Masseter inhibitory reflex threshold: a novel technique for electrophysiological investigation of trigeminal nerve lesions

JOURNAL

OF

THE

NEUROLOGICAL SCIENCES ELSEVIER

Sciences 134 (1995) 1977202

Journal of the Neurological

Short report

Masseter inhibitory reflex threshold: a novel technique for electrophysiological investigation of trigeminal nerve lesions S. HaBfeld

a, V. Schuchardt

a Department of Oromaxillofacial b Department of Neurology, Received

30 December

b, H. Geisler

b, H.-M. Meinck

b,*

Surgery, University of Heidelberg, Heidelberg, Germany University of Heidelberg, Heidelberg, Germany 1994; revised

7 June 1995; accepted

21 June 1995

Abstract The masseter inhibitory reflex was investigated in 60 healthy volunteers, in 38 patients with trigeminalnerve branchlesionsin the majority due to dental surgery, and in 9 patients with facial sensory loss and weakness caused by brain hemisphere lesions. The reflex threshold (TR) was almost symmetric both in normal subjects and in the patients with hemisphere lesions. In peripheral trigeminal hypaesthesia, elevation of TR on the lesioned side proved the most sensitive electrophysiological parameter. There was, moreover, some correspondence between the degree of sensory loss in hypaesthesic skin areas and elevation of TR, and recovery from the lesion was associated with TR normalization. Supratentorial lesions, in contrast, may influence the reflex pattern rather than reflex excitability. Keywords:

Masseter

inhibitory

reflex;

Jaw opening

reflex;

Cutaneo-muscular

1. Introduction Disturbance of facial sensibility is a frequent challenge to the neurologist, and is sometimes difficult to assess by

means of electrodiagnostic techniques such as trigeminal nerve evoked cortical potentials, or orbicularis oculi blink reflexes. We therefore investigated the diagnostic value of the masseter inhibitory

reflex

(MIR).

This cutaneo-muscu-

lar reflex is elicited by electrical or mechanical stimulation of the lower part of the face on either side. It consists transient suppression of the electromyographic (EMG)

of a ac-

tivity of the masseter(as well as temporalis) muscles on both sidesfor about 80 ms (Hoffmann and Tiinnies, 1948). Afferent volleys are conveyed in A beta fibres of the 2nd and 3rd trigeminal nerve branches, and are relayed via two paths on different levels through the pons to the motor nuclei of the jaw-closing muscleswhere they causebiphasic inhibition of the ongoing EMG. In patients with complete trigeminal sensory root lesions, MIR fails to be elicited from the lesioned side (Ongerboer de Visser and Goor, 1976); interruption of the central paths in the pons leads to characteristic abnormalities of the reflex pattern (Ongerboer de Visser et al., 1990). * Corresponding author. At: Neuenheimer Feld 400, D-69120 567507; Fax: (06221) 565348.

Neurologische Heidelberg,

0022-510X/95/$09.50 0 1995 Elsevier SSDI 0022-510X(95)00224-3

Science

Universitatsklinik, Im Germany. Tel.: (06221)

B.V. All rights reserved

reflex;

Exteroceptive

suppression;

Trigeminal

nerve;

Sensory

disturbance

Experiments with increasing stimulus strength and with xylocaine block (Godaux and Desmedt, 1975) suggestthat the onset latency of MIR might not be a sensitive indicator of a nerve lesion. Moreover, the latency of the individual reflex responsesinevitably jitters with fluctuations of the pre-innervation. We therefore investigated the reflex threshold (TR) as the target parameter. Some preliminary data have been published elsewhere(Hal3feld and Meinck, 1992; H&feld and Meinck, 1993).

2. Subjects

and Methods

Investigations were performed in a total of 47 patients aged between 16 and 68 years. Sixty healthy volunteers (31 female, 29 male) aged between 23 and 82 years served as controls. Informed consent was obtained from each subject. Among the patient group, 38 were admitted because of trigeminal branch lesions, most often in the course of dental surgery 1 week to 4 years prior to admission, and 9 were hospitalized with facial weakness and sensory loss after large hemispherestrokes (n = 8) or tumours (n = l), confirmed by magnetic resonanceimaging. MIR was elicited by short (0.1 ms) electrical square wave shocks delivered via a standard bipolar stimulator electrode (Medelec SBS 523050) at intervals pseudo-ran-

198

S. Hapfeld

A

et al. /Journal

of the Neurological

Normal

B

Sciences 134 (1995)

Peripheral

Nerve

197-202

Lesion

h4r IO.2 mV MI

0

1OOms

C

0

Hemispheral

Stim r lower

100ms

Lesion

lip

Stim I lower

lip h4r

25

0.2 mV

IO.2 mV MI

0

100 ms

0

100ms

0

100 ms

Fig. 1. MIR recording specimens from a normal subject (A), a patient with trigeminal nerve branch lesion (B), and a stroke patient (0. A: MIR in the right (r) and left (1) masseter muscles (M) of a normal subject elicited by stimulation of the lower lip with increasing stimulus strengths (from the top to the bottom; given in mA). Vertical and horizontal calibrations are valid for all registrations. Stimulus applied at 0 ms. B: abnormal MIR threshold in a patient with a small (diameter 6 cm) hypaesthesic and neuralgic skin area on the left cheek 2 years after fracture of the zygomatic bone. Labelling as in A. C: abnormal MIR pattern in a patient with large right subcortical infarction. Stimulus intensity on both sides was 42 mA ( = 3.0 X TR). Stimulation of the lower lip on the paretic and hypaesthetic left (1) side evokes an abnormally short S2 component. Note asymmetric innervation due to central weakness. Labelling as in A.

domly varied between 5 and 10 s. Stimuli were applied to the upper and lower lips within the areas supplied by the 2nd and 3rd trigeminal nerve branches on either side, respectively, as well as to the skin over the supra- and infraorbital grooves. In the patients with brain lesions, only the lower lips were stimulated. The EMG was recorded simultaneously from the bilateral masseter or temporalis muscles, or both, with surface electrodes placed 2-3 cm apart over the muscle bellies. Filter settings were 70-2000 Hz (-3 dB). Subjects were instructed to forcibly close their jaws for some seconds during which a single stimulus was applied. After a 5-10 s pause, this procedure was repeated until four identical responses were elicited at a constant stimulus strength. TR was defined as the lowest stimulus intensity by which complete EMG suppression in the time domain of MIR (lo-90 ms) was reproducibly evoked. TR was deterTable 1 Masseter inhibitory

reflex

normative

data. n = 40; Q , 23; 6, 17; 23-82 Infraorbital

Detection threshold (TD) (mA) Side-to-side difference TD (mA) Branch-to-branch difference TD (mA) Reflex threshold (TR) (mA) Side-to-side difference TR (mA) Branch-to-branch difference TR (mA) Onset latency Sl (ms) Duration Sl (ms)

4.52 + 0.87 0.33 f 0.47 0.55 f 0.64 14.50 * 4.40 1.15 f 1.59 2.10+ 1.91 13.99 f 0.95 11.10+1.68

mined separately for each stimulus site by randomly in- or decreasing stimulus strength in l-2 mA steps. With the same procedure the stimulus strength was determined which was required to evoke a just detectable sensation under the cathode (detection threshold, TD). Thus TD, in contrast to TR, is based on the patients subjective experience and on his attentive cooperation, and might reflect cutaneous sensibility of the area stimulated (Jelasic, 1983; Laitinen and Eriksson, 1985).

3. Results

3.1. MIR in normal subjects In all normal subjects, unilateral stimulation of the upper and lower lips as well as over the infraorbital groove

years. Upper

lip

4.87 + 0.82 0.38 f 0.49 15.21+ 3.27 1.05 f 1.20 14.47 f 1.33 10.73 f 1.33

Lower

lip

4.71+ 0.81 0.35 + 0.58 0.56kO.61 13.72k3.10 0.95 + 0.88 2.35 + 1.79 14.87 + 1.63 1.71+ 1.83

S. Hapfeld

et al. /Journal

of the Neurological

evoked a bilateral MIR. Supraorbital groove stimulation, even if performed with even tolerable strength, evoked MIRs in only 10 of 60 normal subjects and will therefore not be considered further in this paper. With stepwise increasing stimulus strength, two EMG suppression phases were evoked with onset latencies around 12 (Sl) and 45 ms (S2), respectively (Fig. 1A). In most (25 of 40) subjects investigated, Sl at all stimulus sites had a somewhat lower threshold than S2. As differential reflex thresholds for Sl and S2 were a frequent finding, any

Sciences 134 (199.5) 197-202

consistent EMG suppression within the time domain of lo-90 ms was accepted as equivalent for determination of TR. The specimen recording (Fig. 1A) shows some attenuation but no suppression of the ongoing EMG activity at 17 mA, and at 18 mA two clear-cut EMG suppressions at about 22 (Sl) and 65 ms (S2), respectively. So TR in the present case was defined as 18 mA. Although both TD and TR scattered considerably (TD: 3-8 mA, TR: 8-27 mA; Fig. 2A, B), their symmetry within one individual was a striking feature. Mean side-

B ‘“1

. l l .

l

A+* i:i .*+ . . . l l

l l

l * .* l .+*

:+::**+ w?“::

10

0

detection

stimulus

20 threshold

0

r

strength

40 threshold

(mA)

60

60

1

r (mA)

B ‘am ?! c, E120 g! Q 10

$20 g, 0

+

z E40

40

0

20 reflex

?EW 3 20 2

. ..+

OV

20 (mA)

strength

. ::

:

l *.*...*

.g

+ .

++‘::e$.

10 stimulus

199

20 detection

40

60

threshold

60 (mA)

0 100

0

10

20 weeks

30

41

Fig. 2. A: correlation of TD- (rectangles) and TR-values (crosses) on the right (horizontal scale) and left (vertical scale) sides. Pooled data from all stimulus sites in 40 normal subjects. B: correlation of TD-(rectangles) and TR (crosses) values with lower (horizontal scale) versus upper lip (vertical scale), and with upper lip (horizontal scale) versus infraorbital groove (vertical scale) stimulation. Pooled data from both sides in 40 normal subjects. C, D: correlation of TD (C) and TR (D) on the right (r) and left (1) sides in patients with hyp- and paraesthesia due to trigeminal nerve branch lesions. Hexagon area indicates normal range. Please notice condensed scale in D. E: correlation of TD (horizontal) and TR (vertical) in 40 normal subjects (open triangles; y = - 8.83864 + 3.1434 X R2 = 0.557) and in 36 patients with trigeminal nerve branch lesions (filled rectangles; y = 16.905 + 1.8525 X R2 = 0.339). F: recovery of abnormal side-to-side differences for TR in 19 patients with 21 trigeminal nerve branch lesions. Each line represents the side-to-side difference (D) of TR in one patient over time of follow-up.

200

S. Hapfeld

et al. /Journal

of the Neurological

to-side differences ranged below 0.4 mA for TD, and around 1 mA for TR (Table 1, Fig. 2A). Mean differences for both TD and TR between upper and lower lips and between upper lip and infraorbital groove stimulation on one side (“branch-to-branch difference”) were only slightly greater (below 0.6 mA for TD, and 2.4 mA for TR; Fig. 2B1, allowing to calculate the 99.9% confidence limits for side-to-side differences of TR as 6 mA, and for branch-to-branch differences as 8 mA (Table 1). Moreover, a linear relationship was found between TD and TR (Fig. 2E, open triangles). The impedance of the stimulator electrode ranged between 2.8 and 7.2 k0. In 10 subjects, a stepwise increase of the electrode impedance by means of an intercalated resistor resulted in a successive increase of TR in all subjects with a mean TR elevation of 2 mA with additional 10 kR. Both onset latency and duration of Sl were found almost constant with increasing stimulus strength. Mean onset latencies with different stimulus sites ranged between 13.99 and 14.87 ms, and mean durations of Sl between 10.73 ms and 11.71 ms (Table 11. The 99.9% confidence limit for a normal Sl latency was calculated as 19.8 ms on the basis of data obtained from lower lip stimulation. Pilot recordings from the temporalis muscles in 15, and from the buccinator and orbicularis oris muscles in 7 normal subjects revealed almost identical reflex patterns and TRs in both temporalis and masseter, but absence of responses in the VIIth cranial nerve muscles. 3.2. MIR in subjects with trigeminal

nerve branch lesions

In all patients investigated, the procedure for elicitation and registration of MIR was tolerated well, and recordings were free of artifacts. Moreover, stimulation of unaffected skin area in this group elicited MIRs normal with respect to pattern and TR. 3.2.1. Patients with anaesthesia Two patients had anaesthesia in the cutaneous supply of the mental branch due to surgical transection of the inferior alveolar nerve. MIR was to be elicited from the anaesthetic zone in none, even with 100 mA stimulation. Six months after transplant of a sural nerve autograft, one patient reported recovery of sensibility over the right chin and lower lip, and MIR was to be elicited from the latter with abnormal TR (46 mA). 3.2.2. Patients with hypaesthesia A total of 32 patients reported hypaesthesia in the supplies of the Vth nerve lower branches. They all had increased values for both TD and TR in the hypaesthesic areas, their side-to-side differences for TR ranging from 9 to 66 mA (Fig. lB, 2C-D). However, the average TR/TD ratio in the hypaesthesic areas was 2.5 and thus only slightly lower than with normal skin sensibility (3.1; Fig.

Sciences 134 (1995)

197-202

2E). In one patient with traumatic lesions of both inferior alveolar nerves, the branch-to-branch differences for TR between the upper and lower lips were 59 mA (right) and 15 mA (left). Delayed onset latencies of MIR were observed in only 3 patients (4 nerves) and ranged between 20 and 27 ms, with side-to-side differences for TR between 9 and 21 mA. Four patients were investigated with unilateral hypaesthesia of the tongue. In these cases, mouth props were inserted on both sides, and patients were instructed to put out their tongue whilst forcibly biting onto the mouth props. Electrical stimulation of the anterolateral surface of the tongue evoked MIRs identical with those evoked from the facial skin, and with similar figures for TD and TR. On the lesioned side, both TD and TR were elevated resulting in side-to-side differences of 3-19 mA (TD) and 5-34 mA (TR), respectively. 3.2.3. Patients with paraesthesia Five patients presented with paraesthesia associated with minor hypaesthesia as sequelae of otitis (n = 2), trauma (n = 2), or sinusitis (n = 11, respectively. In these patients, the absolute values for both TD and TR as well as their side-to-side differences were within normal limits. 3.2.4. Follow-up investigations In 19 patients (21 nerves>, a total of 30 follow-up investigations was performed over 8-40 weeks. Control recordings on their normal sides yielded a mean intraindividual inter-test difference for TR of 1.7 k 2.1 mA (k SD). In most cases the side-to-side differences for TR tended to normalization corresponding to recovery of sensory functions (Fig. 2F). 3.3. MIR in subjects with brain hemisphere lesions In patients with hemisphere lesions, MIR was investigated only with lower lip stimulation. TR on the paretic and hypaesthesic sides was 15.6 k 6.08 mA (*SD> as compared to 18.0 + 5.38 mA on the intact sides, the mean side-to-side difference being 2.0 + 2.76 mA. In all patients, there was some attenuation of EMG activity on their paretic side. In 1 case, asymmetric innervation was associated with a shortening of the S2 component (Fig. 10. In all patients of this group, and on both sides, onset latencies for Sl were within normal limits.

4. Discussion Our data show that the threshold of MIR (TR) is a sensitive electrophysiological parameter in the analysis of facial sensory loss due to mechanical nerve lesions. TR in normal subjects is almost symmetrical (Fig. 2A, B), but is elevated in hypaesthesic skin areas (Fig. 2C, D). There is, moreover, some correspondence between the increase of

S. Hapfeld

et al. /Journal

of the Neurological

TR and the degree of peripheral sensory disturbance as determined by both clinical grading and the treshold for detection of electrical impulses (TD; Fig. 2E). Recovery from trigeminal nerve branch lesions is associated with normalization of the TR asymmetry (Fig. 2F). Pilot studies suggest that the technique described can be applied also to the neurophysiological analysis of sensory disturbances of the tongue. Weakness and sensory loss due to hemispheral lesions, however, may influence the reflex pattern rather than MIR excitability (Fig. lC>. The technique might be relevant not only for diagnostic, but also for therapeutic questions: Surgical revision and reconstruction is more and more considered if complete sensory loss after a mechanical lesion persists over more than three months, with best results in cases with short intervals between trauma and reconstruction (Gregg, 1990a, b). The key parameter TR was defined as the lowest stimulus intensity by which complete suppression of the ongoing EMG activity is reproducibly evoked. One might argue that both “lowest intensity” and “complete suppression” are not very well defined. With the technique described, however, the EMG is suppressed to the baseline level by a stimulus intensity 1-3 mA higher than the one evoking a just visible EMG attenuation (Fig. 1A; see also figure 2 in Godaux and Desmedt (1975)). This suggests that the error induced by such lack of precision is reasonably small. Another source of inaccuracy is the poor definition of the inhibitory reflex component-S1 or S2-used for determination of TR. However, Ongerboer de Visser et al. (1990) have shown that the primary afferents responsible for Sl and S2 are identical. This means that sufficient excitation of these afferents may be indicated by the occurrence of either Sl or S2, or both. Delayed reflex responses were seen only in a minority of patients which suggests that the onset latency of MIR is not a sensitive parameter for detection of a traumatic trigeminal nerve branch lesion. This corresponds to acute pressure palsies of other peripheral nerves which in most instances cause only mildly reduced nerve conduction velocities often confined, moreover, to the compressed nerve segment (Gilliatt and Thomas, 1970; Trojaborg, 1970). It remains, however, to be investigated as to whether inflammatory or immune-mediated trigeminal nerve disorders are associated with a delayed reflex onset. The procedure was tolerated well by all patients and subjects, and proved to be not affected by stimulus artifacts, incomplete relaxation, or spread of voluntary activity from neighbouring muscles. Methodological problems may arise from disturbances of jaw closure such as malocclusion or biting tremor. If, however, one masseter muscle is improperly activated, the other one most often is not and may sufficiently serve as indicator. Another frequent problem may arise from the electrode used for stimulation and its position within the area of disturbed sensibility: Although not explicitly investigated, the size of the stimulator electrode, as well as the position of either cathode or

Sciences

134 (1995)

197-202

201

anode outside a given hypaesthesic area, can be expected to influence the parameters TD and TR. It should be reminded that not the trigeminal nerve branches, but skin areas supplied by the latter were stimulated. Major bias of TR due to asymmetric impedance of the skin under the stimulator electrode can be safely excluded. Correlation of TR with TD persists after trigeminal branch lesions, the correlation coefficients corresponding to those of the normal group (Fig. 2E). Such similarity suggests tight input-output relationship in this cutaneomuscular reflex, provided intact central control of the reflex pathways. The mild lowering of the correlation coefficient TR/TD suggests central adaptation of the reflex relays to afferent disturbance. Altered central processing of afferent impulses obviously accounts for the shortening of S2 after contralateral hemisphere stroke (Fig. lC>. This type of reflex pattern abnormality is not seen in patients with brain stem lesions (Ongerboer de Visser et al., 1990). It might reflect dis-facilitation of the reflex relays on the disturbed side, possibly related to attenuation of the R2 component of the orbicularis oculi blink reflex after contralateral rostra1 lesions (Fisher et al., 1979; Kimura et al., 1985). The physiological mechanisms underlying dis-facilitation of the ponto-medullary reflex paths are not yet fully understood: Shortening or even complete loss of the S2 component of MIR was described in patients having tension-type (Schoenen et al., 1988) or posttraumatic headaches (Keidel et al., 1994). This suggests that these polysynaptic reflex paths are subject to powerful functional influences. Acknowledgements The authors are indebted to Mrs. Barbara Collins, Esther Tauberschmidt, Uta Huckle, and Heike Jakobs as well as to Mr. D. Windfuhr for expert technical assistance. References Fisher, M.A., Shahani, B.T. and Young, R.R. (1979) Assessing segmental excitability after acute rostra1 lesions. II. The blink reflex. Neurology, 29: 45-50. Gilliatt, R.W. and Thomas, P.K. (1970) Changes in nerve conduction in ulnar nerve lesions at the elbow. J. Neurol. Neurosurg. Psychiat., 23: 312-320. Godaux, E. and Desmedt, J.E. (1975) Exteroceptive suppression and motor control of the masseter and temporalis muscles in normal man. Brain Res., 85: 447-458. Gregg, J.M. (1990a) Studies of traumatic neuralgia in the maxillofacial region: symptom complexes and response to microsurgery. J. Oral Maxillofac. Surg., 48: 135-140 Gregg, J.M. (1990b) Studies of traumatic neuralgias in the maxillofacial region: surgical pathology and neural mechanisms. J. Oral Maxillofac. Surg., 48: 228-237. Ha!3feld, S. and Meinck, H.-M. (1992) Der Kieferoffnungsreflex: Eine neue elektrophysiologische Methode zur objektiven Untersuchung . trigeminaler Senstbtlrtatsst6rungen.I. Methodik und Normwerte. Z. EEG-EMG, 23: 184-189.

202

S. Ha@feld et al. /Journal

of rhe Neurological

HaDfeld, S. and Meinck, H.-M. (1993) Der Kieferoffmmgsreflex: Eine neue elektrophysiologische Methode zur objektiven Untersuchung trigeminaler Sensibilit~tsstGrungen. II. Befunde bei Patienten mit Astllsionen des N. trigeminus. Z. EEG-EMG, 24: 147-154. Hoffmann, P. and Tonnies, J.F. (1948) Nachweis des vijllig konstanten Vorkommens des Zungen-Kieferreflexes beim Menschen. Pfliiger’s Arch., 250: 103-108. Jelasic, F. (1983) Quantitative Bestimmung der Hautsensibilitlt. Dtsch. Med. Wschr., 108: 419-421. Keidel, M., Rieschke, P., Jiiptner, M., Diener, H.C. (1994) Pathologischer Kieferoffnungsreflex nach HWS-Beschleunigungsverletzung. Nervenarzt, 65: 241-249. Kimura, J., Wilkinson, T., Damasio, H., Adams, H.R., Shivapour, E. and Yamada, T. (1985) Blink reflex in patients with hemispheric cerebrovascular accident (CVA). J. Neurol. Sci., 67: 15-28.

Sciences 134 (1995)

197-202

Laitinen, L.V. and Eriksson, A.T. (1985) Electrical stimulation in the measurement of cutaneous sensibility. Pain, 22: 139-150. Ongerboer de Visser, B.W. and Goor, C. (1976) Cutaneous silent period in masseter muscles: a clinical and electrodiagnostic evaluation. J. Neurol. Neurosurg. Psychiat., 39: 674-679. Ongerboer de Visser, B.W., Cruccu, G., Manfredi, M. and Koelman, J.H.T.M. (1990) Effects of brainstem lesions on the masseter inhibitory reflex. Functional mechanisms of reflex pathways. Brain, 113: 781-792. Schoenen, J., Jamart, B., Phy, G., Lenarduzzi, P. and Delwaide, P.J. (1988) Exteroceptive suppression of temporalis muscle activity in chronic headache. Neurology, 37: 1834-1836. Trojaborg, W. (1970) Rate of recovery of motor and sensory fibres of the radial nerve: clinical and electrophysiological aspects, J. Neurol. Neurosurg. Psychiat., 33: 625-638.