- Email: [email protected]
Blink reflex in patients with hemispheric cerebrovascular accident (CVA)
Journal of the Neurological Sciences, 1985, 67:15-28
15
Elsevier JNS 02437
Blink Reflex in Patients with Hemispheric Cerebrovascular Accident (CVA) Blink Reflex in CVA J u n Kimura, J. Terry Wilkinson, H a n n a D a m a s i o , H a r o l d R. A d a m s Jr., Essatollah Shivapour and T h o r u Y a m a d a Division of Clinical Electrophysiology, Stroke Unit and Division of Behavioral Neurology, Department of Neurology, University of lowa Hospitals and Clinics, Iowa City, IA 52242 (U.S.A.)
(Received 19 June, 1984) (Revised, received 31 July, 1984) (Accepted 5 August, 1984)
SUMMARY A blink reflex consists of an early unilateral component, R~, and a late bilateral component, R 2. During an acute phase of hemispheric cerebrovascular accident, R 1 and R 2 were abnormal in 30 and 50 of 66 patients, respectively. Paired stimuli usually corrected R 1 but not Rz, which was profoundly suppressed. The discrepancy between polysynaptic R 2 and oligosynaptic R~ indicates a greater disfacilitation at the level of interneurons than at the motoneuron, which serves as the final common path. Abnormality of R z occurred bilaterally with stimulation on the affected side of face and contralaterally after stimulation on the normal side in 31 patients. This finding suggests a diffuse loss of internuncial excitability, contralateral to the hemispheric lesion. Changes of R 2 implicated the brainstem pathways forming the afferent and efferent arc of the reflex in 7 and 8 patients, respectively. The remaining 4 comatose patients had no R 2 irrespective of stimulus sites. Clinical localization of the hemispheric lesion showed no consistent correlation with the type of blink reflex abnormalities. The CT scans revealed widely scattered changes in 29 patients with abnormal blink reflex but with a tendency to overlap in the inferior Rolandic area. This contrasted with conspicuous sparing of the inferior postcentral region in 10 patients with normal blink reflex. These f'mdings suggest the presence of crossed facilitation to this reflex from wide areas of the cortex but most prominently from the sensory representation of the face.
Key words: B l i n k reflex - Brainstem - Disfacilitation - Hemispheric cerebrovascular accident - Internuncial excitability
0022-510X/85/$03.30 © 1985 Elsevier Science Publishers B.V. (BiomedicalDivision)
16 INTRODUCTION Absence or diminution of the corneal reflex usually indicates a lesion in the brainstem if the trigeminal and facial nerves are intact. However, early clinical studies suggested abnormalities of this reflex in patients with hemispheric lesions (Wolff 1913; Oliver 1952; Magladery and Teasdall 1961 ; Richwien 1966; De Jong 1967; Ross 1972). More recent studies have confirmed this view by means of quantitative electrophysiologic techniques. Most commonly employed to test reflex changes in patients with cerebrovascular accidents (CVA) have been electrical or mechanical stimuli to the supraorbital nerve alone or in conjunction with tactile stimuli to the cornea. (Messina and Quattrone 1973; Kimura 1974; Dehen et al. 1976; Fisher et al. 1979; Ongerboer de Visser 1981, 1983; Berardelli et al. 1983). The electrically or mechanically elicited blink reflex consists of two separate responses, R 1 and R 2 (Kugelberg 1952; Rushworth 1962; Kimura et al. 1969; Shahani 1970; Hiraoka and Shimamura 1977). With unilateral stimulation, R1 occurs only ipsilaterally whereas R 2 is elicited bilaterally (Fig. 1, Table 1). Of the two components, R~ is relatively stable in latency since it is mediated by a simple reflex arc between the trigeminal and facial nerves via a pontine relay. In contrast, R 2 is inherently more variable presumably reflecting the fluctuating excitability ofinterneurons during different stages of arousal (Shahani 1968; Lyon et al. 1972; Boelhouwer and Brunia 1977) and habituation (Kimura 1974; Desmedt and Godaux 1976). However, the analysis of R 2 helps localize the lesion (Kimura 1975). In an afferent pattern, R 2 is abnormal bilaterally when the affected side of face is stimulated, whereas in an efferent pattern, R 2 is altered on the paretic side with ipsilateral or contralateral stimulation (Kimura I974, 1975). Investigators agree that the blink reflexes alter substantially in patients with hemispheric lesions but each series has emphasized different patterns of abnormalities, possibly reflecting patient selection and the duration of illness at the time of the study. The purpose of this communication is to (1)confirm the cause-effect relationship between hemispheric lesions and blink reflex abnormalities, (2) describe the patterns of reflex change attributable to acute cerebral dysfunction, and (3) delineate cortical areas exerting an excitatory drive to the reflex pathway in the brainstem. In our earlier investigation (Kimura 1974), patients were seen from one to several weeks after the onset of illness. In contrast, the current series deals with abnormalities during the first few days after the ictus.
MATERIALS AND METHODS We studied 66 patients aged 21-87 years (mean age, 60) within 5 days of a unilateral hemispheric CVA. In 32 patients, the test was repeated weekly. The normal range of R~ and R2 was determined in 23 control subjects aged 40-77 years (mean age, 57) for electrical stimulation and in 21 others aged 50-68 (mean age, 58) for a glabellar tap. An informed consent was obtRined from the patients or their guardians following a full explanation of the procedure.
17 An independent investigator, blind to the electrophysiologic data, retrieved clinical information from chart review for clinical localization of lesion (Table 2). Another investigator analyzed CT scans without knowledge of either the clinical or the electrophysiologic status of the patients. Using standard sets of templates (Damasio and Damasio 1980; Damasio 1983), the lesions were transferred onto a lateral view of the brain to determine the area of major overlap. We elicited the blink reflex according to the standard technique (Kimura 1983). Subjects lay supine with the active electrodes (G~) placed on the upper lateral aspect of the orbicularis oculi muscle and a reference electrode (G2) on the lateral surface of the nose. A ground electrode was located around the ann. The supraorbital, infraorbital and mental nerves were stimulated at the respective foramina on one side. The use of a specially designed amplifier with a short blocking time and a low internal noise minimized the problem of stimulus artifact (Walker and Kimura 1978). Frequency response ranged from 20 Hz to 32 KHz. If R 1 was difficult to elicit with single shocks of appropriate intensities, we used paired stimuli with an interstimulus interval of 5 ms (Kimura 1974). A subthreshold conditioning shock subliminally facilitated the reflex pathway. A supramaximal test stimulus then triggered a sweep and evoked the response, allowing determination of its latency from the second shock. A set of recordings on each side consisted of 8 R~ responses, 4 with single and 4 with paired stimuli (see Fig. 2) as well as 4 R 2 responses at each site of stimulation, 2 with single and 2 with paired shocks (see Fig. 3). Stimuli were separated by a time interval of 30 s or longer to minimize interaction between successive trials (Kimura 1983). For mechanical stimulation, a specially constructed reflex hammer closed a built-in microswitch on impact and triggered a sweep. In contrast to unilateral electrical stimulation, a midline glabellar tap elicited R 1 and R 2 bilaterally (see Fig. 2) allowing instantaneous comparison between the two sides (Fisher et al. 1979). Each set consisted of 4 pairs of R 1 and R 2 recorded simultaneously on both sides.
RESULTS
Finding in normals Table 1 shows the normal range of electrically elicited R~ and R 2 in 23 healthy subjects and mechanically induced R 1 in another 21 normal subjects. Electrical stimulation of the supraorbital nerve always elicited both R~ and R E. Glabellar taps elicited R~ in all subjects but R 2 only inconsistently. Electrical stimulation of the infraorbital nerve evoked R 2 in all, but R~ only in 8 subjects. Stimulation of the mental nerve gave r i s e t o R 2 in 21 cases and RI in one instance. The latencies of R~ and R 2 w e r e similar with stimulation of either the supraorbital or infraorbital nerve. However, R 2 elicited by stimulation of the mental nerve was considerably longer in latency. Based on these observations, the quantitative data analyzed in the patient group included R 1 and R 2 elicited by stimulation of the supraorbital nerve, R 2 following stimulation of the infraorbital nerve and R~ induced by glabellar tap.
18
cs
o,I c5 i
~
'd
r-:
+t
+t
4-1
c5
c~
d
o
+1
+1
+1
+1
~
M
,14
+1
+1
+1
o
2
~
N
'
f
i
"'
I
O e'-,
¢ +1
+1
+1
+1
+I
+1
,~
o el 0 0 . ~0
b
+1
+1
+1
+1
£
2
N
N
,-I
©
Z O g:::l i,.1,.1 m r~ [<
4-1
+1
-H
-H
4-1
-H
+l
4'-I
¢5
,~
o
,~
....g,
@ <
ed Z < <
"6 .0 e,,)
Z
[,,.
~
.,-..
o0
"
rO
'd
-.-ag ~v
F [.. Z
N <
Z
z ,.~.N M
0
19 The corresponding R~ and R2 elicited by stimulation on the right and left side showed no significant differences in latency, amplitude or duration. However, the ipsilateral R 2 was occasionally shorter in latency, larger in amplitude and longer in duration than the simultaneously recorded contralateral R2. Using the mean plus three standard deviations in the control groups as the upper limit of normal ,electrically elicited R~ is abnormal if the latency difference between the two sides exceeds 1.3 ms. Similarly, the latency difference of R 1 evoked by a glabellar tap should be less than 1.6 ms between the two sides. Stimulating either the supraorbital or infraorbital nerve unilaterally, the simultaneously recorded ipsilateral and the contralateral R 2 should not vary more than 5 ms in latency. A latency difference between R 2 evoked by right-sided stimulation and corresponding R 2 evoked by left-sided stimulation should not exceed 7 ms. In preliminary experiments we attempted to electrically rectify and integrate the response in assessing the size of R 1 and R 2. This approach, however, proved to be less accurate than might be expected on theoretical ground because a slight baseline shift induced by stimulus artifact and eye movements precluded a selective measurement of the area under the waveform. Therefore, the size of response was estimated simply as the product of the amplitude and duration (Fig. 1). Its asymmetry was considered abnormal with reduction of R 1 or R 2 to less than one third of the comparable response on the opposite side. Lt.
Rt.
& ~ L ~
~
t-~ Amplifier ~ - ~ t O s c l l l ° u c o p e ~
1. Rt
Amplitude -
C'ere
~--J
2. Ipsilateral R 2 Average Amplitude
-- . . . . . . .
Rt. - - T . ~ : ~ i
I..,-°,
:~ Lt.--" ,
t stimulus llrlifict
-
Latency
,
,,,
~ I .......... ' ~
ou,.,,o. -
,: -
i ....
I
Av.r.ge
Amplitude
3. Contralateral R 2
Fig. 1. Stimulation and recording arrangement for the blink reflex. The presumed pathway of R~ through the pons (1) and direct and consensual R 2 through the pons and lateral medulla (2 and 3) are schematically illustrated. Dotted lines indicate the presence of multiple interneurons. The primary afferents of R] and R 2 are shown as one fiber. This and other details of central connections of these reflexes are not known. Bottom: a typical oscilloscope recording of the blink reflex after right-sided stimulation. Note an ipsilateral R 1 response and bilateral simultaneous R 2 responses (Reprinted by permission of John Wiley & Sons, Inc. from J. Kimura (1982) Conduction abnormalities of the facial and trigeminal nerves in polyneuropathy, Muscle & Nerve, 5: S139-S144).
20 Findings in patients Electrically elicited R~ was normal in 36 patients but was absent with single or paired stimuli in 5. In the remaining 25 subjects, R~ was absent or reduced in size in 21 and delayed in 4 others contralateral to the lesion after a single stimulus, but in all, paired shocks corrected the abnormalities (Figs. 2-5). When the affected side of the face was compared to the normal side, R~ was significantly smaller in amplitude (P < 0.01) and shorter in duration (P < 0.01) but unchanged in latency (P > 0.05). The glabellar tap elicited a normal R~ in 42 of 58 patients tested. In the remaining 16 patients, the response on the affected side of the face was absent in 8 and delayed in the other 8 (Fig. 2). Statistical analyses showed no significant (P > 0.05) difference between the two sides in latency, amplitude or duration of the mechanically elicited R1.
A.H. 12-4-80 1{ --'X--~J ~ - ' ' ~
. . . . . . Rt {
-
,_~Lt {
~
glabellar 2{ ~ Tap
3 { ~ ~
~ - ' - ~
.--,,---~ ~ L
V.W. 1-30-81 - Rt t
{ _ -~/~/~1~~
oL.smv 5ms 1 -~l\~"~'/~J~
"~_
Rt
'~t!\~X.~
~
Rt
Rt(SO)
L-
-
-
J
I _I~,\~--~ Lt(SO)
4- ~
L _ _ _ -J O~.5mV 2ms Lt
I L~
Lt
~
L__-I
L - _I
oL.smV 2m8
Fig. 2. Mechanically and electrically elicited RI in a 60-year-old man with hematoma in the rightfTontat lobe (left side &the figure) and in a 59-year-oldman with a large infarct in the left hem/sphere ( ~ side of the figure). From top to bottom, bilateral Rl elicited by gtahellar tap (small arrows), electrically~ t e d R~ (dotted brackets) with single (trials 1-4) and paired shocks (trials 5-8) on the right, ~ ~ a b t e R~ with stimulation on the left. In both patients, R ! elicited b y ~ tap was increased:in latency on the paretic side. Electrically elicited R i was absent with $in~e $ ~ s but normal with p ~ d ~ k s on the affected side. SO = supraorbital nerve.
21 O v e r a l l , R 2 was abnormal in 50 of 66 patients whether tested with single or paired stimuli of the supraorbital or infraorbital nerve. In 4 comatose patients, R 2 was totally absent bilaterally regardless of the side of stimulation (coma type). In 31 patients (Fig. 3), R 2 was abnormal bilaterally when the affected side of the face was stimulated, but with stimulation of the normal side it was only abnormal contralaterally (mixed type). Afferent (Fig. 4) and efferent (Fig, 5) types of changes described earlier occurred in 7 and 8 patients, respectively. In general, R 2 in patients was significantly (P < 0.01) increased in latency, reduced in amplitude and decreased in duration. However, the ipsilateral R 2 after stimulation on the normal side of the face remained unchanged in
Stimulation on Left
R1(SO)
Stimulation on R i g h t
R.T. 4-3-80
4
L. _ _ J
L._ _.J
10.5mV 2ms
(!
2{ R 2 (SO)
3{ 4{
....... Rt Lt
Lv~
k f I
I
oL•,5mV 10ms Rt
....
2{
_. _ ~
R2 (IO) 3{
{7~ - -
Lt ,
_~ - - - ~ -
¢
t
I
0L~_.5mV 10ms
Fig. 3. Mixed type of abnormality in a 55-year-old man, 3 days after an infarction in the left posterior frontal region. From top to bottom, R) (dotted brackets) by single (trials 1-4) and paired shocks (trials 5-8) of the supraorbital nerve (SO), R z (solid brackets) by single (trials 1 and 2) and paired shocks (trials 3 and 4) of the supraorbital nerve (SO) and R 2 by single (trials 1 and 2) and paired shocks (trials 3 and 4) of the infraorbital nerve (IO)). When the right side was stimulated, R 1 was absent with single shocks and reduced in amplitude with paired shocks, whereas R 2 was absent bilaterally with stimulation on the right and contralaterally with stimulation on the left.
22 Stimulation on Right
Stimulation on Left 1 : ~ . . . . . ~
RI(SO)
B.S. 1,9-80
Lt
41 -
_
-
-
-
8 I.-_ - /
L - - -J
1{
.
.
.
.
.
.
.
m
a{
.,=
_
{ I
R
I
i
-
i
4{
R 2 (IO)
Lt
=
H
1{ ~
Rt ,
i
R 2 (SO)
0L0.~.5mV 2ms
_
i
m
i
,
0.5mV 10ms
t
a{ 4{ I
I
oLsmv 10ms
Fig. 4. Afferent type of abnormality in a 73-year-old woman 1 day after an infarction in the distribution of the leR middle cerebral artery. The same arrangement as in Fig. 3. When the right side was stimulated, Ri was absent with single shocks and reduced in amplitude with paired shocks, whereas R 2 was absent bilaterally after stimulation of either the SO or IO on the right,
amplitude when compared to the corresponding Rz in the controls (P > 0.05). Within the patient group, the ipsilateral R2 after stimulation on the normal side of the face was significantly (P < 0.01) greater in amplitude and shorter in latency compared to the simultaneously recorded contralateral R 2 or to ipsilateral or contralateral R2 elicited by separate stimulation on the affected side (Table 1). Repeat studies performed in 32 patients showed gradual improvement of R~ and R 2. Excitability of R~ usually returned to normal by the second week. The rate of recovery of R 2 was variable and s o m e ~ of abnormality persisted even when the clinical condition had been stable for several weeks. ~ R 2 w a s totally a b s ~ in a comatose patient (Lyon et al. 1972), the return of the reflex response occurred in a predictable order, i.e. ipsilateral before contralateral R 2 with stimulation on thenormal
23
12 ~ R1 (SO)
65;
Stimulation on Left
Stimulation on Right
Lt ~
\ ~
M.S. a-lg-s0 Rt
' ~ L _ _ J
L _ _ -- J
1 { ~...__~.~.f' ..,
Rt { - - ~ ~ ~
~0.5mV 2ms
Lt
R2 (SO) I
_ l
_ I
I
1{ ^ ~ ~ L t
{
I
~
L
oL~..5mv lOms
t
2{ " L - ~ ~ ~ ' ~ ' ' ~ - - ~ Rt { ~ , . , ~ .
I
I
I
I
10.SmV
10ms Fig. 5. Efferent type of abnormality in a 32-year-old woman with moyamoya syndrome 4 days after an infarction in the distribution of the right middle cerebral artery. The same arrangement as in Fig. 3. Stimulating the left side of face, R 1 was absent with single shocks and reduced in amplitude with paired shocks. With stimulation on either right or left side, R 2 recorded on the left was consistently reduced in size.
side of the face followed by ipsilateral before contralateral affected side.
R2
with stimulation on the
Correlation with clinical data
Based on clinical findings the responsible lesion predominantly, though not necessarily selectively, affected frontal, temporal, parietal and occipital lobes in 12, 7, 5 and 2 patients, respectively. In 11 others, the lesion was primarily subcortical (Table 2). The remaining 29 patients had multiple sites of involvement or lesions that we could not localize. Analysis of blink reflex findings in each category revealed no consistent relationship between the type of electrophysiologic abnormalities and the major site of hemispheric involvement (Table 2).
24
Z o
oo
<
L~
©
o (-. <
Z
Z
< o <
z M
Z <
0
0
]
M <
N © z < N M
z
o
M
© ~ . ~
~
< ~
0
~
25 Correlation with CT scan abnormalities
In 27 patients, CT scans, analyzed without knowledge of clinical fmdings, showed only equivocal changes or no abnormalities at all. The remaining 39 patients had CT scans which allowed accurate localization of pathology often involving more than one lobe as described below. In this group, the blink reflex was abnormal in 29 and normal in 10 cases. In the 29 patients, the CT scans showed single or multiple lesions involving the frontal, parietal, temporal and occipital lobes in 16, 9, 6 and 6 patients, respectively. In 7 others the abnormality was exclusively subcortical. The premotor regions were involved in 15 patients but in 7 of these the lesion extended into the parietal structures. The basal ganglia were affected in 5 and the deep white matter close to the corpus callosum in another 2. A composite picture obtained by superimposition of the cortical abnormalities in the CT scans suggested considerable scatter, but with a tendency to overlap in the inferior Rolandic area at the level of face representation (Fig. 6). In the remaining 10 patients without abnormalities of the blink reflex, lesions involved the frontal, parietal, temporal and occipital lobes and basal ganglia in 5, 3, 2, 1 and 1 patient, respectively. A comparable scatter of lesions was noted in similar proportions across all lobes. The post-central Rolandic area was conspicuously spared but the normal group was much smaller and an entirely valid comparison was not possible. DISCUS SION
Recent reports have described several patterns of abnormality of the blink reflex in hemispheric CVA (Messina and Quattrone 1973; Kimura 1974; Dehen et al. 1976; Fisher et al. 1979; Berardelli et al. 1983). In our previous study (Kimura 1974), the most common reflex change in patients with hemifacial hypesthesia was that of an afferent pattern with diminution of R2 bilaterally after stimulation of the affected side of the face. Berardelli et al. (1983) reached similar conclusions. These findings are in contrast to the results reported by Dehen and associates (1976), who emphasized an efferent pattern with diminution of R 2 on the paretic side of the face and enhanced response on the opposite side in patients with hemiplegia. When studied after the patients' condition had stabilized, R 1was normal in latency although reduced (Kimura 1974; Dehen et al. 1976) or enhanced (Messina and Quattrone 1973; BerardeUi et al. 1983) in amplitude on the affected side. Fisher and associates (1979) reported a significant increase in latency of mechanically elicited R 1 during acute stages of CVA. The present study in a large number of patients confirmed that an abnormality of the blink reflex is the rule rather than the exception in patients with acute hemispheric vascular events. In agreement with an earlier observation (Fisher et al. 1979), glabellar tap elicited a delayed R~ contralateral to the hemispheric lesion in a few patients. Similarly, R 1 evoked by single electric shock was occasionally increased in latency. However, in these cases, the amplitude of R1 was substantially reduced, indicating partial activation of the reflex pathway. Indeed, when a maximal response was recorded by paired stimuli R~ was normal in latency. Alteration of R~ does not necessarily imply
26
Fig. 6. A composite picture obtained by superimposition of the CT scans. A: in 29 patients with abnormal blink reflexes, the lesions were scattered considerably although they tended to overlap in the inferior Rolandic area at the level of face representation, both in the pre- and post-central region. B: in 10 patients without abnormalities in the blink reflex, the lesions conspicuously spared the inferior post-central area but the normal group was too small to draw a definite conclusion.
a pathologic process of the reflex arc itself since indirect effects of lesions outside the primary reflex pathway may also compromise its excitab~ty (Kimura 1975; Fisher oral. 1979). When fully activated, however, the latency of Rt usually indicates the conduction characteristics of the reflex pathway (Kimura et al. 1969). The polysynaptic R2 was more profoundly suppressed than the ol~synaptic R~ in acute hemispheric lesions. This finding suggests that interruption of de=~nding pathways prhnarily results in disfacilitation of brainstem intorneurons and, to a lesser degree, motor neurons. Based on increased amplitude of R l, Dehen and others (1976)
27 reported hyperexcitability of the motor neurons ipsilateral to the hemispheric lesions, a finding not confirmed in this study. This discrepancy may be in part related to the time of study following the CVA because excitability changes are influenced by the interval from the onset of the ictus. In the present study, R z was commonly absent or diminished bilaterally when the affected side of the face was stimulated. This observation alone is consistent with reduced excitability in neurons in the area of the spinal nucleus of the trigeminal nerve or interneurons forming the afferent arc (Ongerboer de Visser and Kuypers 1978). However, an additional abnormality of the contralateral R2 after stimulation of the normal side of the face suggests reduced excitability of the efferent arc, i.e. either the crossed interneuronal pathway relaying the impulse from the opposite spinal nucleus or the facial nucleus itself on the paretic side. Thus, the basic R z abnormality probably results from diffuse disfacilitation involving both the afferent and efferent arcs of the reflex contralateral to the hemispheric lesion. Analyses of the composite CT scan templates in our series suggest that blink reflex abnormalities most consistently occur with lesions involving the inferior post-central area (Fig. 6). However, careful review of individual scans indicates that the responsible lesions are widely scattered and sometimes confined to the lower premotor and motor regions or to other areas entirely outside the Rolandic region. Of interest are CT scan abnormalities in patients with normal blink reflex, showing conspicuous sparing of the sensory representation of the face. Although this particular group was too small to draw a definite conclusion, the finding is consistent with the recent result by Ongerboer de Visser and associates (1981) who, studying the corneal reflex, observed that the lower post-central region had an excitatory influence upon the lateral reticular formation. Afferent or efferent patterns may result from the interruption of specific descending pathways. However, no consistent relationship has emerged in correlating the type of blink reflex changes to the location of an acute hemispheric lesion. We conclude that (1) the blink reflex abnormality is the rule rather than the exception in an acute hemispheric CVA, (2) the polysynaptic R 2 is more affected than the oligosynaptic RI, suggesting a greater suppression of brainstem interneurons than motor neurons, (3) contralateral disfacilitation usually occurs diffusely, involving both afferent and efferent reflex pathways, and (4)the excitatory drive to the brainstem pathway of the blink reflex probably originates in wide areas of the cortex but most conspicuously in the inferior post-central area, coinciding with the sensory representation of the face. ACKNOWLEDGEMENTS
The authors wish to thank Sheila R. Mennen, Deborah A. Gevock and Lesa A. Bowles for technical assistance.
28 REFERENCES Berardelli, A., N. Accornero, G. Cruccu, F. Fabiano, V. Guerrisi and M. Manfredi (1983) The orbicularis oculi response at~er hemispheral damage, J. Neurol. Neurosurg. Psychiat., 46: 837-843. Boelhouwer, A. J. W. and C. H. M. Brunia (1977) Blink reflexes and the state of arousal, J. Neurol. Neurosurg. Pyschiat., 40: 58-63. Damasio, H. (1983) A C T guide to the identification of cerebrovascular territories, Arch. NeuroL, 40: 138-142. Damasio, H. and A.R. Damasio (1980) The anatomical basis of conduction aphasia, Brain, 103: 337-350. Dehen, H., J.C. Willer, N. Bathien and J. Cambier (1976) Blink reflex in hemiplegia, Electroenceph. Clin. Neurophysiol., 40: 393-400. De Jong, R.N. (1967) The Neurological Examination, 3rd edition, Harper and Row, New York, p. 226. Desmedt, J. E. and E. Godaux (1976) Habituation ofexteroceptive suppression and of exteroceptive reflexes in man as influenced by voluntary contraction, Brain Res., 106: 21-29. Fisher, M.A., B.T. Shahani and R.R. Young (1979) Assessing segmental excitability after acute rostral lesions, Part 2 (The blink reflex), Neurology (N.Y.), 29: 45-50. Hiraoka, M. and Shimamura, M. (1977) Neural mechanisms of the corneal blinking reflex in cats, Brain Res., 125: 265-275. Kimura, J. (1974) Effect of hemispheral lesions on the contralateral blink reflex, Neurology (Minneap), 24: 168-174. Kimura, J. (1975) Electrically elicited blink reflex in diagnosis of multiple sclerosis-review of 260 patients over a seven-year period, Brain, 98: 413-426. Kimura, J. (1982) Conduction abnormalities of the facial and trigeminal nerves in polyneuropathy+ Muscle & Nerve, 5: S139-S144. Kimura, J. (1983) Electrodiagnosis in Diseases of Nerve and Muscle - - Principles and Practice, F.A. Davis, Philadelphia. Kimura, J., J.M. Powers and M.W. Van Allen (1969) Reflex response of orbicularis oculi muscle to supraorbital nerve stimulation m Study in normal subjects and in peripheral facial paresis, Arch. Neurol., (Chic.), 21: 193-199. Kugelberg, E. (1952) Facial reflexes, Brain, 75: 385-396. Lyon, L.W., J. Kimura and W.F. McCormick (1972) Orbicularis oculi reflex in coma - - Clinical, electrophysiological, and pathological correlations, J. Neurol. Neurosurg. Psychiat., 35: 582-588. Magladery, J.W. and R. D. Teasdall (1961) Corneal reflexes - - An electromyographic study in man, Arch. Neurol. (Chic), 5: 269-274. Messina, C. and A. Quattrone (1973) Comportamento dei riflessi trigemino-facciati in soggetti con tesioni emisferiche, Pa'v. Neurol., 43: 379-386. Oliver, L.C. (1952) The supranuclear arc of the corneal reflex, Acta Psychiat. Scand., 27: 329-333. Ongerboer de Visser, B.W. (1981 ) Corneal reflex latency in lesions of the lower posteentral region, Neurology (New York), 31: 701-707. Ongerboer de Visser, B.W. (1983) Comparative study of corneal and blink reflex latencies in patients with segmental or with cerebral lesions. In: J.E. Desmedt (Ed.), Motor Control Mechanisms in Health and Disease, Raven Press, New York, pp. 757-772. Ongerboer de Visser, B.W. and H.G.J.M. Kuypers (1978) Late blink reflex changes in lateral medullary lesions - - An electrophysiological and neuroanatomieal study of Wallenberg's syndrome, Brain, 101 : 285-294. Richwien, R. (1966) Zur Diagnose der zentralen St/Srungen des Cornealreflexes in afferenten Schenkel, Dtsch. Z. Nervenheilk., 189: 191-195. Ross, R.T. (1972) Corneal reflex in hemisphere disease, J. Neurol. Neurosurg. Pyschiat., 35: 877-880. Rushworth, G. (t962) Observations on blink reflexes, J. Neurol. Neurosurg. Psychiat., 25: 93-108. Shahani, B.T. (1968) Effects of sleep on human reflexes with a double component, J. Neurol. Neurosurg. Psychiat., 31: 574-579. Shahani, B.T. (1970) The human blink reflex, J. Neurol. Neurosurg. Psychiat., 33: 792-800. Walker, D. D. and J. Kimura (1978) A fast-recovery electrode amplifier for electrophysiology, Electroenceph. Clin. Neurophysiol., 45: 789-792. Wolff, R. (1913) Die Areflexie der Cornea bei organischen Nervenkrankheiten, Arch. Psychiat., 52:716-747.
15
Elsevier JNS 02437
Blink Reflex in Patients with Hemispheric Cerebrovascular Accident (CVA) Blink Reflex in CVA J u n Kimura, J. Terry Wilkinson, H a n n a D a m a s i o , H a r o l d R. A d a m s Jr., Essatollah Shivapour and T h o r u Y a m a d a Division of Clinical Electrophysiology, Stroke Unit and Division of Behavioral Neurology, Department of Neurology, University of lowa Hospitals and Clinics, Iowa City, IA 52242 (U.S.A.)
(Received 19 June, 1984) (Revised, received 31 July, 1984) (Accepted 5 August, 1984)
SUMMARY A blink reflex consists of an early unilateral component, R~, and a late bilateral component, R 2. During an acute phase of hemispheric cerebrovascular accident, R 1 and R 2 were abnormal in 30 and 50 of 66 patients, respectively. Paired stimuli usually corrected R 1 but not Rz, which was profoundly suppressed. The discrepancy between polysynaptic R 2 and oligosynaptic R~ indicates a greater disfacilitation at the level of interneurons than at the motoneuron, which serves as the final common path. Abnormality of R z occurred bilaterally with stimulation on the affected side of face and contralaterally after stimulation on the normal side in 31 patients. This finding suggests a diffuse loss of internuncial excitability, contralateral to the hemispheric lesion. Changes of R 2 implicated the brainstem pathways forming the afferent and efferent arc of the reflex in 7 and 8 patients, respectively. The remaining 4 comatose patients had no R 2 irrespective of stimulus sites. Clinical localization of the hemispheric lesion showed no consistent correlation with the type of blink reflex abnormalities. The CT scans revealed widely scattered changes in 29 patients with abnormal blink reflex but with a tendency to overlap in the inferior Rolandic area. This contrasted with conspicuous sparing of the inferior postcentral region in 10 patients with normal blink reflex. These f'mdings suggest the presence of crossed facilitation to this reflex from wide areas of the cortex but most prominently from the sensory representation of the face.
Key words: B l i n k reflex - Brainstem - Disfacilitation - Hemispheric cerebrovascular accident - Internuncial excitability
0022-510X/85/$03.30 © 1985 Elsevier Science Publishers B.V. (BiomedicalDivision)
16 INTRODUCTION Absence or diminution of the corneal reflex usually indicates a lesion in the brainstem if the trigeminal and facial nerves are intact. However, early clinical studies suggested abnormalities of this reflex in patients with hemispheric lesions (Wolff 1913; Oliver 1952; Magladery and Teasdall 1961 ; Richwien 1966; De Jong 1967; Ross 1972). More recent studies have confirmed this view by means of quantitative electrophysiologic techniques. Most commonly employed to test reflex changes in patients with cerebrovascular accidents (CVA) have been electrical or mechanical stimuli to the supraorbital nerve alone or in conjunction with tactile stimuli to the cornea. (Messina and Quattrone 1973; Kimura 1974; Dehen et al. 1976; Fisher et al. 1979; Ongerboer de Visser 1981, 1983; Berardelli et al. 1983). The electrically or mechanically elicited blink reflex consists of two separate responses, R 1 and R 2 (Kugelberg 1952; Rushworth 1962; Kimura et al. 1969; Shahani 1970; Hiraoka and Shimamura 1977). With unilateral stimulation, R1 occurs only ipsilaterally whereas R 2 is elicited bilaterally (Fig. 1, Table 1). Of the two components, R~ is relatively stable in latency since it is mediated by a simple reflex arc between the trigeminal and facial nerves via a pontine relay. In contrast, R 2 is inherently more variable presumably reflecting the fluctuating excitability ofinterneurons during different stages of arousal (Shahani 1968; Lyon et al. 1972; Boelhouwer and Brunia 1977) and habituation (Kimura 1974; Desmedt and Godaux 1976). However, the analysis of R 2 helps localize the lesion (Kimura 1975). In an afferent pattern, R 2 is abnormal bilaterally when the affected side of face is stimulated, whereas in an efferent pattern, R 2 is altered on the paretic side with ipsilateral or contralateral stimulation (Kimura I974, 1975). Investigators agree that the blink reflexes alter substantially in patients with hemispheric lesions but each series has emphasized different patterns of abnormalities, possibly reflecting patient selection and the duration of illness at the time of the study. The purpose of this communication is to (1)confirm the cause-effect relationship between hemispheric lesions and blink reflex abnormalities, (2) describe the patterns of reflex change attributable to acute cerebral dysfunction, and (3) delineate cortical areas exerting an excitatory drive to the reflex pathway in the brainstem. In our earlier investigation (Kimura 1974), patients were seen from one to several weeks after the onset of illness. In contrast, the current series deals with abnormalities during the first few days after the ictus.
MATERIALS AND METHODS We studied 66 patients aged 21-87 years (mean age, 60) within 5 days of a unilateral hemispheric CVA. In 32 patients, the test was repeated weekly. The normal range of R~ and R2 was determined in 23 control subjects aged 40-77 years (mean age, 57) for electrical stimulation and in 21 others aged 50-68 (mean age, 58) for a glabellar tap. An informed consent was obtRined from the patients or their guardians following a full explanation of the procedure.
17 An independent investigator, blind to the electrophysiologic data, retrieved clinical information from chart review for clinical localization of lesion (Table 2). Another investigator analyzed CT scans without knowledge of either the clinical or the electrophysiologic status of the patients. Using standard sets of templates (Damasio and Damasio 1980; Damasio 1983), the lesions were transferred onto a lateral view of the brain to determine the area of major overlap. We elicited the blink reflex according to the standard technique (Kimura 1983). Subjects lay supine with the active electrodes (G~) placed on the upper lateral aspect of the orbicularis oculi muscle and a reference electrode (G2) on the lateral surface of the nose. A ground electrode was located around the ann. The supraorbital, infraorbital and mental nerves were stimulated at the respective foramina on one side. The use of a specially designed amplifier with a short blocking time and a low internal noise minimized the problem of stimulus artifact (Walker and Kimura 1978). Frequency response ranged from 20 Hz to 32 KHz. If R 1 was difficult to elicit with single shocks of appropriate intensities, we used paired stimuli with an interstimulus interval of 5 ms (Kimura 1974). A subthreshold conditioning shock subliminally facilitated the reflex pathway. A supramaximal test stimulus then triggered a sweep and evoked the response, allowing determination of its latency from the second shock. A set of recordings on each side consisted of 8 R~ responses, 4 with single and 4 with paired stimuli (see Fig. 2) as well as 4 R 2 responses at each site of stimulation, 2 with single and 2 with paired shocks (see Fig. 3). Stimuli were separated by a time interval of 30 s or longer to minimize interaction between successive trials (Kimura 1983). For mechanical stimulation, a specially constructed reflex hammer closed a built-in microswitch on impact and triggered a sweep. In contrast to unilateral electrical stimulation, a midline glabellar tap elicited R 1 and R 2 bilaterally (see Fig. 2) allowing instantaneous comparison between the two sides (Fisher et al. 1979). Each set consisted of 4 pairs of R 1 and R 2 recorded simultaneously on both sides.
RESULTS
Finding in normals Table 1 shows the normal range of electrically elicited R~ and R 2 in 23 healthy subjects and mechanically induced R 1 in another 21 normal subjects. Electrical stimulation of the supraorbital nerve always elicited both R~ and R E. Glabellar taps elicited R~ in all subjects but R 2 only inconsistently. Electrical stimulation of the infraorbital nerve evoked R 2 in all, but R~ only in 8 subjects. Stimulation of the mental nerve gave r i s e t o R 2 in 21 cases and RI in one instance. The latencies of R~ and R 2 w e r e similar with stimulation of either the supraorbital or infraorbital nerve. However, R 2 elicited by stimulation of the mental nerve was considerably longer in latency. Based on these observations, the quantitative data analyzed in the patient group included R 1 and R 2 elicited by stimulation of the supraorbital nerve, R 2 following stimulation of the infraorbital nerve and R~ induced by glabellar tap.
18
cs
o,I c5 i
~
'd
r-:
+t
+t
4-1
c5
c~
d
o
+1
+1
+1
+1
~
M
,14
+1
+1
+1
o
2
~
N
'
f
i
"'
I
O e'-,
¢ +1
+1
+1
+1
+I
+1
,~
o el 0 0 . ~0
b
+1
+1
+1
+1
£
2
N
N
,-I
©
Z O g:::l i,.1,.1 m r~ [<
4-1
+1
-H
-H
4-1
-H
+l
4'-I
¢5
,~
o
,~
....g,
@ <
ed Z < <
"6 .0 e,,)
Z
[,,.
~
.,-..
o0
"
rO
'd
-.-ag ~v
F [.. Z
N <
Z
z ,.~.N M
0
19 The corresponding R~ and R2 elicited by stimulation on the right and left side showed no significant differences in latency, amplitude or duration. However, the ipsilateral R 2 was occasionally shorter in latency, larger in amplitude and longer in duration than the simultaneously recorded contralateral R2. Using the mean plus three standard deviations in the control groups as the upper limit of normal ,electrically elicited R~ is abnormal if the latency difference between the two sides exceeds 1.3 ms. Similarly, the latency difference of R 1 evoked by a glabellar tap should be less than 1.6 ms between the two sides. Stimulating either the supraorbital or infraorbital nerve unilaterally, the simultaneously recorded ipsilateral and the contralateral R 2 should not vary more than 5 ms in latency. A latency difference between R 2 evoked by right-sided stimulation and corresponding R 2 evoked by left-sided stimulation should not exceed 7 ms. In preliminary experiments we attempted to electrically rectify and integrate the response in assessing the size of R 1 and R 2. This approach, however, proved to be less accurate than might be expected on theoretical ground because a slight baseline shift induced by stimulus artifact and eye movements precluded a selective measurement of the area under the waveform. Therefore, the size of response was estimated simply as the product of the amplitude and duration (Fig. 1). Its asymmetry was considered abnormal with reduction of R 1 or R 2 to less than one third of the comparable response on the opposite side. Lt.
Rt.
& ~ L ~
~
t-~ Amplifier ~ - ~ t O s c l l l ° u c o p e ~
1. Rt
Amplitude -
C'ere
~--J
2. Ipsilateral R 2 Average Amplitude
-- . . . . . . .
Rt. - - T . ~ : ~ i
I..,-°,
:~ Lt.--" ,
t stimulus llrlifict
-
Latency
,
,,,
~ I .......... ' ~
ou,.,,o. -
,: -
i ....
I
Av.r.ge
Amplitude
3. Contralateral R 2
Fig. 1. Stimulation and recording arrangement for the blink reflex. The presumed pathway of R~ through the pons (1) and direct and consensual R 2 through the pons and lateral medulla (2 and 3) are schematically illustrated. Dotted lines indicate the presence of multiple interneurons. The primary afferents of R] and R 2 are shown as one fiber. This and other details of central connections of these reflexes are not known. Bottom: a typical oscilloscope recording of the blink reflex after right-sided stimulation. Note an ipsilateral R 1 response and bilateral simultaneous R 2 responses (Reprinted by permission of John Wiley & Sons, Inc. from J. Kimura (1982) Conduction abnormalities of the facial and trigeminal nerves in polyneuropathy, Muscle & Nerve, 5: S139-S144).
20 Findings in patients Electrically elicited R~ was normal in 36 patients but was absent with single or paired stimuli in 5. In the remaining 25 subjects, R~ was absent or reduced in size in 21 and delayed in 4 others contralateral to the lesion after a single stimulus, but in all, paired shocks corrected the abnormalities (Figs. 2-5). When the affected side of the face was compared to the normal side, R~ was significantly smaller in amplitude (P < 0.01) and shorter in duration (P < 0.01) but unchanged in latency (P > 0.05). The glabellar tap elicited a normal R~ in 42 of 58 patients tested. In the remaining 16 patients, the response on the affected side of the face was absent in 8 and delayed in the other 8 (Fig. 2). Statistical analyses showed no significant (P > 0.05) difference between the two sides in latency, amplitude or duration of the mechanically elicited R1.
A.H. 12-4-80 1{ --'X--~J ~ - ' ' ~
. . . . . . Rt {
-
,_~Lt {
~
glabellar 2{ ~ Tap
3 { ~ ~
~ - ' - ~
.--,,---~ ~ L
V.W. 1-30-81 - Rt t
{ _ -~/~/~1~~
oL.smv 5ms 1 -~l\~"~'/~J~
"~_
Rt
'~t!\~X.~
~
Rt
Rt(SO)
L-
-
-
J
I _I~,\~--~ Lt(SO)
4- ~
L _ _ _ -J O~.5mV 2ms Lt
I L~
Lt
~
L__-I
L - _I
oL.smV 2m8
Fig. 2. Mechanically and electrically elicited RI in a 60-year-old man with hematoma in the rightfTontat lobe (left side &the figure) and in a 59-year-oldman with a large infarct in the left hem/sphere ( ~ side of the figure). From top to bottom, bilateral Rl elicited by gtahellar tap (small arrows), electrically~ t e d R~ (dotted brackets) with single (trials 1-4) and paired shocks (trials 5-8) on the right, ~ ~ a b t e R~ with stimulation on the left. In both patients, R ! elicited b y ~ tap was increased:in latency on the paretic side. Electrically elicited R i was absent with $in~e $ ~ s but normal with p ~ d ~ k s on the affected side. SO = supraorbital nerve.
21 O v e r a l l , R 2 was abnormal in 50 of 66 patients whether tested with single or paired stimuli of the supraorbital or infraorbital nerve. In 4 comatose patients, R 2 was totally absent bilaterally regardless of the side of stimulation (coma type). In 31 patients (Fig. 3), R 2 was abnormal bilaterally when the affected side of the face was stimulated, but with stimulation of the normal side it was only abnormal contralaterally (mixed type). Afferent (Fig. 4) and efferent (Fig, 5) types of changes described earlier occurred in 7 and 8 patients, respectively. In general, R 2 in patients was significantly (P < 0.01) increased in latency, reduced in amplitude and decreased in duration. However, the ipsilateral R 2 after stimulation on the normal side of the face remained unchanged in
Stimulation on Left
R1(SO)
Stimulation on R i g h t
R.T. 4-3-80
4
L. _ _ J
L._ _.J
10.5mV 2ms
(!
2{ R 2 (SO)
3{ 4{
....... Rt Lt
Lv~
k f I
I
oL•,5mV 10ms Rt
....
2{
_. _ ~
R2 (IO) 3{
{7~ - -
Lt ,
_~ - - - ~ -
¢
t
I
0L~_.5mV 10ms
Fig. 3. Mixed type of abnormality in a 55-year-old man, 3 days after an infarction in the left posterior frontal region. From top to bottom, R) (dotted brackets) by single (trials 1-4) and paired shocks (trials 5-8) of the supraorbital nerve (SO), R z (solid brackets) by single (trials 1 and 2) and paired shocks (trials 3 and 4) of the supraorbital nerve (SO) and R 2 by single (trials 1 and 2) and paired shocks (trials 3 and 4) of the infraorbital nerve (IO)). When the right side was stimulated, R 1 was absent with single shocks and reduced in amplitude with paired shocks, whereas R 2 was absent bilaterally with stimulation on the right and contralaterally with stimulation on the left.
22 Stimulation on Right
Stimulation on Left 1 : ~ . . . . . ~
RI(SO)
B.S. 1,9-80
Lt
41 -
_
-
-
-
8 I.-_ - /
L - - -J
1{
.
.
.
.
.
.
.
m
a{
.,=
_
{ I
R
I
i
-
i
4{
R 2 (IO)
Lt
=
H
1{ ~
Rt ,
i
R 2 (SO)
0L0.~.5mV 2ms
_
i
m
i
,
0.5mV 10ms
t
a{ 4{ I
I
oLsmv 10ms
Fig. 4. Afferent type of abnormality in a 73-year-old woman 1 day after an infarction in the distribution of the leR middle cerebral artery. The same arrangement as in Fig. 3. When the right side was stimulated, Ri was absent with single shocks and reduced in amplitude with paired shocks, whereas R 2 was absent bilaterally after stimulation of either the SO or IO on the right,
amplitude when compared to the corresponding Rz in the controls (P > 0.05). Within the patient group, the ipsilateral R2 after stimulation on the normal side of the face was significantly (P < 0.01) greater in amplitude and shorter in latency compared to the simultaneously recorded contralateral R 2 or to ipsilateral or contralateral R2 elicited by separate stimulation on the affected side (Table 1). Repeat studies performed in 32 patients showed gradual improvement of R~ and R 2. Excitability of R~ usually returned to normal by the second week. The rate of recovery of R 2 was variable and s o m e ~ of abnormality persisted even when the clinical condition had been stable for several weeks. ~ R 2 w a s totally a b s ~ in a comatose patient (Lyon et al. 1972), the return of the reflex response occurred in a predictable order, i.e. ipsilateral before contralateral R 2 with stimulation on thenormal
23
12 ~ R1 (SO)
65;
Stimulation on Left
Stimulation on Right
Lt ~
\ ~
M.S. a-lg-s0 Rt
' ~ L _ _ J
L _ _ -- J
1 { ~...__~.~.f' ..,
Rt { - - ~ ~ ~
~0.5mV 2ms
Lt
R2 (SO) I
_ l
_ I
I
1{ ^ ~ ~ L t
{
I
~
L
oL~..5mv lOms
t
2{ " L - ~ ~ ~ ' ~ ' ' ~ - - ~ Rt { ~ , . , ~ .
I
I
I
I
10.SmV
10ms Fig. 5. Efferent type of abnormality in a 32-year-old woman with moyamoya syndrome 4 days after an infarction in the distribution of the right middle cerebral artery. The same arrangement as in Fig. 3. Stimulating the left side of face, R 1 was absent with single shocks and reduced in amplitude with paired shocks. With stimulation on either right or left side, R 2 recorded on the left was consistently reduced in size.
side of the face followed by ipsilateral before contralateral affected side.
R2
with stimulation on the
Correlation with clinical data
Based on clinical findings the responsible lesion predominantly, though not necessarily selectively, affected frontal, temporal, parietal and occipital lobes in 12, 7, 5 and 2 patients, respectively. In 11 others, the lesion was primarily subcortical (Table 2). The remaining 29 patients had multiple sites of involvement or lesions that we could not localize. Analysis of blink reflex findings in each category revealed no consistent relationship between the type of electrophysiologic abnormalities and the major site of hemispheric involvement (Table 2).
24
Z o
oo
<
L~
©
o (-. <
Z
Z
< o <
z M
Z <
0
0
]
M <
N © z < N M
z
o
M
© ~ . ~
~
< ~
0
~
25 Correlation with CT scan abnormalities
In 27 patients, CT scans, analyzed without knowledge of clinical fmdings, showed only equivocal changes or no abnormalities at all. The remaining 39 patients had CT scans which allowed accurate localization of pathology often involving more than one lobe as described below. In this group, the blink reflex was abnormal in 29 and normal in 10 cases. In the 29 patients, the CT scans showed single or multiple lesions involving the frontal, parietal, temporal and occipital lobes in 16, 9, 6 and 6 patients, respectively. In 7 others the abnormality was exclusively subcortical. The premotor regions were involved in 15 patients but in 7 of these the lesion extended into the parietal structures. The basal ganglia were affected in 5 and the deep white matter close to the corpus callosum in another 2. A composite picture obtained by superimposition of the cortical abnormalities in the CT scans suggested considerable scatter, but with a tendency to overlap in the inferior Rolandic area at the level of face representation (Fig. 6). In the remaining 10 patients without abnormalities of the blink reflex, lesions involved the frontal, parietal, temporal and occipital lobes and basal ganglia in 5, 3, 2, 1 and 1 patient, respectively. A comparable scatter of lesions was noted in similar proportions across all lobes. The post-central Rolandic area was conspicuously spared but the normal group was much smaller and an entirely valid comparison was not possible. DISCUS SION
Recent reports have described several patterns of abnormality of the blink reflex in hemispheric CVA (Messina and Quattrone 1973; Kimura 1974; Dehen et al. 1976; Fisher et al. 1979; Berardelli et al. 1983). In our previous study (Kimura 1974), the most common reflex change in patients with hemifacial hypesthesia was that of an afferent pattern with diminution of R2 bilaterally after stimulation of the affected side of the face. Berardelli et al. (1983) reached similar conclusions. These findings are in contrast to the results reported by Dehen and associates (1976), who emphasized an efferent pattern with diminution of R 2 on the paretic side of the face and enhanced response on the opposite side in patients with hemiplegia. When studied after the patients' condition had stabilized, R 1was normal in latency although reduced (Kimura 1974; Dehen et al. 1976) or enhanced (Messina and Quattrone 1973; BerardeUi et al. 1983) in amplitude on the affected side. Fisher and associates (1979) reported a significant increase in latency of mechanically elicited R 1 during acute stages of CVA. The present study in a large number of patients confirmed that an abnormality of the blink reflex is the rule rather than the exception in patients with acute hemispheric vascular events. In agreement with an earlier observation (Fisher et al. 1979), glabellar tap elicited a delayed R~ contralateral to the hemispheric lesion in a few patients. Similarly, R 1 evoked by single electric shock was occasionally increased in latency. However, in these cases, the amplitude of R1 was substantially reduced, indicating partial activation of the reflex pathway. Indeed, when a maximal response was recorded by paired stimuli R~ was normal in latency. Alteration of R~ does not necessarily imply
26
Fig. 6. A composite picture obtained by superimposition of the CT scans. A: in 29 patients with abnormal blink reflexes, the lesions were scattered considerably although they tended to overlap in the inferior Rolandic area at the level of face representation, both in the pre- and post-central region. B: in 10 patients without abnormalities in the blink reflex, the lesions conspicuously spared the inferior post-central area but the normal group was too small to draw a definite conclusion.
a pathologic process of the reflex arc itself since indirect effects of lesions outside the primary reflex pathway may also compromise its excitab~ty (Kimura 1975; Fisher oral. 1979). When fully activated, however, the latency of Rt usually indicates the conduction characteristics of the reflex pathway (Kimura et al. 1969). The polysynaptic R2 was more profoundly suppressed than the ol~synaptic R~ in acute hemispheric lesions. This finding suggests that interruption of de=~nding pathways prhnarily results in disfacilitation of brainstem intorneurons and, to a lesser degree, motor neurons. Based on increased amplitude of R l, Dehen and others (1976)
27 reported hyperexcitability of the motor neurons ipsilateral to the hemispheric lesions, a finding not confirmed in this study. This discrepancy may be in part related to the time of study following the CVA because excitability changes are influenced by the interval from the onset of the ictus. In the present study, R z was commonly absent or diminished bilaterally when the affected side of the face was stimulated. This observation alone is consistent with reduced excitability in neurons in the area of the spinal nucleus of the trigeminal nerve or interneurons forming the afferent arc (Ongerboer de Visser and Kuypers 1978). However, an additional abnormality of the contralateral R2 after stimulation of the normal side of the face suggests reduced excitability of the efferent arc, i.e. either the crossed interneuronal pathway relaying the impulse from the opposite spinal nucleus or the facial nucleus itself on the paretic side. Thus, the basic R z abnormality probably results from diffuse disfacilitation involving both the afferent and efferent arcs of the reflex contralateral to the hemispheric lesion. Analyses of the composite CT scan templates in our series suggest that blink reflex abnormalities most consistently occur with lesions involving the inferior post-central area (Fig. 6). However, careful review of individual scans indicates that the responsible lesions are widely scattered and sometimes confined to the lower premotor and motor regions or to other areas entirely outside the Rolandic region. Of interest are CT scan abnormalities in patients with normal blink reflex, showing conspicuous sparing of the sensory representation of the face. Although this particular group was too small to draw a definite conclusion, the finding is consistent with the recent result by Ongerboer de Visser and associates (1981) who, studying the corneal reflex, observed that the lower post-central region had an excitatory influence upon the lateral reticular formation. Afferent or efferent patterns may result from the interruption of specific descending pathways. However, no consistent relationship has emerged in correlating the type of blink reflex changes to the location of an acute hemispheric lesion. We conclude that (1) the blink reflex abnormality is the rule rather than the exception in an acute hemispheric CVA, (2) the polysynaptic R 2 is more affected than the oligosynaptic RI, suggesting a greater suppression of brainstem interneurons than motor neurons, (3) contralateral disfacilitation usually occurs diffusely, involving both afferent and efferent reflex pathways, and (4)the excitatory drive to the brainstem pathway of the blink reflex probably originates in wide areas of the cortex but most conspicuously in the inferior post-central area, coinciding with the sensory representation of the face. ACKNOWLEDGEMENTS
The authors wish to thank Sheila R. Mennen, Deborah A. Gevock and Lesa A. Bowles for technical assistance.
28 REFERENCES Berardelli, A., N. Accornero, G. Cruccu, F. Fabiano, V. Guerrisi and M. Manfredi (1983) The orbicularis oculi response at~er hemispheral damage, J. Neurol. Neurosurg. Psychiat., 46: 837-843. Boelhouwer, A. J. W. and C. H. M. Brunia (1977) Blink reflexes and the state of arousal, J. Neurol. Neurosurg. Pyschiat., 40: 58-63. Damasio, H. (1983) A C T guide to the identification of cerebrovascular territories, Arch. NeuroL, 40: 138-142. Damasio, H. and A.R. Damasio (1980) The anatomical basis of conduction aphasia, Brain, 103: 337-350. Dehen, H., J.C. Willer, N. Bathien and J. Cambier (1976) Blink reflex in hemiplegia, Electroenceph. Clin. Neurophysiol., 40: 393-400. De Jong, R.N. (1967) The Neurological Examination, 3rd edition, Harper and Row, New York, p. 226. Desmedt, J. E. and E. Godaux (1976) Habituation ofexteroceptive suppression and of exteroceptive reflexes in man as influenced by voluntary contraction, Brain Res., 106: 21-29. Fisher, M.A., B.T. Shahani and R.R. Young (1979) Assessing segmental excitability after acute rostral lesions, Part 2 (The blink reflex), Neurology (N.Y.), 29: 45-50. Hiraoka, M. and Shimamura, M. (1977) Neural mechanisms of the corneal blinking reflex in cats, Brain Res., 125: 265-275. Kimura, J. (1974) Effect of hemispheral lesions on the contralateral blink reflex, Neurology (Minneap), 24: 168-174. Kimura, J. (1975) Electrically elicited blink reflex in diagnosis of multiple sclerosis-review of 260 patients over a seven-year period, Brain, 98: 413-426. Kimura, J. (1982) Conduction abnormalities of the facial and trigeminal nerves in polyneuropathy+ Muscle & Nerve, 5: S139-S144. Kimura, J. (1983) Electrodiagnosis in Diseases of Nerve and Muscle - - Principles and Practice, F.A. Davis, Philadelphia. Kimura, J., J.M. Powers and M.W. Van Allen (1969) Reflex response of orbicularis oculi muscle to supraorbital nerve stimulation m Study in normal subjects and in peripheral facial paresis, Arch. Neurol., (Chic.), 21: 193-199. Kugelberg, E. (1952) Facial reflexes, Brain, 75: 385-396. Lyon, L.W., J. Kimura and W.F. McCormick (1972) Orbicularis oculi reflex in coma - - Clinical, electrophysiological, and pathological correlations, J. Neurol. Neurosurg. Psychiat., 35: 582-588. Magladery, J.W. and R. D. Teasdall (1961) Corneal reflexes - - An electromyographic study in man, Arch. Neurol. (Chic), 5: 269-274. Messina, C. and A. Quattrone (1973) Comportamento dei riflessi trigemino-facciati in soggetti con tesioni emisferiche, Pa'v. Neurol., 43: 379-386. Oliver, L.C. (1952) The supranuclear arc of the corneal reflex, Acta Psychiat. Scand., 27: 329-333. Ongerboer de Visser, B.W. (1981 ) Corneal reflex latency in lesions of the lower posteentral region, Neurology (New York), 31: 701-707. Ongerboer de Visser, B.W. (1983) Comparative study of corneal and blink reflex latencies in patients with segmental or with cerebral lesions. In: J.E. Desmedt (Ed.), Motor Control Mechanisms in Health and Disease, Raven Press, New York, pp. 757-772. Ongerboer de Visser, B.W. and H.G.J.M. Kuypers (1978) Late blink reflex changes in lateral medullary lesions - - An electrophysiological and neuroanatomieal study of Wallenberg's syndrome, Brain, 101 : 285-294. Richwien, R. (1966) Zur Diagnose der zentralen St/Srungen des Cornealreflexes in afferenten Schenkel, Dtsch. Z. Nervenheilk., 189: 191-195. Ross, R.T. (1972) Corneal reflex in hemisphere disease, J. Neurol. Neurosurg. Pyschiat., 35: 877-880. Rushworth, G. (t962) Observations on blink reflexes, J. Neurol. Neurosurg. Psychiat., 25: 93-108. Shahani, B.T. (1968) Effects of sleep on human reflexes with a double component, J. Neurol. Neurosurg. Psychiat., 31: 574-579. Shahani, B.T. (1970) The human blink reflex, J. Neurol. Neurosurg. Psychiat., 33: 792-800. Walker, D. D. and J. Kimura (1978) A fast-recovery electrode amplifier for electrophysiology, Electroenceph. Clin. Neurophysiol., 45: 789-792. Wolff, R. (1913) Die Areflexie der Cornea bei organischen Nervenkrankheiten, Arch. Psychiat., 52:716-747.