Journal of the Neurological Sciences, 109 (1992) 148-155 © 1992 Elsevier Science Publishers B.V. All rights reserved 0022-510X/92/$05.00
148 JNS 03760
Contralateral early blink reflex in patients with facial nerve palsy: indication for synaptic reorganization in the facial nucleus during regeneration W i l h e l m N a c i m i e n t o a, Klaus Podoll a, M a n u e l B. G r a e b e r b,l, R u d o l f T 6 p p e r a, E c k h a r d M/Sbius a, H a r t m u t O s t e r m a n n a, J o h a n n e s N o t h a a n d G e o r g W. K r e u t z b e r g b a Aifried-Krupp-Hospital, Dept. of Neurology with Clinical Neurophysiology, Alfried-Krupp-Str. 21, D-4300 Essen, Germany, and b Max-Planck-lnstitute of Psychiatry, Dept. of Neuromorphology, Am Klopferspitz 18a, D-8033 Martinsried, Germany
(Received 31 July, 1991) (Revised, received 11 December, 1991) (Accepted 16 December, 1991)
Key words: Blink reflex; Facial nerve palsy; Contralateral early response; Regeneration; Facial nucleus; Synaptic organization
Summary
Fifty patients with Bell's palsy and 30 patients with etiologically different symptomatic peripheral facial nerve palsy were studied by means of electrically evoked blink reflexes 1-23 days after onset of paresis. Their results were compared with a normal control group of 30 healthy subjects. In a significant number of patients (64% in Bell's palsy and 53% in symptomatic facial nerve palsy) a contralateral early blink reflex response (R1) could be elicited upon stimulation of the normal side as compared to 13% in the control group. It is suggested that this result may be explained by synaptic reorganization of the facial nucleus leading to functional unmasking of pre-existing crossed trigemino-facial reflex pathways during regeneration. This view is in line with previous experimental data in animals on the time course of structural changes in the facial nucleus after lesioning of the ipsilateral facial nerve.
Introduction
Electrical stimulation of the supraorbital nerve results in two responses: an early component (R1) ipsilateral to the stimulus which is transmitted through an oligosynaptic arc in the ports (Tanaka et al. 1971; Hirakao and Shimamura 1977) and a bilateral late component (R2) which follows a long polysynaptic pathway travelling through the spinal nucleus of the trigeminal nerve (Kugelberg 1952; Kimura 1973; Hirakao and Shimamura 1977). Contralateral R1 responses can be elicited in normal subjects by facilitating maneuvers, such as contraction of the orbicularis oculi muscle or electrical stimuli of the median nerve
I Present address: Children's Hospital and Harvard Medical School, Boston, MA 02115, USA. Correspondence to: Prof. J. Noth, Alfried-Krupp-Hospitai, Dept. of Neurology with, Clinical Neurophysiology, AIfried-Krupp-Str. 21, D-4300 Essen, Germany. Tel.: (201) 434-2528; Fax: (201) 434-2399.
(Wilier et al. 1984; Soliven et al. 1988). This early crossed response is mediated by direct anatomical connections of the trigeminal sensory nucleus and the contralaterai facial nucleus (Ramon y Cajal 1909; Wilier et al. 1984; Soliven et al. 1988). Under standard recording conditions, the synaptic connections producing the contralateral R1 response are irresponsive to supraorbital nerve stimulation in the majority of normal subjects (Penders and Delwaide 1974; Trontelj and Trontelj 1978; Csecsei 1982; Wilier et al. 1984; Soliven et al. 1988). Therefore, the blink reflex may provide insight into synaptic organization and plasticity of the facial nucleus. The latter is of particular interest in pathological cases (Kimura and Lyon 1972; Bratzlavsky and vander Eecken 1977; Esteban and Molina-Negro 1986; Moeller and Jannetta 1986; Horowitz 1987; Agostino et al. 1988; Laskawi et al. 1990). In patients with facial mass contractions following facial nerve palsy, for instance, the appearance of contralateral R1 has been attributed to alterations in the synaptic organization of the facial nucleus (Bratzlavsky and vander Eecken
149 1977). In this study we present evidence that contralateral R1 also occurs in the early phase of facial nerve regeneration in a significant number of patients. Possible clinical relevance of these findings is discussed in the light of accumulating morphological evidence in animals which demonstrate substantial changes in the synaptic organization of the facial nucleus following peripheral lesioning of the ipsilateral facial nerve (Blinzinger and Kreutzberg 1968; Graeber and Kreutzberg 1990). Methods
Patients and subjects All subjects gave their informed consent to participate in the study after having been informed about its objectives and the procedures involved.
Patients with idiopathic peripheral facial nerve paresis (Bell's palsy) The 50 patients (24 males, 26 females) with Bell's palsy had a mean age of 42.3 years with a range from 15 to 81 years. Paresis was on the left side in 23 patients and on the right side in 27 patients. Apart from Bell's palsy, neurological status was normal in all patients except for an unrelated mild hypacusis in 4 cases. In 30 cases cranial compute~i tomography was performed which was normal in all but 3 eldery patients who showed mild cerebral atrophy. Patients were studied between the 1st and 17th day from onset of Bell's palsy.
Patients with symptomatic peripheral facial nerve paresis The 30 patients (17 males, 13 females) with symptomatic peripheral facial paresis due to different etiologies comprised 5 patients with posttraumatic palsy after fractures of the petrous or mastoid portion of the temporal bone, 4 patients with a postoperative facial paresis after operations for cholesteatoma or chronic otitis media, 1 patient with chronic otitis media and marked tympanic effusion, 10 patients with herpes zoster auricularis (Ramsay Hunt syndrome), 1 patient with HIV infection of the nervous system, 1 patient with neuroborreliosis (Lyme disease), 4 patients with Guillain-Barr~ syndrome and 4 patients with diabetic polyneuropathy. Paresis was left-sided in 16 patients and right-sided in 14 patients. The mean age was 53.1 years with a range from 19 to 86 years. Patients were studied between the 2nd and 23rd day from onset of facial palsy.
Control group Thirty healthy subjects (12 men, 18 women), with a mean age of 41.7 years (range 16-83 years), were studied as a control group.
Blink reflex recording All recordings were performed in a quiet room with the temperature kept between 20 and 22°C. Subjects lay supine and were instructed to close their eyes gently. In 10 patients displaying a contralateral R1 response upon stimulation of the normal side, blink reflexes were additionally recorded with open eyes in order to determine the influence of voluntary effort (that is contraction of the eyelids) on this response. The blink reflex was elicited transcutaneously by electrical stimulation of the supraorbital nerve, with the cathode positioned near the supraorbital foramen and the anode 3 cm above it (at least 2.5 cm away from the midline). Using a constant-current unit, the stimuli consisted of rectangular pulses of 0.2 ms duration and 8-12 mA intensity, delivered randomly with interstimulus intervals of 15-60 s to avoid habituation. Stimuli were kept constant during the recording session and were of such intensity that reflex responses were just maximum. Electromyographic (EMG) activity was recorded from the ipsi- and contralateral orbicularis oculi muscle (pars inferior) by surface electrodes set on the midline of the lower lid and bridge of the nose. The EMG signals were fed into amplifiers, bandpass filtered (band-width 20-3000 Hz) and stored on floppy discs for subsequent evaluation. The sweep time was 100 ms with 1024 addresses per channel, the sampling time being 98 ~s. In every subject, 5 responses were obtained from each side. Measurements were made by visual inspection using a cursor in the optical display. Latencies of the blink reflex responses were determined from the mean of the five measurements, and amplitudes were taken as the maximum peak-to-peak values.
Electroneurography Direct motor potentials were obtained in orbicularis oculi muscle with a supramaximal stimulation of the facial nerve at the stylomastoid foramen. Constant-current stimuli consisted of rectangular pulses of 0.2 ms duration and 30-80 mA intensity. The position of recording electrodes was the same as in blink reflex examinations.
Statistical analysis Mann-Whitney U-tests were used for between-group comparison of different variables. Chi-square analysis was performed for between-group comparisons of the frequencies of different nominally scaled variables. Differences were considered to be significant with a P value < 0.05.
150 Results Control group In all subjects of the control group, electrical stimulation of the supraorbital nerve elicited an early R1 response on,the ipsilateral side and a late R2 response on the ±psi- and contralateral side. An additional contralateral* R1 response was observed bilaterally in 13% (4/30) of control subjects. Since no significant differences were observed between blink reflex responses to left and right side stimulation, the mean of both sides was calculated and used for comparison with the patient groups (Table 1).
i t numberof patients
Patients with Bell's palsy
ml numberofpatients
In the group of 50 patients with Bell's palsy, the ipsilateral R1 and R2 responses to stimulation of the normal side and the contralateral R2 response to stimulation of the paretic side were recorded in all patients and showed no significant differences in latency and amplitude compared to controls except for a slight delay of the ipsilateral R2 latency (P < 0.05; Table 1). The ipsilateral R1 and R2 responses to stimulation of the paretic side and the contralateral R2 response to stimulation on the normal side were obtained at a lower rate than in controls (statistically significant for ipsilateral R1 and contralateral R2 with P < 0.001 and P < 0.025 respectively). In those cases where blink reflex responses were preserved on the paretic side, latencies were markedly prolonged and amplitudes
1
1
~
5
5
ContrnlateraX Contralsteral Contralater81
10
15
20
25
10
15 20 25 daysfromonsetof paresis
daysfromonsetof pnreslo
R1 u p o n s t l n u I a t i o n o f t h e not-maX s i 4 e R1 u p o n s t i m u l a t i o n o£ both sides RI absent on both sides
Fig. 1. Rate of appearance of contralateral R1 responses in the time course from onset of Bell's palsy (A) and of symptomatic facial nerve palsy (B).
were reduced as compared to the control group (P < 0.0001; Table 1). A contralateral R1 response, which was seen in 13% of the age matched control subjects, was observed with
TABLE 1 BLINK REFLEX PARAMETERS IN CONTROL SUBJECTS AND PATIENTS WITH IDIOPATHIC AND SYMPTOMATIC FACIAL PALSY Means + SD. Control group
Patients with Bell's palsy
Symptomatic facial palsy
mean of both sides
paretic side stimulation
normal side stimulation
paretic side stimulation
normal side stimulation
Ipsilateral R1 frequency (%) latency (ms) amplitude (/~V)
100 (30/30) 10.8 4- 0.8 551 4- 170
46 (23/50) 12.5 + 1.7 135 + 85
100 (50/50) 10.6 + 1.2 489 + 229
30 (9/30) 14.0 + 3.4 122 ± 95
100 (30/30) 10.9 4- 1.1 538 4- 238
Contralateral R1 frequency (%) latency (ms) amplitude(/~V)
13 (4/30) 11.4 4- 1.1 150 4- 79
12 (6/50) 12.0 4- 0.9 267 +103
64 (32/50) 11.9 ± 1.2 94 4- 56
10 (3/30) 11.2 4- 0.6 167 + 94
53 (16/30) 13.5 4- 6.0 113 4-45
~teral R2 frequency (%) latency (ms) amplitude (/tV)
100 (30/30) 30.5 4- 1.9 552 4- 154
90 (45/50) 36.6 + 5.9 166 + 144
100 (50/50) 32.0 4. 3.4 534 + 163
63 (19/30) 39.2 + 4.5 155 + 80
100 (30/30) 34.1 4- 3.2 509 4- 175
Contralateral R2 frequency (%) latency (ms) amplitude(/~V)
100 (30/30) 30.2 4- 2.5 489 4-135
100 (50/50) 32.0 4- 4.2 476 4-180
84 (42/50) 36.3 4- 4.8 147 ± 87
97 (29/30) 33.4 ± 4.0 504 +224
73 (22/30) 38.4 ± 4.2 136 4- 80
151 TABLE 2 COMPARISON OF BLINK REFLEX PARAMETERS 1N PATIENTS WITH BELL'S PALSY DISPLAYING AND LACKING CONTRALATERAL R1 RESPONSES
Mean + SD.
Stimulation of paretic side Ipsilateral RI frequency (%) latency (ms) amplitude (/tV) Ipsilateral R2 frequency (%) latency (ms) amplitude (pV) Stimulation of normal side Contralateral R2 frequency (%) latency (ms) amplitude (/zV) Facial nerve motor electroneurography Amplitude reduction (%)
Patients with contralateral R1 responses (n = 32)
Patients lacking contralateral R1 responses (n = 18)
Comparison P
59 (19/32) 12.3+ 1.5 132 + 74
22 (4/18) 13.4+ 2.0 149 + 24
< 0.025 NS NS
100 (32/32) 35.7+ 5.0 178 +158
72 (13/18) 38.9+ 7.0 136 + 96
< 0.01 NS NS
94 (30/32) 35.2 + 4.3 147 + 80
67 (12/18) 38.9 + 5.1 145 +101
< 0.025 < 0.05 NS
69 + 21
60 + 26
NS
NS: no significant difference.
a significantly (P < 0.001) higher frequency of 64% (32/50) in the patient group. In 52% (26/50) of these cases, contralateral R1 was recorded unilaterally following stimulation on the normal side and in 12% (6/50) bilaterally following stimulation of either side. A contralateral R1 response upon stimulation of the normal side was less likely to occur in cases in which severe alterations of other blink reflex components were present on the paretic side. Thus, in the subgroup
LEFT
RZGHT
R1
of the 18 patients lacking contralaterai R1 responses on the paretic side, opposed to the 32 patients displaying contralaterai R1, ipsilaterai R1 and R2 responses to paretic side stimulation and the contralateral R2 response to stimulation of the unaffected side were abolished at a significantly higher rate (P < 0.025). Additionally, the contralateral late R2 component on the paretic side was also significantly prolonged in these cases (P < 0.05; Table 2).
R2
; Fig. 2. Blink reflexes (5 superimposed records) of a 35-year-old woman, 3 days from onset of Bell's palsy on the left side: electrical stimulation of ipsilateral (A) and contralaterai (B) supraorbital skin. Eyes were gently closed during examination. A contralateral RI response (arrow) could be elicited upon stimulation of the normal side. Calibration: horizontal: 10 ms; vertical: 100/zV.
152 TABLE 3 COMPARISON OF BLINK REFLEX PARAMETERS (MEAN + SD) IN PATIENTS WITH SYMPTOMATIC PERIPHERAL FACIAL PALSY DISPLAYING AND LACKING CONTRALATERAL R1 RESPONSES
Patients with contralateral RI responses (n = 16) Stimulation of paretic side lpsilaterai RI frequency (%) latency (ms) amplitude (/tV)
Ipsilateral R2 frequency (%) latency (ms) amplitude (~V)
Stimulation of normal side Contralateral R2 frequency (%) latency (ms) amplitude (/~V)
Patients lacking contralateral RI
Comparison P
responses (n = 14)
38 (6/16) 14.6 + 3.7 83 +37
21 (3/14) 12.9 + 2.4 200 +122
NS NS NS
88 (14/16) 39.8 + 5.0 137 +59
36 (5/14) 37.4 + 2.4 205 +108
NS NS
94 (15/16) 39.0 _+ 4.7 118 +55
50 (7/14) 37.1 _+ 2.7 174 +107
< 0.01 NS NS
< 0.01
Facial nerve motor electroneurography Amplitude reduction (%)
43
+19
49 + 20
NS
NS: no significant difference.
There was no significant difference with respect to sex, mean age and mean duration of the palsy (5.8 vs. 7.7 days, respectively; z = - 1.35, n.s.) between patients lacking contralateral R1 responses and patients exhibiting this blink reflex component. The relationship between the occurrence of contralateral R1 responses and the duration of facial nerve palsy, as assessed by cross-sectional data analysis, is graphically
RIGHT
A
illustrated in Fig. 1A. Disregarding cases with contralateral R1 on both sides, which could be due to physiological facilitation by voluntary contraction of the eyelids, a contralateral R1 response on the paretic side was found at a substantial rate beginning with the second day after onset of facial palsy without showing further significant changes of rate of appearance in the subsequent time course (Fig. 1A). A typical example of LEFT
I
Fig. 3. Blink reflexes (5 superimposed records) of a 56-year-old woman, 17 days from onset of Bell's palsy on the left side: electrical stimulation of ipsilateral (A) and contralateral (B) supraorbital skin. Eyes were open during examination. A contralateral R1 response (arrow) could be elicited upon stimulation of the normal side, in spite of advanced recovery from paresis. Calibration: horizontal: 10 ms; vertical: 200/LV (right) and 100 ~V (left).
153 a contralateral R1 response upon stimulation on the normal side in a patient with Bell's palsy is shown in Fig. 2. The comparison of recording with open and closed eyes in 10 patients showed that the contralateral R1 response on the affected side is not abolished during recording with open eyes. Blink reflex responses recorded with open eyes in a patient with Bell's palsy are shown in Fig. 3.
Patients with symptomatic peripheral facial nerve palsy In the group of 30 patients with symptomatic peripheral facial nerve palsy, the results on blink reflex examination were essentially the same as those encountered in Belrs palsy (Table 1). A contralateral R1 response was found in 53% (16/30) of these patients: in 43% (13/30) to stimulation on the normal side and in 10% (3/30) to stimulation on either side. In the 14 patients lacking contralateral R1 responses, the other blink reflex components on the paretic side were absent at a significantly higher rate than in the 16 patients displaying contralaterai R1 (P < 0.01; Table 3). Thus, as in the group with Bell's palsy there appeared to be an inverse relationship between the presence of contralateral R1 responses and the severity of blink reflex alterations on the paretic side. Contralateral R1 responses were observed as early as 3 days after onset of paresis (Fig. 1B).
Discussion Numerous studies indicate that electrophysiological examination of the blink reflex is of great value in predicting clinical outcome of patients with peripheral facial nerve palsy (Dumitru et al. 1988; Heath et al. 1988; Ghonin and Gavilan 1990). Examination of the blink reflex is especially useful during early stages of facial nerve regeneration, where the persistence or early return of an initially absent ipsilateral R1 component may suggest a qualitatively satisfactory recovery of facial nerve function (Heath et al. 1988). In turn, if this response is absent during the first 2 weeks after onset of facial nerve palsy, a rather unfavorable prognosis is given (Heath et ai. 1988). The results of the present study demonstrate that in a significant number of patients with peripheral facial nerve palsy an additional crossed early blink reflex can be elicited upon stimulation of the normal side during the early phase of regeneration. Interestingly, the occurrence of this phenomenon appears to be unrelated to the etiology of facial nerve paresis. This may suggest that central rather than peripheral mechanisms are involved in the generation of this response, a view which is in line with previous studies indicating that a contralateral R1 response may be mediated by a crossed trigemino-facial
reflex pathway (Wilier et al. 1984; Soliven et al. 1988). Yet, it should be noted that a contralaterai R1 response can be produced by normal subjects on a central basis via corticonuclear facilitation due to contraction of the orbicularis oculi muscle (Wilier et al. 1984; Soliven et al. 1988) In the present study all subjects were instructed to close their eyes gently. The contralateral R1 response in patients could be due to a stronger voluntary effort to contract the eyelids on the paretic side as compared to the normal side. However, the fact that the contralateral R1 response could be elicited on the paretic side during recording with both open and gently closed eyes, as proved in 10 patients, renders this possibility unlikely. Furthermore, the similar rate of appearance of contralateral R1 responses upon stimulation on both sides in patients (12% in Bell's palsy, 10% in symptomatic facial nerve palsy) and in normals (13%) argues against such a bias. Another possible methodological pitfall has to be considered: Moving the stimulation electrode toward the midline may also evoke a contralateral R1 response which can be abolished by blocking the contralateral supraorbital nerve (Soliven et al. 1988), confirming its peripheral elicitation. In order to prevent such mechanisms of contralaterai R1 generation, the stimulation electrode was positioned at a distance of at least 2.5 cm away from the midline. Crossed early blink reflex responses occurred on the affected side in the majority of both patients with Bell's palsy and patients with symptomatic facial nerve paresis (Table 1). In patients lacking a contralateral R1 response on the paretic side the other blink reflexes recorded on the affected side were often also absent (see Results). This can be explained by the severity of facial nerve lesion precluding any blink reflex response on the paretic side. It seems therefore possible that a crossed trigemino-facial reflex pathway, which is normally present but apparently ineffective in most normal subjects under stable routine recording conditions (Wilier et al. 1984; Soliven et al. 1988), may become functional during the regeneration process of the peripherally injured facial nerve. Bratzlavsky and vander Eecken (1977) have previously reported on the occurrence of a crossed early blink reflex in patients with postfacial palsy mass contraction. The authors suggested that, in addition to peripheral causes like misguidance of axons, aberrations at the nuclear level may contribute to faulty restoration of post-lesion facial motility. Furthermore, there is experimental evidence suggesting that disturbance of complex fine motor movements as occurring in hemifacial spasm may be caused by a state of increased excitability of facial motor neurons (Esteban and Molina-Negro 1986; Horowitz 1987; Vails-Sole and Tolosa 1989; Moeller and Sen 1990). Clinically, impairment of fine motor functions is frequently observed as
154 a symptom of incomplete recovery following peripheral facial palsy (Fowler 1939; Bratzlavsky and vander Eecken 1977). Thus, the question arises whether functional and/or structural changes at the nuclear level may accompany the regenerative process of the facial nerve and, if so, whether such alterations may persist even after peripheral reinnervation of the musculature has been achieved. Recent studies employing axotomy of the rat facial nerve have in fact demonstrated that there are massive changes in the synaptic organization of the regenerating motor nucleus. Glial cells, i.e. microglia and astrocytes, appear to be directly involved in this process. While microglia engage in the early detachment and displacement of afferent axonal endings from neuronal surface membranes ("synaptic stripping"; Blinzinger and Kreutzberg 1968), astrocytes appear to maintain this state of synaptic deafferentation of the regenerating motor neuron for longer periods, i.e. months (Graeber and Kreutzberg 1990). Our electrophysiological data would be compatible with these anatomical alterations, particularly since inhibitory input to facial motor neurons has been described to be predominantly affected in "synaptic stripping" phenomenon (Lux and Schubert 1975). "Synaptic stripping" in rats is initiated during the first 4 days post-axotomy (Blinzinger and Kreutzberg 1968), a time course which corresponds to the appearance of the contralateral R1 response after onset of facial nerve palsy in man. Displacement of inhibitory synapses from regenerating motor neurons could thus be the mechanism by which a preexistent crossed trigemino.facial reflex pathway may be unmasked during early stages of paresis. The size of contralateral R1 responses was low in the small number of normal subjects displaying this blink reflex component (Table 1). This suggests that the reflex pathway which mediates the contralateral R1 response may have a lower degree of functional synaptic connectivity with the target motor neurons as compared to the pathways which mediate the other blink reflex components. As a result, in facial nerve palsy the contralateral R1 response would be expected to decrease or even disappear. However, as Table 1 shows, this is not the case: size of R1 did not differ significantly from that in normal subjects. Surprisingly, a considerable number of patients had a contralateral R1 response upon stimulation of the normal side, despite the absence of an ipsilaterai R1 component following stimulation of the paretic side (Tables 2 and 3). This finding suggests a reinforcement of transmission in the corresponding crossed trigemino-facial reflex pathway, which in turn would lend further support to the notion that synaptic modifications occur in the motor nucleus during facial nerve regeneration. Preliminary data in a follow-up study of our patients further indicate that the contralateral early blink reflex
may persist, even during advanced stages of recovery from facial palsy. Systematic studies correlating impairment of fine motor movements with the chronic persistence of a contralateral R1 response are under way. Since the interpretation of our electrophysiological data is based on analogies with anatomical experiments in rat, the validity of transposition to man has to be proven. Therefore, post-mortem analysis of human brainstem will be needed for confirmation of our hypothesis. Acknowledgements The authors thank Dr. E. Berg-Dammer, Prolessor A.C. Nacimiento, and Professor K. Poeck for valuable comments on the manuscript. Part of this work was supported by a grant of the Deutsche Forschungsgemeinschaft (SFB 200/134).
References Agostino. R., A. Berardelli, G. Cruccu, G. Pauletti, F. Stocchi and M. Manfredi (1988) Correlation between facial involuntary movements and abnormalities of blink and corneal reflexes in Huntington's chorea. Mov. Disord., 3: 281-289. Blinzinger, K. and G.W. Kreutzberg (1968) Displacement of synaptic terminals from regenerating motoneurons by microglial cells. Z. Zellforsch., 85: 145-157. Bratzlavsky, M. and H. vander Eecken (1977) Altered synaptic organization in facial nucleus following facial nerve regeneration: an electrophysiological study in man. Ann. Neurol., 2: 71-73. Csecsei, G. (1982) Facial reflexes of short latency Electromyogr. Clin. Neurophysiol., 22: 39-44. Dumitru, D., N.E. Walsh and L.D. Porter (1988) Electrophysiologic evaluation of the facial nerve in Bell's palsy Am. J. Phys. Med. Rehabil., 67: 137-144. Esteban, A. and P. Molina-Negro (1986) Primary hemifacial spasm: a neurophysiological study. J. Neurol. Neurosurg. Psychiat., 49: 58-63. Fowler, E.P. (1939) Abnormal movements following injury to facial nerve. JAMA, 113: 1003-1008. Ghonin, M.R. and C. Gavilan (1990) Blink reflex: prognostic value in acute peripheral facial palsy. J. Otorhinolaryngoi. Relat. Spec., 52: 75-79. Graeber, M.B. and G.W. Kreutzberg (1990) Astrocytic reactions accompanying motor neuron regeneration. In: Seil, F.J. (Ed.), Advances in Neural Regeneration Research, Alan R. Liss, New York, pp. 215-224. Heath, J.P., R.E. Cull, I.M. Smith and J.A. Murray (1988) The neurophysiological investigation of Bell's palsy and the predictive value of the blink reflex. Ciin. Otolaryngol., 13: 85-92. Hirakao, M. and M. Shimamura (1977) Neural mechanisms of the corneal blinking reflex in cats. Brain Res., 125: 265-275. Horowitz, S.H. (1987) Hemifacial spasm and facial myokymia: Electrophysiological findings. Muscle Nerve, 10: 422-427. Kimura, J. and L.W. Lyon (1972) Orbicularis oculi reflex in Wallenberg syndrome: alteration of the late reflex by lesions of the spinal tract and nucleus of the trigeminal nerve. J. Neurol. Neurosurg. Psychiat., 35: 228-233. Kimura, J. (1973) Blink reflex as a test for brainstem and higher central nervous system function. In: Desmedt, J.E. (Ed.), New Developments in Electromyography and Clinical Neurophysiology, Vol. 3, Karger, Basel, pp. 682-691. Kugelberg, E. (1952) Facial reflexes. Brain, 76: 385-396. Laskawi, R., W. Damenz, P. Roggenkamper, M. Schroder and J. Brauneis (1990) The trigemino-facial reflex in patients with es-
155 sential blepharospasm. A contribution to the etiology of the disease picture. Laryngorhinootologie, 69: 45-41. Lux, S.D. and P. Schubert (1975) Some aspects of the electroanatomy of dendrites. In: Kreutzberg, G.W. (Ed.), Advances in Neurology, Vol. 12, Raven, New York, pp. 29-44. M011er, A.R. and C.lq. Sen (1990) Recordings from the facial nucleus in the rat: signs of abnormal facial muscle response. Exp. Brain Res., 81: 18-24. M¢ller, A.R. and P.J. Jannetta (1986) Blink reflex in patients with hemifacial spasm. Observations during microvascular decompression operations. J. Neurol. Sci., 72: 171-182. Penders, C.A. and P.J. Deiwaide (1973) Physiologic aproach to the human blink reflex. In: Desmedt, J.E. (Ed.), New Developments in Electromyography and Clinical Neurophysiology, Vol. 3, Karger, Basel, pp. 649-657.
Ramon y Cajal, $. (1909) Histologie du Syst~me Nerveux de rHomme et des Vertebras, Maloine, Paris, pp. 859-888. Soliven, B., J. Meet', A. Uncini, J. Petajan and R. Lovelace (1988) Physiologic and anatomic basis for contralateral R1 in blink reflex. Muscle Nerve, 11: 848-851. Tanaka, T., H. Yu and S.T. Kitai (1971) Trigeminal and spinal inputs to the facial nucleus. Brain Res., 33: 504-508. Trontelj, M.A. and J.V. Trontelj (1978) Reflex arc of the first component of the human blink reflex: a single motoneuron study. J. Neurol. Neurosurg. Psychiat., 41: 538-547. Vails-Sole, J. and E.S. Tolosa (1989) Blink reflex excitability cycle in hemifacial spasm. Neurology, 39: 1061-1066. Wilier, J.C., P. Boulu and M. Bratzlavsky (1984) Electrophysiological evidence for crossed oligosynaptic trigemino-facial connections in normal man. J. Neurol. Neurosurg. Psychiat., 47: 87-90.
148 JNS 03760
Contralateral early blink reflex in patients with facial nerve palsy: indication for synaptic reorganization in the facial nucleus during regeneration W i l h e l m N a c i m i e n t o a, Klaus Podoll a, M a n u e l B. G r a e b e r b,l, R u d o l f T 6 p p e r a, E c k h a r d M/Sbius a, H a r t m u t O s t e r m a n n a, J o h a n n e s N o t h a a n d G e o r g W. K r e u t z b e r g b a Aifried-Krupp-Hospital, Dept. of Neurology with Clinical Neurophysiology, Alfried-Krupp-Str. 21, D-4300 Essen, Germany, and b Max-Planck-lnstitute of Psychiatry, Dept. of Neuromorphology, Am Klopferspitz 18a, D-8033 Martinsried, Germany
(Received 31 July, 1991) (Revised, received 11 December, 1991) (Accepted 16 December, 1991)
Key words: Blink reflex; Facial nerve palsy; Contralateral early response; Regeneration; Facial nucleus; Synaptic organization
Summary
Fifty patients with Bell's palsy and 30 patients with etiologically different symptomatic peripheral facial nerve palsy were studied by means of electrically evoked blink reflexes 1-23 days after onset of paresis. Their results were compared with a normal control group of 30 healthy subjects. In a significant number of patients (64% in Bell's palsy and 53% in symptomatic facial nerve palsy) a contralateral early blink reflex response (R1) could be elicited upon stimulation of the normal side as compared to 13% in the control group. It is suggested that this result may be explained by synaptic reorganization of the facial nucleus leading to functional unmasking of pre-existing crossed trigemino-facial reflex pathways during regeneration. This view is in line with previous experimental data in animals on the time course of structural changes in the facial nucleus after lesioning of the ipsilateral facial nerve.
Introduction
Electrical stimulation of the supraorbital nerve results in two responses: an early component (R1) ipsilateral to the stimulus which is transmitted through an oligosynaptic arc in the ports (Tanaka et al. 1971; Hirakao and Shimamura 1977) and a bilateral late component (R2) which follows a long polysynaptic pathway travelling through the spinal nucleus of the trigeminal nerve (Kugelberg 1952; Kimura 1973; Hirakao and Shimamura 1977). Contralateral R1 responses can be elicited in normal subjects by facilitating maneuvers, such as contraction of the orbicularis oculi muscle or electrical stimuli of the median nerve
I Present address: Children's Hospital and Harvard Medical School, Boston, MA 02115, USA. Correspondence to: Prof. J. Noth, Alfried-Krupp-Hospitai, Dept. of Neurology with, Clinical Neurophysiology, AIfried-Krupp-Str. 21, D-4300 Essen, Germany. Tel.: (201) 434-2528; Fax: (201) 434-2399.
(Wilier et al. 1984; Soliven et al. 1988). This early crossed response is mediated by direct anatomical connections of the trigeminal sensory nucleus and the contralaterai facial nucleus (Ramon y Cajal 1909; Wilier et al. 1984; Soliven et al. 1988). Under standard recording conditions, the synaptic connections producing the contralateral R1 response are irresponsive to supraorbital nerve stimulation in the majority of normal subjects (Penders and Delwaide 1974; Trontelj and Trontelj 1978; Csecsei 1982; Wilier et al. 1984; Soliven et al. 1988). Therefore, the blink reflex may provide insight into synaptic organization and plasticity of the facial nucleus. The latter is of particular interest in pathological cases (Kimura and Lyon 1972; Bratzlavsky and vander Eecken 1977; Esteban and Molina-Negro 1986; Moeller and Jannetta 1986; Horowitz 1987; Agostino et al. 1988; Laskawi et al. 1990). In patients with facial mass contractions following facial nerve palsy, for instance, the appearance of contralateral R1 has been attributed to alterations in the synaptic organization of the facial nucleus (Bratzlavsky and vander Eecken
149 1977). In this study we present evidence that contralateral R1 also occurs in the early phase of facial nerve regeneration in a significant number of patients. Possible clinical relevance of these findings is discussed in the light of accumulating morphological evidence in animals which demonstrate substantial changes in the synaptic organization of the facial nucleus following peripheral lesioning of the ipsilateral facial nerve (Blinzinger and Kreutzberg 1968; Graeber and Kreutzberg 1990). Methods
Patients and subjects All subjects gave their informed consent to participate in the study after having been informed about its objectives and the procedures involved.
Patients with idiopathic peripheral facial nerve paresis (Bell's palsy) The 50 patients (24 males, 26 females) with Bell's palsy had a mean age of 42.3 years with a range from 15 to 81 years. Paresis was on the left side in 23 patients and on the right side in 27 patients. Apart from Bell's palsy, neurological status was normal in all patients except for an unrelated mild hypacusis in 4 cases. In 30 cases cranial compute~i tomography was performed which was normal in all but 3 eldery patients who showed mild cerebral atrophy. Patients were studied between the 1st and 17th day from onset of Bell's palsy.
Patients with symptomatic peripheral facial nerve paresis The 30 patients (17 males, 13 females) with symptomatic peripheral facial paresis due to different etiologies comprised 5 patients with posttraumatic palsy after fractures of the petrous or mastoid portion of the temporal bone, 4 patients with a postoperative facial paresis after operations for cholesteatoma or chronic otitis media, 1 patient with chronic otitis media and marked tympanic effusion, 10 patients with herpes zoster auricularis (Ramsay Hunt syndrome), 1 patient with HIV infection of the nervous system, 1 patient with neuroborreliosis (Lyme disease), 4 patients with Guillain-Barr~ syndrome and 4 patients with diabetic polyneuropathy. Paresis was left-sided in 16 patients and right-sided in 14 patients. The mean age was 53.1 years with a range from 19 to 86 years. Patients were studied between the 2nd and 23rd day from onset of facial palsy.
Control group Thirty healthy subjects (12 men, 18 women), with a mean age of 41.7 years (range 16-83 years), were studied as a control group.
Blink reflex recording All recordings were performed in a quiet room with the temperature kept between 20 and 22°C. Subjects lay supine and were instructed to close their eyes gently. In 10 patients displaying a contralateral R1 response upon stimulation of the normal side, blink reflexes were additionally recorded with open eyes in order to determine the influence of voluntary effort (that is contraction of the eyelids) on this response. The blink reflex was elicited transcutaneously by electrical stimulation of the supraorbital nerve, with the cathode positioned near the supraorbital foramen and the anode 3 cm above it (at least 2.5 cm away from the midline). Using a constant-current unit, the stimuli consisted of rectangular pulses of 0.2 ms duration and 8-12 mA intensity, delivered randomly with interstimulus intervals of 15-60 s to avoid habituation. Stimuli were kept constant during the recording session and were of such intensity that reflex responses were just maximum. Electromyographic (EMG) activity was recorded from the ipsi- and contralateral orbicularis oculi muscle (pars inferior) by surface electrodes set on the midline of the lower lid and bridge of the nose. The EMG signals were fed into amplifiers, bandpass filtered (band-width 20-3000 Hz) and stored on floppy discs for subsequent evaluation. The sweep time was 100 ms with 1024 addresses per channel, the sampling time being 98 ~s. In every subject, 5 responses were obtained from each side. Measurements were made by visual inspection using a cursor in the optical display. Latencies of the blink reflex responses were determined from the mean of the five measurements, and amplitudes were taken as the maximum peak-to-peak values.
Electroneurography Direct motor potentials were obtained in orbicularis oculi muscle with a supramaximal stimulation of the facial nerve at the stylomastoid foramen. Constant-current stimuli consisted of rectangular pulses of 0.2 ms duration and 30-80 mA intensity. The position of recording electrodes was the same as in blink reflex examinations.
Statistical analysis Mann-Whitney U-tests were used for between-group comparison of different variables. Chi-square analysis was performed for between-group comparisons of the frequencies of different nominally scaled variables. Differences were considered to be significant with a P value < 0.05.
150 Results Control group In all subjects of the control group, electrical stimulation of the supraorbital nerve elicited an early R1 response on,the ipsilateral side and a late R2 response on the ±psi- and contralateral side. An additional contralateral* R1 response was observed bilaterally in 13% (4/30) of control subjects. Since no significant differences were observed between blink reflex responses to left and right side stimulation, the mean of both sides was calculated and used for comparison with the patient groups (Table 1).
i t numberof patients
Patients with Bell's palsy
ml numberofpatients
In the group of 50 patients with Bell's palsy, the ipsilateral R1 and R2 responses to stimulation of the normal side and the contralateral R2 response to stimulation of the paretic side were recorded in all patients and showed no significant differences in latency and amplitude compared to controls except for a slight delay of the ipsilateral R2 latency (P < 0.05; Table 1). The ipsilateral R1 and R2 responses to stimulation of the paretic side and the contralateral R2 response to stimulation on the normal side were obtained at a lower rate than in controls (statistically significant for ipsilateral R1 and contralateral R2 with P < 0.001 and P < 0.025 respectively). In those cases where blink reflex responses were preserved on the paretic side, latencies were markedly prolonged and amplitudes
1
1
~
5
5
ContrnlateraX Contralsteral Contralater81
10
15
20
25
10
15 20 25 daysfromonsetof paresis
daysfromonsetof pnreslo
R1 u p o n s t l n u I a t i o n o f t h e not-maX s i 4 e R1 u p o n s t i m u l a t i o n o£ both sides RI absent on both sides
Fig. 1. Rate of appearance of contralateral R1 responses in the time course from onset of Bell's palsy (A) and of symptomatic facial nerve palsy (B).
were reduced as compared to the control group (P < 0.0001; Table 1). A contralateral R1 response, which was seen in 13% of the age matched control subjects, was observed with
TABLE 1 BLINK REFLEX PARAMETERS IN CONTROL SUBJECTS AND PATIENTS WITH IDIOPATHIC AND SYMPTOMATIC FACIAL PALSY Means + SD. Control group
Patients with Bell's palsy
Symptomatic facial palsy
mean of both sides
paretic side stimulation
normal side stimulation
paretic side stimulation
normal side stimulation
Ipsilateral R1 frequency (%) latency (ms) amplitude (/~V)
100 (30/30) 10.8 4- 0.8 551 4- 170
46 (23/50) 12.5 + 1.7 135 + 85
100 (50/50) 10.6 + 1.2 489 + 229
30 (9/30) 14.0 + 3.4 122 ± 95
100 (30/30) 10.9 4- 1.1 538 4- 238
Contralateral R1 frequency (%) latency (ms) amplitude(/~V)
13 (4/30) 11.4 4- 1.1 150 4- 79
12 (6/50) 12.0 4- 0.9 267 +103
64 (32/50) 11.9 ± 1.2 94 4- 56
10 (3/30) 11.2 4- 0.6 167 + 94
53 (16/30) 13.5 4- 6.0 113 4-45
~teral R2 frequency (%) latency (ms) amplitude (/tV)
100 (30/30) 30.5 4- 1.9 552 4- 154
90 (45/50) 36.6 + 5.9 166 + 144
100 (50/50) 32.0 4. 3.4 534 + 163
63 (19/30) 39.2 + 4.5 155 + 80
100 (30/30) 34.1 4- 3.2 509 4- 175
Contralateral R2 frequency (%) latency (ms) amplitude(/~V)
100 (30/30) 30.2 4- 2.5 489 4-135
100 (50/50) 32.0 4- 4.2 476 4-180
84 (42/50) 36.3 4- 4.8 147 ± 87
97 (29/30) 33.4 ± 4.0 504 +224
73 (22/30) 38.4 ± 4.2 136 4- 80
151 TABLE 2 COMPARISON OF BLINK REFLEX PARAMETERS 1N PATIENTS WITH BELL'S PALSY DISPLAYING AND LACKING CONTRALATERAL R1 RESPONSES
Mean + SD.
Stimulation of paretic side Ipsilateral RI frequency (%) latency (ms) amplitude (/tV) Ipsilateral R2 frequency (%) latency (ms) amplitude (pV) Stimulation of normal side Contralateral R2 frequency (%) latency (ms) amplitude (/zV) Facial nerve motor electroneurography Amplitude reduction (%)
Patients with contralateral R1 responses (n = 32)
Patients lacking contralateral R1 responses (n = 18)
Comparison P
59 (19/32) 12.3+ 1.5 132 + 74
22 (4/18) 13.4+ 2.0 149 + 24
< 0.025 NS NS
100 (32/32) 35.7+ 5.0 178 +158
72 (13/18) 38.9+ 7.0 136 + 96
< 0.01 NS NS
94 (30/32) 35.2 + 4.3 147 + 80
67 (12/18) 38.9 + 5.1 145 +101
< 0.025 < 0.05 NS
69 + 21
60 + 26
NS
NS: no significant difference.
a significantly (P < 0.001) higher frequency of 64% (32/50) in the patient group. In 52% (26/50) of these cases, contralateral R1 was recorded unilaterally following stimulation on the normal side and in 12% (6/50) bilaterally following stimulation of either side. A contralateral R1 response upon stimulation of the normal side was less likely to occur in cases in which severe alterations of other blink reflex components were present on the paretic side. Thus, in the subgroup
LEFT
RZGHT
R1
of the 18 patients lacking contralaterai R1 responses on the paretic side, opposed to the 32 patients displaying contralaterai R1, ipsilaterai R1 and R2 responses to paretic side stimulation and the contralateral R2 response to stimulation of the unaffected side were abolished at a significantly higher rate (P < 0.025). Additionally, the contralateral late R2 component on the paretic side was also significantly prolonged in these cases (P < 0.05; Table 2).
R2
; Fig. 2. Blink reflexes (5 superimposed records) of a 35-year-old woman, 3 days from onset of Bell's palsy on the left side: electrical stimulation of ipsilateral (A) and contralaterai (B) supraorbital skin. Eyes were gently closed during examination. A contralateral RI response (arrow) could be elicited upon stimulation of the normal side. Calibration: horizontal: 10 ms; vertical: 100/zV.
152 TABLE 3 COMPARISON OF BLINK REFLEX PARAMETERS (MEAN + SD) IN PATIENTS WITH SYMPTOMATIC PERIPHERAL FACIAL PALSY DISPLAYING AND LACKING CONTRALATERAL R1 RESPONSES
Patients with contralateral RI responses (n = 16) Stimulation of paretic side lpsilaterai RI frequency (%) latency (ms) amplitude (/tV)
Ipsilateral R2 frequency (%) latency (ms) amplitude (~V)
Stimulation of normal side Contralateral R2 frequency (%) latency (ms) amplitude (/~V)
Patients lacking contralateral RI
Comparison P
responses (n = 14)
38 (6/16) 14.6 + 3.7 83 +37
21 (3/14) 12.9 + 2.4 200 +122
NS NS NS
88 (14/16) 39.8 + 5.0 137 +59
36 (5/14) 37.4 + 2.4 205 +108
NS NS
94 (15/16) 39.0 _+ 4.7 118 +55
50 (7/14) 37.1 _+ 2.7 174 +107
< 0.01 NS NS
< 0.01
Facial nerve motor electroneurography Amplitude reduction (%)
43
+19
49 + 20
NS
NS: no significant difference.
There was no significant difference with respect to sex, mean age and mean duration of the palsy (5.8 vs. 7.7 days, respectively; z = - 1.35, n.s.) between patients lacking contralateral R1 responses and patients exhibiting this blink reflex component. The relationship between the occurrence of contralateral R1 responses and the duration of facial nerve palsy, as assessed by cross-sectional data analysis, is graphically
RIGHT
A
illustrated in Fig. 1A. Disregarding cases with contralateral R1 on both sides, which could be due to physiological facilitation by voluntary contraction of the eyelids, a contralateral R1 response on the paretic side was found at a substantial rate beginning with the second day after onset of facial palsy without showing further significant changes of rate of appearance in the subsequent time course (Fig. 1A). A typical example of LEFT
I
Fig. 3. Blink reflexes (5 superimposed records) of a 56-year-old woman, 17 days from onset of Bell's palsy on the left side: electrical stimulation of ipsilateral (A) and contralateral (B) supraorbital skin. Eyes were open during examination. A contralateral R1 response (arrow) could be elicited upon stimulation of the normal side, in spite of advanced recovery from paresis. Calibration: horizontal: 10 ms; vertical: 200/LV (right) and 100 ~V (left).
153 a contralateral R1 response upon stimulation on the normal side in a patient with Bell's palsy is shown in Fig. 2. The comparison of recording with open and closed eyes in 10 patients showed that the contralateral R1 response on the affected side is not abolished during recording with open eyes. Blink reflex responses recorded with open eyes in a patient with Bell's palsy are shown in Fig. 3.
Patients with symptomatic peripheral facial nerve palsy In the group of 30 patients with symptomatic peripheral facial nerve palsy, the results on blink reflex examination were essentially the same as those encountered in Belrs palsy (Table 1). A contralateral R1 response was found in 53% (16/30) of these patients: in 43% (13/30) to stimulation on the normal side and in 10% (3/30) to stimulation on either side. In the 14 patients lacking contralateral R1 responses, the other blink reflex components on the paretic side were absent at a significantly higher rate than in the 16 patients displaying contralaterai R1 (P < 0.01; Table 3). Thus, as in the group with Bell's palsy there appeared to be an inverse relationship between the presence of contralateral R1 responses and the severity of blink reflex alterations on the paretic side. Contralateral R1 responses were observed as early as 3 days after onset of paresis (Fig. 1B).
Discussion Numerous studies indicate that electrophysiological examination of the blink reflex is of great value in predicting clinical outcome of patients with peripheral facial nerve palsy (Dumitru et al. 1988; Heath et al. 1988; Ghonin and Gavilan 1990). Examination of the blink reflex is especially useful during early stages of facial nerve regeneration, where the persistence or early return of an initially absent ipsilateral R1 component may suggest a qualitatively satisfactory recovery of facial nerve function (Heath et al. 1988). In turn, if this response is absent during the first 2 weeks after onset of facial nerve palsy, a rather unfavorable prognosis is given (Heath et ai. 1988). The results of the present study demonstrate that in a significant number of patients with peripheral facial nerve palsy an additional crossed early blink reflex can be elicited upon stimulation of the normal side during the early phase of regeneration. Interestingly, the occurrence of this phenomenon appears to be unrelated to the etiology of facial nerve paresis. This may suggest that central rather than peripheral mechanisms are involved in the generation of this response, a view which is in line with previous studies indicating that a contralateral R1 response may be mediated by a crossed trigemino-facial
reflex pathway (Wilier et al. 1984; Soliven et al. 1988). Yet, it should be noted that a contralaterai R1 response can be produced by normal subjects on a central basis via corticonuclear facilitation due to contraction of the orbicularis oculi muscle (Wilier et al. 1984; Soliven et al. 1988) In the present study all subjects were instructed to close their eyes gently. The contralateral R1 response in patients could be due to a stronger voluntary effort to contract the eyelids on the paretic side as compared to the normal side. However, the fact that the contralateral R1 response could be elicited on the paretic side during recording with both open and gently closed eyes, as proved in 10 patients, renders this possibility unlikely. Furthermore, the similar rate of appearance of contralateral R1 responses upon stimulation on both sides in patients (12% in Bell's palsy, 10% in symptomatic facial nerve palsy) and in normals (13%) argues against such a bias. Another possible methodological pitfall has to be considered: Moving the stimulation electrode toward the midline may also evoke a contralateral R1 response which can be abolished by blocking the contralateral supraorbital nerve (Soliven et al. 1988), confirming its peripheral elicitation. In order to prevent such mechanisms of contralaterai R1 generation, the stimulation electrode was positioned at a distance of at least 2.5 cm away from the midline. Crossed early blink reflex responses occurred on the affected side in the majority of both patients with Bell's palsy and patients with symptomatic facial nerve paresis (Table 1). In patients lacking a contralateral R1 response on the paretic side the other blink reflexes recorded on the affected side were often also absent (see Results). This can be explained by the severity of facial nerve lesion precluding any blink reflex response on the paretic side. It seems therefore possible that a crossed trigemino-facial reflex pathway, which is normally present but apparently ineffective in most normal subjects under stable routine recording conditions (Wilier et al. 1984; Soliven et al. 1988), may become functional during the regeneration process of the peripherally injured facial nerve. Bratzlavsky and vander Eecken (1977) have previously reported on the occurrence of a crossed early blink reflex in patients with postfacial palsy mass contraction. The authors suggested that, in addition to peripheral causes like misguidance of axons, aberrations at the nuclear level may contribute to faulty restoration of post-lesion facial motility. Furthermore, there is experimental evidence suggesting that disturbance of complex fine motor movements as occurring in hemifacial spasm may be caused by a state of increased excitability of facial motor neurons (Esteban and Molina-Negro 1986; Horowitz 1987; Vails-Sole and Tolosa 1989; Moeller and Sen 1990). Clinically, impairment of fine motor functions is frequently observed as
154 a symptom of incomplete recovery following peripheral facial palsy (Fowler 1939; Bratzlavsky and vander Eecken 1977). Thus, the question arises whether functional and/or structural changes at the nuclear level may accompany the regenerative process of the facial nerve and, if so, whether such alterations may persist even after peripheral reinnervation of the musculature has been achieved. Recent studies employing axotomy of the rat facial nerve have in fact demonstrated that there are massive changes in the synaptic organization of the regenerating motor nucleus. Glial cells, i.e. microglia and astrocytes, appear to be directly involved in this process. While microglia engage in the early detachment and displacement of afferent axonal endings from neuronal surface membranes ("synaptic stripping"; Blinzinger and Kreutzberg 1968), astrocytes appear to maintain this state of synaptic deafferentation of the regenerating motor neuron for longer periods, i.e. months (Graeber and Kreutzberg 1990). Our electrophysiological data would be compatible with these anatomical alterations, particularly since inhibitory input to facial motor neurons has been described to be predominantly affected in "synaptic stripping" phenomenon (Lux and Schubert 1975). "Synaptic stripping" in rats is initiated during the first 4 days post-axotomy (Blinzinger and Kreutzberg 1968), a time course which corresponds to the appearance of the contralateral R1 response after onset of facial nerve palsy in man. Displacement of inhibitory synapses from regenerating motor neurons could thus be the mechanism by which a preexistent crossed trigemino.facial reflex pathway may be unmasked during early stages of paresis. The size of contralateral R1 responses was low in the small number of normal subjects displaying this blink reflex component (Table 1). This suggests that the reflex pathway which mediates the contralateral R1 response may have a lower degree of functional synaptic connectivity with the target motor neurons as compared to the pathways which mediate the other blink reflex components. As a result, in facial nerve palsy the contralateral R1 response would be expected to decrease or even disappear. However, as Table 1 shows, this is not the case: size of R1 did not differ significantly from that in normal subjects. Surprisingly, a considerable number of patients had a contralateral R1 response upon stimulation of the normal side, despite the absence of an ipsilaterai R1 component following stimulation of the paretic side (Tables 2 and 3). This finding suggests a reinforcement of transmission in the corresponding crossed trigemino-facial reflex pathway, which in turn would lend further support to the notion that synaptic modifications occur in the motor nucleus during facial nerve regeneration. Preliminary data in a follow-up study of our patients further indicate that the contralateral early blink reflex
may persist, even during advanced stages of recovery from facial palsy. Systematic studies correlating impairment of fine motor movements with the chronic persistence of a contralateral R1 response are under way. Since the interpretation of our electrophysiological data is based on analogies with anatomical experiments in rat, the validity of transposition to man has to be proven. Therefore, post-mortem analysis of human brainstem will be needed for confirmation of our hypothesis. Acknowledgements The authors thank Dr. E. Berg-Dammer, Prolessor A.C. Nacimiento, and Professor K. Poeck for valuable comments on the manuscript. Part of this work was supported by a grant of the Deutsche Forschungsgemeinschaft (SFB 200/134).
References Agostino. R., A. Berardelli, G. Cruccu, G. Pauletti, F. Stocchi and M. Manfredi (1988) Correlation between facial involuntary movements and abnormalities of blink and corneal reflexes in Huntington's chorea. Mov. Disord., 3: 281-289. Blinzinger, K. and G.W. Kreutzberg (1968) Displacement of synaptic terminals from regenerating motoneurons by microglial cells. Z. Zellforsch., 85: 145-157. Bratzlavsky, M. and H. vander Eecken (1977) Altered synaptic organization in facial nucleus following facial nerve regeneration: an electrophysiological study in man. Ann. Neurol., 2: 71-73. Csecsei, G. (1982) Facial reflexes of short latency Electromyogr. Clin. Neurophysiol., 22: 39-44. Dumitru, D., N.E. Walsh and L.D. Porter (1988) Electrophysiologic evaluation of the facial nerve in Bell's palsy Am. J. Phys. Med. Rehabil., 67: 137-144. Esteban, A. and P. Molina-Negro (1986) Primary hemifacial spasm: a neurophysiological study. J. Neurol. Neurosurg. Psychiat., 49: 58-63. Fowler, E.P. (1939) Abnormal movements following injury to facial nerve. JAMA, 113: 1003-1008. Ghonin, M.R. and C. Gavilan (1990) Blink reflex: prognostic value in acute peripheral facial palsy. J. Otorhinolaryngoi. Relat. Spec., 52: 75-79. Graeber, M.B. and G.W. Kreutzberg (1990) Astrocytic reactions accompanying motor neuron regeneration. In: Seil, F.J. (Ed.), Advances in Neural Regeneration Research, Alan R. Liss, New York, pp. 215-224. Heath, J.P., R.E. Cull, I.M. Smith and J.A. Murray (1988) The neurophysiological investigation of Bell's palsy and the predictive value of the blink reflex. Ciin. Otolaryngol., 13: 85-92. Hirakao, M. and M. Shimamura (1977) Neural mechanisms of the corneal blinking reflex in cats. Brain Res., 125: 265-275. Horowitz, S.H. (1987) Hemifacial spasm and facial myokymia: Electrophysiological findings. Muscle Nerve, 10: 422-427. Kimura, J. and L.W. Lyon (1972) Orbicularis oculi reflex in Wallenberg syndrome: alteration of the late reflex by lesions of the spinal tract and nucleus of the trigeminal nerve. J. Neurol. Neurosurg. Psychiat., 35: 228-233. Kimura, J. (1973) Blink reflex as a test for brainstem and higher central nervous system function. In: Desmedt, J.E. (Ed.), New Developments in Electromyography and Clinical Neurophysiology, Vol. 3, Karger, Basel, pp. 682-691. Kugelberg, E. (1952) Facial reflexes. Brain, 76: 385-396. Laskawi, R., W. Damenz, P. Roggenkamper, M. Schroder and J. Brauneis (1990) The trigemino-facial reflex in patients with es-
155 sential blepharospasm. A contribution to the etiology of the disease picture. Laryngorhinootologie, 69: 45-41. Lux, S.D. and P. Schubert (1975) Some aspects of the electroanatomy of dendrites. In: Kreutzberg, G.W. (Ed.), Advances in Neurology, Vol. 12, Raven, New York, pp. 29-44. M011er, A.R. and C.lq. Sen (1990) Recordings from the facial nucleus in the rat: signs of abnormal facial muscle response. Exp. Brain Res., 81: 18-24. M¢ller, A.R. and P.J. Jannetta (1986) Blink reflex in patients with hemifacial spasm. Observations during microvascular decompression operations. J. Neurol. Sci., 72: 171-182. Penders, C.A. and P.J. Deiwaide (1973) Physiologic aproach to the human blink reflex. In: Desmedt, J.E. (Ed.), New Developments in Electromyography and Clinical Neurophysiology, Vol. 3, Karger, Basel, pp. 649-657.
Ramon y Cajal, $. (1909) Histologie du Syst~me Nerveux de rHomme et des Vertebras, Maloine, Paris, pp. 859-888. Soliven, B., J. Meet', A. Uncini, J. Petajan and R. Lovelace (1988) Physiologic and anatomic basis for contralateral R1 in blink reflex. Muscle Nerve, 11: 848-851. Tanaka, T., H. Yu and S.T. Kitai (1971) Trigeminal and spinal inputs to the facial nucleus. Brain Res., 33: 504-508. Trontelj, M.A. and J.V. Trontelj (1978) Reflex arc of the first component of the human blink reflex: a single motoneuron study. J. Neurol. Neurosurg. Psychiat., 41: 538-547. Vails-Sole, J. and E.S. Tolosa (1989) Blink reflex excitability cycle in hemifacial spasm. Neurology, 39: 1061-1066. Wilier, J.C., P. Boulu and M. Bratzlavsky (1984) Electrophysiological evidence for crossed oligosynaptic trigemino-facial connections in normal man. J. Neurol. Neurosurg. Psychiat., 47: 87-90.
