d. Cpstic anter-poserio andmulateral. views. A-Internal Carotid. v c B-Anterior Cerebral. C-MiddleCerebral. D-AnteriorCommunicating. a-Vertebral. b-Basilar. c-Posterior Communicating

156 BULL: Cerebral Angiography March 1950

%%

I%

AnrId.

B~~Bnero Ceerl.7a

CMCa

CAnterior Communicating. Cpsti

omaininraertebral.arere as seni

U Sc

b-Fas.laN. asetio

c anter-poserio andmulateral. views.

A-Internal Carotid. v c

B-Anterior Cerebral.C-Middle Cerebral.D-Anterior Communicating.a-Vertebral.b-Basilar.c-Posterior Communicating.d-Posterior Cerebral.

Protected by copyright.

on Decem

ber 20, 2020 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

'57

CEREBRAL ANGIOGRAPHYBy J. W. D. BULL, M.A., M.D., M.R.C.P., D.M.R.

Assistant Radiologist, St. George's Hospital. Assistant Radiologist, The National Hospital, Queen Square.Radiologist, Maida Vale Hospital

IntroductionCerebral angiography has now become an

essential diagnostic procedure in certain neuro-logical diseases and a valuable pre-operative pro-cedure in assisting the neurosurgeon to plan hisoperation. Largely as a result of angiography,vascular neurosurgery has been made possible.Moniz (I927) introduced the method nearly a

quarter of a century ago. In the early days theprocedure could not be lightly undertaken for itinvolved an open operation on the neck to exposethe carotid artery and the contrast substances usedwere not without danger. In recent years thedirect or percutaneous puncture of the vessel hasbeen satisfactorily performed and an almost be-nign contrast substance (diodone) has been used.Further than this the vertebral artery can now besuccessfully punctured through the skin, thusallowing all the main intracranial vessels to beexamined radiologically. Previously vertebralarteriography was very seldom undertaken as theapproach was almost a major operation in itself.

AnatomyA detailed knowedge of the vascular tree is

essential before undertaking angiography. Notonly must one know the disposition of the mainvessels, but also the common anomalies. One ofthe fascinations of angiography is the greatvariations of normal. The variations are muchgreater than the cerebral ventricular system andthe subarachnoid basal cisterns. Further it isimportant to appreciate the limitations of angio-graphy. In some respects it is far more crude thanhistological or clinical methods. It must be con-stantly borne in mind that only the main vesselscan be outlined radiologically. Thus certaintumours may fail to reveal their abnormal vascularpattern, which is easily seen under the microscope,but not detectable by ordinary radiologicalmethods. Microradiography, developed by thelate A. E. Barclay, may in the course of timecorrect this defect. Then again the neurologistmay be able to diagnose a limited or even fairlygross cerebral thrombosis which will not bedemonstrable at all at angiography.Four main arteries supply the brain; two in-

ternal carotids and two vertebrals. These vesselsform an anastomosis round the base, called the

Circle of Willis. The two external carotids mustalso be mentioned as they are important, par-ticularly in meningiomas and certain angiomas.

Figs. i and 2 show the internal carotid andvertebral arteries and their main branches. Thoughthe representation is somewhat crude anatomicallyit gives a picture as seen radiologically in the twobasic projections-lateral and anteroposterior.The direction of blood flow in the Circle of

Willis is somewhat variable. For example theposterior cerebral arteries are filled only from thebasilar in about two-thirds of cases, while in theremaining third the posterior communicatingartery is large enough to accept enough blood tofill the posterior cerebral artery. Thus, in aboutone-third of carotid angiograms the posteriorcerebral artery is filled. Again the two posteriorcommunicating arteries may be of different sizes sothat sometimes the posterior cerebral artery isfilled from one internal carotid artery, but not fromthe other on the opposite side.

For further details of the arterial pattern and itsrelation to important structures, the reader isreferred to anatomical textbooks.

TechniqueThere is no space to describe technique in de-

tail. Those interested are referred to Lindgren(i947), British Journal of Radiology. The mostimportant single factor is a good team used toworking together. Introducing the needle intothe vessel requires considerable practice before onebecomes skilled, but after performing about 5opunctures it should be possible to succeed innearly every case.The carotids of young children may be difficult

to puncture as the lumen is small, the vessel isvery mobile and the necks of small children arevery short. Local anaesthesia preceded by basalnarcosis should be adequate in the majority ofcases. Only when the patient is uncooperative forsome reason such as semi-coma, mental deficiencyor undue apprehension, is a general anaestheticnecessary. Patients under about i6 years nearlyalways require a general anaesthetic.

Films should always be taken in at least twoplanes. The most important are the lateral andanteroposterior. Oblique and basal views aresometimes helpful as ancillary projections. Phlebo-


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

158 POSTGRADUTATE MEDICAL JOURNAL March 1950

..... it

FIG. 3.-Large aneurysm of internal carotid artery, seen on AP viewat inner angle of orbit.

grams are essential in tumour diagnosis. In mostcases two phlebograms, taken 2 and 4 secs. afterthe arteriogram, are sufficient. If another series ofpictures is taken in the same projection the timingof the arteriogram and phlebograms can be alteredsuitably, so that in effect six phases, not three, ofthe vascular flow are obtained. Some workers con-sider that many more films should be taken-up to20 in about io secs.-but it is doubtful if additionalinformation is obtained for ordinary clinical pur-poses. There is no doubt that this more elaboratemethod is of physiological value, but for routinework it is very expensive and exposes the patientto a considerable amount of radiation.

The Contrast SubstanceDiodone is used exclusively nowadays. The

closely related substance, iodoxyl, often used forintravenous pyelography, should not be used.Diodone is quite harmless in the tissues, whileiodoxyl is not. Thus if a periarterial injection ofdiodone is given to a patient, no harm results.35 per cent. solution should always be used forlateral views and the greatest concentration per-missible in my opinion is 42.5 per cent. This isobtained by mixing equal quantities of 35 and 50per cent. solutions. About 40 per cent. solution isusually necessary for anteroposterior projections,where a greater thickness of bone and brain hasto be penetrated.

The Dangers of AngiographyThe only danger of any importance is the con-

trast substance. It is hypertonic and the strongerthe solution the more hypertonic it becomes. It isprobable that an oedema of the brain is producedby stronger concentrations, and indeed someoedema may be caused transiently by the lowerconcentrations. Fits, hemiplegia and coma haveall been seen; and in patients with very largetumours, which are made for all practical purposesstill larger by the oedema, death has occurred inthree cases in my experience of over i ,ooo examina-tions. Incidentally this figure probably favoursvery well with ventriculography though not withintravenous pyelography, where the same drug isused, but in different circumstances.On the whole it is fair to say that diodone, when

used in concentrations not greater than 42.5 percent. is a most satisfactory and almost entirely safedrug.

DosageUsually it is not necessary to give more than

four injections, sometimes less. Occasionally eightor nine injections have to be given to achieve thediagnosis. io cc. is used each time but there isabout 3 cc. of ' dead space ' in the injectionapparatus. Thus about 7 cc. actually enters theartery each time.


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

March 1950 BULL: Cerebral Angiography I5,9

*

(A.

FiGs. 4 (A) AND (B).

Parietal angioma supplied by middlecerebral vessels and showing a largevein, which is filled in the arterial phase,draining up to the superior sagittal sinus.

The Applications of Cerebral AngiographyThere are two main pathological groups that

can be investigated:-(i) Vascular and (2) Neo-plastic. The first group may be subdivided into(a) spontaneous vascular disturbances, and (b)traumatic. Both carotid and vertebral angiographymay be used but the scope of the former is by farthe greater.

Spontaneous Vascular DisturbancesI. INTRACRANIAL ANEURYSMS

It has been known for several years that spon-taneously occurring aneurysms of the larger intra-cranial arteries are not uncommon, and that similaraneurysms of arteries elsewhere in the body areexceedingly rare. The large majority are thoughtnowadays to be c6ngenital in origin. That is notto say that they are present at birth, but there is acongenital weakness of the arterial wall which is apotential site for aneurysmal formation. Perhapsthe stresses of life and increase of blood pressurewith age cause the weak wall to bulge and forman aneurysm. It has been found that most suchaneurysms occur at or very near bifurcations ofarteries. This rule does not apply always, par-ticularly in the proximal part of the intracranialportion of the internal carotid artery. When abifurcation is not adjacent to the aneurysm the ()


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

160 . POSTGRADUATE MEDICAL JOURNAL March I950

.4,

-\'-e

.. -l*:.: | ! | =>:*:..'.n,,.. j l|- -

w.<N ..

PE.FIG. 5.-Internal carotid arters thrombosed 2 cm. from origin.

strength of the theory is preserved by postulatingthat branches were present in the particular situa-tion in early embryonic life.A very small proportion of intracranial

aneurysms are arteriosclerotic. These are usuallyfusiform, while the congenital type tend to bepedunculated. An even smaller group aremycotic-those secondary to septic emboli. It isdoubtful whether any are syphilitic.The size varies tremendously, some are so small

that they may not show angiographically. Everygradation of size is seen-from that of a pea seedto a hen's egg.The symptoms and signs depend on two main

factors; (i) whether or not the aneurysm has burstand therefore caused a haemorrhage, and (2) itssize. Further subsidiary factors must be postu-lated to explain the features shown by manyaneurysms. First it must be remembered thatthese larger vessels under discussion lie in thesubarachnoid space and not in brain substance.Thus a rupture of the smallest aneurysm may causea severe or fatal subarachnoid haemorrhage, butgive no other signs. But since the majority lie onor near the Circle of Willis they will be veryadjacent to the upper six cranial nerves (with theexception of the olfactory pair). Pressure on oneor more of these nerves will produce valuable

localizing signs and sometimes symptoms of pain,when the fifth nerve is pressed upon. Manyaneurysms of no more than 0.5 cm. diameter canproduce quite severe signs.When one considers the larger aneurysms, other

features may present themselves. Just as aneurysmsof the descending thoracic aorta may erode bodiesof vertebrae, intracranial aneurysms can erode theskull bones and also naturally the much softertissues of the brain itself. Occasionally one seesmuch or part of the sella or superior orbital fissuredestroyed by carotid aneurysms in the vicinity ofthe cavernous sinus. Aneurysms arising from moreperipheral vessels such as the anterior and posteriorcommunicating artery, anterior, middle and pos-terior cerebral arteries may press upon or erodebrain tissue. Such aneurysms may be quite silent,even when very large, particularly if they invadethe frontal lobe. Again they may simulate tumoursand can only be differentiated by arteriography.Such a case was described by the writer (P.R.S.M.,1949). If such aneurysms burst they may eitherbleed into brain tissue, or if sufficiently adjacent,rupture into the ventricles. Aneurysms of thevertebral or basilar arteries, pressing upon, butnot necessarily invading the pons, may be quiteindistinguishable from pontine tumours except byangiography.


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

March I9-x5 BULL: Cerebral Angiography i6i

FIG. 6. Left Subdural haematoma. Note also displacement ofanterior cerebral vessels.

Pa-rt or all of the wall of some aneurysms cal-cifies and this feature combined with the erosivequalities already described, makes plain X-raysof the skull an essential precursor to angiography.

Presenting Signs and Confirmation of DiagnosisThe above description makes it apparent that

aneurysms present in a variety of ways:(a) Pure subarachnoid haemorrhage. Here the

diagnosis is usually simple by clinical methods,supported by lumbar puncture. However fromthe point of view of treatment it is the localizationwhich is very important. Sometimes one isassisted in lateralizing the lesion by a partialaphasia, lateralized headache, perhaps a transientweakness of one side of the face or a limb. Butin the majority of cases one has to guess and soit is often necessary to perform carotid arterio-grams on both sides. If they are both negative, onevertebral artery must be filled to outline the pos-terior part of the arterial tree. By this meansnearly, but not quite all, aneurysms are localizable.The remainder are presumably either too smallor have thrombosed.A current burning question is when to perform

an angiography on a patient with a known sub-arachnoid haemorrhage ? Some authors say it is

necessary to wait six weeks, others ten days andyet others that no delay is necessary. The writeragrees with the last group and regards the con-dition as a surgical emergency. There seems tobe no valid reason for delay.

(b) Aneurysms pressing on cranial nerves. This isa smaller group than that presenting with sub-arachnoid haemorrhage. In this group theaneurysms do not always bleed but manifest them-selves by pressing either on a motor nerve to theeye, the optic nerve or the fifth nerve. Thus someare associated with intense pain. Aneurysms of thisgroup (Fig. 3) are easier to localize radiologicallyfor they are lateralized clinically and often thevessel from which they arise can be predictedclinically-for example carotid, anterior com-municating, posterior communicating and so on.Only one carotid artery needs to be punctured insuch cases. Furthermore it is in this group thatsigns are seen on the straight X-rays, such aslocalized bone erosion or calcification.

(c) Aneurysms invading brain substance and pre-senting as ' tumours.' This is the most difficultgroup to diagnose clinically. In the other twogroups the clinical diagnosis is relatively easy, butthe localization may be difficult or impossible. Inthis last group the clinical localization may bemoderately accurate but the pathological diagnosis


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

162 POSTGRADUATE MEDICAL JOURNAL March x9S°

..... ----

.- . . -l! | |i | - | | l |

- - | * |J - - | * * -

&: - - | * * -

:::g1 1111i Rl- R l.: - | l - * I * * rX*:s Illlt E l - * I * * E.. . o,: 1 .. l '.df | * l * * - JR- IB;W | | - * **:|- 11111l X 2 | l | |*: . . - . l . - - ....... ..^ l | | * l l l l - :f'.E.:- E E - - . - . -- | s - - . - . .'

- - e 111-- | - * -a}:: Sf:.:-- S el - - | - * -B iai3

*.e- - K * - - | - * -- - | * - - | - * -

r:N* -- - * CE6w.. :- ll | * - - | - * .. R v:.<a.-;;i-6

*'l111l1111E. 11 I * | 11 I- ..sSill- lySw-- w--...- - . l 1.3-il-mn sx l

ij|;:¢rJ-

FIG. 7.-Parietal meningioma outlined in venous phase.

is often incorrect, and indeed aneurysm may noteven be considered in the differential diagnosis.

Unless an angiogram is performed the diagnosismay not be revealed until surgical exploration.

(d) ' Silent' aneurysms. There is no doubt that'silent' aneurysms exist and some possibilitythat they are not uncommon. They may perhapsbe likened to diverticula of the colon which are sooften ' silent ' or symptomless.The writer has seen one case of aneurysm which

had bled and was well localized clinically and con-firmed arteriographically. There were two silentaneurysms on the tree of the opposite internal caro-tid artery. Another case was of a malignantglioma of the parieto-occipital region shown angio-graphically. There was a 'silent' aneurysm of theinternal carotid artery at the site of entry into thesubarachnoid space.

2. ARTERIOVENOUS FISTULASuch lesions can occur between the intra-

cavernous portion of the internal carotid arteryand the cavernous sinus. Such cases are very rareand the writer has seen only one example.

3. VASCULAR MALFORMATIONS (angioma)Until recently they were thought to be rare

within the skull. However with greater use of

cerebral angiography many more are being re-vealed. The writer has seen over 50 in the lastthree years. Quite a large proportion have a bruitwhich often can only be heard by very carefulauscultation. Rather less than one-third showchanges on the plain X-rays which may be obviousor very difficult to detect. The obvious ones showgross enlargement of the diploic vessels, while theless obvious may only show an enlargement of theforamen spinosum on the affected side and minimalvascular changes in the skull vault. Some casessuffer from epilepsy and others have localizingcortical signs. Advanced cases may show mentaldeterioration.Pneumography often reveals no abnormality, or

perhaps a little dilatation of the lateral ventricle onthe affected side, without displacement of theseptum lucidum. Until recently such cases wereusually diagnosed as cortical atrophy.There is a very small group which first manifest

themselves by subarachnoid haemorrhage. Therehave previously been no physical signs or symp-toms, and the angioma is not suspected until it isseen on the angiogram. Such a case was recentlyseen in which the angioma lay along the medialsurface of the hemisphere, being derived from thepericallosal branch of the anterior cerebralartery.


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

March 1950 BULL: C7erebral Angiography i63

FI.8-Lagpotro tepoalob tuou dipacn Sylvia:vssl upwars.1S.

Radiological features of angiomas (arteriovenousaneurysms)The majority of these tumours are situated on

the distribution of the middle cerebral artery(Figs. 4a and 4b), but some lie on the medialsurface of the brain and are supplied by the anteriorcerebral artery. Both these groups may receivea secondary blood supply from the posteriorcerebral artery. These tumours vary very muchin their size, shape and depth, but nearly alwayspart of the tumour lies on the surface.The external carotid artery also takes part in

supplying some angiomas, and this feature isusually apparent clinically. Tortuous vesselssomewhere over the skull with or without anaevus may point to the lesion, but without anyangiogram it is impossible to state how much ofthe lesion is intracranial.The following features are usually seen on the

angiograms -

i. The hypertrophied carotid vessels on theaffected side.

2. The hypertrophied vessel or vessels directlysupplying the angioma.

3. The tortuosity of the supplying vessels.4. The poor filling of the vessels not supplying

the angioma. For example, if the lesion is suppliedby a branch of the middle cerebral artery, theanterior cerebral artery may not fill at all.

5. The huge efferent veins leading to the venoussinuses. These vessels are filled in the arterialphase, a feature not seen in other lesions.

6. The capillaries of the brain are by-passed.7. The veins are usually seen already to have

emptied in the phlebograms.

TreatmentRecently surgeons have been extirpating the

more accessible and not too large angiomas. Someof the results are excellent. It is important toperform a post-operative arteriogram in order tosee if all the tumour vessels have been removed.

4. THROMBOSIS OF THE INTERNAL CAROTIDARTERY

This condition is less uncommon than is gener-ally supposed, and the writer has seen eight casesdemonstrated by arteriography in the last threeyears. Moniz states that the possibility of a caro-tid thrombosis should be considered in any obscurecase of hemiplegia.The classical and almost constant site for the

thrombosis is about 2 cm. distal to the origin of the


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

164 POSTGRADUATE MEDICAL JOURNAL March I950

internal carotid artery (Fig. 5). Unlike femoralarterial thrombosis, the condition is often seenin younger patients in the third or fourth decades.When performing an arteriogram it is very im-

portant to show the bifurcation of the carotid arteryon the radiographs. This may be difficult withshort-necked people. If the bifurcation is notshown it is sometimes wrongly assumed that theinjection has been made into the external carotidartery, for that vessel alone is seen on the radio-graph.

5. EMBOLUSOnly emboli of the larger arteries can be shown

by angiography.

6. INTRACRANIAL HAEMORRHAGEIf intracranial haemorrhage is suspected in a

case, an immediate arteriogram may reveal thediagnosis by displacement of vessels and so pro-vide the surgeon with sufficient evidence oflocalization to enable him to enucleate the clot.This procedure has only been practised recentlyand insufficient experience has been gained toassess its value.

7. CHRONIC SUBDURAL HAEMATOMAThis is a not uncommon condition and the diag-

nosis is not always obvious as quite frequently thereis no history of trauma. Arteriography provides apathognomonic picture in the anteroposteriorposition and shows the middle cerebral vessels dis-placed medially away from the inner table of theskull (Fig. 6). This is an example of the absolutenecessity of taking anteroposterior pictures, for thelesion cannot be shown at all in the lateral view

The Diagnosis of Intracranial Tumours byAngiographyRecently this method has been used more and

more as a diagnostic method and for the localiza-tion of intracranial tumours. Valuable though themethod is, it must not be forgotten that ventri-culography is an extremely accurate method andcannot be dispensed with. The two investigationsare complementary and it is very important todecide which method should be used first in anygiven case. In order to save the patient un-necessary investigations, irteriography should pre-cede ventriculography whenever:-(i) a vascularlesion is suspected, (2) a supratentorial lesion issuspected which has been localized either clinicallyor on the plain X-rays, as for example a pinealshift, a patch of pathological calcification orabnormal vascular channels in the skull.On the other hand ventriculography should pre-

4ede angiography when a posterior fossa or mid-brain expanding lesion is suspected.

Although vertebral angiography has been usedin the diagnosis of posterior fossa lesions, it cannotcompare in accuracy with ventriculography.Lesions in or around the sella may require bothcarotid angiography and ventriculography. Ventri-culography usually provides a more accuratepicture of the size of the lesion, while angio-graphy may be helpful in determining the path-ology, particularly if the lesion is an aneurysm, andthe majority of aneurysms are situated in theregion of the sella.

For the most part then, angiography is of par-ticular value in the diagnosis of tumours on ornear the surface of the cerebral hemispheres.These tumours may be revealed in two ways. (i)By retaining the contrast substance in their vesselsand thus outlining themselves on the angiograms,or (2) by displacement of the anterior or middlecerebral arteries and thus allowing the localizationto be deduced by the same methods as one appliesfor ventriculography. When the vessels in thetumour are shown on the angiogram one can statethe size, shape, position and morbid anatomy of thetumour. Thus a more complete picture is ob-tained than by ventriculography. Unfortunatelymany tumours do not have sufficiently largevessels to reveal themselves in this way.

Tumours Revealed by Angiography(a) Meningioma. Though not the commonest

brain tumour, it is one of the most important tothe neurosurgeon in that it is nearly always be-nign. Furthermore the majority of these tumoursare surgically accessible. Whenever a supra-tentorial meningioma is suspected, angiography isindicated. About 50 per cent. of such tumours casta shadow on the angiograms. The quality ofshadow varies from case to case, but the classicalappearance is the so-called ' blush' or ' stain.'The picture is not unlike that of the kidney some-times seen at pyelography. The organ stands outhomogeneously stained, the so-called nephrogram.It is caused by the contrast substance being loadedin the renal capillaries. Similarly the contrastsubstance becomes loaded in the capillaries of ameningioma. For this reason the 'blush' issometimes not seen on the radiograph taken in thearterial phase. This is an example where phlebo-grams are so necessary. In some meningiomaslarger vessels are filled and the tumour is revealedin the arteriogram. After experience of angio-graphy in about 6o supratentorial meningiomas,one has the impression that the greater part ofthe blood supply is via the internal carotid artery,rather than the external, even in cases where theplain X-rays show quite large branches of themiddle meningeal artery. Very occasionally thebulk of the blood supply is from the external caro-


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

March 1950 BULL: Cerebral Angiography i65

tid artery. In such cases it is desirable to puncturethe external carotid artery in the neck. This isdifficult as the vessel cannot be palpated separatelyfrom the internal carotid artery in the neck. Onecannot yet say why certain meningiomas give a' blush' and others do not. It probably dependspartly on the blood supply and partly on thesituation of the tumour (Fig. 7).

(b) Glioma. The pathological classification ofthe gliomas is difficult and ideas are constantlychanging. These tumours show almost everygradation, from the very rapidly growing, highlymalignant types, to the very slow growing, almostbenign types. There seems to be some relation-ship between the degree of malignancy and thevascularity. If the two extremes are considered,one finds that most of the very malignant tumoursare highly vascular and that a large number of thevessels are very primitive, of large capacity, andoften have an arteriovenous angiomatous formationin miniature. At the other end of the scale-thevery benign tumours-there is a paucity of bloodvessels.

It is reasonable to expect that angiography canbe helpful in the more malignant types, and thisis in fact true. Various writers have given differentfigures of the percentage of malignant gliomaswhose vessels can be seen on angiography, butmost of the figures are about 50 per cent.The typical picture, sometimes seen only in the

arterial phase, sometimes in the venous and some-times in both, shows a large number of rathershort, often slightly tortuous vessels, frequently ofrelatively large calibre.

In a small proportion of cases an almost purecapillary ' blush' is seen, and the author has mis-diagnosed three malignant gliomas as meningio-mas. This error is not usually serious as allsurgeons operate on meningiomas, but it is possibleto see the reverse state of affairs where a men-ingioma gives an angiographic pattern indisting-uishable from a malignant glioma. For this reasonthe diagnosis of tumour pathology made by angio-graphy should always be accepted with some re-serve and histological confirmation, either by aneedle biopsy or section of part of the tumourshould be attempted in every case.

(c) Metastasis. Some intracerebral metastasesare sufficiently vascular to be demonstrated angio-graphically. If more than one tumour is seen onthe angiogram the diagnosis is virtually certain.It is sometimes difficult to differentiate the singlemetastasis from a malignant glioma as the patternof the blood vessels may be very similar. Howevermetastases tend to be discrete and some-times show a surrounding marginal vein, whilegliomas tend to be more diffuse and it is difficultto delineate their margin on the radiograph.

Most intracerebral metastases come from thebronchus and a chest radiograph, combined withbronchoscopy, is usually sufficient to confirm thediagnosis, though occasionally the primary tumourmay be so small as to elude both the radiologistand bronchoscopist.

The Angiographic Diagnosis of Tumours withoutPathological Vessels

Angiography is valuable in the localization ofmany such lesions. Most pneumoencephalo-graphic diagnosis is dependent upon the correctinterpretation of the deformity of the ventricularsystem. Lysholm developed this to such a highdegree that he could accurately localize nearly allintracerebral expanding lesions, provided theywere of sufficient size to cause some displacementor distortion of the ventricular system.The same methods can be applied to the main

cerebral vessels and their branches. The accuracyof the method is not so great for two reasons:(a) the vessels are less susceptible to displacementthan the ventricular system, and (b) the variationsof the normal arterial tree are very much greaterthan the ventricular system, which is very con-stant. Thus considerable experience with angio-graphy is required before it is possible, in difficultcases, to assert that the pattern is or is not withinnormal limits.Having localized a tumour by angiography one

cannot accurately predict its pathology if itsvessels do not fill on the angiograms. But here oneis no worse off than with pneumoencephalographywhere again the morbid anatomy cannot usuallybe predicted. The advantage of angiography isthat there is quite a considerable chance of pre-dicting the pathology and outlining the exact ex-tent of the tumour, as described above.

Although it has been stated that angiogiaphy isa less accurate method of localization than pneumo-graphy, tumours in certain situations nearlyalways displace main vessels sufficiently to allow aconfident diagnosis of localization to be made.For example, subfrontal tumours regularly displacethe proximal portion of the anterior cerebral ar-tery upwards. Nearly all temporal tumours dis-place the middle cerebral artery or its mainbranches upwards (Fig. 8). Low lateral parietaltumours displace the middle cerebral vessels down-wards and occipital tumours displace them for-wards. Tumours in other situations are less re-liable in causing a characteristic vascular displace-ment.

In this last section the word ' tumour' is usedin its widest sense to include all types of expandinglesion.

BIBLIOGRAPHYBULL, J. W. D. (I949), Proc. Roy. Soc. Med., 52, 880.LINDGREN, E. (I947), Brit. Journ. of Radiology, 20, 326.


on Decem


j.bmj.com

/P


j.26.293.156 on 1 March 1950. D

ownloaded from

http://pmj.bmj.com/

Documents

d. Cpstic anter-poserio andmulateral. views. A-Internal Carotid. v c B-Anterior Cerebral. C-MiddleCerebral. D-AnteriorCommunicating. a-Vertebral. b-Basilar. c-Posterior Communicating