5
Journal of Neurology, Neurosurgery, and Psychiatry, 1977, 40, 209-213 Impact of computerised axial tomography on the management of posterior fossa tumours in childhood E. BOLTSHAUSER', H. HAMALATHA, D. N. GRANT, AND K. TILL From the Department of Neurosurgery, The Hospital for Sick Children, Great Ormond Street, London SUMMARY Computerised axial tomiography (CAT) has profoundly altered the management of most children with posterior fossa tumours. Fifty such children were operated on from October 1973 to December 1975, 20 of whom were explored after investigation by CAT only. Most recent experience suggests that CAT need be the only investigation in the majority of children suspected of having an expanding lesion in the posterior fossa. 4! Computerised axial tomography (CAT) is a rela- tively new method of radiological diagnosis which has the particular advantages of being rapid, non- invasive, and free from hazards or discomfort to the patient. However, heavy sedation or general anaesthesia, which might be necessary in investi- gating non-cooperative children, and the regular use of contrast enhancement carry a small but defined risk (New and Scott, 1975). The-reliability of CAT in demonstrating a wide range of intra- cranial disorders is well established (Baker et al., 1974; Gawler et al., 1974; Paxton anrd Ambrose, 1974; Ambrose et al., 1975; New et al., 1975) and it is not our intention to compare the CAT results with other invasive neuroradiological procedures. Already in many neurosurgical departments where CAT is available, angiography, ventriculography, and encephalography are far less often performed (Baker, 1975; Epstein et al., 1976). An increasing number of patients are operated upon without con- trast radiology, particularly patients with head injury and intracranial haematoma (Baker et al., 1974; Gawler et al., 1975), avoiding in the remain- ing patients unnecessary and possibly harmful intervention. In evaluating CAT findings in the posterior fossa, some important points which have been clearly emphasised and illistraed by New and Scott (1975) should be kept in mind. Small fourth ventricles are often difficult or impossible to identify, and the lateral recesses are rarely demonstrated clearly by CAT. The normal aqueduct is usually not visible. The, cerebdllar vermis frequently appears denser in comparison 'Address for correspondence: E. B., Children's Hospital, Steinwiess- trasse 75, CH- 8032 Zurich, Switzerland. Accepted 21 October 1976 with the hemispheres and this should not be misinterpreted as a midline cerebellar lesion of increased density. The cerebellar grey matter frequently is of greater density compared with white matter. Occasionally,; ithe vein of Galen and/or parts of the straight sinus are visible on plain CAT. After injection of contrast medium these vascular structures are seen more frequently, as are sometimes dural sinuses of the falx and tentorium. Head movement often produces multiple artefacts, particularly in the basal CAT sections, making interpretation difficult, or even impossible. A tran'Verse dark band of factitiously low absorption across the pons between the dense bone of the anterior portions of the petrous pyramids is a commonly recurring artefact (New and Scott, 1975). The purpose of this paper is to analyse the im- pact of CAT on the management of children with posterior fossa neoplasms. Case material From October 1973 to December 1975 20 children with symptoms and signs of an infratentorial tumour have undergone posterior fossa explora- tion after investigation only by plain skull radio- graphs and CAT. There were nine boys and 11 girls aged from 3 years and 3 months to 13 years. CAT was performed with an EMI scanner at the National Hospital for Nervous Diseases, Queen Square; Attention was paid to obtaining appro- priate sections through. the posterior fossa by centering the x-ray beam approximately 3 cm above the orbitomeatal line with the patient's head flexed 15 or 20 degrees (Gawler et al., 1975). 209 Protected by copyright. on May 31, 2020 by guest. http://jnnp.bmj.com/ J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. Downloaded from

Impact of computerised tomography on the management of … · Computerised axial tomography (CAT) is a rela-tively newmethodof radiological diagnosis which has the particular advantages

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Journal ofNeurology, Neurosurgery, and Psychiatry, 1977, 40, 209-213

Impact of computerised axial tomography on themanagement of posterior fossa tumours in childhoodE. BOLTSHAUSER', H. HAMALATHA, D. N. GRANT,AND K. TILL

From the Department of Neurosurgery, The Hospital for Sick Children, Great Ormond Street, London

SUMMARY Computerised axial tomiography (CAT) has profoundly altered the managementof most children with posterior fossa tumours. Fifty such children were operated on fromOctober 1973 to December 1975, 20 of whom were explored after investigation by CAT only.Most recent experience suggests that CAT need be the only investigation in the majority ofchildren suspected of having an expanding lesion in the posterior fossa.

4!

Computerised axial tomography (CAT) is a rela-tively new method of radiological diagnosis whichhas the particular advantages of being rapid, non-invasive, and free from hazards or discomfort tothe patient. However, heavy sedation or generalanaesthesia, which might be necessary in investi-gating non-cooperative children, and the regularuse of contrast enhancement carry a small butdefined risk (New and Scott, 1975). The-reliabilityof CAT in demonstrating a wide range of intra-cranial disorders is well established (Baker et al.,1974; Gawler et al., 1974; Paxton anrd Ambrose,1974; Ambrose et al., 1975; New et al., 1975) andit is not our intention to compare the CAT resultswith other invasive neuroradiological procedures.Already in many neurosurgical departments whereCAT is available, angiography, ventriculography,and encephalography are far less often performed(Baker, 1975; Epstein et al., 1976). An increasingnumber of patients are operated upon without con-trast radiology, particularly patients with headinjury and intracranial haematoma (Baker et al.,1974; Gawler et al., 1975), avoiding in the remain-ing patients unnecessary and possibly harmfulintervention. In evaluating CAT findings in theposterior fossa, some important points which havebeen clearly emphasised and illistraed by Newand Scott (1975) should be kept in mind. Smallfourth ventricles are often difficult or impossibleto identify, and the lateral recesses are rarelydemonstrated clearly by CAT. The normalaqueduct is usually not visible. The, cerebdllarvermis frequently appears denser in comparison'Address for correspondence: E. B., Children's Hospital, Steinwiess-trasse 75, CH- 8032 Zurich, Switzerland.Accepted 21 October 1976

with the hemispheres and this should not bemisinterpreted as a midline cerebellar lesion ofincreased density. The cerebellar grey matterfrequently is of greater density compared withwhite matter. Occasionally,; ithe vein of Galenand/or parts of the straight sinus are visible onplain CAT. After injection of contrast mediumthese vascular structures are seen more frequently,as are sometimes dural sinuses of the falx andtentorium. Head movement often producesmultiple artefacts, particularly in the basal CATsections, making interpretation difficult, or evenimpossible. A tran'Verse dark band of factitiouslylow absorption across the pons between the densebone of the anterior portions of the petrouspyramids is a commonly recurring artefact (Newand Scott, 1975).The purpose of this paper is to analyse the im-

pact of CAT on the management of children withposterior fossa neoplasms.

Case material

From October 1973 to December 1975 20 childrenwith symptoms and signs of an infratentorialtumour have undergone posterior fossa explora-tion after investigation only by plain skull radio-graphs and CAT. There were nine boys and 11girls aged from 3 years and 3 months to 13 years.CAT was performed with an EMI scanner at theNational Hospital for Nervous Diseases, QueenSquare; Attention was paid to obtaining appro-priate sections through. the posterior fossa bycentering the x-ray beam approximately 3 cmabove the orbitomeatal line with the patient'shead flexed 15 or 20 degrees (Gawler et al., 1975).

209

Protected by copyright.

on May 31, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. D

ownloaded from

E Boltshauser, H. Hamalatha, D. N. Grant, and K. Till

The first two children of this series were investi-gated with the 80X80 matrix, the remainder withthe 160X 160 matrix. Younger patients were oftensedated to reduce movements artefacts to a

minimum. Contrast medium (sodium iothala-mate, Conray 420) was injected in most of thechildren in order to enhance tissue density(Ambrose et al., 1975). Unfortunately, in many

children movement artefact was increased afterthe injection. The interpretation and reports ofthe EMI scans were based on the appearance ofthe Polaroid photographs of the cathode ray tubedisplay. The digital computer printout of theabsorption coefficients was not analysed routinelyand was often not available at the time of opera-tion. For comparison the total number ofposterior fossa explorations with their preopera-

tive investigations, during the period under review,are given in Table 1. The children with un-

explored brain stem tumours are listed in Table 2.

Table 1 Posterior fossa explorations for tumoursand preoperative investigations

No. Pre-operative investigations

CATonly CATplus No CATother neuro-radiol. inv.

1973 7 1 1 5

(Oct.-Dec.)1974 19 2 2 151975 24 17 2 5

Total 50 20 5 25(100%) (40%) (10%) (50%)

Table 2 Brain stem tumours (not explored)

Number

1973 (Oct.-Dec.) 11974 5

1975 7

Total 13

During the same period 13 posterior fossa explora-tions for non-neoplastic lesions were carried out.The limited availability of the scanner in theadjacent hospital was the reason for not examin-ing all suitable children by this method.

Restilts

The 20 children investigated pre-operatively byEMI scan alone were studied by comparing theCAT appearances with the findings recorded inthe description of the operation. It was found thatCAT and the findings at operation correlated well

in 19 cases as far as localisation and extent of thespace occupying lesion were concerned. (The dis-crepancy in the remaining case is discussed below).The location of the tumours is summarised inTable 3. Clinical signs suggesting midline or

Table 3 Localisation of posterior fossa tumoursinvestigated preoperatively by CAT only

Number

Cerebellar vermis 5Cerebellar hemisphere 8Fourth ventricle (mainly) 5Brain stem 1Total 19*

*Because a technically unsatisfactory EMI scan was misinterpreted anegative posterior fossa exploration was performed in a child in whomsubsequently a suprasellar tumour was found.

mainly unilateral lesions correlated closely withthe CAT findings in 15 children. In three casesclinical examination suggested a midline lesion,whereas both CAT and operation revealed alargely right cerebellar mass, just crossing themidline. In one patient a left cerebellar tumourwas suspected on clinical grounds but CAT sug-gested a mass within the fourth ventricle whichwas confirmed by operative exploration. In all theEMI scans, mild to marked dilatation of the thirdand lateral ventricles was seen.A cystic cavity was found at operation in eight

cases, having been suspected in seven of these onthe scan. In two further cases a cyst was suspectedon CAT but not confirmed at operation.

Particularly important is a case in whom atechnically unsatisfactory EMI scan was misin-terpreted. The child had a history of vomitingand headache and was found to have bilateralpapilloedema with gross gait and limb ataxia. TheEMI scan showed marked ventricular dilatationbut movement artefacts prevented detailedrecognition of abnormality in the posterior fossa.A possible high density lesion in the vermis wasinterpreted as a tumour but exploration failed toreveal an abnormality in the posterior fossa.Scanning after operation and further conventionalneuroradiological investigations showed the boyto have a small suprasellar mass, presumably acraniopharyngioma.The findings on plain skull radiographs are

shown in Table 4.The CAT appearances before and after in-

jection of contrast medium are compared withthe histological findings in Table 5, in which fourother cases investigated by CAT and other meanshave been included. The lesions have been classi-

210

Protected by copyright.

on May 31, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. D

ownloaded from

Impact of computerised axial tomography

Table 4 Interpretation of plain skull radiographs of20 cases with suspected posterior fossa tumoursinvestigated by CAT only

Number

Normal 3Signs of increased intracranial pressure 12Unilateral thinning of floor of posterior fossa 3Midline post. fossa calcification, signs of increased intra-

cranial pressure 1Not done 1

Table 5 Posterior fossa tumours: comparison ofhistology with density and contrast enhancement onCAT

Histology Density of Contrastlesion enhancement

Astrocytoma grade 1 low +Astrocytoma grade I low +Astrocytoma grade 1 lowAstrocytoma grade I low +Astrocytoma grade 1 low +Astrocytoma grade 1 low NEAstrocytoma grade 1 high +

(haemorrhagic)Astrocytoma grade 2 low +Astrocytoma grade 2 lowAstrocytoma grade 2 low +Astrocytoma grade 2 lowAstrocytoma grade 3 low +Astrocytoma grade 4 high +

(haemorrhagic)Medulloblastoma high +Medulloblastoma high +Medulloblastoma high +Medulloblastoma highMedulloblastoma high +Medulloblastoma, cystic low +Ependymoma high NEEpendymoma high +Ependymoma high NEEpendymoma high NE

+ contrast enhanced.- not contrast enhanced.NE not examined.

fied as either of high or low density in comparisonwith normal surrounding brain tissue as seen onthe Polaroid photograph. It is again emphasisedthat the digital printout of the absorptioncoefficients had not been analysed routinely beforeoperation. However, it is noteworthy that astro-cytomas as a rule were of low radio density (Fig.1). There was a uniformly low density area evidentin nine proven astrocytomas on plain CAT, whilein two cases a rim and a mural nodule respectivelyof higher density were adjacent to the low densityzone. Contrast enhancement was positive in mostastrocytomas (Fig. 2), demonstrating a solidnodule in three additional cases. All ependymomashad generally high absorption values (Fig. 3)although they were not always of uniform highdensity. As these tumours have a tendency tocalcify, measurement of the absorption values or

Fig. 1 A typical unenhanced scan from a patientwith a cerebellar astrocytoma showing an extensivelow density area in the right half of the posteriorfossa encroaching on the mid-brain.

Fig. 2 The same patient showing markedenhancement following injection of sodiumiothalamate.

analysis of the digital printout might be helpful insuggesting this diagnosis. Medulloblastomas weregenerally of high density; in two cases a 'mottled'(predominantly) increased and decreased patternwas apparent. Only one exception was encoun-

211

Protected by copyright.

on May 31, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. D

ownloaded from

E Boltshauser, H. Hamalatha, D. N. Grant, and K. Till

Fig. 3 Unenhanced scan showing a high densitylesion (an ependymoma) projecting into the fourthventricle.

tered, a cystic medulloblastoma in the cerebellarhemisphere, itself an unusual finding (Fig. 4).The reasons for using other neuroradiological

investigations in addition to CAT in the five cases(Table 1) were as follows: one patient investi-

Fig. 4 The only low density medulloblastomaencountered. The low density in the right half ofthe posterior fossa proved to be a cyst.

gated with the 80X 80 matrix had a low densityarea in the region of the fourth ventricle but theexact site was in doubt. Pneumoencephalographyshowed a low vermis lesion which proved to be asolid astrocytoma. In one infant, the scan wastechnically unsatisfactory because the lowerposterior fossa was not included, although markedhydrocephalus was evident. Pneumoencephalo-graphy showed a low vermis lesion which provedto be a medulloblastoma. The scan of the thirdpatient clearly showed a large midline high densitylesion in the region of the vermis pushing thefourth ventricle forwards. In view of an unusuallyloud cranial bruit, vertebral angiography wasperformed but no vascular malformation wasdemonstrated. Operation revealed a large medul-loblastoma. In the remaining two cases CATshowed marked abnormality in the upper brainstem with possible extra-axial extension; moreaccurate diagnosis by further neuroradiologicalinvestigation was needed.

In the period under review, five presumed brainstem gliomas were investigated by CAT as theinitial study and subsequently were confirmed bylumbar pneumoencephalography. In three alesion anterior to a posteriorly displaced fourthventricle was clearly evident on CAT (Fig 5). In anadditional child, the scan showed a large dis-placed fourth ventricle but it was not possible tobe certain about localisation of the lesion. In

Fig. 5 Unenhanced scan showing low density areaanterior to the posteriorly displaced fourth ventricle.Grade 3 astrocytoma was found at exploration withinthe expanded pons.

212

Protected by copyright.

on May 31, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. D

ownloaded from

Impact of computerised axial tomography

another patient with multiple lower cranial nerveabnormalities, the important low section of thescan was marred by movement of the head andpneumoencephalography was required to show thestem expansion. In the four technically satis-factory scans of brain stem neoplasms, two wereof low and two of high density.

Discussion

The approach to diagnosis and surgical manage-ment of posterior fossa tumours has been greatlychanged by CAT in this department. In 1975, 17out of 24 (70%) such children had an EMI scanas the only preoperative investigation. It is pro-bable that easier access to CAT and greaterexperience will result in a larger proportion ofchildren being managed in this way. The correla-tion between CAT and operative findings has beenvery satisfactory. One scan, however, with muchmovement artefact was misinterpreted. It mustbe emphasised that the scan should be of goodquality and must include sections of the wholeof the posterior fossa. Sedation of younger childrenis often unavoidable. If a posterior fossa tumouris suspected, contrast enhancement should beinvestigated routinely, as plain scans may appearnegative in occasional cases of neoplasms (Newand Scott, 1975). In addition, recognition of aneoplastic pattern may be easier when absorptionis increased by injected contrast medium. In ad-dition to the size and the localisation, the natureof the lesion (Table 5) may be suggested to someextent and further experience should improve our

ability to make more accurate preoperativediagnosis. It may well be that detailed analysis ofthe absorption coefficients will help in thisrespect. In our limited experience cerebellar astro-cytomas have generally shown low density areaswhile medulloblastomas and ependymomas wereof higher density compared with normal sur-rounding brain tissue. Of course, there arenumerous exceptions to this rule of thumb (forillustrated examples see New and Scott, 1975;Baker and Houser, 1976; Epstein et al., 1976).

In this department a vertical midline incisionis made in all children believed to have a cere-bellar tumour and the extent of bony decompres-sion and dural incision is uniform. It has beenpossible therefore to adopt the following policy.Plain skull radiographs and CAT are consideredto be the only necessary preoperative investigationif a lesion in the vermis or cerebellar hemispheresis thus revealed. When the lesion is suspected ofinvolving the cerebellopontine angle, when a vas-cular malformation is suspected, or when the

symptoms and signs suggest the presence of ex-tension of tumour into the brain stem then furtherneuroradiological examination is carried out.

rhis approach to the management of posteriorfossa neoplasms in childhood has been found tobe safe, reliable, and has the great advantage ofavoiding preoperative invasive investigations inthe majority of these cases. We feel encouragedto proceed along these lines.

We are grateful to Dr J. Bull and Professor G. DuBoulay for allowing our patients to be examined atthe National Hospital on the EMI scanner. Dr R. D.Hoare (Neuroradiologist, The Hospital for SickChildren) has provided invaluable help and advice.E.B. was supported by a grant from the SwissNational Foundation for Scientific Research.

References

Ambrose, J., Gooding, M. R., and Richardson, A. E.(1975). Sodium iothalamate as an aid to diagnosisof intracranial lesions by computerized transverseaxial scanning. Lancet, 2, 669-674.

Baker, H. L. (1975). The impact of computed tomo-graphy on neuroradiologic practice. Radiology,116, 637-640.

Baker, R. L., Campbell, J. K., Houser, 0. W., Reese,D. F., Sheedy, P. F., Holman, C. B., and Kurland,R. L. (1974). Computer assisted tomography of thehead. An early evaluation. Mayo Clinic Proceed-ings, 49, 17-27.

Baker, H. L., and Houser, 0. W. (1976). Computedtomography in the diagnosis of posterior fossalesions. Radiologic Clinics of North America,14, 129-147.

Epstein, F., Naidich, T. P., Chase, N. E., Kricheff,I. I., Lin, J. P., and Ransohoff, J. (1976). Role ofcomputerized axial tomography in diagnosis andtreatment of common neurosurgical problems ofinfancy and childhood. Child's Brain, 2, 111-131.

Gawler, J., Bull, J. W. D., Du Boulay, G. H., andMarshall, J. (1975). Computerized axialtomography:the normal EMI scan. Journal of Neurology, Neuro-surgery and Psychiatry, 38, 935-947.

Gawler, J., Du Boulay, G. H., Bull, J. W. D., andMarshall, J. (1974). Computer assisted tomography(EMI scanner). Its place in investigation of sus-pected intracranial tumours. Lancet, 2, 419-423.

New, P. F. J., and Scott, W. R. (1975). ComputedTomography of the Brain and Orbit (EMI Scan-ning). Williams and Wilkins: Baltimore.

New, P. F. J., Scott, W. R., Schnur, J. A., Davis,K. R., Taveras, J. M., and Hochberg, F. H. (1975).Computed tomography with the EMI scanner in thediagnosis of primary and metastatic intracranialneoplasms. Radiology, 114, 75-87.

Paxton, R., and Ambrose, J. (1974). The EMIscanner. A brief review of the first 650 patients.British Journal of Radiology, 47, 530-565.

213

Protected by copyright.

on May 31, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.40.3.209 on 1 March 1977. D

ownloaded from