An Afferent Vessel of Arteriovenous Malformation and Noninvasive Estimation of Its Functional Eloquence

V.B. Semenyutin, V.. Aliev, V.P. Bersnev, *A. Patzak, P.I. Nikitin, V.A.Pak, A.V. Kozlov

Polenov Research Neurosurgical Institute, Saint Petersburg, Russia
*Johannes-Mueller Institute of Physiology University Hospital Charite, Humboldt- University of Berlin, Germany

 

The most informative method of diagnosing functional eloquence of an afferent vessel of cerebral arteriovenous malformation (AVM) is a barbiturate test [9, 15] and direct measurement of blood pressure in an afferent vessel [4-7 16]. The limitations of these methods are invasiveness and their possible use only during endovascular intervention. Decision making on operation expediency is dependent on catheterization of an afferent vessel (an invasive and expensive procedure). If it is not functionally eloquent, then an AVMs compartment can be embolized via this afferent vessel. In case of its high functional eloquence, such embolization is impossible because of a high risk of postoperative neurological complications. This situation demands elaborating noninvasive methods of diagnosis of an afferent vessel functional eloquence, which would allow to determine indications for endovascular intervention.

One can suppose, that there is primordial impairment of autoregulation in a region of an afferent vessel, feeding AVM, and adjacent brain areas. This supposition is based on the fact, that AVMs vessels have no autoregulation due to changes in a histological structure of their vascular wall [13, 18].

Thus, the study goal was qualitative estimation of autoregulation in the region of an AVMs afferent vessel as a means of noninvasive diagnosis of its functional eloquence, ensuring safety of endovascular operation.

Materials and Methods. There were 38 cases with cerebral AVM, aged 20-55 (22 males and 16 females). Their clinical description is given in Table 1.

Table 1

Clinical Description of Cases

Classification according to Spetzler-Martin [17]

I

2

II

2

III

24

IV

10

Localization (Lobes)

Frontal

6

Frontal, parietal

4

Parietal, temporal

3

Parietal, occipital

2

Temporal

4

Parietal

7

Occipital

5

Cerebellum and brain stem

2

Cerebral lobes, subcortical structures

5

A Size

Up to 3 cm

7

From 3 up to 6 cm

21

More than 6 cm

10

A Course Type

Hemorrhagic

28

Epileptic

8

Mixed

2

Conventional endovascular operations were performed in all cases under conditions of sedation-analgesia. A glue composition (histoacryl with lipoiodol) was used for embolization of an AVMs compartment.

When an afferent vessel was reached, its functional eloquence was estimated with the help of direct measurement of blood pressure (BP) within it [16] and a barbiturate test [15]. Barbiturates with a short-term effect were used (thiopental sodium 40-50 mg). Diagnosis of functional eloquence of an afferent vessel allowed to determine indications for embolizing a certain compartment of AVM. Embolization was performed, when it was low. If functional eloquence of an afferent vessel was high, embolization of an AVMs compartment via it was declined.

BP in an afferent vessel was measured directly with the M-34 device (Siemens, Sweden).

Cerebral autoregulation (CAR) was estimated with a cuff test [1] and cross-spectral analysis [3] of slow oscillations of systemic blood pressure (SBP) and linear blood flow velocity (LBFV) in major intracranial arteries. It was done within the range of systemic M-waves before and after operation.

Intracranial major arteries were examined transcranially with the Multi Dop X system (DWL, Germany). LBFV on an AVMs side was recorded either in an artery, whose direct continuation was an afferent vessel, or in the vessel proper, if it was technically possible.

Transcutaneous plethysmographic examination was carried out for noninvasive recording of SBP on a finger. It was done with the help of the Finapres-2000 apparatus (Ohmeda, USA).

Results of a cuff test were used for calculation of CAR speed (RoR), whose normal value is 20±3 %/sec and depends on CO2 tension, reducing in hypercapnia and rising in hypocapnia. CAR disorders can be accompanied by RoR decrease up to zero.

Statistica 6.0 for Windows (Time Series and Prognostication) was used for cross-spectral analysis.

LBFV in major intracranial arteries and SBP were monitored with the Multi Dop X device (DWI, Germany) for carrying out cross-spectral analysis. During monitoring a patient was in a horizontal position with his head tilted up to 30o. Continuous recording was carried out during 5 minutes. It was done at rest and against a background of preserved spontaneous breathing, corresponding to normal ventilation. A phase shift (PS) between slow oscillations of LBVF and SBP within the range of M-waves was calculated in radians (rad). While calculating PS, we used a high coherence criterion, i.e. a PS value was chosen at that frequency, where a coherence index of M-waves of LBFV and SBP was maximum (0.6-08).

Data of CAR estimation were compared with results of intraoperative diagnosis of functional eloquence of an afferent vessel and indices of a postoperative neurological status.

Data were processed with applying conventional statistical programs (Statistica 6.0 for Windows, Excel). Parametric (Student) and non-parametric (Kolmogorov-Smirnov) criteria were used. Difference was considered to be reliable in p<0.05.

The protocol of the study, carried out in volunteers and patients, was approved by the Ethical Committee of the Polenov Research Neurosurgical Institute. Participation in the study was possible only after receiving a patients written consent.

Results. Table 2 demonstrates distribution of cases, dependent on functional eloquence of an AVMs afferent vessel, diagnosed with the help of invasive intraoperative and noninvasive preoperative tests.

Table 2

Distribution of Cases, Dependent on Functional Eloquence of an Afferent Vessel,
Estimated with Intraoperative and Preoperative Tests

Estimation Methods

Functional Eloquence

 

Low

High

A barbiturate test

33

5

Blood pressure in an afferent vessel

32

6

CAR speed in an afferent vessel

31

7

It should be noted, that a barbiturate test played the leading role in determination of indications for embolization, as it is the most widespread method of estimating functional eloquence of an afferent vessel (FEAV), used in endovascular neurosurgery [13, 15].

Superselective embolization of cerebral AVM was performed in 33 out of 38 cases. In 31 cases results of intraoperative testing of FEAV and data of preoperative estimation of CAR speed in an afferent vessel were identical. Embolzation of an AVMs compartment did not lead to any postoperative complications in them. In other 2 cases in spite of a negative barbiturate test, embolization of an AVMs compartment was accompanied by development of persistent postoperative neurological complications. The result of BP estimation was positive in 1 case, being an indication of high functional eloquence of an afferent vessel. It was false-positive in the second case. As for these two cases, preoperative values of RoR and PS were very close to those, observed in patients with high functional eloquence. They were 15 %/sec and 0.8 rad and 12 %/sec and 0.61 rad, respectively.

High functional eloquence, confirmed by intraoperative tests and preoperative estimation of CAR in an afferent vessel region, was a cause of embolization refusal in the rest 5 cases.

Mean values of LBFV and CAR speed on the side of AVM, dependent on functional eloquence of an afferent vessel, are given in Table 3. It should be noted, that the above two cases were also included into the group of patients with high functional eloquence with taking into account preoperative CAR speed and neurological complications, which developed after embolization.

Table 3

Mean Values of LBFV, RoR and PS on the side of AVM and Their Dependence
on Functional Eloquence of an Afferent Vessel

Indices

Functional Eloquence

 

Low (n=31)

High (n=7)

LBFV (cm/sec)

160±30

115±41

RoR (%/sec)

5±3

17.1±1.2

PS (rad)

0.32±0.17

0.68±0.16

Mean values of RoR and PS in cases with high functional eloquence of an afferent vessel were much bigger, as compared to patients with low functional eloquence.

Low Functional Eloquence of an Afferent Vessel

(Fig.1) presents the results, watched in a female patient with AVM of convexital segments of the left parietal lobe.

AVM was fed by hypertrophied branches of the left MCA at the level of M3-M4 segments. Preoperative examination with transcranial Doppler (TCD) revealed the following shunt pattern in the left MCA: increase of LBFV up to 171 cm/sec and decrease of the pulsation index up to 0.38. LBFV (65 cm/sec) and the pulsation index (0.83) in the right MCA were within normal range. Cross-spectral analysis of spontaneous oscillations of SBP and LBFV showed normal values of a phase shift (1.2±0.1 rad) in the right MCA region and its considerable reduction in the region of the left MCA, participating in AVMs blood supply. According to the data of a cuff test, CAR speed in the right and left MCA was 40 %/sec and 2 %/sec, respectively.

Superselective embolization was performed via the left MCA region. A microcatheter was introduced into the afferent vessel of AVM. A barbiturate test was negative. Blood flow in the afferent vessel was 600 ml/min, while blood pressure in it was 30 mm Hg, constituting 32 % in SBP of 93 mm Hg. The afferent vessel was considered to have no functional eloquence and embolization was made. Control angiography revealed no filling of AVM. There was no augmentation of neurological symptoms.

The shunt pattern changed postoperatively: LBFV in the left MCA reduced up to 94 cm/sec, while the pulsation index increased up to 0.64. Cross-spectral analysis of spontaneous oscillations of SBP and LBFV on the AVM side demonstrated increase of a phase shift between oscillations of LBFV (0.8±0.2 rad) and SBP within the range of M-waves; a cuff test showed a growth of CAR speed up to 30 %/sec. It was indicative of considerable improving of CAR in the left MCA region after intravascular operation.

(Fig.2) illustrates the results, obtained in a female case with AVM of medial segments of the left frontal lobe.

AVM was supplied by hypertrophied branches of the left anterior cerebral artery (ACA) at the level of A3-A4 segments. The shunt pattern in the left ACA, estimated with the help of transcranial Doppler, was characterized by increase of LBFV up to 100 cm/sec and decrease of the pulsation index up to 0.64. LBFV in the right ACA was within normal range (67 cm/sec); the pulsation index reduced (0.62). Cross-spectral analysis of spontaneous oscillations of SBP and LBFV demonstrated normal values of a phase shift in the range of M-waves in the right ACA (0.9±0.2 rad) and its considerable decrease (0.4±0.1 rad) in the left ACA, feeding AVM. According to the cuff test data, CAR speed in the right and left ACA was 16 %/sec and 5 %/sec, respectively. Superselective embolization of AVM was performed via the left ACA region. A microcatheter was introduced into the AVMs afferent vessel. A barbiturate test was negative. Blood flow in the afferent vessel was 420 ml/min; blood pressure reached 38 mm Hg, constituting 39% in SBP of 98 mm Hg. The afferent vessel was regarded as functionally ineloquent and embolization was made. Control angiography revealed subtotal exclusion of AVM from blood circulation. There was no augmentation of neurological symptoms.

Postoperative changes of the shunt pattern manifested themselves in LBFV reduction in the left ACA up to 51 cm/sec and increase of the pulsation index up to 0.75. Cross-spectral analysis showed, that a phase shift between oscillations of LBFV and SBP within the range of M-waves, watched on the AVM side, increased up to 0.7±0.2 rad. According to the results of a cuff test, CAR grew up to 18 %/sec. These findings allowed to speak of CAR improvement in the left ACA region after intravascular operation.

Thus, low functional eloquence of an afferent vessel can be diagnosed preoperatively on the basis of noninvasive estimation of CAR in a region of a vessel, feeding AVM.

High Functional Eloquence of an Afferent Vessel

(Fig.3) is a demonstration of the results, observed in a male patient with AVM of the left frontal lobe.

This AVM was fed by hypertrophied branches of the left ACA at the level of A segment. Preoperative TCD-examination showed, that a shunt pattern in the left ACA was characterized by increase of LBFV up to 96 cm/sec and reduction of the pulsation index up to 0.60. As for the right ACA, LBFV was normal (75 cm/sec) and the pulsation index was slightly reduced (0.75).

Cross-spectral analysis of spontaneous oscillations of SBP and LBFV was indicative of normal values of a phase shift within the range of M-waves in the right ACA (1.2±0.2 rad) and its moderate reduction in the left ACA (0.78±0.11 rad), supplying AVM. The results of a cuff test demonstrated, that CAR speed in the right and left ACA was 37 %/sec and 21 %/sec, respectively.

Endovascular intervention, performed under conditions of sedation-analgesia, was not accompanied by considerable changes of LBFV and the pulsation index. A barbiturate test was negative. Blood flow in the afferent vessel was 400 ml/min. Blood pressure in it was 53 mm Hg, constituting more than 54 % in SPB of 97 mm Hg. Decision on embolization of the AVMs compartment was based on estimation of functional eloquence of the afferent vessel (a negative barbiturate test). The patient developed motor aphasia and severe right-side hemiparesis some minutes after embolization. They demanded long-term restorative treatment.

The postoperative shunt pattern was characterized by reduction of LBFV up to 56 cm/s and increase of the pulsation index up to 0.75. Cross-spectral analysis revealed reduction of a phase shift between oscillations of LBFV and SBP within the range of M-waves (0.75±0.18 rad) on the AVM side; the results of a cuff test showed decrease of CAR speed up to 19 %/sec. It was indicative of a smaller speed of CAR, which was probably caused by worsening of brain perfusion in the areas, adjacent to AVM, after embolization of its compartment.

Thus, high functional eloquence of the AVMs afferent vessel was diagnosed with the help of noninvasive preoperative estimation of its CAR speed. As for a barbiturate test, it turned out to be false-negative.

(Fig.4) illustrates the results, observed in a female patient with AVM of subcortical ganglia of the left temporal lobe.

AVM was fed by hypertrophied branches of the left MCA at the level of M2-M3 segments. Preoperative TCD-examination demonstrated, that a shunt pattern in the left MCA was characterized by reduction of the pulsation index up to 0.56 and a normal value of LBFV (78 cm/sec). LBFV in the right MCA was within normal range too (52 cm/sec). The pulsation index reduced up to 0.72.

Cross-spectral analysis of spontaneous oscillations of SBP and LBFV within the range of M-waves demonstrated normal values of a phase shift in the right MCA (1.1±0.1 rad) and its moderate reduction in the lest MCA (0.7±0.1 rad), feeding AVM. Just as in the previous example, it was a sign of high functional eloquence of the AVMs afferent vessel. According to the results of a cuff test (4-4) CAR speed in the right and left MCA was 30 %/sec and 12 %/sec, respectively.

A microcatheter was introduced into the AVMs afferent vessel intraoperatively via the left MCA region. A barbiturate test was positive, which manifested itself in development of transient focal neurological symptoms. Blood flow in the afferent vessel was 280 ml/min. Blood pressure in it was 48 mm Hg, constituting 51 % in SBP of 95 mm Hg. The afferent vessel was considered to be functionally eloquent and it was decided to abstain from embolization.

High functional eloquence of the afferent vessel of AVM, localized in the left parietal lobe, was diagnosed by means of noninvasive preoperative estimation of CAR speed.

Discussion. The results of intraoperative invasive diagnosis of functional eloquence of an AVMs afferent vessel and preoperative noninvasive estimation of CAR speed in it were compared. This comparison has confirmed the supposition, according to which low functional eloquence of an afferent vessel is characterized by low blood pressure, watched in it, a negative barbiturate test and small indices of CAR speed in a vessel, participating in AVMs blood supply. As for vessels of adjacent areas, their autoregulatory response to cerebral perfusion pressure reduction, conditioned by blood shunting via AVM, manifests itself in compensatory vasodilation. Marked shunting leads to perfusion pressure reduction beyond the lower level of autoregulation. It can cause to complete exhaustion of vasodilation reserve, which results in decrease of cerebral blood flow in these areas. Thus, there are at least two factors, effecting CAR speed in an afferent vessel region: blood shunting via AVM and a degree of circulation compensation in adjacent areas. According to our data, a contribution of CAR of adjacent areas is rather insignificant in comparison with AVMs vessels and can be neglected. More detailed elucidation of this problem demands further study.

According to discriminant analysis, criteria of low functional eloquence are as follows: a phase shift of less than 0.5 rad between M-waves of SBP and LBFV in an AVMs feeder, and CAR speed on the same side, which does not exceed 10 %/sec. Thus, preoperative estimation of CAR in a vessel, feeding AVM, is an informative method, which can be used as an additional criterion for diagnosis of an afferent vessel functional eloquence with the purpose of determining indications for superselective embolization of cerebral AVM. Moreover, it does not demand use of expensive microcatheters and intra-arterial administration of barbiturates with a short-term effect, capable of causing persistent neurological disorders even without embolization.

Preoperative estimation of functional eloquence of an AVMs afferent vessel allows to solve a problem of anesthesiologic support during endovascular intervention. Today there are two main trends of endovascular interventions in cerebral AVM, proceeding from principles of anatomic accessibility and physiologic allowance. Adherents of the first trend, i.e. the so-called anatomic school [10, 11, 14, 19], propagate AVM embolization, based on knowledge of neuroanatomy and architectonics of cerebral vessels without preliminary estimation of functional eloquence of its afferent vessel. Thus, they prefer embolization under conditions of total intravenous anesthesia. Besides, it is considered to be mandatory, when modern embolizing substances are used. Advocates of the second trend or the so-called physiologic school [9 14], declare necessity of intraoperative estimation of the brain functional state in a zone of AVM localization with applying different tests before embolization under conditions of sedation-analgesia. They explain it by considerable variability of a functional zone, which does not always coincide with an anatomic one; it particularly concerns cases with AVM. In our opinion, the second approach is more substantiated and safe, as it allows to observe basic principles of surgery to a greater extent and, in the long run, to achieve effective results in ensuring good life quality. For this reason, a method of functional eloquence estimation, based on preoperative diagnosis of CAR in a vessel, participating in AVMs blood supply, could become a key link in uniting positions of both schools. Besides, it could arm adherents of the anatomic school with knowledge of their opponents, promoting safer embolization and prevention of possible neurological complications during endovascular interventions for cerebral AVM, performed under conditions of total intravenous anesthesia.

Literature of the last yeas contains reports, which illustrate growing potentialities of ultrasonic visualization of cerebral vessels with the help of duplex and triplex scanning [2, 8]. Soon these means will be no less informative, than such expensive methods as MRI and digital angiography. Search for ways of combining TCD-monitoring with ultrasonic visualization of an examined vessel and development of programs for on-line cross-spectral analysis of slow oscillations will allow to reach an absolutely new level of studying cerebral autoregulation and to carry out more precise, as well as noninvasive and safe diagnosis of functional eloquence of an AVMs afferent vessel.

  1. Aaslid R., Lindegaard K.F., Sorteberg W., Nornes H. Cerebral autoregulation dynamics in humans // Stroke. 1989. Vol. 20. N. 1. P. 45 - 52.
  2. Bartels E. Evaluation of Arteriovenous Malformations (AVMs) With Transcranial Color-Coded Duplex Sonography: Does the Location of an AVM Influence Its Sonographic Detection? // J. Ultrasound Med. 2005. Vol. 24. P. 1511 - 1517.
  3. Diehl R.R, Linden D., Lucke D., Berlit P. Phase relationship between cerebral blood flow velocity and blood pressure. A clinical test of autoregulation. // Stroke. 1995. Vol. 26. P. 1801 - 1804.
  4. Fleischer L.H., Young W.L., Pile-Spellman J. et al. Relationship of transcranial Doppler flow velocities and arteriovenous malformation feeding artery pressures // Stroke. 1993. Vol. 24. . 1897-1902.
  5. Handa T.C., Hegoro M., Miyachi S., Sugita K. Evalution of pressure changes in feeding arteries during embolization of intracerebral arteriovenous malformations // J. Neurosurg. - 1993. Vol. 79 3. . 383-389.
  6. Henkes H., Gotwald T.F., Brew S., Kaemmerer F., Miloslavski E., Kuehne D. Pressure measurements in arterial feeders of brain arteriovenous malformations before and after endovascular embolization. // Neuroradiology. 2004 Vol. 46(8) P. 673-677.
  7. Jungreis C.A., Horton J.A., Hecht S.T. Blood pressure changes in feeders to cerebral arteriovenous malformations during therapeutic embolization. // American Journal of Neuroradiology. 1989. Vol. 10 P. 575578.
  8. Klotzsch C. Henkes H.,. Nahser H.C., Kuhne D., Berlit P. Transcranial Color-Coded Duplex Sonography in Cerebral Arteriovenous Malformations // Stroke. 1995. Vol. 26. P. 2298-2301.
  9. Lazar R.M., Marshall R.S., Pile-Spellman J. et al. Anterior translocation of language in patients with left cerebral arteriovenous malformations. // Neurology. 1997;49:802808.
  10. Luginbuhl M., Schroth G., Thomson D. Interventional neuroradiology and minimally invasive neurosurgery. // Curr Opin Anaesthesiol. 1997 Vol. 10. P. 287-296.
  11. Manninen P.H., Gignac E.M., Gelb A.W. et al. Anesthesia for interventional neuroradiology. // J Clin Anesth. 1995 Vol. 7. P. 448-452.
  12. Massoud T. C., Vinuela F. et al. An experimental arteriovenous malformation model in swine: anatomic basis and construction technique. // American Journal of Neuroradiology. 1994 Vol. 15. P.1537-1545.
  13. Moo L.R., Murphy K.J., Gailloud P., Tesoro M., Hart J. Tailored Cognitive Testing with Provocative Amobarbital Injection Preceding AVM Embolization. // American Journal of Neuroradiology. 2002. Vol. 23 P. 416-421.
  14. Ogilvy C.S., Stieg P.E., Awad I., Brown R.D., Kondziolka D., Rosenwasser R., Young W.L., Hademenos G. Recommendations for the Management of Intracranial Arteriovenous Malformations A Statement for Healthcare Professionals From a Special Writing Group of the Stroke Council, American Stroke Association. // Stroke. 2001 Vol. 32 P. 1458-1471.
  15. Rauch R., Vinuela F., Dion J., Duckwiler G., Amos E.C., Jordan S.E., Martin N., Jensen M.E., Bentson J., Thibault L. Preembolization functional evaluation in brain malformation: the superselective Amytal test // American Journal of Neuroradiology. 1992. Vol. 13. P. 303-318.
  16. Semenyutin V.B., Nikitin P.I., Antonov V.I., Bukhaev I.M. Monitoring of intracranial hemodynamics in endovascular exclusion of cerebral arteriovenous malformations from blood circulation//Journal of Regional Blood Circulation and Microcirculation.-2006.-N 3 (9).-P. 4-12.
  17. Spetzler R.F., Martin N.A. A proposed grading system for arteriovenous malformations // J. Neurosurg. - 1986. - V. 65, 4. - P. 476-483.
  18. Wakhloo B. B., Lieber R., Siekmann D. J., Gounis M. J. Acute and Chronic Swine Rete Arteriovenous Malformation Models: Hemodynamics and Vascular Remodeling // American Journal of Neuroradiology. 2005. Vol. 6 P. 1702-1706.
  19. Young W.L., Pile-Spellman J. Anesthetic considerations for interventional neuroradiology. // Anesthesiology. 1994. Vol. 80. P.427-456.