вход по аккаунту


Blood flow changes in arteriovenous malformation during behavioral activation.

код для вставкиСкачать
Blood Flow Changes in Arteriovenous
Malformation Dutrng Behavioral Activation
Georg Deutsch, P h D
Striking task-dependent fluctuations were observed in the cerebral blood flow pattern of a patient with a left
posterior hemispheric arteriovenous malformation (AVM). Two-dimensional measures of regional cerebral flow in
the resting state, using the xenon 133 inhalation technique, revealed a region of high flow coincident with the AVM
seen on the patient’s arteriograms. In subsequent studies, the AVM stood out as a region of high blood flow during a
relaxed state, while it approached normal levels of flow when there was attentional demand. These observations
suggest that focal regulatory mechanisms exist at the AVM or else that very substantial redistributions of blood
flow are taking place which the flow rate in the AVM reflects only passively. Patients considered for embolic
treatment of an AVM would benefit from an assessment of behavioral influences on flow in the AVM.
Deutsch G: Blood flow changes in arteriovenous malformation during behavioral activation.
A n n Neurol 13:38-43, 1983
In 1928, Fulton [5] reported changes in the flow of an
occipital arteriovenous malformation (AVM) during
visual stimulation. His technique was simple: he listened to the bruit with a stethoscope. Visual stimulation increased the sound (putatively, the flow) compared to the resting state. This observation accords
with the traditional attribution of visual function to
posterior areas of the brain and with coupling of
blood flow to metabolism.
Since then, techniques to quantify regional cerebral blood flow (rCBF) patterns in awake, functioning
human subjects have improved markedly. They include the development of a noninvasive procedure
using inhalation of xenon 133 gas [15, 171. Although
the exact mechanisms are not defined, it is now
widely accepted that rCBF is usually controlled by
local neuronal activity and usually reflects local
metabolic rate [20, 2 51. However, the relationship
between flow through pathological vascular formations, such as an AVM, and regional metabolic activity is not well understood. Generally these vascular
anomalies have been viewed as passive circulatory
short circuits.
Striking alterations in the blood flow pattern of the
cerebral hemisphere are reported here in a patient
with a left posterior AVM-changes
that depended
on the patient’s state of mental activation. The
changes were greater than those seen in activation
studies with normal subjects (in which rCBF techniques are used t o monitor cerebral changes during
altered behavioral states). They were paradoxical, in
that certain mental activities normalized a pathological flow pattern seen during the resting state.
From the Dementia Research Service, The Burke Rehabilitation
Center, 785 Mamaroneck Ave, White Plains, NY 10605.
Received Oct 14, 1981, and in revised form Mar 22, 1982. Accepted for publication Mar 26, 1982.
Patient Description
Three months prior to testing in our laboratory, a 52year-old right-handed man, employed as a systems engineer, suffered a sudden loss of speech. Computerized
tomographic (CT) scan and arteriographic studies
documented a hemorrhage i n an AVM centered about the
region of the angular gyrus in the left hemisphere (Fig 1).
The patient initially had an aphasia with dysnomia,
paraphasia, and difficulty with reading, writing, and following commands. He also had a mild right hemiparesis
and extinguished the right stimulus during double simultaneous visual stimulation. He recovered rapidly and was
referred by his neurologist to our institution for outpatient
speech therapy seven weeks later. Examination at that time
revealed normal visual fields and no sensory deficits. All
other cranial nerve function was normal. He had no ataxia,
his gait was normal, and there was n o motor weakness. The
patient was alert and fully oriented. His recent memory was
intact, calculations were done easily, and he could read and
Speech and language evaluation included administration
of parts of the Boston diagnostic aphasia examination [6].
Conversational and expository speech was rated on this test
as “some obvious loss of fluency in speech or facility of
Address reprint requests to D r Deutsch, Division of Neurosurgery E17, The University of Texas Medical Branch, Galvesron, TX 77550.
0364-5 134/83/010038-06$1.50 0 1982 by the American Neurological Association
F i g I . Left carotid and vertebral arteriogrums. The large AVM
in the parietal-temporal-occipital junction zuas sapplied by the
middle and posterior cerebral arteries and a scalp artery (not
comprehension, without significant limitation on ideas expressed or form of expression." Some hesitancy and distortion in multisyllabic word production was noted. A mild
auditory comprehension deficit was revealed by the token
test [21, o n which the patient received a score of 145 out of
Cerebral blood flow testing commenced six weeks after
these outpatient admission evaluations. The patient was reevaluated by a speech pathologist six weeks later (between
rCBF sessions C and D; see Procedure), at which time he
obtained a near-perfect score (162 out of 163) on the token
test and his speech was rated "minimal discernible speech
handicaps" on the Boston diagnostic aphasia examination.
The patient had returned to full-time employment in his
previous engineering capacity and was experiencing no
difficulty carrying out his responsibilities. A repeat C T scan
clearly showed the AVM but no further signs of hemorrhage.
Materials and Methods
The '""Xe inhalation method, described in detail by Obrist
et a1 117, 181, was used to determine rCBF using the
Medimatic Inhamatic 33 system. T h e patient breathed
'""Xe, 3 mCi per liter mixed with air and oxygen, for a
period of one minute through a close-fitting face mask in a
5 liter rebreathing system. T h e buildup and washout of the
radioisotope in the brain was monitored for ten minutes by
16 pairs of collimated sodium iodide scintillation detectors
placed over homologous regions of the two cerebral hemispheres. Head position within the detector array was kept
constant throughout the series of studies through the use of
an aligning template. Resolution at the cortical surface was
approximately 2.5 cm. The curves generated by changes in
the head detector counts could be displayed for inspection
on a color monitor during the data acquisition phase of
each study.
The patient was tested in a reclining chair with his head
approximately 30 degrees from the horizontal. All studies
were conducted in an environment-controlled laboratory
under virtually identical background lighting and noise conditions. The subject gave informed consent.
The patient's expiratory and inspiratory '"Xe level was
continuously monitored by drawing air directly from the
mask through a thin catheter and passing it through an air
detector system. The resulting curve, representing the
isotope input, was used in conjunction with the curves generated by the head detectors to calculate regional washout
rates. (The end-tidal ""Xe concentrations approximate the
arterial levels of the isotope.) The expired air was also continuously monitored for carbon dioxide content.
The rCBF values were calculated by an on-line computer
system using a two-compartment analysis [18, 231. Values
calculated for individual detectors were considered valid if
(1) there was no xenon leakage from the mask, (2) the head
curve did not show unusual discontinuities or artifacts, and
(3) the computing algorithm achieved a good fit within
fewer than 30 iterations. Calculated variables included Fg
(gray matter flow derived from the fast compartment), IS1
(initial slope index measured between the second and third
minute, indicative of overall flow but dominated by gray
matter flow), Wg (relative gray matter weight), and K2 (exponent describing the washout rate of the slow compartment, mainly white matter and extracerebral flow).
The patient was tested ten times over the course of four
months (Table). All studies except for the first and last
Deutsch: Blood Flow Changes in AVM
Gray Matter Blood Flow a t Various Testing Conditions versus Resta
Study No. and
Testing Condition
Session A (10/22180)
1 Resting
Session B (11/12/80)
2 Resting
3 Visual feedback
Session C (11/19/80)
4 Autosuggestion
5 Resting
Session D (12115/80)
6 Resting
7 Autosuggestion
Session E (1121/81)
8 Vigilance
7 Resting
Session F (2116181)
10 Restina
a. Left
b. Right
c. Left
d . Right
e. L-R
f. Left
g. Right
h. L - R
- 1.2
aTesting conditions and resulting flow values (mU100 gm/min), per hemisphere, as indicated by shaded detector positions. Shown are overall
mean flow for 16 detectors, mean flow value for 4 posterior detectors, and mean flow value for 4 anterior detectors. Right hemisphere values
are subtracted from homologous left hemisphere values to form the posterior and anterior "difference" columns.
were conducted in pairs, one resting state measurement
and one measurement involving some form of directed
mental activity. There were a total of six sessions. T h e
order of the rest and activation state was alternated. T h e
resting state measurement was conducted under low ambient lighting and with auditory stimulation limited to the
constant low-level background noise of the laboratory instruments. The patient was instructed to lie back in the recliner and relax. He reported at the end of these sessions
that he had felt comfortable and relaxed despite the presence of the face mask and head detectors.
Resting state measurements of cerebral flow were conducted in all six sessions. Measurements involving some
form of directed mental activity were conducted in sessions
B, C, D, and E. These were as follows.
Session B-visual
feedback condition: T h e patient
viewed o n a television monitor the formation of four
curves representing counts measured by 4 left hemisphere
detectors. Two of the curves were generated by detectors
pointing at the AVM region (as established by the restingstate study). The patient was instructed to concentrate on
keeping these curves from dropping too rapidly, i.e., to
prolong the washout of xenon. Two curves from frontal
detectors were also displayed for comparison.
condition [ 141:
Sessions C and D-"autosuggestion"
The patient concentrated on imagining that the back of his
head was very cold and covered with ice, while the front
was hot and active. He started this attempt at internal self-
40 Annals of Neurology
Vol 13 No 1 January 1983
regulation one minute before xenon inhalation and continued through the entire ten-minute measurement period.
Session E-vigilance task: The patient silently counted
the number of rimes the word and occurred in a passage
recited out loud by the examiner. Reading was started one
minute before xenon inhalation and continued through the
entire measurement period. The patient correctly identified the number as 31.
The 133Xewashout curves recorded in all regions, including that of the AVM, were valid by conventional
criteria [17, 18, 211. In the initial resting state, a region of very high flow (>170 m1/100 gm/min) was
observed in the parietooccipital junction of the left
hemisphere. During attention-demanding nonvisual
activities, flow in this region decreased to less than
100 m1/100 gm/min (Fig 2).
Examination of interstudy regional flow changes is
conventionally facilitated by expressing all regional
values as deviations from the mean flow of the study
in which they were recorded [lo, 19, 22, 241. In our
patient, the fluctuations in the AVM flow contributed
substantially to fluctuations in the overall mean flow.
Since examination of AVM flow change was our goal,
it was more appropriate to normalize overall fluctua-
F i g 2. Computer-generated, color-scaled imageJ of gray matter
flow ( F g ) i n a resting state and during performance of a vigilance task, derived from 32 detectorpositions by interpolating
values between all adjacent detectors. Both studies are from Jession D . The scale represents milliliters per 100 grams of tissue
per minute.
tions by using homologous regions of the right hemisphere as control values for the left hemisphere (the
site of the AVM).
There was a high flow in the posterior region
of the patient’s left hemisphere during all rest
states (see the Table). All nonvisual activation studies
showed a substantially smaller left-right difference in
the posterior region (- 1.5 to 11% greater in the left)
than did the resting-state studies (2 1 to 385% greater
in the left) (Table; Fig 3 ) . Study no. 3 (visual feedback) showed an increase in the posterior flow of
both hemispheres (Table, columns c and d) as well as
maintenance of the marked left-right difference (high
AVM flow) seen in all resting state measures. If
the regional mean flow calculation for the three nonvisual activation studies and for the three restingstate measures accompanying them is averaged, only
the change in the posterior left hemisphere in the
resting Versus active state is nonoverlapping and
significant across these sessions. An independent
check of the statistical significance of fluctuations in
detector values was performed by computing the
distribution of regional flow values for all detectors in
each hemisphere for each study in the sessions involving both a resting state and nonvisual activation
(sessions C , D, and E). T h e 9995 confidence interval
for each distribution was determined. Detector values exceeding the confidence interval (p < 0.01) OCcurred only in the resting state and involved only the
2 detectors located at the AVM.
Despite some variation in task, all sessions involving
nonvisual activation ( C , D , and E) showed a consistent pattern: the AVM stood out as a region of high
blood flow during a relaxed state, while it approached
normal flow levels when there was attentional demand.
All resting-state studies revealed a high flow in the
AVM region. T h e one activation study which also
showed a high flow in that region (study 3) involved
concentrated visual activity. This study in essence reproduces Fulton’s observation and is compatible with
conventional views of localization of function as well
as expected changes in regional metabolic activity. It
was, however, a weaker effect than that observed in
the nonvisual activation conditions.
Deutsch: Blood Flow Changes in AVM
- c
n u
10 1
% @
The patient’s overall mean flow fluctuated considerably in the earlier studies but became more stable
as the four-month period of testing progressed. Such
fluctuations in overall flow are not ordinarily seen in
control subjects [lo, 19, 241 and, in addition to the
obvious contribution of changes in AVM flow itself,
may be due to more general effects of the AVM and
the previous hemorrhage on cerebral blood flow [ 11,
The state-dependent changes in AVM flow, especially the fact that certain mental activity “normalizes” the pathological flow pattern seen at rest,
raise theoretical and practical issues. Focal regulatory
mechanisms may exist in the AVM, a pathological
formation usually viewed only as a passive shunt or
short circuit [ 4 , 261. On the other hand, the AVM
may simply be passively reflecting substantial redistributions of blood flow in other parts of the cerebrum. If the first alternative is correct, then it
suggests a neural regulating mechanism spanning a
large distance between the AVM site and the cortical
regions traditionally assumed to be involved in attention: the frontal lobes. Simple attentional tasks are
not ordinarily thought to involve posterior cortical
areas [7,131.
The second alternative accords with documented
examples of substantial shifts in blood flow, such as
the “subclavian steal” syndrome [ 1,4, 161. Activation
of frontal regions by an attentional task may in effect
“steal” blood from posterior regions, especially if the
posterior flow is abnormally high to begin with. The
absolute changes in frontal flow observed between
rest and activation states support this explanation
only partially, because study 8 (the vigilance task)
42 Annals of Neurology
Vol 13 No 1 January 1983
F i g 3 . State-dependent regional ftou’ juctuations of the lefi
hemisphere relatiue to right hemisphere$ow. I n the upper
graph. the mean ftow measured by 4 right posterior detectors is
subtracted from the mean jou’for 4 homologous left detectors
tthe location of the AVM)for each stu4y. In the lower graph,
the mean f t o w for 4 right unterior detectors is subtracted from
the mean ftoul of 4 homologous left detectors for each study. The
values for a normal control in a resting State and while performing the vigilance task are included at the right (published
values for regional cerebral blood $ow .thou) no significant
left-right difference.sfo. normai subjects in the resting state
and only smal, if any, a.tymmetrie.s during various activation
tasks [ 19, 22, 241).
showed a smaller frontal flow compared to the subsequent resting state measurement. Normalizing for
the fluctuations in overall mean flow is not entirely
legitimate, since, as mentioned earlier, AVM flow
contributes significantly to mean flow in the resting
state and, as a result of the lesion’s posterior position,
decreases the relative hyperfrontality [8] of the resting state. However, the normalized values d o support
the “steal” explanation. Subcortical flow changes
certainly cannot be ruled out by two-dimensional
measurements. The vertebral artery system supplies
the mesencephalic reticular formation [28], and increased reticular activity during attentional activation
might divert blood from the basilar supply of the
Caution concerning the flow values calculated for
the AVM region is needed. The model for calculating
cerebral blood flow via gas washout [9, 21, 291 assumes free diffusion of xenon into tissues and an
equilibrating tendency between the xenon concen-
tration in tissue and blood. The flow calculation is
dependent on both the observed washout rate and
the partition coefficient, a constant representing the
rate of xenon exchange between tissue and blood.
The model may not hold for the detector(s) at the
AVM location, for they are viewing a large pool of
blood. The determined flow values should not be regarded as precise quantification of flow in the AVM,
but rather as estimates whose changes provide relative information (such as the relative differences in
flow observed under varying experimental conditions).
A practical implication of the state-dependent
fluctuations in AVM flow concerns surgical procedures involving AVM embolization [3, 12,271. Some
of these techniques depend on the high flow to the
AVM to direct emboli from an upstream catheter to
locations in the AVM itself. Patients considered for
embolic treatment of an AVM would benefit from
careful assessment of cerebral blood flow to determine whether changes in behavior influence flow in
the AVM. If the patient's behavioral state alters this
rate of flow, then behavioral control might optimize
flow through the AVM during the procedure.
Supported by the Christian Johnson Endeavor Fund, the Will
Rogers Foundation, and the Winifred Masterson Burke Relief
Presented in part at the Fourth European Meeting of the International Neuropsychological Society, June 29-July 2, 1981, Bergen,
It is a pleasure to thank Dr John Blass, Dr Alec Zemcov, Dr
Fletcher McDoweil, Dr James Tweedy, and Mr Oscar Perilla for
helpful discussions and support.
1. Barnett HJM, Wortzman G , Gladstone RM, et al: Diversion
and reversal of cerebral blood flow: external carotid artery
"steal." Neurology (Minneap) 2O:l-14, 1970
2. De Renzi E, Vignolo LA: The token test: a sensitive test to
detect receptive disturbances in aphasics. Brain 85:665-678,
3. Drake CG: Advances in the neurosurgical treatment of
aneurysms, arteriovenous malformations and hematomas of
the vertebral circulation. Adv Neurol 16:2 11-225, 1977
4. Fisher CM: The anatomy and pathology of the cerebral vasculature. In Meyer JS (ed): Modern Concepts of Cerebrovascular Disease. New York, Spectrum, 1975, pp 1-41
5. Fulton JF: Observations upon the vascularity of the human
occipital lobe during visual activity. Brain 51:310, 1928
6. Goodglass H , Kaplan E: Assessment of Aphasia and Related
Disorders. Philadelphia, Lea & Febiger, 1972
7. Hecaen H , Albert ML: Human Neuropsychology. New York,
Wiley, 1978, pp 354-378
8. Ingvar D H : Hyperfrontal distribution of the cerebral grey
matter Bow in resting wakefulness: on the functional anatomy
of conscious state. Acta Neurol Scand 60:12-25, 1979
9. Kety SS: Theory of blood tissue exchange and its application
to measurement of blood flow. Methods Med Res 8~223-227,
10. Lassen NA, Roland PE, Larsen B, et al: Mapping of human
cerebral functions: a study of the rCBF pattern during rest, its
reproducibility and the activation seen during basic sensory
and motor functions. Acta Neurol Scand 56 [Suppl 641:262263, 1977
1. Leszczynski S, Rajszys R, Miernowski S: Analysis of the
dynamics of cerebral blood circulation in cases of arteriovenous fistula of the brain, with the aid of roentgen cinematography. Fortschr Geb Roentgenstr Nuklearmed 98:7 14-724,
2. Luessenhop AJ, Presper JP: Surgical embolization of cerebral
arteriovenous malformations through internal carotid and
vertebral arteries: long term results. J Neurosurg 42:443451, 1975
3 . Luria AR: Higher Cortical Functions in Man. New York,
Basic Books, 1966, pp 218-295
4. Luthe W: Autogenic Therapy. New York, Grune 81 Stratton,
5 . Mallett BL, Veall N : Measurement of regional cerebral clearance rates in man using xenon-133 inhalation and extracranial
recording. Clin Sci 29:179-191, 1965
6. Mannock JA, Suter LG, Hume DM: The "subclavian steal"
syndrome. JAMA 182:254, 1961
17. Obrist WD, Thompson HK, King CH, et al: Determination of
regional cerebral blood flow by inhalation of ""xenon. Circ
Res 20:124-135, 1967
18. Obrist WD, Thompson HK, Wang HS, et al: Regional cerebral blood flow estimated by I33xenon inhalation. Stroke
6:245-256, 1975
19. Prohovnik I, Hakansson K, Risberg J: Observations on the
functional significance of regional cerebral blood flow in
"resting" normal subjects. Neuropsychologia 18:203-2 17,
20. Raichle ME, Grabb RL, Gad0 MH, et al: Correlation between
regional cerebral blood flow and oxidative metabolism. Arch
Neurol 33:523-536, 1976
21. Reivich M, Slater R, Sano N: Further studies on experimental
models of cerebral clearance curves. In Brock M, et a1 (eds):
Cerebral Blood Flow. New York, Springer-Verlag, 1969, pp
22. Risberg J : Regional blood flow measurement by 1"3xenoninhalation: methodology and applications in neuropsychology
and psychiatry. Brain Lang 9:9-34, 1980
23. Risberg J, Ali 2, Wilson EM, et 21: Regional cerebral blood
flow by '38xenon inhalation. Stroke 6:142-148, 1975
24. Risberg J, Halsey JH, Wills EL, et al: Hemispheric specialization in normal man studied by bilateral measurement of the
rCBF-a study with the 133xenoninhalation technique. Brain
98:511-524, 1975
25. Sokoloff L: Circulation and energy metabolism of the brain. In
Siege1 GJ, Albers RW, Agranoff BW, Katzman R (eds): Basic
Neurochemistry. Third edition. Boston, Little, Brown, 1981
26. Stein BM, Wolpert SM: Arteriovenous malformations of the
brain, I: Current concepts and treatment. Arch Neurol
37:l-5, 1980
27. Stein BM, Wolpert SM: Arteriovenous malformations of the
brain, 11: Current concepts and treatment. Arch Neurol
37:69-75, 1980
28. Truex RC, Carpenter MB: Human Neuroanatomy. Sixth e d tion. Baltimore, Williams & Wilkins, 1969, pp 74-78
29. Veall N , Mallett BL: The partition of trace amounts of xenon
between human blood tissues at 37°C. Phys Med Biol
Deutsch: Blood Flow Changes in AVM
Без категории
Размер файла
639 Кб
flow, behavior, change, malformations, arteriovenous, activation, blood
Пожаловаться на содержимое документа