ahn ,2010

Introduction

One uncommon type of intracerebral hemorrhage (ICH),

the primary intraventricular hemorrhage (PIVH), is a non-

traumatic hemorrhage confined to the ventricular system,

without any bleeding into the cerebral parenchyma, and its

incidence is very rare.1)3)13)19) Sanders described PIVH first,

more than a century ago, but few researchers have studied it

since then. Prior to the introduction of the computed

tomography (CT) scan, knowledge concerning PIVH came

from autopsy series, which usually included patients with

massive intraventricular hemorrhages who had had poor

prognoses, or, more rarely, from a few patients undergoing

surgery.

The definition of PIVH is “a hematoma either confined

completely within the ventricular system or arising within 15

mm of the ventricular wall.”1)13)19) The presence of

intraventricular blood is a source of general alarm to

neurosurgeons due to the possibility of hydrocephalus and

ensuing brain damage. Although small amounts of blood

resolve quickly, with no sequelae, the neurosurgeon often

encounters a patient who has moderate or severe IVH

involving all the ventricles. Relatively little is known about

IVH's clinical and imaging features, and even less about its

prognosis and mortality predictors. The present study aimed

to analyze the clinical and imaging features, etiology,

prognosis, and predictors of outcome in PIVH.

KOR J CEREBROVASCULAR SURGERYSeptember 2010 Vol. 12 No 3, page 159-164

Non-Traumatic Primary Intraventricular Hemorrhage

Department of Neurosurgery, College of Medicine, Ewha Womans University, Seoul, Korea

Sang Wook Ahn, MD·Sung-Kyun Hwang, MD

ABSTRACT

Objective : Non-traumatic primary intraventricular hemorrhage (PIVH) in adults has rarely been reported. This study aimed to identifyclinical features, risk factors, and neurological outcomes of PIVH in adults. Methods : We retrospectively reviewed the clinical data,complementary examinations, outcomes, and computed tomography (CT) scans for non-traumatic PIVH occurring between 2002 and 2008at our institutions. We defined PIVH as “a hematoma caused by non-traumatic factors, either confined completely within the ventricularsystem or arising within 15 mm of the ventricular wall.” Result : Among the 19 PIVH patients treated in our institutions, the mean age was55.5 years (range 19-81), and the male to female ratio was 0.72. Symptom onset was abrupt in 18 patients and progressive in 1. The mostfrequent complaint was headache (47%) followed by nausea and vomiting (38%). Seven patients (38%) had mental status impairment,ranging from stupor to semicoma. The causes of PIVH were unknown, although arteriography showed a suspicious vascular malformation in1 patient. Seven (37%) patients died or entered a vegetative state. All survivors became functionally independent. No patients receivedventriculoperitoneal shunts. Two of the surviving patients had limb weakness. Overall, 7 patients (37%) had GOS scores 1 to 2 and 12 (63%)had GOS scores 3 to 5. Conclusion : Although present technology allows clinicians to reach a diagnosis in many PIVH patients, thecondition's etiology remains unknown, and its clinical manifestations vary, due to the rarity of these cases. Overall results and prognosesseem relatively better than in secondary intraventricular hemorrhage. However, evaluating and clarifying the nature of PIVH requiresexperience with more cases. (Kor J Cerebrovascular Surgery 12(3):159-164, 2010)

KEY WORDS : Intraventricular hemorrhage·Risk factor·Outcome·Stroke

159

논문접수일 : 2009년 12월 27일심사완료일 : 2010년 06월 21일교신저자 : Sung-Kyun Hwang, Department of Neurosurgery,

Mokdong Hospital, 911-1 Yangcheon-gu, Seoul, 158-710, South Korea

전화 : (02) 2650-2872•전송 : (02) 2650-0948

E-mail : [email protected]

Patients and Method

We retrospectively reviewed the clinical data,

complementary examinations, outcomes, and computed

tomography scans for non-traumatic PIVH occurring

between 2002 and 2008 at our institutions. We defined PIVH

as “a hematoma caused by non-traumatic factors, either

confined completely within the ventricular system or arising

within 15 mm of the ventricular wall.” Evaluations of the

clinical features included type of onset, presenting type and

symptoms, risk factors, and treatment methods. We

evaluated all CT scans in detail for the location of blood in

the ventricle and for hydrocephalus and reviewed other

radiological examinations for any abnormalities.

We evaluated patient outcomes as death directly related to

PIVH, survival with a significant neurologic deficit, or

survival with minimal neurologic deficits and functional

independence. The patients' follow-up periods ranged from 3

month to 2 years.

Results

Patient age ranged from 19-81 years (mean age 55.5

years), with the majority (79%) between the ages of 38 and

71 years. There were 19 patients: 8 females and 11 males.

Hematoma sizes ranged between 2 and 42 cc, with 7 under

10 cc, 6 between 10 and 30 cc, and 6 between 30 and 50 cc.

Symptom onset was abrupt in 18 patients and progressive

in 1 patient. The most frequent complaint was headache

(47%), followed by nausea and vomiting (38%), and 7

patients (38%) suffered from impaired mental statuses,

ranging from stupor to semicoma. One patient had cranial

nerve abnormalities, consisting of extraocular movement

abnormalities (dysconjugate gaze) and facial droop. Motor

deficits and abnormal reflexes (hyperreflexia, pathologic

reflexes) were present in 2 patients.

Ten patients had elevated blood pressure, with systolic

pressure ranging from 150-230 and diastolic pressure from

80-120 mm Hg. Intraventricular hemorrhage was present in

the lateral ventricles in 9 scans (7 bilateral, 2 unilateral),

while 10 scans showed blood in all ventricles. Of the 9

patients presenting with or developing, hydrocephalus, 8

underwent extraventricular drainage. The remaining 1 was

treated medically (Table 1).

The associated risk factors were hypertension in 9

patients, diabetes mellitus in 3 patients, and end stage renal

disease, angina, and valvular heart disease in 1 patient each.

Angiography was performed on 8 patients. These revealed

no aneurysms, definite arteriovenous malformations, or other

vascular anomalies except for a suspicious vascular

malformation in 1 patient.

Regarding outcomes, 7 (37%) patients died or entered a

vegetative state, of whom 4 (58%) had hematomas less than

30 cc and 3 (42%) had hematomas greater than 30 cc. Of the

remaining 12, all became functionally independent; 2 (16%)

had hematomas greater than 30 cc and 10 (84%) has

hematomas less than 30 cc. These results suggest that

hematoma volume could correlate with prognosis, but

demonstrating this would require a large-scale study needed.

Of the functionally independent patients, 2 had limb

weaknesses. Overall, 7 patients (37%) had Glasgow

Outcome Scale (GOS) scores of 1 to 2 and 12 (63%) scored

3 to 5. No permanent CSF diversions were performed.

Discussion

1. Clinical Features

Excluding premature infants, PIVH is very rare and

accounts for only 3.1% of all non-traumatic central nervous

system hemorrhages. Considering only the adult population,

PIVH is still relatively rare, comprising only 3.1% of all

intracranial hemorrhages, and clinicians have reported only a

small number of cases.1)9)13)19) Although PIVH is rarely

encountered clinically, the real frequency is unknown, as

PIVH may in fact be the cause of any sudden death, in

particular, any death occurring before the patient reaches the

hospital or receives a CT scan. Hospital series may thus be

biased toward more benign cases, as current diagnostic

imaging capabilities allow the detection of less severe cases.

CT can now confirm a diagnosis of PIVH easily and rapidly,

depicting blood when it occurs in the ventricular system

only, even when it is a small clot in a circumscribed portion

of the ventricular system (usually a lateral ventricle).

PIVH affects all age groups, with a mean onset age of 60

years, similar to that reported in other series.13)19) Earlier

PIVH diagnoses relied on a sudden onset, with coma or

decreased consciousness level, nausea and/or vomiting,

nuchal rigidity, severe headache, and bilateral extensor


160 Kor J Cerebrovascular Surgery 12(3):159-164, 2010

plantar responses, followed by a severe and often fatal

outcome.

Nearly all our patients complained of headache. Nausea

and/or vomiting, nuchal rigidity, and unilateral or bilateral

extensor plantar responses were frequent, but focal motor

symptoms and signs were absent or mild. This probably

correlated with midline hemorrhage without parenchymal

damage, although hematoma's asymmetric features could

explain the focal signs. As noted in other studies,

hydrocephalus was also a frequent complication in our

patients but only occurred when the third or fourth ventricle

was involved.

PIVH without a parenchymal component is unusual,

carrying an average mortality rate of about 40%. When it

occurs in association with aneurysm and a large blood

volume causing hydrocephalus, the mortality rate

approaches 90%.12)13)19) Unlike intraparenchymal hemorrhage

patients, PIVH survivors often left the hospital with no

physical disability, which presumably indicates the absence

of parenchymal damage. On the other hand, residual

cognitive dysfunction is relatively frequent in such patients.

2. Etiology

The etiology of PIVH is varied, and, in some patients,

unknown. Some researchers have argued, by analogy with

subarachnoid hemorrhage, that vascular malformations

adjacent to the ependymal lining or very small

malformations, which either self-destroy because of the

hemorrhage or do not opacify on angiography, may cause

these cases. In some reports, patients whose angiographic

Sang Wook Ahn∙Sung-Kyun Hwang

161Kor J Cerebrovascular Surgery 12(3):159-164, 2010

Table 1. Clinical data

Pt Age CT(IVH) vol. Onset HC DSA Symptom & Sign Risk factor

Tx. Outcome

No. /Sex

1 70/m all 30 A + - semicoma none EVD death

2 49/m all 32 A + - semicoma none mannitol deathvavular heart

3 62/f LV 16 A - - headache mannitol no impairmentdisease

4 44/m LV 23 A + nl stupor none EVD memory deficit

5 45/m LV 19 A - nl headache none mannitol no impairment

6 19/m LV 6 A - - headache none mannitol no impairment

7 76/f all 35 A - - headache HT mannitol no impairment

8 54/m LV 6 A - nl headache HT mannitol no impairment

9 63/m all 32 A + - stupor HT/DM EVD death

10 81/f LV 24 A - - headache HT mannitol no impairment

11 76/f all 9 A + - semicoma HT/DM EVD death

12 74/f all 7 A + - semicoma ESRD/DM EVD death

13 66/f all 39 A + - semicoma none EVD vegetative state

14 45/m all 6 P - nl headache none mannitol no impairment

EOM

15 50/f all 20 A + - abnormality, facial HT EVD death

droop

16 38/m LV 2 A - nl hemiplegia HT mannitol hemiplegia

17 40/f LV 5 A - nl headache HT mannitol no impairment

susp18 32/m all 45 A + hemiplegia none EVD hemiplegia

AVM

19 71/m LV 10 A - nl headache HT/angina mannitol no impairment

Tx.=treatment, HC=hydrocephalus, m=male, f=female, IVH=intraventricular hemorrhage, nl = normal,

LV=Lateral ventricle, 3V=third ventricle, 4V=fourth ventricle, A=acute, P=progressive, HT=hypertension, DM=diabetes mellitus, ESRD=end stage renal

disease, AVM=arteriovenous malformation, EOM=extraocular movement

findings were normal while they lived showed a vascular

malformation at autopsy.13)22) PIVH is frequently caused by

vascular malformations, including cavernous angiomas and

aneurysms. Some researchers have suggested that

malformations are more probable in young patients.13)22)

PIVH was attributed to hypertension in 9 of our patients.

However, we prefer to speak in terms of association rather

than causality, because we cannot prove that hypertension in

fact caused the bleeding. Hypertension was the only

identified risk factor in 50%-70% of patients in the

series,1)11)13) and it was present in all patients reported by

Gates et al.6)

In literature review, the most frequent cause of PIVH was

a very small, parenchymatous, hypertensive hemorrhage,

originating in tissues very close to the ventricular system

(caudate and thalamus), of a type that can go undetected by

CT or arise in the choroid plexus. We therefore believe that

hypertension can induce hemorrhage in choroidal arteries.

This study agrees with previous work regarding the high

probability that thalamic ICH will rupture into the ventricles.

This may be due to the thalamus’s anatomic proximity to the

third ventricle, as well as to blood's tendency to spread

medially. Finally, one of our patients had a suspected

intraventricular tumor, and previous studies have also

reported such. A wide range of other possible causes of

PIVH has been reported in the literature. Recently, CT,

computed tomography angiography (CTA), magnetic

resonance image (MRI), and magnetic resonance

angiography (MRA) have come into use as noninvasive

screening procedures for intracranial vascular abnormalities.

Hence, these noninvasive imaging modalities are useful for

screening for PIVH, particularly in older, hypertensive

patients. MRI is useful for investigating possible vascular

malformations, and MRA and CTA are likewise useful for

aneurysm investigation. However, a negative result cannot

completely exclude a vascular lesion, and a definitive

diagnosis and treatment planning may require conventional

cerebral angiography.22)23)

3. Prognosis & Risk factors

Hydrocephalus is a frequent PIVH complication and is

probably due to the obstruction of cerebrospinal fluid (CSF)

circulation or impairment in meningeal absorption.

Intuitively, it makes sense that IVH impacts clinical

outcomes through several possible mechanisms.20) The

ventricular system can provide an outlet for IVH expansion

with perhaps less resistance than does brain parenchyma,

and IVH volume can independently exert a mass effect on

the surrounding brain tissue. In addition, obstruction of the

cerebrospinal fluid can cause obstructive hydrocephalus,

which can raise intracranial pressure, resulting in global

impairment.7)12)19) Finally, the presence of blood in the CSF

may represent a global injury to the brain, whereas the

original IVH may only result in a focal deficit. It was not

surprising that hematoma volume and degree of midline shift

correlated positively with mortality. If the IVH was close to

the foramen of Monro or within the third or fourth

ventricles, the likelihood of obstructive hydrocephalus

increased, which increased the chance of a lethal result.

Acute hydrocephalus contributed significantly to death and

deterioration in the acute stages. Brott et al.2) speculated that

hemorrhages originating in the periventricular regions are

more likely to expand toward the nearby ventricle because

the fluid spaces are more compressible than the surrounding

brain parenchyma is, as the CSF redistributes extracranially.

The results of Young et al.21) support this. They found that

thalamic hemorrhages, with their proximity to the ventricles,

correlated with larger IVH volumes and worse outcomes8)10)19)

and showed a close correlation between baseline

intraventricular blood volume and outcome. Similarly, they

found deep hemorrhages in association with increased

incidence of obstructive hydrocephalus, which results in

raised intracranial pressure, beyond the direct effects due to

the hemorrhage volume. Prognosis would logically seem

related to the amount of blood; the more severe the

hemorrhage, the worse the prognosis, and our study shows

possibility of this relationship. However, some authors have

not found a lethal volume of intraventricular hematoma or a

correlation between hematoma size and outcome.1)3)13)

The correlation between IVH and increased morbidity is

not clearly understood. Certainly, the CSF outflow blockage

with obstructive hydrocephalus and subsequent increase of

intracranial pressure (and, therefore, the reduction in

cerebral perfusion pressure) plays an important prognostic

role.20) Hydrocephalus is an independent outcome predictor

in spontaneous IVH. The observation that IVH volume may



user

Highlight

user

Highlight

user

Highlight

be associated with a commensurate decrease in global

cerebral blood flow further supports this concept.

Interestingly, in other studies, outcomes did not differ in

those patients treated by ventriculostomy. This may support

the hypothesis that the enhanced morbidity associated with

IVH is due, at least in part, to the pressure the clot exerts on

periventricular structures. This emphasizes the possible

impact of any treatment that prevents IVH or limits further

IVH expansion. Others have correlated total IVH volume

with outcome via a “lethal volume” of IVH, which is greater

than 20 mL.8)21) Proposed mechanisms for this IVH-induced

morbidity include development of hydrocephalus, decreased

consciousness, and IVH-induced inflammation. The greater

the age and the further the blood pressure is out of the

normal range (whether higher or lower), the worse the

prognosis.12) Research has also confirmed that clinical

condition and level of consciousness on admission

significantly influence prognosis.2)13) A good prognosis in

such circumstances seems due to the absence of

parenchymatous damage.13)

Some authors have noted memory difficulties, a common

consequence in such patients.5)13)19) Some have suggested that

these memory problems are due to a lesion of the fornix or

dorsomedian thalamic nuclei. In addition, in patients whose

initial clinical condition is not serious, early treatment

strategies aimed to prevent increases in ventricular

enlargement may improve their prognoses.

4. Treatment

The appropriate treatment for PIVH is not clear, and the

prognosis is variable, with mortality rates previously

reported as 40-83%.1)9)13)19)20) The need of extraventricular

drainage in IVH patients is seven times greater than in other

types of hemorrhage; however, the poor outcome rate for

EVD patients is alarmingly high.8)14)20) While this is likely due

to the larger volume of IVH and hydrocephalus, EVD related

morbidity such as rebleeding and infection may contribute to

the observed poor outcome in these patients, although this

has been recently debated. IVH patients' poor outcomes,

with or without EVD, should encourage the development

and testing of new treatments for IVH, such as

intraventricular instillation of thrombolytic agents.4)15-18)

Clinicians have used for IVH due to a variety of causes,

with good results and no bleeding complications, and,

consequently, can make good use of it. However,

establishing that it improves outcomes or reduces the need

for ventriculoperitoneal shunting would be a complicated

analysis, due to both IVH's variable causes and the many

prognostic factors involved. Perhaps the aggressive

management of hydrocephalus and the use of thrombolytic

agents in the future will improve the outcomes of patients

with primary intraventricular hemorrhage.

Conclusion

Although present technology allows us to reach a

diagnosis of PIVH in many patients, its etiology and clinical

manifestations are varied and remain largely unknown, due

to rarity of such cases. Our study shows a possible

relationship between hemorrhage volume and prognosis,

with overall results and prognoses relatively better than for

secondary IVH. However, evaluating and clarifying the

nature of PIVH requires experience with more cases.

REFERENCES

1) Angelopoulos M, Gupta SR, Azat Kia B. Primaryintraventricular hemorrhage in adults: clinical features, riskfactors, and outcome. Surg Neurol 44:433-7, 1995

2) Brott T, Broderick J, Kothari R, Barsan W,Tomsic T,Sauerbeck L, et al. Early hemorrhage growth in patients withintracerebral hemorrhage. Stroke 28:1-5, 1997

3) Darby DC, Donnan CA, Saling MA, Walsh KW, Bladin PF.Primary intraventricular hemorrhage: clinical andneuropsychological finding in a prospective stroke series.Neurology 38:68-75, 1988

4) Findlay M, Croce MCA, Weir B. Treatment of intraventricularhemorrhage with tissue plasminogen activator. Neurosurgery32:941-7, 1993

5) Fogelholm R, Nuutila M, Vuorela AL. Primary intracerebralhemorrhage in the Jyvaskyla region, central Finland 1985?89:incidence, case fatality rate, and functional outcome. J NeurolNeurosurg Psych 55:546-552, 1992

6) Gates PC, Barnett HJM, Vinters HV, Simonsen RL, Siu K.Primary intraventricular hemorrhage in adults. Stroke 17:872-7, 1986

7) Goh KY, Poon WS. Recombinant tissue plasminogen activatorfor the treatment of spontaneous adult intraventricularhemorrhage. Surg Neurol 50:526-32, 1998

8) Hallevi H, Albright KC, Aronowski J, Barreto AD, Martin-Schild S, Khaja AM, et al. Intraventricular hemorrhage:Anatomic relationships and clinical implications. Neurology70:848-52, 2008

9) Hill MD, Silver FL, Austin PC, Tu JV. Rate of strokerecurrence in patients with primary intracerebral hemorrhage.Stroke 31:123-7, 2000

Sang Wook Ahn∙Sung-Kyun Hwang

163Kor J Cerebrovascular Surgery 12(3):159-164, 2010

10) Little JR, Blomquist GA Jr, Ethier R . Intraventricularhemorrhage in adults. Stroke 8;143-9, 1987

11) Lee KS, Bae HG, Yun IG. Recurrent intracerebralhemorrhage due to hypertension. Neurosurgery 26:586-90,1990

12) Liu Y, Yang Y, Zhang Q, Zhang W, Zhu S, Li X, et al. A studyof classification of spontaneous intraventricular haemorrhage:a report of 324 cases. J Clin Neurosci 5:182-5, 1998

13) Martí-Fábregas J, Piles S, Guardia E, Martí-Vilalta JL.Spontaneous primary intraventricular hemorrhage: clinicaldata, etiology and outcome. J Neurol 246:287-91, 1999

14) Maruishi M, Shima T, Okada Y, Nishida M, Yamane K, Okita S.Clinical findings in patients with recurrent intracerebralhemorrhage. Surg Neurol 44:444-9, 1995

15) Mayer SA, Brun NC, Broderick J, Stephen D, Michael ND,Brett ES, et al. Safety and feasibility of recombinant factorVIIa for acute intracerebral hemorrhage. Stroke 36:74-9, 2005

16) Naff NJ. Intraventricular hemorrhage in adults. Curr TreatOptions Neurol 1:173-8, 1999

17) Naff NJ, Carhuapoma JR, Williams MA, Bhardwaj A,Ulatowski JA, Bederson J, et al. Treatment of intraventricularhemorrhage with urokinase: effects on 30-day survival. Stroke31:841-7, 2000.

18) Naff NJ, Hanley DF, Keyl PM, Tuhrim S, Kraut M, Bederson

J, et al. Intraventricular thrombolysis speeds blood clotresolution: results of a pilot, prospective, randomized, double-blind, controled trial. Neurosurgery 54:577-84, 2004.

19) Passero S, Ulivelli M, Reale F. Primary intraventricularhaemorrhage in adults. Acta Neurol Scand 105:115-9, 2002

20) Steiner T, Diringer MN, Schneider D, Mayer SA, Begtrup K,Broderick J, et al. Dynamics of intraventricular hemorrhage inpatients with spontaneous intracerebral hemorrhage: riskfactors, clinical impact, and effect of hemostatic therapy withrecombinant activated factor VII. Neurosurgery 59:767-74,2006

21) Young WB, Lee KP, Pessin MS, Kwan ES, Rand WM, CaplanLR . Prognostic significance of ventricular blood insupratentorial hemorrhage: a volumetric study. Neurology40:616-9, 1990

22) Wakai S, Kumakura N, Nagai M. Lobar intracerebralhemorrhage. A clinical, radiographic, and pathological studyof 29 consecutive operated cases with negative angiography. JNeurosurg 76:231-8, 1992

23) Zhu XL, Chan MS, Poon WS. Spontaneous intracranialhemorrhage: which patients need diagnostic cerebralangiography? A prospective study of 206 cases and review ofthe literature. Stroke 28:1406-9, 1997



ahn ,2010

Documents