91

Douglas A. Graeb, M.D.

William D. Robertson, M.D.

Jocelyne S. Lapointe, M.D.Robert A. Nugent, M.D.Philip B. Harrison, M.D.

Computed Tomographic Diagnosis

of Intraventricular Hemorrhage

Etiology and Prognosis1

Sixty-eight patients with intraventricularhemorrhage (IVH) diagnosed by computedtomography (CT) were reviewed retro-spectively to determine the etiology andprognosis, relationship to delayed hydro-cephalus, and effect on neurological out-come. The most common causes were aruptured aneurysm, trauma, and hyper-tensive hemorrhage. Ruptured aneurysmsof the anterior communicating arterycan often be predicted from the non-enhanced CT scan. The total mortalityrate was 50%; however, 21% of patientsreturned to normal or had only milddisability. Patients in whom no causewas identified had a better prognosis.Delayed hydrocephalus was related tothe effects of subarachnoid hemorrhagerather than obstruction of the ventricularsystem by blood. IVH per se is seldom amajor factor in the neurological outcome.

Index terms: Aneurysm, cerebral. 1(71.730 #{149}Brain,

hemorrhage #{149}Brain, injuries #{149}Head, computed

tomography. 1[0].l211 . Hydrocephalus. 1[6].4372.

Subarachnoid space. hemorrhage

Radiology 143: 91-96, April 1982

I From the Department of Diagnostic Radiology.

Vancouver Ceneral Hospital, Vancouver, B.C., Canada.Received May 7, 1981; accepted and revision requested

Aug. 11; revision received Sept. 18.Presented in part at the Forty-third Annual Meeting

of the Canadian Association of Radiologists. Montreal,Quebec, Canada, May 11-15, 1980. sjh

B EFORE computed tomography (CT), the clinical diagnosis ofintraventmiculam hemorrhage (IVH) was usually suggested by

sudden coma followed by severe and often fatal brainstem dys-function (1-3). This so-called “classical” presentation was generallyassociated with massive intraventricular hemorrhage. However, CTfrequently demonstrates IVH even when the “classical” clinicalpicture is absent. Fortunately, although it usually occurs in association

with severe brain trauma, even relatively massive IVH may havea good outcome (4) (Fig. 1).

While several reports have documented the ability of CT to showIVH (4-7), to our knowledge only one sizable series has been ana-lyzed, namely that of Little et a!. (8) who observed 54 patients of

whom only 9 survived. Our experience suggests that the prognosisis somewhat better than these statistics indicate. We have reviewed

our own cases to determine the etiology and prognosis and evaluatethe mole of IVH in both delayed hydrocephalus and the neurologicaloutcome.

MATERIALS AND METHODS

Of 77 cases of CT-diagnosed IVH available for study, 9 were dis-carded because of inadequate follow-up, poor scan quality, on becausethe IVH was postoperative, leaving 68 cases for review. Prematureinfants were not included.

All patients had non-contrast scans on an EM! 1O1OS; contrast-

enhanced scans were also obtained in 1 1 cases. In patients with aruptured aneurysm, we attempted to locate it on the non-contrast

scan, based on the site of hematoma formation. In 29 patients scannedmore than once, ventricular size was followed up in order to deter-mine the relationship between subarachnoid and/on intraventriculambleeding and delayed hydmocephalus, defined as an increase inventricular size developing at least one week after the initial hem-omrhage. Any increase in size was also correlated with the presenceof blood in the third and fourth ventricles and subarachnoid space.IVH was graded according to the amount of blood within each yen-tnicle (TABLE I), with a maximum score of 12: 1-4 was called mild, 5-8moderate, and 9-12 severe.

Outcome was assessed using the Glasgow scale (9), in which pa-tients are assigned to one of five categories: normal or mild disability

(I), moderate disability (II), major disability (III), vegetative state (IV),or death (V). Each patient’s clinical record was also reviewed to de-termine the cause of death, on the nature of any residual neurologicaldeficit in the case of survivors.

RESULTS

Age and sex distributions are shown in Figure 2. Males and females

Aneurysm

,a _.

I I I

Trauma

�rJf�

Hypertensive

f///kr/i2ffAhemorrhage0 20 40 60 80

Age

Severe IVH and right frontal hematoma

caused by a ruptured right anterior cerebral

artery aneurysm. The patient was normal on

discharge except for mild paresis of the left leg

secondary to spasm of the vessel.

80

60

U Male (34 cases) 40

� Female (34 cases) 20

20 1 #{176}

16 a� 20

12 0

8 40

4 20

0 I I I I

20 40 60 80Age

a. Age and sexdistribution of 68 patients with IVH.

b. Age distribution of patients with IVH due to a ruptured aneurysm, trauma, or hypertensive

hemorrhage.

Figure 3

25 17 17

13 12

01

0 15Ca)

c�10 5

5 3

‘S

S P’ �

-c 011�n(�O4 I#{231}�5 �t%O�OS �(%‘

- �

Etiology of 68 cases of IVH. The number of cases is indicated at the

top of each column.

Figure 4 were affected in equal numbers, with

a peak incidence in the sixth and sev-enth decades which was composed

mainly of patients with ruptured an-

eurysm on hypertensive hemorrhage.Patients with trauma showed a bi-phasic distribution, with peaks at ages10-30 and 60-80. The most frequentcauses of IVH were ruptured aneu-rysm, trauma, and hypertensive hem-

omrhage (Fig. 3).Aneurysms of the anterior cerebral

Non-enhanced scan shows a ruptured anteriorcommunicating artery (ACA) aneurysm with

IVH and subarachnoid bleeding. Note the

localized hematoma in the anterior inter-

hemispheric fissure.

Figure 1 Figure 2

92 April 1982 Volume 143, Number 1 Graeb and Others

TABLE I: System for Grading Severity

of IVH

Lateral VentriclesScore: 1 = trace of blood or mild bleeding

2 = less than half of the ventriclefilled with blood

3 = more than half of the ventriclefilled with blood

4 = ventricle filled with blood andexpanded

(Each lateral ventricle is scored separately)

Third and Fourth VentriclesScore: 1 = blood present, ventricle size

normal2 = ventricle filled with blood and

expandedTotal Score (maximum 12)

TABLE II: Accuracy of CT Diagnosis

of Aneurysms at DifferentSites

CorrectNo.of CT

Proved Diag-Site of Aneurysm Cases nosis

Anterior communi-catingartery 6 4

Anterior cerebralartery 2 1

Azygous artery I 0Pericallosal artery 1 0Callosomarginal artery 1 0Internal carotid artery 3 0Middle cerebral artery 1 0Posterior communi-

catingartery I 0Posterior inferior

cerebellar artery 1 0

U,

U,

0U

0

a)�0E

z

Figure 6

Figure 5

CT Diagnosis of Intraventnicular Hemorrhage NEURORADIOLOGY 93

Cerebral trauma with severe IVH (Glasgow scores: left ventricle 4; right ventricle 3; third ventricle 2; fourth ventricle 2; total 1 1)., Note the hemorrhage in the subarachnoid space and the soft tissues of the scalp and face. The dilatation of the lateral ventricles is probably

due to intraventricular obstruction of CSF flow.

artery and its branches, including theanterior communicating artery (ACA),

were the most common, constituting 11out of 17 cases (65%) (TABLE II). Onlyruptures involving the ACA could

be predicted, being seen as localizedbleeding in the anterior intenhemi-

spheric fissure, just anterior to the lam-

ma termina!is, in 4 out of 6 cases (Fig.

4).

The diagnosis of trauma was evident

from the history and physical findings

and was confirmed by autopsy in 11

out of 17 patients. Swelling on hemon-

rhage in the extracranial soft tissue was

present in 13 patients (76%). Paren-

chymal hematomas were found in 12

(71%). Additional findings included

ventricular enlargement, midline shift,

subarachnoid hemorrhage, acute sub-

dunal hematoma, cerebral contusion,

skull fracture, and intracranial air (Fig.

5).

Hypertensive hemorrhage, seen in

13 cases, was established by hyperten-

sion �180/100 (mean, 220/120) and a

deep parenchymal hematoma on the

CT scan. In 3 patients, angiographyshowed no aneurysm; 2 patients had

hypertensive hemorrhage at autopsy.

In 7 cases (54%), there was bleedinginto the thalamus or basal ganglia

b (Fig. 6). Three had bleeding into thetemporal lobe, 1 into the frontal lobe,

and 1 into the pons; another patienthad multiple hemorrhages throughout

TABLE III: Correlation BetweenDelayed Hydrocephalus

and Hemorrhage (29 cases)

DelayedNo. of Hydrocephalus

Patients No. %

IVH in third or fourth ventriclePresent 19 9 47Absent 10 3 30

Subarachnoid hemorrhagePresent 15 10 67Absent 14 2 14

both hemispheres and the cerebellum.

Only 4 had CT evidence of subanach-

noid bleeding. Other findings includ-

ed ventricular enlargement and mid-line shift.

Anteniovenous malformation (AVM)

was diagnosed by angiognaphy in all 5

cases. Intracerebral hematomas were

associated with midline shift in 3 and

ventricular enlargement in 2.

Three patients had a hemorrhagic

diathesis established by coagulation

studies. Two were ovenmedicated with

oral anticoagulants and 1 had profound

thrombocytopenia due to chemothen-

apy for acute pnomyelocytic leuke-

mia. No characteristic CT features were

found.In 1 patient (Fig. 7), IVH was pre-

sumably caused by metastases from

renal-cell carcinoma (which had also

Hypertensive hemorrhage into the right basal

ganglia in a patient with mild IVH. (Glasgow

scores: left ventricle 1; right ventricle 2;

third ventricle = 1; fourth ventricle 0; total

= 4). The low-density area lateral to the left

caudate nucleus is consistent with an old

infarct or hemorrhage.

spread to the lungs and bone), a!-

though we have no pathological

proof.

In 12 of the 68 patients, the etiology

was undetermined. Angiography in 8

failed to show the bleeding site, as didautopsy in 1 case. Two patients died

Figure

16

e

a 12I-)

Figure 9

Figure 10

Outcome of IVH (68 cases). The number of

cases is indicated at the top of each column.

a)at0Ca)UI-

a)a-

Outcome of IVH in four main etiologic categories. The number of cases is indicated at the top

of each column.

Figure 7

94 April 1982 Volume 143, Number 1 Graeb and Others

a. b.

a. Scan shows a round area of predominantly increased density posteromedial to the right trigone (arrows), with bleeding into both lateral

ventricles.

b. Vertebral angiogram reveals abnormal vessels (arrowheads) arising from the posterior choroidal arteries; this is consistent with vascular

metastases, though this was not proved. Note the mass effect causing anterior displacement of the lateral posterior choroidal artery (short

arrow) and caudal displacement of the posterior penicallosal artery (long arrow).

1.3 7.14 15.28 33

Days

Length of CT follow-up (29 cases).

Outcome

I II IN IV V

a) 40010Ca)U 14

- - - �5 20103164186 Q� 10

without undergoing angiography, and

34 autopsy was not performed. Two el-derly patients recovered satisfactorilyand were not investigated because oftheir age. There were no CT featureswhich distinguished these 12 patients,nor was there any significant differ-ence in the severity of IVH betweenthis group of 12 patients and the 56 inwhom the etiology was undetem-mined.

In most of the 29 patients with repeatscans, follow-up lasted one to four

weeks (Fig. 8). While the association of

delayed hydrocephalus with blood in

the third on fourth ventricle was notstatistically significant (P > 0.3), its

association with subanachnoid hem-

ornhage was (P 0.004) (TABLE III).Outcome is shown in Figure 9. There

was only 1 patient in category IV;

however, as a permanent vegetativestate is as serious as death for the pa-

tient and family, categories IV and Vare grouped together here. The out-

comes in the four categories are shownin Figure 10. Compared to the mortal-

ity rates in trauma (71%), hypertensive

hemorrhage (62%), and ruptured an-

eurysm (53%), patients with bleeding

of undetermined cause had a consid-enably better prognosis: the mortalityrate was only 25%, while 50% survivedwith no deficit or only mild disability.

If the 4 patients who were not investi-

TABLE IV: Additional CT

Abnormalities in IVH

80

Outcome

I II Ill IV-V

IVH Grade

Severe

Figure 11

100

a)at0Ca)Ua)

a-

Relationship of outcome to severity of IVH. The number of cases

is indicated at the top of each column.

Figure 12

100 UNil

��.JSAH only

It �:0 0 Other abnormalities 2 � 2 � 28

I II Ill IV-VOutcome

Relationship of outcome to other CT abnormalities. The number

of cases is indicated at the top of each column.

frequent finding which can be used topredict the site of the aneurysm basedon the CT scan, something which wasnot possible with ruptures at other

sites.Although subarachnoid bleeding

from cerebral metastases (particularlymalignant melanoma) is known tooccur (11, 13), intnaventniculan bleed-ing from metastases has not been re-ported previously to our knowledge.While the CT appearance was not spe-cific for metastases, the angiographicdemonstration of tumor vessels sup-ports this diagnosis (Fig. 7).

CT Diagnosis of Intnaventniculan Hemorrhage NEURORADIOLOGY 95

No. %

Parenchymal hematomaMidline shift

4222

6232

Subdural hematoma 5 7Transtentorial herniation 3 4Contusion 1 1

Recent infarct 1 1Subarachnoid bleedingNone

119

1613

gated and hence may have a discov-enable cause of IVH are excluded, themortality rate was only 1 in 8 (12.5%).The more severe the IVH, the worse

the prognosis; indeed, 90% of patientswith severe IVH died (Fig. 11).

IVH is seldom an isolated event. In87% of our cases, additional abnor-

malities were seen on CT (TABLE IV).Patients with a poor outcome also hada higher incidence of associated ab-nonmalities (Fig. 12). Although only 9of our patients (13%) had no additionalabnormalities, 7 had a good on excellentoutcome (category I on II). In all cases,factors other than IVH were sufficientto explain the neurological outcome.In 63 cases the cause was a pamenchy-mal hematoma, midline shift, angio-graphically demonstrated spasm, oninjury to other organs. Of the 5 me-maining patients, 3 suffered massivere-bleeding and died before a repeatCT scan could be obtained, while 2died of postoperative neurologicalcomplications.

DISCUSSION

Although aneurysms of the anteriorcerebral artery and its branches made

up 65% of our series, they account foronly 36% of all ruptured aneurysmswith subarachnoid bleeding (10).

Other authors have noted the increasedlikelihood of intraventniculam rupturefrom aneurysms of these vessels, par-ticulanly the ACA (8, 1 1). Aneurysms of

the ACA may readily rupture into thelateral ventricles, as only the relativelythin olfactory and panatenminal gyniseparate the aneurysms from the

frontal horns.Our experience supports Yock’s me-

port of the CT appearance of a rupturedACA aneurysm (12). The most charac-tenistic feature is a hematoma localizedto the anterior interhemisphemic fis-sure, just anterior to the lamina tenmi-nalis. Although previously describedas an uncommon event (11), CT

suggests that it is actually a relatively

The overall mortality rate of 50% inour series (Fig. 9) is considerably lowerthan the 83% reported by Little et a!. (8),

which is only partly explained by thegreater percentage of cases of unde-temmined cause in our series (18% vs.

4%). Fourteen patients (21%) returnedto normal on had only mild disability,while 10 (15%) survived with moderatedisability.

As demonstrated with subamachnoidbleeding of undetermined etiology (14,15), patients with IVH of unknowncause have a significantly better prog-nosis. Just as microscopic vascular

96 April 1982 Volume 143, Number 1 Gnaeb and Others

malformations have been postulated asa cause of subanachnoid bleeding ofundetermined etiology (16), similarmalformations adjacent to the epen-dymal lining of the ventricles may beresponsible for some cases of IVH inwhich no cause is found.

Approximately 50% of our patientshad mild enlargement of the lateralventricles on the initial CT scan. As noprevious scans were available andmany patients were at an age when at-nophy could be expected, this findingwas difficult to interpret; however, anyincrease in ventricular size on subse-quent scans was considered significant.Allcock (17) has stated that develop-ment of hydrocephalus after IVH maybe due to obstruction of the flow of#{231}erebrospinal fluid (CSF) at the foma-men of Monno or aqueduct of Sylvius;although this is likely to be true inacute cases, our study shows that hy-drocephalus which develops at leastone week after IVH can be correlatedwith subarachnoid hemorrhage on theinitial CT scan rather than with IVH inthe third on fourth ventricle. In suchcases, the hydmocephalus is more likelyto be communicating (due to oblitema-tive amachnoiditis in the posterior fossaon impaired resorption of CSF in thearachnoid granulations) than obstruc-tive (18, 19).

The effect of IVH on outcome couldperhaps be assessed most accurately ina prospective fashion by comparisonwith a control group without IVH;however, obtaining an unbiased con-trol group presents a problem in thatthe ultimate severity of brain damageis difficult if not impossible to evaluate

prospectively. Our approach was tocorrelate the clinical record and the CT

scan with outcome in each case. In ourseries, outcome was determined byassociated brain injury rather thanby IVH itself. Factors such as pamenchy-ma! hematoma, midline shift, cerebraledema, and vasospasm had predictableeffects which in each case were suffi-cient to explain the neurological out-come. Patients with IVH and no addi-tiona! CT abnormalities had a muchbetter outcome, and those in whompanenchymal damage was confined to“silent” areas of the brain (such as thefrontal lobes) also did well. Our find-ings indicate that IVH per se is seldom

a major contributor to the neurologicaloutcome in these patients.

As CT is now capable of disclosingIVH in many cases where it could notbe diagnosed clinically, this finding by

itself should not be overemphasized. In

the great majority of cases, the eventualoutcome will be determined by thepresence or absence of associated brainabnormalities. Even severe IVH neednot be cause for a gloomy prognosis

in all cases.

SUMMARY

Sixty-eight patients with CT-diag-nosed IVH were reviewed retrospec-tively. The most frequent causes werea ruptured aneurysm, trauma, and hy-pertensive hemorrhage; less frequentcauses included bleeding from an am-teriovenous malformation, a hemon-rhagic diathesis, on probable metasta-ses. In 12 patients the cause was un-

determined. Ruptured aneurysms ofthe anterior cerebral artery and itsbranches (including the ACA) causedIVH more often than other aneurysmsand could frequently be predicted bytheir appearance on non-enhancedCT scans.

The overall mortality rate was 50%.However, 21% of patients returned tonormal on had only mild disability,while 15% had only moderate disabil-ity. Patients with IVH of undeterminedcause had a much better prognosis.IVH due to cerebral trauma carried thehighest mortality mate (71%) and wouldappear to be an ominous finding. Hy-drocephalus which develops one week

or more after IVH can be correlatedwith subamachnoid bleeding on the CTscan rather than with blood in the thirdor fourth ventricle and is thereforemore likely to be communicating thanobstructive. IVH per se seldom has amajor effect on the neurological out-come.

Department of Diagnostic RadiologyVancouver General Hospital

855 W. 12th Ave.Vancouver, B.C. V5Z 1M9

Canada

References

1. Gordon A. Primary ventricular hemor-rhage. Further contribution to a character-istic symptom group. Arch Neurol Psychi-

atry 1938; 39:1272-1276.2. McDonald JV. Midline hematomas sim-

ulating tumors of the third ventricle.Neurology 1962; 12:805-809.

3. Ojemann RG, New PFJ. Spontaneous res-

olution of an intraventricular hematoma.Report of a case with recovery. J Neurosurg

1963; 20:899-902.4. Scott WR, New PFJ, Davis KR, Schnur JA.

Computerized axial tomography of intra-cerebral and intraventricular hemorrhage.Radiology 1974; 112:73-80.

5. Jacobs L, Kinkel WR, Heffner RR Jr. Au-topsy correlations of computerized tomog-

raphy: experience with 6,000 CT scans.Neurology 1976; 26:1111-1118.

6. Baker HL Jr. Campbell JK, HouserOW, et al.Computer assisted tomography of the head.An early evaluation. Mayq Clin Proc 1974;

49:17-27.7. Paxton R, Ambrose J. The EM! scanner.

A brief review of the first 650 patients.Br J Radiol 1974; 47:530-565.

8. Little JR. Blomquist GA Jr. Ethier R. Intra-ventricular hemorrhage in adults. Surg

Neurol 1977; 8:143-149.9. Jennett B, Bond H. Assessment of outcome

after severe brain damage. Lancet 1975;1:480-484.

10. Locksley HB. Report on the cooperative

study of intracranial aneurysms and sub-arachnoid hemorrhage. Section V. Part I.

Natural history of subarachnoid hemor-rhage, intracranial aneurysms and arterio-venous malformations. Based on 6368 casesin the cooperative study. J Neurosurg 1966;

25:219-239.11. Heidrich R. Subarachnoid hemorrhage. In:

Vinken PJ, Bruyn GW, eds. Vascular diseases

of the nervous system. Part II. Handbook ofclinical neurology, Vol. 12. New York: El-

sevier, 1972:68-204.12. Yock DH Jr. Larson DA. Computed to-

mography of hemorrhage from anterior

communicating artery aneurysms, withangiographic correlation. Radiology 1980;134:399-407.

13. Madonick MJ, Savitsky N. Subarachnoidhemorrhage in melanoma of the brain. ArchNeurol Psychiatry 1951; 65:628-636.

14. Adams RD. Victor M. Principles of neurol-ogy. New York: McGraw-Hill, 1977:545.

15. Andrioli GC, Salar G, Rigobello L, MingrinoS. Subarachnoid haemorrhage of unknownaetiology. Acta Neurochir (Wien) 1979;

48:217-221.

16. KrayenbOhl H, Ya�argil G. L’an#{234}vrismec#{233}r#{233}bral.Series Chirurgia M4. Basel: Geigy,1959: 65. [Cited by Andrioli et al. (15)1

17. Allcock JA. Aneurysms. In: Newton TH,Potts DG, eds. Radiology of the skull and

brain: Angiography (Vol. 2, Book 4). St.Louis, Mo.: Mosby. 1974:2435-2489.

18. Shulman K, Martin BF, Popoff N, Ransohoff

J. Recognition and treatment of hydro-cephalus following spontaneous subarach-noid hemorrhage. J Neurosurg 1963; 20:1040- 1047.

19. Kibler RF, Couch RSC, Crompton MR.Hydrocephalus in the adult followingspontaneous subarachnoid haemorrhage.

Brain 1961; 84:45-61.