Asymmetry of Intracranial Hemodynamics as an Indicator of Mass Effect in Acute Intracerebral Hemorrhage
A Transcranial Doppler Study
Background and Purpose Hematoma volume is an important determinant of outcome and predictor of clinical deterioration in patients with intracerebral hemorrhage. In many cases, worsening results from herniation due to compartmentalized pressure gradients. We used transcranial Doppler sonography (TCD) to assess the impact of hematoma volume on symmetry of intracranial hemodynamics in patients with acute intracerebral hemorrhage. The goal was to evaluate TCD as a noninvasive method for monitoring compartmentalized mass effect.
Methods TCD was performed an average of 1.1 days (range, 0 to 3 days) after onset in 30 patients with supratentorial intracerebral hemorrhage. Hematoma, hematoma+edema, and intraventricular hemorrhage volumes were calculated from admission CT scans using computerized planimetry and were compared with combined TCD values from the middle cerebral and internal carotid arteries.
Results Ipsilateral pulsatility indexes were consistently elevated and mean velocities consistently depressed when intracerebral hemorrhage volumes exceeded 25 mL. Compared with patients with small hemorrhages, those with large hemorrhages (≥25 mL, n=10) had significantly higher ipsilateral pulsatility indexes (1.72 versus 1.13, P<.0001) and higher ratios of ipsilateral-to-contralateral pulsatility (1.29 versus 1.06, P=.001). The ratio of ipsilateral-to-contralateral mean velocity was similarly reduced in patients with large versus small hemorrhages (0.87 versus 1.06, P=.01), but this effect was less pronounced. In a multiple regression analysis, ipsilateral and contralateral pulsatility indexes correlated primarily with intraventricular hemorrhage volume (P<.001), whereas the ratio of ipsilateral-to-contralateral pulsatility correlated with total hemispheric lesion (hematoma+edema) volume (P=.003).
Conclusions Asymmetry of intracranial hemodynamics as assessed by TCD occurs when intracerebral hemorrhage volumes exceed 25 mL. Alterations of pulsatility index reflect intracranial lesion volume more reliably than mean velocity. Although pulsatility is strongly influenced by the presence of intraventricular blood, elevated ratios of ipsilateral-to-contralateral pulsatility correlate primarily with hemispheric lesion volume and may reflect compartmentalized intracranial pressure gradients.
Hematoma volume is an important determinant of outcome1 2 3 4 5 and predictor of neurological deterioration6 7 in patients with ICH. Large space-occupying lesions lead to compartmentalization of mass effect and ICP gradients.8 9 Herniation of brain tissue related to ICP gradients may be an important mechanism of worsening after ICH.
Given the importance of lesion volume in ICH, a noninvasive method for detecting and monitoring compartmentalized mass effect might be of considerable clinical utility. TCD allows noninvasive assessment of intracranial hemodynamics and has been shown to reflect both global10 11 12 and focal13 14 15 increases in ICP. As ICP increases, cerebral perfusion pressure falls, diastolic and mean blood FVs decrease, and PI increases.10 11 12 Studies in patients with traumatic intracranial hemorrhage13 14 15 have shown that these effects are typically more pronounced in the ipsilateral hemisphere, presumably reflecting ICP gradients related to compartmentalized mass effect.
The goal of the present study was to evaluate TCD as a method for detecting hemodynamic disturbances related to compartmentalized mass effect in acute nontraumatic ICH. Our specific aims were (1) to analyze the relationship between ICH volume and lateralized alterations of mean FV and PI, (2) to determine a “cutoff” ICH volume above which these changes consistently occur, and (3) to test the hypothesis that TCD asymmetries in ICH primarily reflect hemispheric lesion volume.
Subjects and Methods
Seventy-two patients with CT-documented ICH admitted to the Columbia-Presbyterian Neurological Intensive Care Unit between July 1992 and January 1995 were assessed for entry into the study. Eligibility was based on supratentorial ICH and admission within 24 hours of onset. Exclusion criteria included (1) deep coma (Glasgow Coma Scale score ≤5); (2) surgical hematoma evacuation or ventriculostomy; (3) history of ischemic stroke or carotid stenosis; or (4) hemorrhage related to trauma, neoplasm, aneurysm, arteriovenous malformation, or coagulopathy. On the basis of these criteria, 35 patients were eligible. Adequate TCD windows could not be obtained in 5 patients (14%), leaving a study population of 30 patients (20 men, 10 women; age range, 41 to 97 years). The study protocol was reviewed and approved by the hospital institutional review board.
ICH, ICH+edema, and IVH volumes were calculated from admission CT scans by planimetry. Lesion areas on each slice were calculated by tracing the perimeter of the appropriate high- or low-attenuation zone on the CT console (General Electric 9800); these values were then multiplied by slice thickness to yield single-plane lesion volumes, which were summed to yield total lesion volume.4 TCD examination was performed on either the first or second hospital day with a Medasonics CDS 2-MHz pulsed-wave device, using a handheld probe in the temporal window. The MCA was insonated at a depth of 45 to 60 mm and the ICA at 65 to 70 mm when bidirectional flow was detected. Peak FV, mean FV, and PI were recorded bilaterally in each vessel at the level of maximal mean FV. PI was calculated according to the formula of Gosling as follows: PI=(Peak Systolic FV−End Diastolic FV)/Mean FV. Systolic and diastolic blood pressure were recorded at the time of TCD examination; all patients were normotensive or hypertensive during the examination.
Patients were evaluated clinically on hospital days 1, 2, 3, 7, 14, and 30 (or at discharge). Neurological status was evaluated using the Glasgow Coma Scale,16 National Institutes of Health Stroke Scale,17 and Stroke Data Bank weakness scale.7 Neurological deterioration was defined as a change from the initial neurological examination of at least one of the following: (1) a decrease of 2 points or more in the Glasgow Coma Scale score or (2) an increase of 1 point or more in the Stroke Data Bank weakness scale.7 Functional outcome at 30 days or discharge was assessed using the Glasgow Outcome Scale.18 For statistical analysis, outcome was classified as dead, dependent (vegetative or severely disabled), or independent (moderately or minimally disabled).
MCA and ICA TCD values in each patient were combined in the statistical analysis. Differences in proportions were compared using the χ2 test or Fisher's exact test. Mean values were compared using the two-tailed t test or the Mann-Whitney U test for nonparametric data. Multiple linear regression was performed to identify significant independent clinical or radiological determinants of ipsilateral and contralateral mean FV and PI, and the ipsilateral-to-contralateral ratios of these values. The following variables were screened in a univariate analysis using simple linear regression and were included in the multivariate model if a significant association (P<.01) was found with at least one TCD value or ratio (dummy values of 0 or 1 were arbitrarily assigned to categorical variables): age, sex (men=0, women=1), admission Glasgow Coma Scale score, ICH location (deep=0, lobar=1), interval in days from onset to TCD examination, mean arterial blood pressure, pulse pressure, ICH volume, ICH+edema volume, and IVH volume. TCD values and ratios were coded as the dependent variable, and clinical/radiographic features coded as independent variables. Significance was judged at the P<.01 level.
The Figure⇓ shows scatterplots depicting the relationship between ICH volume and ipsilateral mean FV, ipsilateral/contralateral mean FV, ipsilateral PI, and ipsilateral/contralateral PI. Based on visual analysis, a cutoff value of 25 mL was identified as the ICH volume beyond which mean FVs and ipsilateral-to-contralateral ratios were consistently depressed and PIs and ipsilateral-to-contralateral ratios consistently elevated.
The clinical features and TCD findings in patients with large (≥25 mL) versus small (<25 mL) ICH are compared in Table 1⇓. Patients with large ICH were older, more frequently had lobar hemorrhage, and had a higher frequency of neurological deterioration than those with small ICH. With combined MCA and ICA values, ipsilateral mean FVs were lower in patients with large versus small ICH (35 versus 44 cm/s), but this difference was nonsignificant, and contralateral FVs in the two groups were similar (41 versus 43 cm/s). By contrast, the ipsilateral-to-contralateral mean FV ratio was significantly depressed in patients with large versus small ICH (0.87 versus 1.06, P=.01). PIs in large-ICH patients were significantly elevated both ipsilaterally (1.72 versus 1.13, P<.0001) and contralaterally (1.34 versus 1.05, P=.003). The ipsilateral-to-contralateral pulsatility ratio was also significantly higher in patients with large versus small ICH (1.29 versus 1.06, P=.001). There was no significant difference between ipsilateral-to-contralateral ratios of mean FV or PI obtained at the MCA and ICA.
Six clinical and radiographic variables were found to be significantly associated (P<.01) with at least one TCD value or ratio in a univariate analysis, and they were included in the multiple regression model (Table 2⇓). Independent determinants of ipsilateral PI (r=.88, R2=.78, P<.0001 for the entire model) included IVH volume (P<.0001) and to a lesser extent ICH+edema volume (P=.02). Contralateral PI (r=.76, R2=.58, P<.0001) was highly correlated with IVH volume (P<.001) and systemic pulse pressure (P=.01). The ratio of ipsilateral-to-contralateral PI (r=.65, R2=.42, P=.0002) correlated primarily with ICH+edema volume (P=.003). The multiple regression model was not significant for estimating ipsilateral mean FV (r=.49, R2=.24, P=.03), contralateral mean FV (r=.46, R2=.22, P=.06), or ipsilateral-to-contralateral mean FV ratios (r=.41, R2=.17, P=.19).
Our results indicate that hematoma volumes in excess of 25 mL are consistently associated with asymmetrical intracranial hemodynamics as reflected by TCD in the MCA or ICA, and suggest that the ratio of ipsilateral-to-contralateral pulsatility may provide a clinically useful noninvasive measure of compartmentalized mass effect in patients with acute ICH.
TCD has an established role in the management of patients with ischemic stroke and subarachnoid hemorrhage. By contrast, the role of TCD in the management of ICH is currently limited to noninvasive screening for medium to large arteriovenous malformations.19 We performed this study to assess the potential utility of TCD for detecting and monitoring compartmentalized mass effect and ICP gradients in patients with acute supratentorial ICH. Although TCD asymmetries have been previously demonstrated in patients with traumatic subdural hemorrhage,13 14 15 to our knowledge this is the first investigation of this phenomenon in spontaneous ICH.
Elevated ICP has been shown to reduce FV mainly during diastole, resulting in reduced mean FV and increased PI.10 11 12 Cardoso and Kupchak13 analyzed side-to-side differences in the TCD waveform in 11 patients with chronic or subacute traumatic subdural hematoma and found lateralized depression of mean FV and elevation of PI that normalized after surgical evacuation. These differences were attributed to ICP gradients resulting from compartmentalized mass effect, with reduced regional cerebral perfusion pressure primarily accounting for the diminution in mean FV, and increased cerebral microvascular resistance from tissue compression accounting for the elevation in PI. Shigemori and associates14 15 have reported similar results in clinical and experimental studies of traumatic subdural and epidural hematoma.
Our results confirm that lateralized alterations of the TCD waveform also occur with spontaneous ICH and offer further insights that might be applicable to intracranial mass lesions in general. As might be expected, the impact of ICH on intracranial hemodynamics was volume dependent, with 25 mL emerging as the critical level beyond which lateralized TCD readings occurred consistently. This finding is in accordance with data showing that the risk of neurological deterioration and poor outcome (death or severe disability) increases dramatically as ICH volume exceeds 20 to 30 mL.4 5 6 7 In addition to the direct effects of increased lesion volume, superimposed disturbances of blood flow hemodynamics, including impairment of autoregulation,20 may contribute to the poor prognosis associated with hemorrhages larger than 25 mL. As the Figure⇑ demonstrates, hemodynamic asymmetries occur in some but not all patients with large ICH. This may reflect equalized pressures from globally increased ICP or perhaps variations in local tissue pressure related to the etiology of bleeding (hypertensive versus amyloid).
Our findings also indicate that global and lateralized alterations of TCD pulsatility are more accurate than mean FV for reflecting the effects of intracranial lesion volume. Although large ICH (≥25 mL) was associated with depressed ratios of ipsilateral-to-contralateral mean FV, the magnitude of this effect on ipsilateral FV was small. By contrast, PI was significantly increased in patients with large ICH both ipsilaterally and contralaterally, and the effect of large ICH volume on ipsilateral-to-contralateral PI ratios was statistically more pronounced (Table 1⇑). Although multiple regression analysis revealed that both ipsilateral and contralateral PI values were highly influenced by IVH volume (P<.001), ipsilateral-to-contralateral PI ratios correlated primarily with ICH+edema volume (P=.003). Single-photon emission CT imaging studies have shown that in acute ICH, perilesional edema corresponds with large zones of reduced perfusion surrounding the hematoma.21 22 23 Because TCD pulsatility generally reflects small-vessel cerebrovascular resistance distal to the vessel being insonated, it makes sense that PI asymmetries correlate more strongly with the volume of the entire hemispheric lesion (hematoma+edema) than with ICH volume alone.
The superiority of PI over mean FV for reflecting lesion volume in ICH patients is in accordance with experimental TCD studies showing that pulsatility is more sensitive than mean FV to variations in ICP.24 25 The impact of extraneous factors that might influence velocity more strongly than pulsatility must also be considered. According to Poiseuille's law, mean FV depends on the pressure gradient between the two ends of the vessel being insonated, its radius, and the viscosity of the circulating blood. A large hemorrhage would be expected to result in reduced ipsilateral mean FV to the extent that locally increased ICP results in decreased perfusion pressure. However, direct compression of the insonated artery by a deeply situated hematoma might tend to counteract this effect, as might hyperemia in regions distant from the hemorrhage, which we have documented in some patients with single-photon emission CT.21 PI, calculated as systolic FV minus diastolic FV divided by mean FV, generally reflects re sistance to flow distal to the insonated vessel and may be less susceptible to these and other factors that influence mean velocity measurements, such as hematocrit level, Pco2, coexisting arterial stenosis, and angle of insonation.26 27
A number of limitations of our study should be emphasized. Because we excluded patients with ventriculostomy or craniotomy, our findings apply only to patients with an intact intracranial vault. We analyzed TCD values only as a function of lesion volume; however, lobar hemorrhage was overrepresented in the large ICH group (≥25 mL), and it remains possible that hemodynamic features unique to lobar ICH explain the observed differences between large and small hemorrhages. The TCD asymmetries detected in our patients most likely reflect compartmentalized pressure gradients, but they cannot be taken as evidence of such. Perilesional reduction of blood flow in ICH might be explained by microvascular compression from brain tissue hypertension, but it might also result from metabolic depression with associated vasoconstriction. Finally, because we did not record ICP and cerebral perfusion pressure in our patients, the relative importance of these variables in relation to lesion volume for determining TCD values could not be assessed.
In summary, our findings demonstrate the utility of TCD for detecting asymmetrical intracranial hemodynamics in patients with acute ICH exceeding 25 mL. Although PI values are strongly influenced by the presence of IVH, elevated ratios of ipsilateral-to-contralateral pulsatility correlate strongly with hemispheric lesion volume and may reflect compartmentalized ICP gradients. Further study is needed to clarify the time course and prognostic significance of TCD waveform asymmetries in patients with large ICH.
Selected Abbreviations and Acronyms
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
|TCD||=||transcranial Doppler sonography|
This work was supported by a National Stroke Association research fellowship award (Dr Mayer) and by National Institutes of Health grant RR00645 (Dr Diamond). The authors thank James H. Halsey, MD, and J.P. Mohr, MD, for critical review of the manuscript.
Reprint requests to Stephan A. Mayer, MD, Division of Critical Care Neurology, Neurological Institute, Box 39, New York, NY 10032. E-mail email@example.com.
- Received February 29, 1996.
- Revision received July 16, 1996.
- Accepted July 16, 1996.
- Copyright © 1996 by American Heart Association
Portenoy RK, Lipton RB, Berger AR, Lesser ML, Lantos G. Intracerebral hemorrhage: a model for the prediction of outcome. J Neurol Neurosurg Psychiatry. 1987;50:976-979.
Daverat P, Castel JP, Dartigues JF, Orgogozo JM. Death and functional outcome after spontaneous intracerebral hemorrhage: a prospective study of 166 cases using multivariate analysis. Stroke. 1991;22:1-6.
Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G. Volume of intracerebral hemorrhage (ICH): a powerful and easy-to-use predictor of 30-day mortality. Stroke. 1993;24:987-993.
Lisk DR, Pasteur W, Rhoades H, Putnam RD, Grotta JC. Early presentation of hemispheric intracerebral hemorrhage: prediction of outcome and guidelines for treatment allocation. Neurology. 1994;44:113-139.
Qureshi AI, Safdar K, Weil EJ, Barch C, Bliwise DL, Colohan AR, MacKay B, Frankel MR. Predictors of early deterioration and mortality in black Americans with spontaneous intracerebral hemorrhage. Stroke. 1995;26:1764-1767.
Mayer SA, Sacco RL, Shi T, Mohr JP. Neurologic deterioration in noncomatose patients with supratentorial intracerebral hemorrhage. Neurology. 1994;44:1379-1384.
Cardoso ER, Kupchak JA. Evaluation of intracranial pressure gradients by means of transcranial Doppler sonography. Acta Neurochir. 1992;55[suppl]:1-5.
Shigemori M, Kikuchi N, Tokutomi T, Ochiai S, Harada K, Kikuchi T, Kuramoto S. Monitoring of severe head-injured patients with transcranial Doppler (TCD) ultrasonography. Acta Neurochir (Wien). 1992;55[suppl]:6-7.
Goldstein LB, Bertels C, Davis JN. Interrater reliability of the NIH stroke scale. Arch Neurol. 1989;39:638-643.
Mast H, Mohr JP, Thompson JLP, Osipov A, Trocio SH, Mayer S, Young WL. Transcranial Doppler ultrasonography in arteriovenous malformations: diagnostic sensitivity and association of flow velocity with spontaneous hemorrhage and focal neurological deficit. Stroke. 1995;26:1024-1027.
Klingelhöfer J, Sander D. Doppler CO2 test as an indicator of cerebral vasoreactivity and prognosis in severe intracranial hemorrhages. Stroke. 1992;23:962-966.
Mayer SA, Kessler DB, Van Heertum RL, Thomas CE, Nour R, Fink ME, Esser PD. Perilesional and global cerebral blood flow in acute intracerebral hemorrhage: a SPECT study. J Nucl Med. 1995;36(suppl):52P. Abstract.
Mayer SA, Nour R, Van Heertum RL, Fink ME, Thomas CE, Brannagan T, Rogers J. Perilesional blood flow and edema formation in acute intracerebral hemorrhage: a SPECT study. Neurology. 1995;45(suppl 4):A411. Abstract.
Grotta J, Sills C, Villar C, Pasteur W, Ramirez A. Demonstration of ischemia in patients with intracerebral hemorrhage. Neurology. 1995;45(suppl 4):A411. Abstract.
De Witt LD, Rosengart A, Teal PA. Transcranial Doppler ultrasonography: normal values. In: Babikian VL, Wechsler LR, eds. Transcranial Doppler Ultrasonography. St Louis, Mo: Mosby-Year Book Inc; 1993:29-38.
Gosling RG, King DH. Arterial assessment by Doppler-shift ultrasound. Proc R Soc Med. 1974;67:447-449.