HEMISPHERIC ISCHEMIC STROKE: PROGNOSTICATION ALGORITHM BASED ON DATA OF TRANSCRANIAL DOPPLEROGRAPHY

Tikhomirova O.V., Sorokoumov V.A., Mashkova N.P., Skoromets T.A.

Elizavetinskaya Municipal Hospital, Chair of Neurology and Neurosurgery, Saint Petersburg State Medical University named after acad. Pavlov, Saint Petersburg

 

Prognostication of the disease outcome, development of complications and recurrences is of crucial importance for determining optimum terms and tactics of treatment. Transcranial dopplerography (TCD), being an adequate and noninvasive modality, allows to estimate short- and long-term prognosis in cerebrovascular diseases. A conventional diagnostic algorithm comprises evaluation of such dopplerographic parameters as linear blood flow velocity (BFV), pulsatility index (PI), coefficients of cerebrovascular reactivity, etc. and discriminant analysis for determination of critical values of these indices and singling out groups with favorable and unfavorable prognosis. According to some reports, a value of BFV in the middle cerebral artery (MCA) watched during the first hours after the disease onset can be used for prediction of functional outcome of hemispheric ischemic stroke [6, 11, 14]. Reduction of cerebrovascular reactivity has been proved to result in a higher risk of development of recurrent acute cerebral ischemia in patients with a hemodynamically significant lesion of the internal carotid artery (ICA) [2, 7, 10, 13] and affects functional outcome of stroke [5]. It has been found out that value of BFV, PI and reactivity coefficients are important from the point of view of prognostication of hyperemic complications after removal of arteriovenous malformations (AVM) [3]. Thus, there is no doubt that TCD can be used for prognosticating a degree of restoration of neurologic deficit, development of complications and recurrences in various cerebrovascular diseases. The authors of the above-mentioned studies estimated prognostic significance of absolute values of some dopplerographic indices for the whole group of patients. We consider this approach to be appropriate only for diseases characterized by one pathogenetic mechanism and one type of hemodynamic disorders which corresponds to it. For example, dopplerographic diagnosis of AVM is based on registering a shunt pattern; one can watch the following: the more a malformation size, the higher BFV and the lower values of peripheral resistance and cerebrovascular reactivity. However, cerebrovascular diseases, at least the majority of them, are heterogenic both in their pathogenesis and hemodynamic picture. In this case one and the same absolute value of a dopplerographic parameter can have different prognostic meaning which depends on a type of cerebral hemodynamics.

The goal of the present study is determination of an algorithm which would allow to use TCD for prediction of outcome of hemispheric ischemic stroke being a heterogenic cerebrovascular disease.

Material and Methods

Dopplerographic examination was carried out in 447 patients in an acute period of hemispheric ischemic stroke (the first 5 days). It was done with the help of Biomed-P apparatus (BIOSS). BFV, PI, asymmetry coefficient (AC) of BFV in MCA, spectral parameters, signs of activation of collateral blood flow were studied. Estimation of autoregulation of cerebral circulation was based on results of a compression test [8] with determining overshoot coefficient (OC) [4]. Complex analysis made it possible to classify all dopplerograms according to hemodynamics types. CT examination (Somaton AR.C, SIEMENS) was carried out in 370 cases. Its results were used for verification of data on an ischemic focus, its size and localization of an infarction zone. A functional state watched by the 21st day of stay in hospital was evaluated in compliance with Bartel’s index (247 patients).

The results were processed with the help of Origin and SPSS programs. Statistical description of variables, correlation, regression and discriminant analyses were used. Reliability of variables (p<0.05) in samples was estimated on the basis of Student’s T-criterion. The final goal of regression and discriminant analyses was obtaining reliable patterns of process prognostication.

Results

A symmetrical major type of blood flow in MCA was registered in 122 out of 447 patients (27%). Indices of BFV, PI, OC were within normal range. AC did not exceed 15%. There were no changes in spectral parameters of flow and signs of activation of collateral circulation. The majority of patients of this group (85%) had lacunar or minor cortical infarctions.

An asymmetrical major type of blood flow in MCA with reduction of BFV on a stroke side was watched in 8% (36 out of 447 cases). AC varied from 15 up to 50% The rest indices of blood flow were within normal range. Lacunar and minor cortical infarctions were diagnosed in a greater part of patients of this group too.

Dopplerographic signs of occlusion of intracranial arteries were present in 62 out of 447 cases. Occlusions of MCA (M1 or M2 segments) and the intracranial portion of ICA were observed in 59 and 3 patients respectively. The main dopplerographic criterion of MCA occlusion was absence or marked reduction of BFV in it with simultaneous increase of BFV in the ipsilateral anterior cerebral artery (ACA). Reduction or absence of vasodilation reserve in the MCA area on a stroke side was a typical phenomenon. Registration of dopplerographic signs of persistent occlusion of intracranial arteries was indicative of gross infarctions (total, cortical-subcortical, striatal-capsular) in 97% of cases.

Signs of hyperperfusion with increase of BFV and reduction of indices of peripheral resistance and cerebrovascular reactivity in the MCA area on a stroke side were revealed in 8% of cases (37 out of 447 patients). Hyperperfusion was characteristic mainly of gross infarctions (total, cortical-subcortical and striatal-capsular infactions were diagnosed in 87% of cases).

A hemodynamically significant lesion of the ICA extracranial portion with residual flow in MCA was registered in 66 out of 447 patients (15%).

Dopplerographic signs of MCA stenosis were observed in 68 out of 447 cases (15%).

Hampered perfusion with reduction of diastolic blood flow velocity and increase of peripheral resistance was typical of 55 out of 447 patients (12%). A combination of hampered perfusion with changes of BFV synchronous to an act of breathing as well as with clinical manifestation s of brain edema took place in 12 patients with gross infarctions. Increase of intracranial pressure is considered to be a cause of hampered perfusion in these cases. Reduction of BFV and growth of PI in the rest 43 patients were not accompanied by symptoms of brain edema; they were associated with lacunar and minor cortical infarctions. This version of hampered perfusion can be a result of peripheral vasoconstriction and arteriolosclerosis.

Angiospasm of MCA was diagnosed in 1 patient with extensive cardioembolic infarction and secondary subarachnoid-parenchymal hemorrhage.

Presence of symmetrical and asymmetrical major types of flow and signs of hampered perfusion against a background of peripheral vasoconstriction and arteriolosclerosis was watched mainly in patients with minor cortical and lacunar strokes, a mild course of the disease and good restoration of neurologic functions.

Signs of persistent occlusion of intracranial arteries, hyperperfusion or hampered perfusion against a background of increased intracranial pressure were observed in 94% of cases with gross infarctions (total, cortical-subcortical, striatal-capsular), a severe course and a high risk of fatal outcome. These changes of cerebral hemodynamics were reliably predominant in patients with a higher risk of cardioembolism and absence of hemodynamically significant stenosis. It is indicative of a cardioembolic origin of stroke in such cases. Hyperperfusion syndrome took place in embolic occlusion with early spontaneous recanalization and was a more favourable predictive sign than persistent occlusion.

In case of residual flow in MCA and signs of stenosis of intracranial arteries the most probable cause of stroke is an atherosclerotic lesion of major arteries which leads to stroke characterized by a hemodynamic mechanism of development or resulting from arterio-arterial embolism. Patients with isolated stenosis of intracranial branches of ICA usually had small infarctions with a mild course and good restoration. Middle-sized infarctions with a more severe course were watched in a hemodynamically significant lesion of the extracranial portion of ICA very often.

The correlation analysis of some dopplerographic indices of cerebral hemodynamics (BFV, PI, OC), a dopplerographic pattern and outcome was carried out with the purpose of determining the most informative predictive criteria. Before it dopplerographic patterns were ranked on the basis of values of Bartel’s index watched by the 21st day of the disease (table 1).

Table 1. Functional Outcome of Stroke by the 21st Day of the Disease and Its Dependence on a Type of Cerebral Hemodynamics in an Acute Period

A type of cerebral hemodynamics Bartel’s index
M±m
Rank
Symmetrical major blood flow 18.2±1.0 1
Hampered perfusion against a background of peripheral vasoconstriction and arteriolosclerosis 18±0.9 2
Asymmetric major blood flow 17.2±1.5 3
Intracranial stenosis 17±0.6 4
Hyperperfusion 15±1.5 5
Residual flow 12±1.3 6
Intracranial occlusion 6.6±1.1 7
Hampered perfusion in increased intracranial pressure 3.5±1.1 8

Patients were classified into 4 groups based on the disease outcome:

  1. Patients with minimum neurologic deficit watched by the 21st day of the disease (Bartel’s index equal to 16-20).
  2. Patients with moderate neurologic deficit watched by the 21st day of the disease (Bartel’s index equal to 11-15).
  3. Patients with severe neurologic deficit watched by the 21st day of the disease (Bartel’s index of 10 and lower).
  4. Patients who died during the first month of the disease.

The correlation analysis showed that a dopplerographic pattern played the most important part in outcome prognostication (r=0.74) (fig.1). Coefficients of correlation of specific dopplerographic parameters with the disease outcome were smaller: 0.58 for OC (fig.2), 0.51 for BFV (fig.3) and 0.37 for PI. Thus, complex estimation of dopplerograms with determining a type of cerebral hemodynamics had greater predictive importance in comparison with specific parameters of blood flow.

Prognostication of favorable (presence of mild and moderate neurologic deficit) and unfavorable (presence of severe neurologic deficit and death) outcome based on the discriminant analysis demonstrated that OC was the most significant predictive factor among specific dopplerographic indices. It allowed to predict unfavorable outcome in 76% of cases regardless its values of less than 1.21. Precision of prognosis based on data of BFV and PI was lower (table 2). The most precise prognosis was obtained when the disease outcome was evaluated on the basis of a dopplerographic pattern (table 3).

 

Table 2. Dopplerographic Indices in Groups with Favorable (1) and Unfavorable Stroke Outcomes. Critical Values of Dopplerographic Indices for Prognostication of Unfavorable Outcome.

Groups of patients BF (cm/s) PI OC
1 (n=180) 60±22 0.9±0.22 1.3+0.17
2 (n=67) 41+26 1.05+0.49 1.1+0.11
A critical value <50 >0.98 <1.21
Precision 66% 62% 76%
P <0.01 <0.05 <0.001

Table 3. Different Dopplerographic Patterns and Probability of Favorable Outcome in Patients with Ischemic Stroke in the Carotid Area

N Dopplerographic pattern Probability of favorable outcome
1. Symmetrical major blood flow (n=66) 88%
2. Hampered perfusion against a background of peripheral vasoconstriction and arteriolosclerosis (n=20) 90%
3. Asymmetrical major blood flow (n=23) 91%
4. MCA stenosis (n=36) 94%
5. Hyperperfusion (n=20) 80%
6. Residual flow in MCA (n=33) 63%
7. MCA occlusion (M1 and M2 segments) (n=39) 31%
8. Hampered perfusion against a background of increased intracranial pressure (n=10) 0%

However, probability of precise prediction of the disease outcome is not the same in different patterns. Table 3 shows that if a dopplerogram is attributed to one of the first 5 types, then favorable outcome is supposed to be characterized by a high degree of probability. Unfavorable outcome is typtcal of all patients with hampered perfusion against a background of edema. There are two groups (patients with residual flow and signs of intracranial occlusion) in which precision of prognosis based only on a dopplerographic pattern is low. The discriminant analysis carried out in each of them was aimed at obtaining critical values for the most significant dopplerographic parameters (tables 4 and 5).

Table 4. Dopplerographic Indices in Patients with Signs of Proximal and Distal Occlusion of MCA in Groups with Favorable (1) and Unfavorable (2) Outcome of Stroke. Critical Values of Dopplerograhic Indices in Prognostication of Unfavorable Outcome.

Groups of patients BFV in MCA, cm/s PI OC
1 (n=12) 41+16 0.96+0.2 1.17+0.05
2 (n=27) 29+12 1.1+0.4 1.03+0.1
A critical level <35 - <1.1
Prescision 60% - 89%
P <0.05 >0.05 <0.001

Table 5. Dopplerographic Indices in Patients with Residual Flow in MCA in Groups with Favorable (1) and Unfavorable (2) Outcome of Stroke. Critical Values of Dopplerograhic Indices in Prognostication of Unfavorable Outcome.

Groups of patients BFV in MCA, cm/s PI OC
1 (n=21) 46+ 12.6 0.65+ 0.13 1.1+ 0.13
2 (n=12) 45.6+ 14.9 0.5+ 0.1 1.06+0.09
A critical level - <0.55 -
Prescision - 70% -
P >0.05 <0.05 >0.05

According to table 4 criteria of unfavorable outcome in a group of patients with signs of proximal and distal occlusion of MCA include reduction of BFV and OC below 36cm/s and 1.1 respectively. OC had the greatest diagnostic significance. PI was of no importance for differential diagnosis in this group.

As for patients with residual flow, BFV and OC had no significant effect on prognosis in an acute period of ischemic stroke. PI of less than 0.55 can be regarded as a criterion of unfavorable outcome.

The obtained results allow to propose an algorithm for prediction of an outcome of hemispheric ischemic stroke. It consists in complex evaluation of dopplerograms with identification of a type of cerebral hemodynamics at the first stage and determination of critical values of specific dopplerographic parameters in a pattern at the second stage. The latter concerns cases when precision of prognostication based only on identification of a type of cerebral hemodynamics is rather low (fig.4). Precision of prediction of stroke outcome with the help of the algorithm under discussion made up 87%. It was higher than that obtained in application of any specific dopplerographic parameter (fig.5).

Discussion

The results of comparison prove that dopplerographic criteria used for estimation of the disease outcome are highly informative. Many scientists studying changes of cerebral hemodynamics in the acutest period of hemispheric ischemic stroke with the help of dopplerography arrived at the same conclusion [1,6,9,11,12,14,17]. Reduction of BFV in CMA below a certain limit was considered to be the main criterion of prediction in hemispheric ischemic stroke. Critical values of BFV given in various articles varied within a wide range. Classifications, based on the analysis of BFV as the only index, distinguished subgroups with stenotic flow, normal flow, asymmetry with reduction of BFV on a stroke side and absence of blood flow in MCA. Their use resulted in ranking the majority of cases among patients with asymmetrical flow which, as well as absence of blood flow in MCA, was a criterion of unfavorable outcome in all studies [9,12,15].

Our study revealed dependence of prognostic significance of specific dopplerographic parameters and their critical values on a pattern. Asymmetry of blood flow with reduction of BFV on a stroke side can be a sequela of stenosis located proximally and occlusion located distally. Besides, there is a great number of unspecified asymmetries when flow looks like a major one. Prognostic value of each of these groups differs greatly. Thus, their unification into one category and an attempt to determine a value of critical BFV, which would be common for all groups and indicative of unfavorable outcome, reduce prognostic capabilities of doppplerography to a considerable extent.

The classification of dopplerograms, used in our study and taking into account a complex of criteria, is rather simple. At the same time it increases diagnostic and prognostic effectiveness of TCD. Simultaneous use of the above-mentioned complex results in determination of a dopplerographic pattern and is of higher predictive significance than any of indices taken separately. The correlation between discovered types of cerebral hemodynamics, results of tomographic examinations and etiologic causes of stroke made it possible to get pathogenetic description of each pattern. Determination of critical values for the most important dopplerographic parameters can enhance diagnosis precision. However, importance of specific dopplerographic parameters and their critical values are dependent on a pattern. For example, a marked disorder of blood flow autoregulation (reduction of OC) turned out to be decisive in cases with intracranial occlusion but it was inessential in patients with residual flow.

The results of our study are indicative of necessity to estimate the most important predictive dopplerographic criteria of blood flow and their critical values, but it should be done at the second stage, i.e. after classification of dopplerograms according to their types. Such specifying use of critical values of BFV, PI and OC within the limits of a pattern differs greatly from attempts to determine common critical values of these parameters for the whole continuum of data. This approach can be applied to prognostication of outcome of various cerebrovascular diseases.

fig1.gif (4883 bytes)

Fig. 1. Dependence of stroke outcome on changes of cerebral hemodynamics.

The X-axis: 1 - symmetrical major flow; 2 – hampered perfusions against a background of peripheral vasoconstriction and arteriolosclerosis; 3 – asymmetric major blood flow; 4 – MCA stenosis; 5 – hyperperfusion; 6 – residual flow; 7 – occlusion of MCA (M1 and M2 segments); 8 – hampered perfusion and edema.

The Y-axis: 1 – presence of minimum neurologic deficit by the 21 day after the disease onset (Bartel’s index of 16-20); 2 – presence of moderate neurologic deficit (Bartel’s index of 11-15); 3 – presence of severe neurologic deficit (Bartel’s index of 10 and lower); 4 – fatal outcome during the first month of the disease.

fig2.gif (4900 bytes)

Fig. 2. Dependence of stroke outcome on a state of blood flow autoregulation in an acute period of the disease.

The X-axis: values of overshoot coefficient.

The Y-axis: 1 – presence of minimum neurologic deficit by the 21 day after the disease onset (Bartel’s index of 16-20); 2 – presence of moderate neurologic deficit (Bartel’s index of 11-15); 3 – presence of severe neurologic deficit (Bartel’s index of 10 and lower); 4 – fatal outcome during the first month of the disease.

fig3.gif (5994 bytes)

Fig. 3. Dependence of stroke outcome on mean values of BFV in MCA in an acute period of the disease.

The X-axis: vales of mean BFV in MCA on a stroke side.

The Y-axis: 1 – presence of minimum neurologic deficit by the 21 day after the disease onset (Bartel’s index of 16-20); 2 – presence of moderate neurologic deficit (Bartel’s index of 11-15); 3 – presence of severe neurologic deficit (Bartel’s index of 10 and lower); 4 – fatal outcome during the first month of the disease.

fig4.GIF (16920 bytes)

Fig. 4. The algorithm of predicting outcome of hemispheric ischemic stroke on the basis of TCD findings.

fig5.gif (7260 bytes)

Fig. 5. Comparative characteristic of prognostic significance of specific dopplerographic indices and an algorithmic approach.

 

References

  1. Aleksandrov A.V., Marochnik C., Pullichino P.M. et al. Clinical significance of transcranial dopplerography and photon-emissive tomography in acute cerebral ischemia //Intern. Symp. on Transcranial Doppler and Intraoperative Monitoring: Sci. Rep.- St. Petersburg, 1995.-P.9 (Rus.).
  2. Barkhatov D.Yu., Dzhibladze D.N., Nikitin Yu.M. Functional evaluation of cerebral hemodynamics with the help of transcranial Doppler and a nitroglycerin test in occlusion of internal carotid artery//Zhurn. nevropatol. and psikhiatr.-1998.-V.98, N 1.-P.31-33 (Rus.).
  3. Gaidar B.V., Parfenov V.E., Svistov D.V. et al. Choice of tactics of surgical treatment of arteriovenous malformations based on data of a minimum invasive diagnostic complex//Transactions of the Vth Intern. Symp.: Brain Injuries (Minimum Invasive Methods of Diagnosis and Treatment.-St. Petersburg, 1999.-P.311-320 (Rus.).
  4. Svistov D.V. Perioperative transcranial dopplerography in arteriovenous malformations of brain: Author’s abstract of thesis for a degree of a candidate of medical science.-St. Petersburg.-1993 (Rus.).
  5. Khilko V.A., Shulev Yu.A., Ivanova N.E., Bikmullin V.N. Diagnosis and treatment of chronic cerebrovascular insufficiency in atherosclerotic stenosis or occlusion of carotid artery//In: Vascular Pathology of the Nervous System.-St. Petersburg.-1998.-P.50-52 (Rus.).
  6. Blaster T., Krueger S., Kross R. etal. Time dependent relevance of transcranial color-coded duplex sonography in acute stroke//Cerebrovasc. Dis.-2000.-Vol. 10, suppl. 2.-P.69.
  7. Cullinane M., Markus H. Severely impaired cerebrovascular reactivity predicts stroke risk in carotid stenosis and occlusion: a prospective study//Cerebrovasc. Dis.-2000.-Vol. 10, suppl. 2.-P.7.
  8. Giller C.A. A bedside test for cerebral autoregulation usinf transcranial Doppler ultrasound//Acta Neurochir..-1991.-Vol. 108, Fasc. 1-2.-P.7-14.
  9. Goertler M., Kross R., Baeumer M. et al. Diagnostic impact and prognostic relevance of early contrast-enhanced transcranial color-coded duplex sonography in acute stroke//Stroke.-1998.-Vol. 29, N 5.-P.955-962.
  10. Gur A.Y., Bova I., Bornstein N.M. Is impaired cerebral vasomotor reactivity a predictive factor of stroke in asymptomatic patients//Stroke.-1996.-Vol. 27, N 12.-P.2188-2190.
  11. Halsey J.H. Prognosis of acute hemiplegia estimated by transcranial Doppler ultrasonography//Stroke.-1988. Vol. 19.-P.648-649.
  12. Kushner M.J., Zanette E.M., Bastianello S. et al. Transcranial Doppler in acute hemispheric brain infarction//Neurology.-1991.-Vol. 41, N 1.-P.109-113.
  13. Markus H., Harrison S. Estimation of cerebrovascular reactivity using transcranial Doppler, including the use of breath-holding as the vasodilatatory stimulus//Stroke.-1992.-Vol. 23.-P.668-673.
  14. Ni X.S., Horner S., Fazekas F., Niedenkorn K. Serial transcranial Doppler somography in ischemic strokes in middle cerebral artery territory//J. Neuroimaging.-1994.-Vol.4, N 4.-P. 232-236.
  15. Toni D., Fiorelli M., Zanette E.M. et al. Early spontaneous improvement and deterioration of ischemic stroke patients. A serial study with transcranial Doppler sonography//Stroke.-1998.Vol. 29, N 6.-P.1144-1148.
  16. Viola S., Antonacci R., D’Annuzio S. et al. Three-dimensional transcranial Doppler in acute ischemic stroke in the territory of the middle cerebral artery: clinical and CT correlation//Ital. J. Neurol. Sci.-1991.-Vol. 12, N 6.-P.545-555.
  17. ViolaS, Tenalgia M.G., De Leonardis E. et al. Acute hemispheric stroke” correlation between three-dimensional transcranial Doppler, MR-angiography, CT and clinical findings//Ital J. Neurol. Sci.-1993-Vol. 14, N 3.-P.225-232.