Acta Neurochir (Wien) (2005) 147: 621–626
DOI 10.1007/s00701-005-0501-y

Clinical Article
Clinical features of postoperative cerebral venous infarction
H. Nakase, Y. Shin, I. Nakagawa, R. Kimura, and T. Sakaki
Department of Neurosurgery, Nara Medical University, Kashihara, Nara, Japan
Received August 12, 2004; accepted January 14, 2005; published online March 18, 2005
# Springer-Verlag 2005

Summary
There is a potential risk of sacrificing the cortical vein during neurosurgical operations, particularly in the interhemispheric or subtemporal
approach. An impaired cortical vein might cause cerebral venous circulatory disturbances (CVCDs) resulting in venous infarction. In this
article, we have reviewed the management and results of eight cases
with symptomatic postoperative venous infarction.
We have encountered eight cases with symptomatic postoperative
venous infarction (0.3%) during the past 5 years. The series is composed
of 3 males and 5 females, with ages that ranged from 43 to 76 years
(mean age of 58.1 years), and consisted of five brain tumors, one
cavernoma, one dural AVF, and one trigeminal neuralgia. Initial symptoms occurred intra-operatively in two, on 0 day after the operation in
one, 1 day in three, 3 days in one, and 4 days in one case. The symptoms
were intra-operative brain edema in two cases, disorientation in one,
cerebellar signs in one, hemiparesis in one, aphasia in two, and headache
in one case. Two cases required surgical intervention. The results were a
good outcome in 6 and a fair outcome in 2 cases.
In conclusion, there are two types of postoperative venous infarction;
severe onset (severe type) and gradual onset (mild type). The former
needs immediate treatment from the intra-operative period onward, and
the prevention of the ongoing venous thrombosis is essential in the latter.
Keywords: Venous infarction; complication; operation.

Introduction
Recently extensive operations such as skull base surgery have become popular in the neurosurgical field.
As compared with previous times, sinuses and bridging
veins are more often exposed with an increased risk
of postoperative complications following an interruption
of the venous circulation [9]. Moreover, neurosurgical
operations are performed more frequently in elderly patients and mild cerebral venous circulation disturbances
(CVCDs), which are without consequence in young patients, often unexpectedly cause severe complications in

the elderly. Therefore, attention has been paid to the study
of cerebral injury that follows CVCDs [3–19].
On the other hand, the clinical symptoms and the prognosis of intra-operative occlusion of cerebral veins are
well known to be quite variable, ranging from no symptoms at all to severe venous infarction. Because of this
variety of symptoms, a low incidence of venous infarction, and the frequent anatomical variations of the cerebral venous system, little information is actually available
about CVCDs.
We reviewed the management and results of 8 cases
with symptomatic postoperative venous infarction, and
examined the clinical features of venous infarction after
the intra-operative sacrifice of cortical veins.

Material and methods
We have encountered 8 cases with symptomatic postoperative venous
infarction (0.3%) during the past 5 years. Emergency operations were
excluded because of multiple influencing factors. The series is composed
of 3 males and 5 females, with ages ranging from 43 to 76 years (mean
age of 58.1 years), and consisted of five brain tumors (2 falx meningiomas, 2 acoustic neurinomas, and one metastatic brain tumor), one
cavernoma, one dural AVF, and one trigeminal neuralgia. The affected
side is the left in 6 and the right in 2 cases (Table 1).

Results
Three cases (cases 1–3) were of severe onset (severe
type), and five cases (cases 4–8) of gradual onset (mild
type). In the severe type (cases 1–3), the initial symptoms occurred intra-operatively in two cases and 0 day
after the operation in one case. The symptoms were
intra-operative brain edema in two cases and cerebellar

622

H. Nakase et al.

Table 1. Summary of the cases with postoperative venous infarction
No

Case

Diagnosis

Sacrificed vein

Symptoms

Interval

Treatment

Outcome

1.
2.

48F
43M

Trigeminal neuralgia (left)
Acoustic neurinoma (left)

Petrosal vein
Petrosal vein

0 day
0 day

Ventricular drainage
Internal decompression

Good
Fair

3.

43F

Acoustic neurinoma (left)

Petrosal vein

0 day

Barbiturate therapy

Good

4.

76F

Cortical vein

3 days

Conservative

Good

5.
6.

48M
69F

Dural AVF
(left transvers–sigmoid)
Metastatic tumor (right frontal)
Meningioma (left falx)

Cerebellar sign
Brain edema
(intra-operative)
Brain edema
(intra-operative)
Aphasia
Drowsy
Headache

1 day
4 days

Conservative
Conservative

Good
Good

7.
8.

74F
64M

Meningioma (left falx)
Cavernom (left frontal)

Aphasia, Hemiparesis
Disorientation

1 day
1 day

Conservative
Conservative

Fair
Good

Sylvian vein
Superior
sagittal sinus
Cortical vein
Cortical vein

signs in one case. This type required immediate and
extensive treatment such as internal decompression, ventricular drainage, and barbiturate therapy. The outcome
was good in 2 cases and fair (slight ataxia) in 1 case.
In all 3 cases, the sacrificed vein was the petrosal vein.
On the other hand, in the mild type (cases 4–8), the
initial symptoms occurred on 1 day in 3 cases, on day
3 in one, and on day 4 in one case. The symptoms were

disorientation in one, hemiparesis in one, aphasia in
two, and headache in one case. This group could be
conservatively treated using depressants for cerebrospinal pressure (i.e., steroid, mannitol, and glycerol), and
the outcomes were good except for one case (case 7).
The sacrificed veins were a cortical vein in 3, the
Sylvian vein in 1, and the superior sagittal sinus in 1 case
(Table 1).

Fig. 1. A case of the left acoustic neurinoma (case 3). The petrosal vein
was injured intra-operatively (a, b). The petrosal vein is indicated by
white arrows in (a). The postoperative emergent CT demonstrated a low
density area with the left cerebellar hemorrhage (c), and angiography
showed the occlusion of the vein (black arrow, b). Lt-VAG The left
vertebral artery angiography

623

Clinical features of postoperative cerebral venous infarction

Illustrative cases
Case 3 (severe type)
A 43-year-old woman presented with hearing disturbance in the left ear and gait disturbance. Magnetic
resonance images revealed a left cerebellopontine angle
tumor (diameter 25 mm). Tumor resection was performed through a left suboccipital craniotomy. The
petrosal vein was injured intra-operatively (Fig. 1a, b).
The tumor was completely removed but brain swelling
occurred at the end of the microsurgery, and barbiturate
therapy was given immediately. The postoperative emergent CT demonstrated a low density area with hemorrhage (Fig. 1c), and angiography showed the occlusion
of the petrosal vein (Fig. 1b). The barbiturate therapy
was postoperatively continued for one week. Cerebellar

swelling improved gradually (neuroradiologically and
clinically), and she was discharged without any neurological deficits one month after the operation.

Case 6 (mild type)
A 69-year-old woman presented with a headache.
Magnetic resonance images showed a left falx meningioma. The tumor was completely removed via the left
interhemispheric approach. The deviated superior sagittal
sinus was injured during the operation (Fig. 2a, b). The
patient showed good recovery without any neurological
deficits, and the CT showed good postoperative findings
(Fig. 2c). However, 4 days after the operation she complained of a headache, and the emergency CT demonstrated venous infarction with hemorrhage (Fig. 2d). We

Fig. 2. A case of the left falx meningioma (Case 6). The deviated superior sagittal sinus (SSS) was injured during the operation (a, b). The patient
showed good recovery without neurological deficit, and the CT showed uneventful postoperative findings (c). However, 4 days later CT demonstrated venous infarction with hemorrhage in the left frontal lobe (d). The postoperative angiography showed the occlusion of the SSS (arrow, b).
Lt-CAG The left carotid artery angiography

624

treated conservatively her with depressants for cerebrospinal pressure (steroid and glycerol), and the headache
disappeared 7 days thereafter. The patient was discharged
without any neurological deficits.
Discussion
The incidence of postoperative venous infarction is
difficult to determine due to an unclear definition, the
rare complications, a variety of symptoms, and the inclusion of other factors during the operation itself (i.e. brain
retraction). As cited in the literature, Kageyama et al. [7]
reported postoperative venous infarction in 13% of the
120 cases operated on by them and Saito et al. [21]
reported postoperative venous infarction in 2.6% cases
after the fronto-temporal bridging vein was cut during
the pterional approach, and Al-Mefty and Krisht [1]
showed that brain edema occurred in 10% cases after
sacrifice of the superficial Sylvian vein. Kubota [11]
reported that 4 of 10 patients with vein sacrifice during
an interhemispheric approach suffered from brain
damage. Roberson [20] reported that the complication
rate of venous insufficiency was 1.5 per 1000 cases of
neurological-skull base surgery.
Roberson [20] divided venous infarction into 2 types;
the acute form and the chronic form. The acute form
manifests itself in the postoperative period and can be
life threatening. The chronic form manifests itself
months or years postoperatively with headaches, disequilibrium, and visual changes due to papilloedema. In the
current study, we dealt with 2 additional cases of perioperative venous infarction in their acute form; severe
and mild types. The severe type requires extensive treatment like internal decompression and barbiturate therapy immediately after the operation. The mild type has a
slow clinical deterioration by gradual thrombus evolution and can be treated conservatively. In all the severe
cases of our series, the sacrificed cortical vein was the
petrosal vein. It is believed that sacrificing the petrosal
vein does not cause a serious problem in most cases.
However, it is quite possible that a large petrosal vein
is the main drainage and its occlusion is dangerous. The
potential danger of this vein’s sacrifice has been addressed several times [6, 24]. Strauss et al. [23] reported
that continuous brain stem auditory evoked potential
monitoring during the test clipping of the vein might
be one method of determining whether the vein can be
cut or not. Sindou [22] showed that the safer approach
is to divide a bridging vein to maintain the venous flow
in an unavoidable situation. More attention should be

H. Nakase et al.

paid to the venous drainage system during surgery of
the cerebellopontine angle.
The intra-operative sacrifice of the anastomotic or
bridging vein usually does not lead to venous infarction
due to the absence of cerebral valves. The recruitment of
collateral pathways occurs during the early phase of
venous occlusion. The severity of CVCDs depends on
the availability of individual venous collaterals. Only
when the collateral venous flow is compromised by additional compression or under extraordinary physiological conditions, e.g. intra-operative brain retraction [12],
excessive changes of systemic blood pressure [17] and
an elderly patient [19], venous infarction occurs. We
previously demonstrated that the growth of the thrombus
coincided with decrease in regional cerebral blood flow
and brain damage using fluorescence angiography and
cortical laser Doppler scanning in the vein occlusion
model [13, 15, 16]. Mild CVCDs in aged patients are
known to frequently cause unexpectedly severe postoperative complications in neurosurgical practice [2, 19].
The frequency of postoperative venous infarction following CVCDs is reportedly higher in older patients
than in young patients [2, 26]. Therefore, venous infarction is an increasingly recognized cause of postoperative
complications in older patients. We examined the influence of advancing age on cerebral venous infarction
using experimentally induced CVCDs model and demonstrated an age-related increase in the rate and size of
venous infarction following vein occlusion, suggesting
that the greater vulnerability to CVCDs in the aged brain
might be attributed to early and extensive hypoperfusion
of circumscribed brain areas drained by the occluded
vein [19]. However, the age-related susceptibility to
venous infarction was not observed in this study.
Effective surgical treatment of deeper brain lesions
requires exposure that frequently can be obtained only
by direct brain retraction. Despite these advantages,
brain compression can have disastrous ischemic sequelae in the territory of the compressed brain [12, 25].
We demonstrated that, compared with vein occlusion
or brain compression alone, the accumulated episode
caused severe ischemia and increased the vulnerability
of the tissue to damage in the rat brain. Also, we showed
that the degree of ischemia correlated well with subsequent brain injury [12]. Therefore, it is advisable to
retract the brain with the least force necessary and for
the shortest time possible. Intermittent brain compression could decrease compression-induced ischemia and
could be more beneficial than a constant one during
neurosurgical operations under vein occlusion, provided

625

Clinical features of postoperative cerebral venous infarction

that compressed pressure declines as the process
advances. The potential benefit of intermittent compression is based on the assumption that the blood circulation under the retractor recovers during a period of
release [8].
Recently, the contribution of ischemia mechanisms
to the pathophysiological consequences of CVCDs has
been reported [3–5, 13–16]: the supply of blood may
fall below a critical threshold in circumscribed brain
areas drained by the occluded veins. Brain damage
subsequent to cortical vein occlusion could be predicted
by continuous LCBF monitoring [9], MRI [10], and
repeated angiographic findings. However, an estimate
from only venous anatomy, which had been examined
retrospectively, was not achieved. Previous studies by
our group have shown that venous occlusion is followed
by regional cerebral blood flow (rCBF) decreases in the
adjacent cortical area, and 90% of histologically analyzed rat brains showed evidence of brain damage upon
occlusion of two cortical veins [13, 16]. Additional
experiments in a similar setting proved the lower autoregulation limit to be shifted upward after even after a
single cortical vein occlusion in the rat [17]. With these
results in mind one may consider the brain of patients to
be at a particular risk for the development of neural
damage during early postoperative periods of lowered
cerebral perfusion pressure. Together with previously
reported results, which showed an upward shift in the
lower limit of autoregulation, these findings may have
implications for blood pressure management in patients
with intra-operative coagulation of large cortical veins –
at least during the acute stage. This may be the pathophysiological basis for subsequent hemorrhagic transformation of venous infarction. The fact that a brain with
CVCDs is very fragile has been underestimated so far.
Extreme care should be taken in such an ailing brain.
There are still many problems awaiting solution in
effective prevention, diagnosis, and treatment for patients with postoperative venous infarction. For example,
we do not know whether venous infarction occurs postoperatively if we injure the specific vein during the
operation, whether venous or tissue pressure recording
could allow assessment of tolerance or intolerance, what
can we do if a venous injury has occurred, whether
barbiturates or hypothermia can limit the deleterious
effect, and whether heparin is useful as it is in the setting
of venous thrombosis, and so on. However, there are
several limitations in relation to clinical study only.
Therefore, basic studies using a model simulating the
clinical setting are also required.

References
1. Al-Mefty O, Krisht AF (1996) The danger veins. In: Hakuba (ed).
Surgery of the intracranial venous system. Springer, Berlin Heidelberg New York Tokyo, pp 338–345
2. Eguchi T (1995) Geriatric cerebrovascular surgery. Geriatric Neurosurgery 7: 9–16 (in Japanese)
3. Frerichs KU, Deckert M, Kempski O et al (1994) Cerebral sinus and
venous thrombosis in rats induces long-term deficits in brain function and morphology – evidence for a cytotoxic genesis. J Cereb
Blood Flow Metab 14: 289–300
4. Fries G, Wallenfang T, Hennen J et al (1992) Occlusion of the pig
superior sagittal sinus, bridging and cortical veins; multistep evolution of sinus-vein thrombosis. J Neurosurg 77: 127–133
5. Gotoh M, Ohmoto T, Kuyama H (1993) Experimental study of
venous circulatory disturbance by dural sinus occlusion. Acta
Neurochir (Wien) 124: 120–126
6. Fujimaki T, Kirino T (1999) Coagulation of the petrosal vein for
MVD. J Neurosurg 90(6): 1148
7. Kageyama Y, Watanabe K, Kobayashi S, Nakamura H, Satoh A,
Watanabe Y, Yamaura A (1996) Postoperative brain damage due to
cerebral vein disorders resulting from the pterional approach. In:
Hakuba (ed). Surgery of the intracranial venous system. Springer,
Berlin Heidelberg New York Tokyo, pp 311–315
8. Kaido T, Nakase H, Nagata K, Ohtsuka H, Sakaki T (2001)
Intermittent isometric exposure prevents brain retraction injury
under venous circulatory impairment. Neurol Res 23(7): 739–744
9. Kanno T, Kasama A, Shoda M, Abe M (1992) Intraoperative
monitoring on the occlusion of the venous system. Neurosurgeons
11: 51–59 (in Japanese)
10. Kimura R, Nakase H, Sakaki T, Taoka T, Tsuji T (2003) Vasogenic
edema and VEGF expression in a rat two-vein occlusion model.
Acta Neurochir (Wien) [Suppl] 86: 213–217
11. Kubota M, Ono J, Saeki N, Yamaura A (1996) Postoperative brain
damage due to sacrifice of bridging veins during the anterior
interhemispheric approach. In: Hakuba (ed). Surgery of the intracranial venous system. Springer, Berlin Heidelberg New York
Tokyo, pp 291–294
12. Nagata K, Nakase H, Kakizaki T, Otsuka H, Sakaki T (2000) The
effect of brain compression under venous circulatory impairment.
Neurol Res 22(7): 713–720
13. Nakase H, Heimann A, Kempski O (1996) Local cerebral blood
flow in a rat cortical vein occlusion model. J Cereb Blood Flow
Metab 16: 720–728
14. Nakase H, Heimann A, Kempski O (1996) Alterations of regional
cerebral blood flow and tissue oxygen saturation in a rat sinus-vein
thrombosis model. Stroke 27: 720–727
15. Nakase H, Nagata K, Otsuka H, Sakaki T, Kempski O (1997) An
experimental model of intraoperative venous injury in the rat. Skull
Base Surgery 7(3): 123–128
16. Nakase H, Kempski O, Heimann A (1997) Microcirculation after
cerebral venous occlusion as assessed by laser Doppler scanning.
J Neurosurg 87: 307–314
17. Nakase H, Nagata K, Otsuka H, Sakaki T, Kempski O (1998) Local
cerebral blood flow autoregulation following ‘‘asymptomatic’’
cerebral venous occlusion in the rat. J Neurosurg 89: 118–124
18. Nakase H, Kaido T, Shin Y, Nishioka T, Sakaki T (2002) Clinical
study on postoperative venous infarction. Jpn J Phlebol 13: 181–185
(in Japanese)
19. Otsuka H, Nakase H, Nagata K (2000) Effect of age on cerebral venous circulation disturbances in the rat. J Neurosurg 93:
298–304
20. Roberson JB Jr, Brackmann DE, Fayad JN (2000) Complications of
venous insufficiency after neurotologic-skull base surgery. Am J
Otol 21(5): 701–705

626

H. Nakase et al.: Clinical features of postoperative cerebral venous infarction

21. Saito F, Haraoka J, Ito H, Nishioka H, Inaba I, Yamada Y (1998)
Venous complications in pterional approach; About frontotemporal bridging veins. Surg Cereb Stroke 26: 237–241 (Jpn)
22. Sindou M, Auque J (2000) The intracranial venous system as a
neurosurgeon’s perspective. Adv Tech Stand Neurosurg 26: 131–216
23. Strauss C, Neu M, Bischoff B, Romstock J (2001) Clinical and
neurophysiological observations after superior petrosal vein obstruction during surgery of the cerebellopontine angle: case report.
Neurosurgery 48(5): 1157–1159
24. Tsukamoto H, Matsushima T, Fujiwara S, Fukui M (1993)
Peduncular hallucinosis following microvascular decompression
for trigeminal neuralgia: case report. Surg Neurol 40(1): 31–34
25. Tsutsumi K, Shiokawa Y, Sakai T, Aoki N, Kubota M, Saito I
(1991) Venous infarction following the interhemispheric approach
in patients with acute subarachnoid hemorrhage. J Neurosurg 74:
715–719
26. Wiebers DO, Adams HP Jr, Whisnant JP (1990) Animal model
stroke: are they relevant to human disease? Stroke 21: 1–3

Comments
This is a thoughtful clinical paper by a group with a strong record of
appropriate animal research. It concerns a condition which must be
(more) common, but is, as the authors suggest, confused with retraction
contusion=ischaemia. The infrequency of pre and post operative angiography in current neurosurgical practice contributes to the infrequency
of establishing the correct diagnosis. Non-invasive venography, either

CT or MR venography, is adequate for the diagnosis of venous sinus
thrombosis but is severely restricted in the demonstration of cortical
venous interruption or thrombosis. The authors refer regularly to their
experimental work but admit that they have failed to show an age related
effect in their clinical practice.
E. Teasdale
Glasgow
The authors report their retrospective experience of 8 patients who
suffered postoperative venous infarction after elective surgery. The clinical features of the 8 cases are described and the causal venous injuries
are pinpointed. It is common knowledge that veins should not be sacrificed and that brain retraction should not be used anymore.
Postoperative venous infarction is a serious complication and apart
from trivial primary preventive strategies as mentioned above we do not
know what we can do if a venous injury has occurred. We do not know
whether venous or tissue pressure recording could allow assessment of
tolerance or intolerance. We do not know whether barbiturates or
hypothermia can limit the deleterious impact and we do not know
whether heparin is useful as it is in the setting of venous thrombosis.
H.-J. Steiger
Duesseldorf

Correspondence: Hiroyuki Nakase, Department of Neurosurgery,
Nara Medical University, 840 Shijo-cho, Kashihara, Nara 634-8522,
Japan. e-mail address: nakasehi@naramed-u.ac.jp