing it to the right.The lesion did not enhancewith Gd-DTPA andremained unchangedfor 2years.

the other lesions remained unchanged over a3 year period. Eight of 16 children hadincreased intensity signals on T2 weightedMRI scans compared with 5 of 22 adults(p = 0d16).Four patients were mentally retarded with

a full scale IQ range between 51 and 67. Oneof these subjects had a single increased inten-sity T2 weighted lesion in the region of theright insular cortex as described above.Nineteen patients had an IQ in the range70-89. Increased intensity lesions on T2weighted images were seen in 8 of this groupand were found predominantly in the basalganglia, cerebellum and brainstem. Thirteenpatients had an IQ range 91-112. Six hadincreased intensity T2 weighted lesions in thebasal ganglia, cerebellum, cerebellar pedun-cle, and optic tract. There was therefore nosignificant association between the presenceof T2 weighted lesions and intellectualimpairment (p = 0 79).Twenty five of 38 patients had macro-

cephaly defined as a head circumferencegreater than the 98th centile. Increased inten-sity T2 weighted lesions were present in 10 ofthese patients with macrocephaly and in 3 ofthe 13 with normal head circumferences.Among the 25 patients without T2 weightedabnormalities, 15 had macrocephaly and 10had a normal head circumference. There wasno statistically significant relationshipbetween the presence of increased intensitylesions and macrocephaly.

Post mortem studies were carried out on 2patients, a 59 year old male whose IQ hadbeen 60 and a 61 year old female who hadbeen a lawyer. In the first case a T2 weightedMRI scan had shown multiple high intensitylesions in the brainstem, basal ganglia andperiventricular regions, which were consid-ered to be infarcts, a diagnosis which wasconfirmed at post mortem. The MRI of thesecond patient showed multiple high signallesions on T2 weighted images which wereseen in the periventricular areas, pons andcerebellum. The lesions and clinical historywere consistent with multiple sclerosis whichwas verified by the neuropathological exami-nation. No evidence of heterotopias, gliomasor hamartomas was found in either patient.

DiscussionThe most significant findings were high signalintensity lesions in the basal ganglia, brain-stem and cerebellum on T2 weighted MRIscans in 8 of a selected population of 16 chil-dren with NF1, in agreement with previousstudies.69 All the lesions in our seriesappeared to be asymptomatic. This fact, thelack of enhancement following Gd-DTPA,and the observation that similar lesions areless common in adults (only being present in5 of 22 in our series) would suggest that thelesions are benign and might be hamartoma-tous. However, in 2 of our cases the lesionsproduced mass effect and in another 3 casesthe lesions were associated with gliomas else-where in the brain which suggests that the



j.com/

J Neurol N


ay 1993. Dow

nloaded from


MRI in neurofibromatosis 1

lesions may have malignant potential. It ispossible there are two different types oflesion, one which will regress and one whichmay develop into a glioma. Further follow upand rescanning of our patients should resolvethis question. In the meantime we do notadvocate that patients should be screened forthe presence of these lesions as all of ourpatients have remained asymptomatic.Patients need to be scanned only in accor-dance with clinical indications. When theselesions are encountered in NF1, patients canbe reassured that they are usually benign, butfollow up is necessary. The nature of the highsignal intensity lesions on T2 weighted MRIscans might be resolved by appropriate postmortem studies.We also observed the usual range of abnor-

malities associated with NF1. The high fre-quency of optic nerve gliomas in this studycan be attributed to the selection of cases byreferral to our neurologically orientatedNeurofibromatosis Clinic. However opticnerve gliomas are a recognised associationwith NF1 and were reported in 15% of 217cases from a Neurofibromatosis Clinic with amore general orientation.'4 These optic nervegliomas may be asymptomatic, as in 2 of our6 cases, but may also cause visual failure andhydrocephalus. The three parenchymalgliomas in our series were located in thebrainstem and thalamus: the possibility arisesthat they had originated in the optic tract andthat the glial tissue of the anterior and poste-rior optic pathways has a particular potentialfor gliomatous change. Four patients haddilated ventricles and these were due to aque-duct stenosis, a tumour impinging on theforamen of Munro, cerebral atrophy and idio-pathic ventriculomegaly. We detected occa-sional examples of other abnormalities withNF1 which have not been previously associat-ed and may be coincidental (table 3).Macrocephaly is recognised as a feature ofNFP4 but it did not correlate with the pres-ence of abnormalities found on T2 weightedMRI scans in our series, and it is not associat-ed with intellectual impairment in NFl .4No differences were discovered between

the MRI scans of those patients who werementally retarded and those who were not,although patients with parenchymal gliomasshowed the deterioration in cerebral functionanticipated from the sites of their lesions. Apost mortem study had shown that grey mat-ter heterotopia was common in the brains of

patients with NFl.5 Grey matter heterotopiashave been demonstrated in the non-NF1population with MRI scans.'5 We did notidentify any grey matter heterotopias duringthe study period. However, since then wehave identified bilateral Ti weighted lesionsin the basal ganglia, similar to those describedby Mirowitz et al,'0 in one patient with a lowaverage IQ. We were unable to find a correla-tion between the presence, number, size andsites of T2 weighted abnormalities and theoccurrence of mental retardation in childrenor adults. The cause of mental retardation inNFI deserves further investigation whichmight profitably be pursued with studies ofcerebral metabolism by positron emissiontomography.We thank Dr I Janota for the post mortem examinations, MrsPhilippa Graves for performing the MRI scans, and ProfessorJ Weinman for advice about the psychometric studies.

1 Huson S, Harper PS, Compston DAS. VonRecklinghausen neurofibromatosis: a clinical and popu-lation study in South East Wales. Brain 1988;111,1355-81.

2 National Institutes of Health Consensus DevelopmentConference Statement on Neurofibromatosis.Neurofibromatosis Res Newsletter 1987;3: 1;3-6.

3 Aron AM, Rubenstein AE, Wallace SA, Halperin JC.Learning disabilities in neurofibromatosis. InRubenstein AE, Korf BR, eds. Neurofibromatosis: ahandbook for patients, families, and health-care profes-sionals. New York: Thieme Medical, 1990:55-58.

4 Ferner RE, Intellectual problems in Neurofibromatosis 1.In: Huson SM, Hughes RAC, eds. Neurofibromatosis.Chapman and Hall (in press).

5 Rosman NP, Pearce J. The brain in multiple neurofibro-matosis (von Recklinghausen's disease): a suggestedneuropathological basis for the associated mental defect.Brain 1967;90:829-38.

6 Bognanno JR, Edwards MK, Lee TA, Dunn DW, RoosKL, Klatte EC. Cranial MR imaging in neurofibro-matosis. AmJ Radiol 1988;151:381-8.

7 Duffner PK, Cohen ME, Seidel FG, Shucard DW. Thesignificance of MRI abnormalities in children with neu-rofibromatosis. Neurology 1989;39:373-8.

8 Goldstein SM, Curless MJ, Donovan Post, Quencer RM.A new sign of neurofibromatosis on magnetic resonanceimaging of children. Arch Neurol 1989;46: 1222-4.

9 Sevick RJ, Barkovich AJ, Edwards MS, Koch TK,Lempert T, Norman D. Neurofibromatosis type 1: tem-poral evolution of white matter lesions. Radiology 1991,8(P), 193.

10 Mirowitz SA, Sartor K, Gado M. High-intensity basalganglia lesions on TI-weighted MR images in neurofi-bromatosis. Am Y Radiol 1990;154:369-73.

11 Dunn DW, Roos KL. Magnetic resonance imaging evalu-ation of learning difficulties in neurofibromatosis.Neurofibromatosis 1989;2:1-5.

12 Wechsler, D. Wechsler Intelligence Scale for Children-Revised. New York: The Psychological Corporation,1976.

13 Wechsler, D. Wechsler Adult Intelligence Scale-Revised.New York: The Psychological Corporation, 1981.

14 Lewis RA, Gerson LP, Axelson KA, Riccardi VM,Whitford RP. Von Recklinghausen neurofibromatosis,incidence of optic gliomata. Ophthalmology1984;91:929-35.

15 Barkovich AJ, Chuang SH, Norman D. MR of neuronalmigration anomalies. Am Y Radiol 1988;150: 179-87.



j.com/

J Neurol N


ay 1993. Dow

nloaded from


impairment - bmj

Documents