(Stroke. 1995;26:1399-1403.)
© 1995 American Heart Association, Inc.
Articles |
From the Department of Neurology, University Hospital of the Saarland, Homburg/Saar, and the Department of Psychiatry, Winnenden (G.P.H.), Germany.
Correspondence to Martin Müller, MD, Department of Neurology, University Hospital of the Saarland, Oscar-Orth-str 3, D-66421 Homburg/Saar, Germany.
| Abstract |
|---|
|
|
|---|
Methods Thirty-five unselected patients (24 men, 11 women; mean age, 51±18 years) with bacterial (n=33) or fungal (n=2) meningitis prospectively underwent serial transcranial Doppler sonography recordings of mean blood velocity (MBV) and pulsatility index in the middle (MCA) and anterior (ACA) cerebral arteries, as well as recordings of the ratio of the MBV of the MCA and internal carotid artery (MCA/ICA ratio) on days 1, 3, 5, 8, 14, and 21 after admission. The results were correlated with the Glasgow Coma Scale (days 1 to 14), the occurrence of focal cerebral signs, and the Glasgow Outcome Scale (short-term outcome, day 21). An MCA stenosis was diagnosed by an MBV of 120 cm/s or more or an MCA/ICA ratio of more than 3. An ACA stenosis was diagnosed by an MBV of 100 cm/s or more.
Results Transient stenoses occurred most frequently
between days 3 and 5 and were detected in 18 patients (51%). Seventeen
patients remained without a stenosis. Patients with
stenoses showed a significantly poorer mean Glasgow Coma Scale
score from day 3 (9±4) to day 14 (11±4) than patients without a
stenosis (day 3: 13±4, P<.01 by t test;
day 14: 14±1, P<.05). The mean Glasgow Outcome Scale score
was not significantly different between both groups. The occurrence of
mainly transient focal cerebral signs was significantly related to the
number of narrowed vessels per patient (P<.05,
2 test).
Conclusions Stenoses of the intracranial arteries occur frequently in bacterial meningitis and are associated with a complicated course of the disease.
Key Words: cerebral arteries meningitis ultrasonics vasospasm
| Introduction |
|---|
|
|
|---|
| Subjects and Methods |
|---|
|
|
|---|
Each patient received cranial CT (Siemens Somatom DR) before the first lumbar puncture. Thereafter, a CT scan or MRI was performed when clinically indicated. EEGs were recorded repeatedly during hospitalization. All patients initially received a standard antibiotic therapy consisting of penicillin G, cefotaxime, and gentamycin with adaptation according to the following microbiological susceptibility. Patients with signs of raised ICP or with brain edema on CT scan were treated with hyperosmolar substances. When indicated, mechanical hyperventilation was performed with an arterial blood carbon dioxide tension of 31 mm Hg or more. Dexamethasone therapy was not used.
To correlate the course of the TCD findings with the course of the ICP, the lumbar opening pressure (in millimeters of water) was repeatedly measured by lumbar puncture soon after the TCD examinations in three patients. A fourth patient required surgery for external ventricle drainage due to acute hydrocephalus shortly after admission. In this patient the ICP (in millimeters of water) was recorded continuously with the use of an external ventricle drainage system (EVD Set, PFM), with the drip chamber secured approximately 12 cm above the ventricle level. An ICP of 200 mm H2O or less was considered normal.
According to our study protocol, the first TCD recordings (TC 2-64, EME) were performed within 12 hours after admission of the patients. This included the bilateral insonation of the MCA (at a depth of 50 to 55 mm) and of the ACA (at a depth of 60 to 70 mm) via the temporal approach above the zygomatic arch and of the ICA (at a depth of 35 to 40 mm) via the submandibular approach according to previously published criteria.17 18 The follow-up TCD examinations included the same recordings and were fixed for days 3, 5, 8, 14, and 21 after admission. Additional TCD recordings on the days between were allowed but not necessary. The maximum systolic velocity (Vsystolic), end-diastolic velocity (Vend-diastolic), and MBV were recorded in each artery after blood velocities of 10 cardiac cycles showed a constant velocity steady state. As a vascular resistance index the PI19 was calculated as follows: PI=(Vsystolic-Vend-diastolic)/MBV. Additionally, the recorded blood velocity waveforms were analyzed according to the absence or presence of the diastolic notch, which was found to be present in blood velocity elevation due to a presumed vasospasm but absent in severely increased blood velocities as a result of hyperemia.20
At the time of each TCD recording, the patients were clinically scored with the use of the GCS.21 Additionally, focal cerebral signs such as hemiparesis, aphasia, neglect syndromes, cortical sensory syndromes, focal seizures, and focal slowing or focal epileptic activity on EEG recordings were recorded separately. Because patients with an uncomplicated course of bacterial meningitis have usually recovered after 3 weeks5 and additionally because most hemodynamic disturbances that occur in bacterial meningitis have normalized within 3 weeks,16 22 we chose day 21 after admission to evaluate the short-term outcome using the GOS, as follows: 1, death; 2, persistent vegetative state; 3, severe neurological deficit; 4, slight neurological deficit; and 5, complete recovery.23
Our control group for the TCD recordings consisted of 69 healthy volunteers (34 men, 35 women; mean±SD age, 44±17 years; range, 17 to 80 years) and has been described previously.18 The mean values of MBV and PI for the MCA are 57±13 cm/s and 0.83±0.15, respectively, and for the ACA are 46±11 cm/s and 0.85±0.20, respectively.
Because of the lack of corresponding angiographic and TCD examinations in bacterial meningitis, it is unknown whether a severely increased MBV corresponds to a narrowing of a basal cerebral artery in this condition. In subarachnoid hemorrhage, an MBV in the MCA of 120 cm/s or more or an MCA/ICA ratio of more than 3 corresponds well with angiographically demonstrated vasospasm.24 25 Cerebral hyperemia also elevates the MBV. However, as shown for inspired carbon dioxide, hyperemia increases the MCA/ICA ratio only by approximately 10%.26 In our control subjects, the highest normal MCA/ICA ratio was 3, which is in accordance with a previous report.25 The criterion used for our patients to indicate a narrowing of the MCA was an MBV in the MCA of 120 cm/s or more or an MCA/ICA ratio of more than 3 when the diastolic notch was present in the blood velocity waveform. For an ACA vasospasm, conclusive corresponding data between angiography and TCD do not exist. An MBV of 120 cm/s in the MCA is more than twice the normal MBV of our control subjects. To indicate a stenosis of the ACA, a threshold MBV of 100 cm/s was chosen, which was also more than twice the normal MBV of the ACA in our control subjects. As for the MCA, the diastolic notch had to be present in the ACA blood velocity waveform.
With respect to the total number of narrowed arteries identified during the examination period, the patients were classified into four groups: patients without a stenosis, with stenosis of one artery only, with stenoses of two arteries, and with stenoses of more than two arteries.
Statistical Analysis
All values are reported as mean±SD. The courses of the MCA MBV,
the MCA PI, and the MCA/ICA ratio as well as of the ACA MBV and the ACA
PI were compared with respect to the presence or absence of a
stenosis of the relevant artery by the unpaired t
test. This test was also used to compare the GCS score on days 1 to 14
and the GOS score on day 21 between the patients without and those with
narrowed arteries. Using the
2 test (after
Yates' correction), we analyzed the occurrence of focal
cerebral signs with respect to the number of narrowed basal cerebral
arteries per patient and with relation to the causative pathogen. The
effects of age and of the causative pathogen on the GCS from days 1 to
14 and on the GOS were analyzed by stepwise regression
analysis. Regression analysis was also used to
correlate CSF total protein content with the presence of narrowing of
the MCA.
| Results |
|---|
|
|
|---|
|
An MCA stenosis occurred bilaterally in 9 patients and
unilaterally in an additional 9. An ACA stenosis was bilateral
in 3 patients and unilateral in 4. Of the 10 ACA stenoses, 9
occurred accompanying an MCA stenosis and 1 was isolated.
According to the number of stenoses of the basal cerebral
arteries, 17 patients were without a stenosis, 7 patients
exhibited 1 stenosis, 6 patients exhibited 2, and 5 patients
exhibited more than 2 stenoses. There was no significant
correlation between the number of narrowed arteries per patient and the
causative pathogen (
2 test). By regression
analysis, the occurrence of an MCA stenosis correlated
significantly (P<.01) with an increasing CSF total protein
content.
In 4 patients repeated measurements of ICP allowed a comparison between
the course of the MBV and the course of ICP (Table 2
). A
stenosis in the MCA developed while the ICP was within the
normal range (patient 1) or severely increased. Additionally, the
MCA/ICA ratio can indicate a stenosis of the MCA before the MBV
increase (patients 2 and 3).
|
Clinically, the mean GCS score did not differ between the patients with (10±3) and without (12±4) stenoses of the intracranial arteries on day 1. Thereafter, the patients with stenoses of the intracranial arteries exhibited a significantly poorer mean GCS score from day 3 (9±4) to day 14 (11±4) compared with the patients without any stenosis of a basal cerebral artery (day 3: 13±4, P<.01; day 14: 14±1, P<.05). Additionally, by stepwise regression analysis, a lower GCS score on days 1 to 5 was significantly associated with advancing age (P<.01), while the causative pathogen had no effect on the GCS scores from days 1 to 14.
Twenty-two focal cerebral signs were present in 18 patients. An
isolated hemiparesis was most frequent (12 of the 22 focal hemispheric
signs), followed by aphasia accompanied by a right-sided hemiparesis
(2), focal seizures (2), focal slowing on EEG (2), isolated aphasia
(2), neglect accompanied by left-sided hemiparesis (2), hemiparesis
with accompanying hemisensory disturbances for
touch and pain (1), and isolated hemisensory
disturbances for touch and pain (1). Six of the 22 focal
cerebral signs were already present on day 1, 13 developed between
days 2 and 5, and 3 developed between days 7 and 11. With respect to
the vascular territories, 20 of the 22 focal cerebral signs could be
allocated to the MCA territory and 2 to the ACA territory. By the
2 test, focal cerebral signs were significantly
more frequent in patients with an increasing number of narrowed
arteries (P<.05). Six of the 22 focal cerebral signs were
correlated with pathological changes on CT or MRI scans: small
infarctions within the MCA territory in 3 patients, small watershed
infarctions in the MCA/ACA border zone in 1 patient, a complete
infarction in the MCA territory in 1 patient, and a small
subarachnoid hemorrhage (on day 9) within the sylvian
fissure in 1 patient. Seventeen focal cerebral signs were transient,
disappearing usually within 3 days after their onset. Five focal
cerebral signs remained clinically relevant, and all 5 exhibited a
correlating brain infarction on CT scan and were accompanied by
ipsilateral stenosis of the MCA and/or the ACA.
With respect to the outcome, a poorer GOS score correlated significantly with advancing age (P<.01) but not with the causative pathogen by stepwise regression analysis. The mean GOS score of the patients without a stenosis of an artery (GOS score, 4.05±1.60) did not differ significantly from those with stenoses (GOS score, 3.61±1.50). Including the two patients with fungal meningitis, 6 of the 35 patients died (17%).
| Discussion |
|---|
|
|
|---|
The vascular complications of bacterial meningitis such as stenosis or occlusion of the large or small vessels, widely distributed wall irregularities of the small vessels, and thrombosis of the veins or sinus are well known.9 10 11 12 13 14 15 The real frequency of a stenosis or an occlusion of the basal cerebral arteries in patients with bacterial meningitis is unknown. Because of the risks of angiography, a serial angiographic study of unselected patients with bacterial meningitis to clarify the frequency and the clinical relevance of stenoses of the basal cerebral artery is not justifiable. Our study consisted of 35 unselected patients with bacterial and fungal meningitis. Eighteen of the 35 patients (51%) exhibited a stenosis of at least one basal cerebral artery that occurred regularly between days 3 and 5. Clinically, the occurrence of vessel stenoses was significantly associated with a poorer GCS score between days 3 and 14. Additionally, the presence of focal cerebral signs between days 2 and 5 was significantly associated with an increasing number of narrowed arteries per patient. These findings are in accordance with two serial angiographic studies of selected patients with bacterial meningitis who underwent angiography because of focal complications or unexplained coma.9 12 In both studies the angiographies were usually performed between days 3 and 5 after admission. They demonstrated isolated or multiple stenoses or occlusions of the large intracranial arteries unilaterally or bilaterally.
Apart from the diameter of the insonated artery, other factors may
affect MBV. First, a decrease in ICP has been reported to increase MBV
as an index of cerebral blood flow in patients with bacterial
meningitis.8 22 An increase in ICP leads to an increase in
PI.33 34 35 On days 1 and 3, the inverse relationship between
the pattern of the PI and MBV in our patients may therefore reflect a
decrease of ICP, with a corresponding increase of MBV. Second,
hyperemia increases MBV.20 27 However, the rapid
increase of the MCA/ICA ratio is a strong argument against the
assumption that the highly elevated MBVs represent
hyperemia. Additionally, severe hyperemia should
increase ICP and hence the PI. Our corresponding measurements of the
ICP and TCD parameters (Table 2
) demonstrated that the MBV
increases did not correlate with ICP. Bode and Harders7
described two infants with bacterial meningitis in whom MBV was
severely increased while the repeatedly measured ICP was normal. It is
therefore reasonable to assume that the recorded MBV increases in
our four patients were independent of ICP.
Most of the focal cerebral signs were transient and without a correlation on imaging scans. This might explain why the outcome was not significantly different between patients with and those without vessel stenosis. Cerebral blood flow studies in patients with bacterial meningitis with the use of single-photon emission clearance techniques or stable enhanced xenon CT36 37 38 have demonstrated brain areas with a reduced regional cerebral blood flow as correlates of clinical focal signs. We found a significant relationship between the number of narrowed vessels per patient and the occurrence of focal cerebral signs. This may suggest that cerebral ischemia due to arterial narrowing may worsen critical metabolic disturbances caused by the inflammatory process.6 37 39 40
| Selected Abbreviations and Acronyms |
|---|
|
Received April 6, 1995; revision received May 15, 1995; accepted May 18, 1995.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C.-H. Lu, H.-W. Chang, C.-C. Lui, C.-R. Huang, and W.-N. Chang Cerebral haemodynamics in acute bacterial meningitis in adults. QJM, December 1, 2006; 99(12): 863 - 869. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |