- Email: [email protected]
Cerebral blood flow decrements in chronic head injury syndrome
146
BIOL PSYCHIATRY 1985;20:146-157
Cerebral Blood Flow Decrements in Chronic Head Injury Syndrome Laurie Barclay, Alexander Zemcov, Walter Reichert, and John P. Blass
Global and regional cerebral blood flow (CBF) has been reported to be decreased immediately after head injury. Since neuropsychological deficits in patients with chronic head injury syndrome persist long after the initial trauma, we studied CBF, which reflects cerebral metabolism and activity, and evaluated the results of neuropsychological tests in 12 alert, responsive patients aged 18-26, both 2-13 months after head injury and 1-12 months after they regained consciousness. Global CBF was significantly decreased in patients with head injury relative to age-matched normal controls. Four patients had well-localized injury; in three of these, CBF over the affected region was significantly decreased relative to the same region in the opposite hemisphere and relative to the same region in an age-matched normal control. In one patient with rightfrontal injury, cognitive improvement on repeat testing was associated with a relative increase in right frontal CBF. The remaining eight patients had diffuse bilateral injury. Asymmetry ratios (larger value of mean hemispheric CBF divided by the smaller) were significantly higher for the group with localized dysfunction (1.2 +- 0.08 vs. 1.0 +_ 0.01, p < 0.01), suggesting homolateral decrease in CBF in this group. Three patients with diffuse injury underwent repeat studies 5-14 weeks later; all improved on psychological tests, and two had a significant increase in global CBF. These results suggest that the chronic sequelae of head injury include decreased CBF, presumably reflecting decreased cerebral metabolism, which correlates with the neuropsychological impairment. Introduction In the United States, traumatic brain injuries occur at a rate of 200 new cases per 100,000 population per year; approximately 10% of all head-injured patients who survive the initial trauma have neuropsychiatric disability requiring prolonged hospitalization (Alexander 1982). Patterns of neuropsychological recovery are extremely variable and are not easily predicted from clinical characteristics at the time of injury (Dikmen et al. 1983). Although the computed tomographic (CT) scan gives useful anatomical information about the injury, a physiological correlate of brain function such as cerebral blood flow (CBF) would be useful as a prognostic indicator of neuropsychological recovery. Global CBF is generally
From the Burke Rehabilitation Center, Comell University Medical Center, White Plains, New York 10605. Address reprint requests to: Dr. Laurie Barclay, The Burke Rehabilitation Center, 785 Mamaroneck Avenue, White Plains, New York 10605. Received May 4, 1984; revised August 20, 1984.
© 1985 Society of Biological Psychiatry
0006-3223/85/$03.30
Chronic Head Injury Syndrome
BIOLPSYCHIATRY 1985;20:146-157
147
reduced immediately after acute head injury and tends to increase toward normal shortly thereafter (Fieschi et al. 1974; Overgaard and Tweed 1974). Bruce et al. (1973) reported that only two of eight patients had greatly reduced CBF adjacent to cerebral contusions, whereas Obrist et al. (1979a) reported generally good correlations between evidence of brain damage on CT and CBF. As level of consciousness improved in head injury patients, global CBF generally increased as well (Obrist et al. 1979b). Since neuropsychological deficits in patients with chronic head injury persist long after the initial head trauma, and since CBF reflects cerebral metabolism, we postulated that CBF might be depressed even in chronic head injury syndrome. We now report on global and regional patterns of CBF in 12 patients 2-13 months after head trauma, their correlation with anatomical and neuropsychological deficits, and their normalization with time, paralleling the pattern of neuropsychological recovery.
Methods
Patients Patients were admitted to the Head Trauma Unit at the Burke Rehabilitation Center for rehabilitation of cognitive, motor, and sensory deficits following closed head injury sustained in motor vehicle accidents. All patients were comatose after head trauma for periods ranging from l to 10 weeks, but all were alert and responsive and had regained consciousness 1-12 months preceding CBF study. Three patients had skull fractures, and in three others a subdural hematoma had been removed (Table 1). On the basis of CT scan at injury and neurological examination at the time of CBF study, patients were divided into those with predominantly unilateral dysfunction (Table 2) (n = 4) and those with diffuse bilateral dysfunction (Table 3) (n = 8). Tables 2 and 3 list the interval between head trauma and CBF and neuropsychological testing (range, 2-13 months), as well as medication and neurological findings at the time of study.
Controls Normal controls were cognitively intact, gainfully employed individuals who were free of overt neuropsychiatric disease and who had no significant medical illnesses by history.
CBF Study CBF was estimated by noninvasive la3xe inhalation and measurement of its washout over 10 min in a 32-detector apparatus (Medimatic Corporation, Copenhagen, Denmark). Xenon is an inert gas that diffuses freely and rapidly through all brain tissues. Peak radioactivity counts are achieved at each detector after xenon inhalation through a face mask; counts decline in biexponential fashion during the washout phase, during which the subject breathes room air. The rate of decline of radioactive counts is roughly proportional to CBF, and indices of CBF can be derived from peak values and decay constants of the washout curve. The location of detectors over each hemisphere is given in Figure 1. In this study CBF values measured included F1, or gray matter flow as derived from the rapid phase of the biexponential curve (Ohrist et al. 1975), and the initial slope index (ISI), or slope of the washout curve 2-3 rain after peak values are achieved (Risberg et al. 1975). Informed consent for CBF study was obtained from patients and their parents
148
BIOL PSYCHIATRY 1985;20:146-157
L. Barclay et al.
or directly from normal controls under a protocol approved by the Burke committee on research involving human subjects. Four patients had repeat CBF study at intervals ranging from 6 to 14 weeks.
Psychometric Testing All patients except patient 11, who was combative and resistive to all test procedures, were tested at the time of each CBF study. Tests included standardized tests of list memory (Tweedy et al. 1982); the block design subtest of the Wechsler Adult Intelligence Scale (Wechsler 1955); phonemic and semantic fluency tasks, in which patients were asked to name as many words as possible beginning with the letter G or as many animals as possible in 45 sec (Goodglass and Kaplan 1976); picture recognition (Tweedy et al. 1982); token test for receptive dysphasia (DeRenzi and Vignolo 1962); progressive matrices (Raven 1963); cancellation tests of attention span (Tweedy et al. 1982); and motility index (Stem et al. 1969).
Data Analysis In patients with well-localized dysfunction, mean F1 over the affected region was compared with mean F1 over the same detectors in the unaffected hemisphere and in an agematched control subject. Statistical significance was tested by the paired t-test. In all patients, asymmetry ratios were calculated by determining mean ISI over each hemisphere and by then dividing the larger hemispheric value by the smaller. Mean asymmetry ratios were compared using the unpaired t-test. One patient with well-localized dysfunction had repeat CBF testing 7 weeks later. The ratio of mean F1 over the affected detectors to mean F1 over the same detectors in the opposite hemisphere was calculated at both time points, and an improvement ratio was defined as the initial ratio divided by the repeat ratio. Three patients with diffuse bilateral dysfunction also underwent repeat CBF studies. Mean F2 over all 32 detectors was compared for each patient at initial and at repeat study using the paired t-test. Improvement ratios were calculated by dividing mean F2 at initial study by mean F2 at repeat study. Figure 1. Detector placement. Xenon-133 washout was measured in a 32-detector apparatus, with 16 detectors applied to each hemisphere.
149
8~
.8 r~
r~ Z
ZZ
z
z
.8
z
zz
zi~
z~
°
~z
Z
~
~.~
~
- °
.~ .8
~'~ ~
o
r=
~
"~'~
d
[-r~
o.8
z~
o
z
~z
~
~
zz
e~
2:2 o
~J
u
.-&
150
L. Barclay et al.
BIOL PSYCHIATRY 1985;20:146-157
Results Patients ranged in age from 18 to 26 years; 10 were male and 2 female. All patients were fight-handed except for patient 2. Patients 3, 5, and 6 all had basilar skull fractures; patient 3 had an associated CSF leak, patient 5 had associated fractures of the left zygoma and sphenoid, and patient 8 had associated left frontal and occipital fractures. On CT scan, four patients had subdural hematomas, four had contusions, two had intraventricular hemorrhages, and four had ventricular dilatation. The five patients tested with an electroencephalogram (EEG) showed diffuse slowing. Three patients had their subdural hematomas removed; the other nine patients did not undergo surgery. Duration of coma after trauma ranged from 1 to 10 weeks (Table 1). On the basis of CT findings at injury and neurological and psychometric findings at the time of CBF study, patients were divided into two groups: those with predominantly unilateral dysfunction (patients 1-4) (Table 2) and those with diffuse bilateral cerebral dysfunction (patients 5-12) (Table 3). Patients in the two groups were comparable in terms of mean age (_+SEM) (21.3 +_ 1.7 vs. 20.6 -+ 1.5); male : female ratio (3 : 1 vs. 7 : 1), duration of coma, and percentage with skull fracture or surgery (Table 1). Patients in the two groups were also comparable in terms of interval between trauma and initial CBF study (mean _+SEM; 5.5 _+ 2.6 vs. 4.3 _+ 0.8 months) and medications at the time of CBF study (Tables 2 and 3). Patients with predominantly unilateral dysfunction had few diffuse signs and had focal signs involving only one hemisphere (Tables 2 and 3). All patients were found to have deficits on neuropsychological testing (Table 4). Normal controls were reasonably well matched to patients with head trauma in age and sex. Mean ISI values were comparable in each hemisphere, so that asymmetry ratios in normals approximated 1.0 (Table 5). Mean ISI over 32 detectors was significantly higher in normal controls than in patients with predominantly unilateral dysfunction (mean _+SEM; 50.75 _+ 1.67 vs. 40.81 _+ 4.03, p < 0.01 as determined by unpaired t-test). Patients with diffuse bilateral dysfunction also had lower global ISI than did normal controls (45.25 -+ 2.96 vs. 50.75 _+ 1.67, p < 0.05). Table 2. Clinical Characteristics of Patients with Predominantly Unilateral Dysfunction" Patient number
Interval (months)
l
13
2
5
Medications b
Diffuse signs"
DPH
None
DPH, lnderal,
None
Benadryl 3
2
DPH
4
2 4
None
None
~, attention span; agitation and perseveration; frontal lobe signs
LH signs
RH signs
Exp and rec dysphasia Mild R HP and HS Rec > exp dysphasia Mild R HP Exp and rec dysphasia Mild apraxia and R HP R facial seizures None L
None None None
neglect, HP and HS
ODPH, diphenylhydantoin; exp, expressive; rec, receptive; HP, hemiparesis; HS. hemisensory deficit. ,l, decreased. bAt the time of cerebral blood flow study. CSignsof diffuse bilateral cortical dysfunction. dLeft hemispheric dysfunctions. eRight hemispheric dysfunctions.
151
r~
ve ..A
.~
~.~
.~
~,
e~
~g ,3
oo e~
;.g
---,
~
~
---,---,
<
---, o
z" o
% e~ o
u~ °~
M
it3
~ID
t' ~
~
~'~
~
,-~
f',,I
152
~ ~ ZZZZZZ--
-
~ ZZ~
~ ~ ZZZZZZZZ
@@@@@q@O@q@q@@@@@@@@
ZZZZZZZZZZZZZZZZZZZZ ¢)
~
~
_
~
zzzzzz-
~zz
z-
~ ~ ~ z - ~
~
~
-
~
--
zzz
-z
~
~
-
zzzzz
- ~ ---
~
_
~
F
zzzzzzzzzz~zzz
0 [...,
0
~o 0
z
~-~-~- I ~
I°~
zzzzz
I I I~°~-I
° I I
BIOLPSYCHIATRY
Chronic Head Injury Syndrome
153
1985;20:146-157
T a b l e 5. N o r m a l C o n t r o l s a,b
Number
Age
Sex
ISI L
ISI R
Asymmetry ratio c ISI larger/smaller
IN 2N 3N 4N 5N 6N 7N 8N 9N 10N 1IN 12N
30 19 24 26 22 21 22 30 30 27 27 22
M M M F M M F F M M M F
55.12 46.49 58.44 47.81 47.70 51.73 37.43 53.45 52.08 51.72 52.59 43.85
57.12 49.64 59.97 49.64 48.63 52.26 38.75 56.16 53.03 53.04 54.57 46.77
1.038 1.068 ! .026 1.038 1.019 1.010 1.035 1.051 ! .018 1.026 1.038 1.067
aMean ISI (initial slope index) over the left (L) and right (R) hemispheres was calculated as described by Risherg et al. (1975). #Mean: 1.036 .4- 0.005. CAsymmetry ratio: larger value of hemispheric ISI divided by the smaller value.
In three of four patients with predominantly unilateral dysfunction, mean F1 over the detectors overlying the affected region was significantly decreased relative to mean FI over the same detectors in the opposite hemisphere and relative to mean F1 over the same detectors in an age-matched normal control subject (Table 6). The mean asymmetry ratio in patients with predominantly unilateral dysfunction was significantly increased relative to patients with bilateral diffuse dysfunction, indicating greater hemispheric asymmetry of CBF in the former group (Table 7). Furthermore, mean asymmetry ratio in patients with predominantly unilateral dysfunction was significantly larger than in normal controls (p < 0.001 by unpaired t-test) (Tables 5 and 7).
Table 6. Regional Changes in Cerebral Blood Flow in Patients with Predominantly Unilateral Dysfunction ~ F1
Patient number 1 2 3 4
Affected region L L L R
Detectors
temporal 4,5,6 frontoparietal temporal 4,5,6,8,9,10,13,14,15 frontoparietal temporal 4,5,6,8,9,10,13,14,15 frontoparietal 4,5,8,9,10,13,14,15
Affected hemisphere (ml/100 g-min) 41.00 70.88 59.56 67.50
----- 2.1 ----- 3.8 ± 2.8 ± 3.1
Control (opposite) hemisphere (ml/100 g-rain) 55.33 76.75 65.22 78.25
± ± ± ±
1.8 b 3.0" 2.3 3.8 c
Control (ml/100 g-min) 66.27 82.36 53.96 103.83
± ± ± ±
4.2 d 3.6 c 1.1 4.5 b
"Mean FI, or gray matter flow, was calculated over those detectors thought to represent the most damaged region as determined by CT scan, neurological evaluation, and psychometric testing and was compared with mean FI over the same detectors in the opposite hemisphere, as well as with mean FI over the detectors in the corresponding normal control (see Table 5). Columns represent patient number, affected region as determined by CT and neuropsychoiogical evaluation, detector numbers overlying the affected region, F1 over those detectors in the affected hemisphere, FI over the same det~tors in the opposite hemisphere, and FI over the same detectors in the corresponding age-matched normal control. FI values represent mean -- SEM. Allp values are relative to FI (patient), affected hemisphere, by paired t-test.
~, < 0.001. Cp < 0.05. ~p < 0.0O5.
154
BIOLPSYCHIATRY
L. Barclay et al.
1985;20:146-157
Table 7. Asymmetry Ratiosa Patient number
Dysfunction
1 2 3 4
L L L R
ISI L
ISI R
Ratio ISI larger/smaller
Predominantly unilateral dysfunction 34.00 45.39 41.29 43.43
37.61 50.32 45.60 28.82
1.106 1.109 1.104 1.507
Bilateral diffuse dysfunction 5 6 7 8 9 10 11 12
Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral B il ateral
36.77 41.42 48.08 34.83 43.80 46.96 47.11 60.24
35.51 42.12 48.87 34.33 46.25 49.58 48.02 60.08
1.035 1.017 1.016 1.015 1.056 1,056 1.019 1.003
"Mean value of asymmetry ratios were averaged over the group of patients with predominantly unilateral dysfunction and the group of patients with bilateral diffuse dysfunction, The mean ratio for the former group (1.204 - 0.085) was significantly higher (p < 0.01 by the unpaired t-test) than that for the latter (1.027 ± 0,007), indicating more asymmetry of CBF in the former group.
All four patients for whom a repeat CBF study was done also underwent repeat neuropsychological testing at the same time, and all showed improvements in selected subtests. Mean overall improvement for each patient ranged from 60% to greater than 500% (Table 8). One patient (patient 4) with right frontoparietal dysfunction had a repeat CBF study 7 weeks after initial study. Although mean F1 over the affected detectors decreased on the repeated study, global FI was also decreased, and the ratio of FI over the affected detectors to F1 aver the same detectors in the opposite hemisphere increased on the repeat study (Table 9) suggesting normalization of FI in the affected region. Two of three repeat CBF studies on patients with diffuse bilateral dysfunction showed a significant increase in mean F1 averaged over 32 detectors (Table 9). Discussion Previous studies have indicated that global CBF is depressed immediately after head injury (Fieschi et al. 1974; Overgaard and Tweed 1974); the present study indicates that global CBF is depressed relative to age-matched controls for periods of as long as 13 months following head trauma. Global CBF is persistently decreased regardless of whether patients have well-localized dysfunction or diffuse, bilateral dysfunction. Since this study was done on patients with chronic head injury syndrome at intervals of 1-12 months after consciousness had been regained, the results cannot be extrapolated to patients with acute head injury; specifically we cannot comment on the prognostic significance of CBF in predicting return of consciousness. Three of four patients with well-localized dysfunction on CT scan and neuropsychological testing were shown to have decreased F1 over the affected region, whether compared with the same region in the opposite hemisphere or with the same region in an age-matched control subject. Our findings confirm and extend those of Obrist et al. (1979a), who reported a good correlation between CT scan and CBF in head trauma.
155
+++
+
++
I ++÷t+
I
I
+
+
÷1 +
- - ° ~
+
+~ + +
+
~ mZ
+1 +
+
+ +
+
__
+
+~ i
i +i + ~
÷1
+
÷~ ~~
+
+1
_
~
ZZ
e~
+ + + + + + ÷ 1 1
i +
i
÷1 + + +
+ ÷1 + + + ÷l + +
+ ÷l
~ " o
% ~
-
6
z
0
.~2
+ ÷l + + + +
÷1 + + ~
+ + ÷1 + + + ÷1 + +
+ ÷l
o 't~ 'N
'~'~"
I °
I °°
I I °°
I °
I I
.o.-~
2
o~ Z
=
~
~._~
~ [...
_o
_~o ~
~
~ "i~o
156
BIOL PSYCHIATRY 1985;20:146-157
L. Barclay et al.
Table 9. Repeat Cerebral Blood Flow Studies" Localized Dysfunction b Patient number Interval FI affected side (initial) FI control side (initial) F1 affected/control
4 7 weeks 65.8 - 7.4 82.5 ± 27.9 0.797
(initial) FI affected (repeat) FI control (repeat) F1 ratio (repeat) Ratio initial/repeat
Number
Interval
FI initial
5 6 7
7 wks 6 wks 14 wks
47.0 _ 5.1 57.7 - 9.3 64.0 - 6.8
53.8 -4- 5.9 51.0 ± 4.2 1.054 0.756
Generalized Dysfunction c F I repeat FI repeat-initial 57.6 ± 5.8 73.0 ± 8.5 55.3 -+ 7.7
10.0 .4- 5.7 20.1 ± 7.1 - 7 . 0 _ 8.0
FI initial/repeat
pd
.816 .790 1.159
<0.05 <0.01 N.S.
'rRepeat ~33Xewashout measurement was done 6-14 weeks after initial study in one patient with localized dysfunction (patient 4) and in three patients with generalized dysfunction (patients 5,6, and 7). Gray matter flow (F1) values are given in ml/100 g-min, and represent mean ( 4- SEM). the patient with localized dysfunction, FI over the affected region was averaged over detector numbers 4, 5, 8, 9, 10, 13, 14, and 15 in the fight hemisphere, which correspond to the site of his deficit (right frontoparietal). FI (control) was averaged over the corresponding detectors in the left hemisphere. qn the patients with generalized dysfunction, FI was averaged over all 32 detectors. 'tProbability values (p) of statistically significant differences between initial and repeat values were determined by the paired t-test; N.S. = not significant.
Furthermore, mean asymmetry ratio was increased in patients with well-localized dysfunction relative to patients with diffuse bilateral dysfunction or to age-matched normal controls, suggesting a homolateral decrease in hemispheric CBF in this group. One patient with right frontoparietal dysfunction showed decrease on repeat CBF measurement in F1 over the affected region but showed an increase in the ratio of mean F1 over the affected region to mean FI over the same region in the left hemisphere. This finding suggests normalization of regional CBF on repeat study despite decrease in global F1, which may suggest relative hyperemia at the time of initial CBF study. Three of four patients with diffuse generalized dysfunction showed improvement of global CBF on repeat studies. All these patients demonstrated improved performance on neuropsychological tests. Recovery in CBF seems to parallel recovery in neuropsychological function and is not limited to the acute increase in CBF that mirrors returning level of consciousness (Obrist et al. 1979b). Nevertheless, we do not advocate the use of CBF as a diagnostic tool in chronic head injury, since it merely reflects the regional or global decrease in cerebral activity that one would expect from the CT scan and neuropsychological deficits. Further studies are needed to determine whether CBF in chronic head injury may be of prognostic significance in predicting degree of ultimate neuropsychological recovery. Global and regional patterns of CBF seem to reflect neuropsychological deficits and their pattern of recovery in head-injured patients. The persistence of reduced CBF in patients with chronic neurological sequelae of head injury suggests the presence of a persistent physiological lesion accompanying the neuropsychiaa'ic deficit. Since CBF reflects cerebral metabolism, these findings also suggest a persistent decrease in metabolic activity. Whether this decrease in CBF is the primary disturbance underlying the neu-
Chronic Head Injury Syndrome
BIOLPSYCHIATRY 1985;20:146-157
157
ropsychological deficits is unclear, but it seems more likely to be the secondary effect of decreased cerebral activity due to underlying structural or metabolic changes following head trauma.
References Alexander MP (1982): Traumatic brain injury. In Benson DF, Blumer D (eds), Psychiatric Aspects of Neurologic Disease, Vol 2. New York: Grune & Stratton, pp 219-248. Bruce DA, Langfitt TW, Miller JD, Schutz H, Vapalahti MP, Stanek A, Goldberg H (1973): Regional cerebral blood flow, intracranial pressure, and brain metabolism in comatose patients. J Neurosurg 38:131. DeRenzi E, Vignolo LA (1962): The token test: A sensitive test to detect receptive disturbances in aphasics. Brain 85:665. Dikmen S, Reitan RM, Temkin NR (1983): Neuropsychological recovery in head injury. Arch Neurol 40:333. Fieschi C, Battistini N, Beduschi A, Boselli L, Rossanda M (1974): Regional cerebral blood flow and intraventdcular pressure in acute head injuries. J Neurol Neurosurg Psychiatry 37:1378. Goodglass H, Kaplan H (1976): The Assessment of Aphasia and Related Disorders, Philadelphia: Lea and Febiger. Obrist WD, Thompson HK, Wang HS, Wilkinson WE (1975): Regional cerebral blood flow estimated by 133Xenon inhalation. Stroke 6:245. Obrist WD, Dolinskas CA, Gennarelli TA, Zimmerman RA (1979a): Relation of cerebral blood flow to CT scan in acute head injury. In Popp AJ (ed), Neural Trauma. New York: Raven Press, pp 41-50. Obrist WD, Gennarelli TA, Segawa H, Dolinskas C, Langfitt TW (1979b): Relation of cerebral blood flow to neurological status and outcome in head-injury patients. J Neurosurg 51:292. Overgaard J, Tweed WA (1974): Cerebral circulation after head injury: Cerebral blood flow and its regulation after closed head injury, with emphasis on clinical correlations. J Neurosurg 41:531. Raven JC (1963): Guide to Using the Coloured Progressive Matrices, London: HK Lewis and Company, Ltd. Risberg J, Ali Z, Wilson EM, Wills EL, Halsey J (1975): Regional cerebral blood flow by ~33Xenon inhalation: Preliminary evaluation of an initial slope index in patients with unstable flow compartments. Stroke 6:142. Stern PH, McDowell F, Miller JM, Elkin RD (1969): Quantitative testing of motility defects in patients after stroke. Arch Phys Med Rehabil 50:320. Tweedy J, Reding M, Garcia C, Schulman P, Deutsch G, Antin S (1982): Significance of cortical disinhibition signs. Neurology (NY) 32:169. Wechsler D (1955): Manual for the WechslerAdultlntelligence Scale. New York: The Psychological Corporation.
BIOL PSYCHIATRY 1985;20:146-157
Cerebral Blood Flow Decrements in Chronic Head Injury Syndrome Laurie Barclay, Alexander Zemcov, Walter Reichert, and John P. Blass
Global and regional cerebral blood flow (CBF) has been reported to be decreased immediately after head injury. Since neuropsychological deficits in patients with chronic head injury syndrome persist long after the initial trauma, we studied CBF, which reflects cerebral metabolism and activity, and evaluated the results of neuropsychological tests in 12 alert, responsive patients aged 18-26, both 2-13 months after head injury and 1-12 months after they regained consciousness. Global CBF was significantly decreased in patients with head injury relative to age-matched normal controls. Four patients had well-localized injury; in three of these, CBF over the affected region was significantly decreased relative to the same region in the opposite hemisphere and relative to the same region in an age-matched normal control. In one patient with rightfrontal injury, cognitive improvement on repeat testing was associated with a relative increase in right frontal CBF. The remaining eight patients had diffuse bilateral injury. Asymmetry ratios (larger value of mean hemispheric CBF divided by the smaller) were significantly higher for the group with localized dysfunction (1.2 +- 0.08 vs. 1.0 +_ 0.01, p < 0.01), suggesting homolateral decrease in CBF in this group. Three patients with diffuse injury underwent repeat studies 5-14 weeks later; all improved on psychological tests, and two had a significant increase in global CBF. These results suggest that the chronic sequelae of head injury include decreased CBF, presumably reflecting decreased cerebral metabolism, which correlates with the neuropsychological impairment. Introduction In the United States, traumatic brain injuries occur at a rate of 200 new cases per 100,000 population per year; approximately 10% of all head-injured patients who survive the initial trauma have neuropsychiatric disability requiring prolonged hospitalization (Alexander 1982). Patterns of neuropsychological recovery are extremely variable and are not easily predicted from clinical characteristics at the time of injury (Dikmen et al. 1983). Although the computed tomographic (CT) scan gives useful anatomical information about the injury, a physiological correlate of brain function such as cerebral blood flow (CBF) would be useful as a prognostic indicator of neuropsychological recovery. Global CBF is generally
From the Burke Rehabilitation Center, Comell University Medical Center, White Plains, New York 10605. Address reprint requests to: Dr. Laurie Barclay, The Burke Rehabilitation Center, 785 Mamaroneck Avenue, White Plains, New York 10605. Received May 4, 1984; revised August 20, 1984.
© 1985 Society of Biological Psychiatry
0006-3223/85/$03.30
Chronic Head Injury Syndrome
BIOLPSYCHIATRY 1985;20:146-157
147
reduced immediately after acute head injury and tends to increase toward normal shortly thereafter (Fieschi et al. 1974; Overgaard and Tweed 1974). Bruce et al. (1973) reported that only two of eight patients had greatly reduced CBF adjacent to cerebral contusions, whereas Obrist et al. (1979a) reported generally good correlations between evidence of brain damage on CT and CBF. As level of consciousness improved in head injury patients, global CBF generally increased as well (Obrist et al. 1979b). Since neuropsychological deficits in patients with chronic head injury persist long after the initial head trauma, and since CBF reflects cerebral metabolism, we postulated that CBF might be depressed even in chronic head injury syndrome. We now report on global and regional patterns of CBF in 12 patients 2-13 months after head trauma, their correlation with anatomical and neuropsychological deficits, and their normalization with time, paralleling the pattern of neuropsychological recovery.
Methods
Patients Patients were admitted to the Head Trauma Unit at the Burke Rehabilitation Center for rehabilitation of cognitive, motor, and sensory deficits following closed head injury sustained in motor vehicle accidents. All patients were comatose after head trauma for periods ranging from l to 10 weeks, but all were alert and responsive and had regained consciousness 1-12 months preceding CBF study. Three patients had skull fractures, and in three others a subdural hematoma had been removed (Table 1). On the basis of CT scan at injury and neurological examination at the time of CBF study, patients were divided into those with predominantly unilateral dysfunction (Table 2) (n = 4) and those with diffuse bilateral dysfunction (Table 3) (n = 8). Tables 2 and 3 list the interval between head trauma and CBF and neuropsychological testing (range, 2-13 months), as well as medication and neurological findings at the time of study.
Controls Normal controls were cognitively intact, gainfully employed individuals who were free of overt neuropsychiatric disease and who had no significant medical illnesses by history.
CBF Study CBF was estimated by noninvasive la3xe inhalation and measurement of its washout over 10 min in a 32-detector apparatus (Medimatic Corporation, Copenhagen, Denmark). Xenon is an inert gas that diffuses freely and rapidly through all brain tissues. Peak radioactivity counts are achieved at each detector after xenon inhalation through a face mask; counts decline in biexponential fashion during the washout phase, during which the subject breathes room air. The rate of decline of radioactive counts is roughly proportional to CBF, and indices of CBF can be derived from peak values and decay constants of the washout curve. The location of detectors over each hemisphere is given in Figure 1. In this study CBF values measured included F1, or gray matter flow as derived from the rapid phase of the biexponential curve (Ohrist et al. 1975), and the initial slope index (ISI), or slope of the washout curve 2-3 rain after peak values are achieved (Risberg et al. 1975). Informed consent for CBF study was obtained from patients and their parents
148
BIOL PSYCHIATRY 1985;20:146-157
L. Barclay et al.
or directly from normal controls under a protocol approved by the Burke committee on research involving human subjects. Four patients had repeat CBF study at intervals ranging from 6 to 14 weeks.
Psychometric Testing All patients except patient 11, who was combative and resistive to all test procedures, were tested at the time of each CBF study. Tests included standardized tests of list memory (Tweedy et al. 1982); the block design subtest of the Wechsler Adult Intelligence Scale (Wechsler 1955); phonemic and semantic fluency tasks, in which patients were asked to name as many words as possible beginning with the letter G or as many animals as possible in 45 sec (Goodglass and Kaplan 1976); picture recognition (Tweedy et al. 1982); token test for receptive dysphasia (DeRenzi and Vignolo 1962); progressive matrices (Raven 1963); cancellation tests of attention span (Tweedy et al. 1982); and motility index (Stem et al. 1969).
Data Analysis In patients with well-localized dysfunction, mean F1 over the affected region was compared with mean F1 over the same detectors in the unaffected hemisphere and in an agematched control subject. Statistical significance was tested by the paired t-test. In all patients, asymmetry ratios were calculated by determining mean ISI over each hemisphere and by then dividing the larger hemispheric value by the smaller. Mean asymmetry ratios were compared using the unpaired t-test. One patient with well-localized dysfunction had repeat CBF testing 7 weeks later. The ratio of mean F1 over the affected detectors to mean F1 over the same detectors in the opposite hemisphere was calculated at both time points, and an improvement ratio was defined as the initial ratio divided by the repeat ratio. Three patients with diffuse bilateral dysfunction also underwent repeat CBF studies. Mean F2 over all 32 detectors was compared for each patient at initial and at repeat study using the paired t-test. Improvement ratios were calculated by dividing mean F2 at initial study by mean F2 at repeat study. Figure 1. Detector placement. Xenon-133 washout was measured in a 32-detector apparatus, with 16 detectors applied to each hemisphere.
149
8~
.8 r~
r~ Z
ZZ
z
z
.8
z
zz
zi~
z~
°
~z
Z
~
~.~
~
- °
.~ .8
~'~ ~
o
r=
~
"~'~
d
[-r~
o.8
z~
o
z
~z
~
~
zz
e~
2:2 o
~J
u
.-&
150
L. Barclay et al.
BIOL PSYCHIATRY 1985;20:146-157
Results Patients ranged in age from 18 to 26 years; 10 were male and 2 female. All patients were fight-handed except for patient 2. Patients 3, 5, and 6 all had basilar skull fractures; patient 3 had an associated CSF leak, patient 5 had associated fractures of the left zygoma and sphenoid, and patient 8 had associated left frontal and occipital fractures. On CT scan, four patients had subdural hematomas, four had contusions, two had intraventricular hemorrhages, and four had ventricular dilatation. The five patients tested with an electroencephalogram (EEG) showed diffuse slowing. Three patients had their subdural hematomas removed; the other nine patients did not undergo surgery. Duration of coma after trauma ranged from 1 to 10 weeks (Table 1). On the basis of CT findings at injury and neurological and psychometric findings at the time of CBF study, patients were divided into two groups: those with predominantly unilateral dysfunction (patients 1-4) (Table 2) and those with diffuse bilateral cerebral dysfunction (patients 5-12) (Table 3). Patients in the two groups were comparable in terms of mean age (_+SEM) (21.3 +_ 1.7 vs. 20.6 -+ 1.5); male : female ratio (3 : 1 vs. 7 : 1), duration of coma, and percentage with skull fracture or surgery (Table 1). Patients in the two groups were also comparable in terms of interval between trauma and initial CBF study (mean _+SEM; 5.5 _+ 2.6 vs. 4.3 _+ 0.8 months) and medications at the time of CBF study (Tables 2 and 3). Patients with predominantly unilateral dysfunction had few diffuse signs and had focal signs involving only one hemisphere (Tables 2 and 3). All patients were found to have deficits on neuropsychological testing (Table 4). Normal controls were reasonably well matched to patients with head trauma in age and sex. Mean ISI values were comparable in each hemisphere, so that asymmetry ratios in normals approximated 1.0 (Table 5). Mean ISI over 32 detectors was significantly higher in normal controls than in patients with predominantly unilateral dysfunction (mean _+SEM; 50.75 _+ 1.67 vs. 40.81 _+ 4.03, p < 0.01 as determined by unpaired t-test). Patients with diffuse bilateral dysfunction also had lower global ISI than did normal controls (45.25 -+ 2.96 vs. 50.75 _+ 1.67, p < 0.05). Table 2. Clinical Characteristics of Patients with Predominantly Unilateral Dysfunction" Patient number
Interval (months)
l
13
2
5
Medications b
Diffuse signs"
DPH
None
DPH, lnderal,
None
Benadryl 3
2
DPH
4
2 4
None
None
~, attention span; agitation and perseveration; frontal lobe signs
LH signs
RH signs
Exp and rec dysphasia Mild R HP and HS Rec > exp dysphasia Mild R HP Exp and rec dysphasia Mild apraxia and R HP R facial seizures None L
None None None
neglect, HP and HS
ODPH, diphenylhydantoin; exp, expressive; rec, receptive; HP, hemiparesis; HS. hemisensory deficit. ,l, decreased. bAt the time of cerebral blood flow study. CSignsof diffuse bilateral cortical dysfunction. dLeft hemispheric dysfunctions. eRight hemispheric dysfunctions.
151
r~
ve ..A
.~
~.~
.~
~,
e~
~g ,3
oo e~
;.g
---,
~
~
---,---,
<
---, o
z" o
% e~ o
u~ °~
M
it3
~ID
t' ~
~
~'~
~
,-~
f',,I
152
~ ~ ZZZZZZ--
-
~ ZZ~
~ ~ ZZZZZZZZ
@@@@@q@O@q@q@@@@@@@@
ZZZZZZZZZZZZZZZZZZZZ ¢)
~
~
_
~
zzzzzz-
~zz
z-
~ ~ ~ z - ~
~
~
-
~
--
zzz
-z
~
~
-
zzzzz
- ~ ---
~
_
~
F
zzzzzzzzzz~zzz
0 [...,
0
~o 0
z
~-~-~- I ~
I°~
zzzzz
I I I~°~-I
° I I
BIOLPSYCHIATRY
Chronic Head Injury Syndrome
153
1985;20:146-157
T a b l e 5. N o r m a l C o n t r o l s a,b
Number
Age
Sex
ISI L
ISI R
Asymmetry ratio c ISI larger/smaller
IN 2N 3N 4N 5N 6N 7N 8N 9N 10N 1IN 12N
30 19 24 26 22 21 22 30 30 27 27 22
M M M F M M F F M M M F
55.12 46.49 58.44 47.81 47.70 51.73 37.43 53.45 52.08 51.72 52.59 43.85
57.12 49.64 59.97 49.64 48.63 52.26 38.75 56.16 53.03 53.04 54.57 46.77
1.038 1.068 ! .026 1.038 1.019 1.010 1.035 1.051 ! .018 1.026 1.038 1.067
aMean ISI (initial slope index) over the left (L) and right (R) hemispheres was calculated as described by Risherg et al. (1975). #Mean: 1.036 .4- 0.005. CAsymmetry ratio: larger value of hemispheric ISI divided by the smaller value.
In three of four patients with predominantly unilateral dysfunction, mean F1 over the detectors overlying the affected region was significantly decreased relative to mean FI over the same detectors in the opposite hemisphere and relative to mean F1 over the same detectors in an age-matched normal control subject (Table 6). The mean asymmetry ratio in patients with predominantly unilateral dysfunction was significantly increased relative to patients with bilateral diffuse dysfunction, indicating greater hemispheric asymmetry of CBF in the former group (Table 7). Furthermore, mean asymmetry ratio in patients with predominantly unilateral dysfunction was significantly larger than in normal controls (p < 0.001 by unpaired t-test) (Tables 5 and 7).
Table 6. Regional Changes in Cerebral Blood Flow in Patients with Predominantly Unilateral Dysfunction ~ F1
Patient number 1 2 3 4
Affected region L L L R
Detectors
temporal 4,5,6 frontoparietal temporal 4,5,6,8,9,10,13,14,15 frontoparietal temporal 4,5,6,8,9,10,13,14,15 frontoparietal 4,5,8,9,10,13,14,15
Affected hemisphere (ml/100 g-min) 41.00 70.88 59.56 67.50
----- 2.1 ----- 3.8 ± 2.8 ± 3.1
Control (opposite) hemisphere (ml/100 g-rain) 55.33 76.75 65.22 78.25
± ± ± ±
1.8 b 3.0" 2.3 3.8 c
Control (ml/100 g-min) 66.27 82.36 53.96 103.83
± ± ± ±
4.2 d 3.6 c 1.1 4.5 b
"Mean FI, or gray matter flow, was calculated over those detectors thought to represent the most damaged region as determined by CT scan, neurological evaluation, and psychometric testing and was compared with mean FI over the same detectors in the opposite hemisphere, as well as with mean FI over the detectors in the corresponding normal control (see Table 5). Columns represent patient number, affected region as determined by CT and neuropsychoiogical evaluation, detector numbers overlying the affected region, F1 over those detectors in the affected hemisphere, FI over the same det~tors in the opposite hemisphere, and FI over the same detectors in the corresponding age-matched normal control. FI values represent mean -- SEM. Allp values are relative to FI (patient), affected hemisphere, by paired t-test.
~, < 0.001. Cp < 0.05. ~p < 0.0O5.
154
BIOLPSYCHIATRY
L. Barclay et al.
1985;20:146-157
Table 7. Asymmetry Ratiosa Patient number
Dysfunction
1 2 3 4
L L L R
ISI L
ISI R
Ratio ISI larger/smaller
Predominantly unilateral dysfunction 34.00 45.39 41.29 43.43
37.61 50.32 45.60 28.82
1.106 1.109 1.104 1.507
Bilateral diffuse dysfunction 5 6 7 8 9 10 11 12
Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral B il ateral
36.77 41.42 48.08 34.83 43.80 46.96 47.11 60.24
35.51 42.12 48.87 34.33 46.25 49.58 48.02 60.08
1.035 1.017 1.016 1.015 1.056 1,056 1.019 1.003
"Mean value of asymmetry ratios were averaged over the group of patients with predominantly unilateral dysfunction and the group of patients with bilateral diffuse dysfunction, The mean ratio for the former group (1.204 - 0.085) was significantly higher (p < 0.01 by the unpaired t-test) than that for the latter (1.027 ± 0,007), indicating more asymmetry of CBF in the former group.
All four patients for whom a repeat CBF study was done also underwent repeat neuropsychological testing at the same time, and all showed improvements in selected subtests. Mean overall improvement for each patient ranged from 60% to greater than 500% (Table 8). One patient (patient 4) with right frontoparietal dysfunction had a repeat CBF study 7 weeks after initial study. Although mean F1 over the affected detectors decreased on the repeated study, global FI was also decreased, and the ratio of FI over the affected detectors to F1 aver the same detectors in the opposite hemisphere increased on the repeat study (Table 9) suggesting normalization of FI in the affected region. Two of three repeat CBF studies on patients with diffuse bilateral dysfunction showed a significant increase in mean F1 averaged over 32 detectors (Table 9). Discussion Previous studies have indicated that global CBF is depressed immediately after head injury (Fieschi et al. 1974; Overgaard and Tweed 1974); the present study indicates that global CBF is depressed relative to age-matched controls for periods of as long as 13 months following head trauma. Global CBF is persistently decreased regardless of whether patients have well-localized dysfunction or diffuse, bilateral dysfunction. Since this study was done on patients with chronic head injury syndrome at intervals of 1-12 months after consciousness had been regained, the results cannot be extrapolated to patients with acute head injury; specifically we cannot comment on the prognostic significance of CBF in predicting return of consciousness. Three of four patients with well-localized dysfunction on CT scan and neuropsychological testing were shown to have decreased F1 over the affected region, whether compared with the same region in the opposite hemisphere or with the same region in an age-matched control subject. Our findings confirm and extend those of Obrist et al. (1979a), who reported a good correlation between CT scan and CBF in head trauma.
155
+++
+
++
I ++÷t+
I
I
+
+
÷1 +
- - ° ~
+
+~ + +
+
~ mZ
+1 +
+
+ +
+
__
+
+~ i
i +i + ~
÷1
+
÷~ ~~
+
+1
_
~
ZZ
e~
+ + + + + + ÷ 1 1
i +
i
÷1 + + +
+ ÷1 + + + ÷l + +
+ ÷l
~ " o
% ~
-
6
z
0
.~2
+ ÷l + + + +
÷1 + + ~
+ + ÷1 + + + ÷1 + +
+ ÷l
o 't~ 'N
'~'~"
I °
I °°
I I °°
I °
I I
.o.-~
2
o~ Z
=
~
~._~
~ [...
_o
_~o ~
~
~ "i~o
156
BIOL PSYCHIATRY 1985;20:146-157
L. Barclay et al.
Table 9. Repeat Cerebral Blood Flow Studies" Localized Dysfunction b Patient number Interval FI affected side (initial) FI control side (initial) F1 affected/control
4 7 weeks 65.8 - 7.4 82.5 ± 27.9 0.797
(initial) FI affected (repeat) FI control (repeat) F1 ratio (repeat) Ratio initial/repeat
Number
Interval
FI initial
5 6 7
7 wks 6 wks 14 wks
47.0 _ 5.1 57.7 - 9.3 64.0 - 6.8
53.8 -4- 5.9 51.0 ± 4.2 1.054 0.756
Generalized Dysfunction c F I repeat FI repeat-initial 57.6 ± 5.8 73.0 ± 8.5 55.3 -+ 7.7
10.0 .4- 5.7 20.1 ± 7.1 - 7 . 0 _ 8.0
FI initial/repeat
pd
.816 .790 1.159
<0.05 <0.01 N.S.
'rRepeat ~33Xewashout measurement was done 6-14 weeks after initial study in one patient with localized dysfunction (patient 4) and in three patients with generalized dysfunction (patients 5,6, and 7). Gray matter flow (F1) values are given in ml/100 g-min, and represent mean ( 4- SEM). the patient with localized dysfunction, FI over the affected region was averaged over detector numbers 4, 5, 8, 9, 10, 13, 14, and 15 in the fight hemisphere, which correspond to the site of his deficit (right frontoparietal). FI (control) was averaged over the corresponding detectors in the left hemisphere. qn the patients with generalized dysfunction, FI was averaged over all 32 detectors. 'tProbability values (p) of statistically significant differences between initial and repeat values were determined by the paired t-test; N.S. = not significant.
Furthermore, mean asymmetry ratio was increased in patients with well-localized dysfunction relative to patients with diffuse bilateral dysfunction or to age-matched normal controls, suggesting a homolateral decrease in hemispheric CBF in this group. One patient with right frontoparietal dysfunction showed decrease on repeat CBF measurement in F1 over the affected region but showed an increase in the ratio of mean F1 over the affected region to mean FI over the same region in the left hemisphere. This finding suggests normalization of regional CBF on repeat study despite decrease in global F1, which may suggest relative hyperemia at the time of initial CBF study. Three of four patients with diffuse generalized dysfunction showed improvement of global CBF on repeat studies. All these patients demonstrated improved performance on neuropsychological tests. Recovery in CBF seems to parallel recovery in neuropsychological function and is not limited to the acute increase in CBF that mirrors returning level of consciousness (Obrist et al. 1979b). Nevertheless, we do not advocate the use of CBF as a diagnostic tool in chronic head injury, since it merely reflects the regional or global decrease in cerebral activity that one would expect from the CT scan and neuropsychological deficits. Further studies are needed to determine whether CBF in chronic head injury may be of prognostic significance in predicting degree of ultimate neuropsychological recovery. Global and regional patterns of CBF seem to reflect neuropsychological deficits and their pattern of recovery in head-injured patients. The persistence of reduced CBF in patients with chronic neurological sequelae of head injury suggests the presence of a persistent physiological lesion accompanying the neuropsychiaa'ic deficit. Since CBF reflects cerebral metabolism, these findings also suggest a persistent decrease in metabolic activity. Whether this decrease in CBF is the primary disturbance underlying the neu-
Chronic Head Injury Syndrome
BIOLPSYCHIATRY 1985;20:146-157
157
ropsychological deficits is unclear, but it seems more likely to be the secondary effect of decreased cerebral activity due to underlying structural or metabolic changes following head trauma.
References Alexander MP (1982): Traumatic brain injury. In Benson DF, Blumer D (eds), Psychiatric Aspects of Neurologic Disease, Vol 2. New York: Grune & Stratton, pp 219-248. Bruce DA, Langfitt TW, Miller JD, Schutz H, Vapalahti MP, Stanek A, Goldberg H (1973): Regional cerebral blood flow, intracranial pressure, and brain metabolism in comatose patients. J Neurosurg 38:131. DeRenzi E, Vignolo LA (1962): The token test: A sensitive test to detect receptive disturbances in aphasics. Brain 85:665. Dikmen S, Reitan RM, Temkin NR (1983): Neuropsychological recovery in head injury. Arch Neurol 40:333. Fieschi C, Battistini N, Beduschi A, Boselli L, Rossanda M (1974): Regional cerebral blood flow and intraventdcular pressure in acute head injuries. J Neurol Neurosurg Psychiatry 37:1378. Goodglass H, Kaplan H (1976): The Assessment of Aphasia and Related Disorders, Philadelphia: Lea and Febiger. Obrist WD, Thompson HK, Wang HS, Wilkinson WE (1975): Regional cerebral blood flow estimated by 133Xenon inhalation. Stroke 6:245. Obrist WD, Dolinskas CA, Gennarelli TA, Zimmerman RA (1979a): Relation of cerebral blood flow to CT scan in acute head injury. In Popp AJ (ed), Neural Trauma. New York: Raven Press, pp 41-50. Obrist WD, Gennarelli TA, Segawa H, Dolinskas C, Langfitt TW (1979b): Relation of cerebral blood flow to neurological status and outcome in head-injury patients. J Neurosurg 51:292. Overgaard J, Tweed WA (1974): Cerebral circulation after head injury: Cerebral blood flow and its regulation after closed head injury, with emphasis on clinical correlations. J Neurosurg 41:531. Raven JC (1963): Guide to Using the Coloured Progressive Matrices, London: HK Lewis and Company, Ltd. Risberg J, Ali Z, Wilson EM, Wills EL, Halsey J (1975): Regional cerebral blood flow by ~33Xenon inhalation: Preliminary evaluation of an initial slope index in patients with unstable flow compartments. Stroke 6:142. Stern PH, McDowell F, Miller JM, Elkin RD (1969): Quantitative testing of motility defects in patients after stroke. Arch Phys Med Rehabil 50:320. Tweedy J, Reding M, Garcia C, Schulman P, Deutsch G, Antin S (1982): Significance of cortical disinhibition signs. Neurology (NY) 32:169. Wechsler D (1955): Manual for the WechslerAdultlntelligence Scale. New York: The Psychological Corporation.