Trigemino-cervical reflexes in normal subjects

Journal of Clinical Neurophysiology 18(4):378 –385, Lippincott Williams & Wilkins, Inc., Philadelphia © 2001 American Clinical Neurophysiology Society

Trigeminocervical Reflexes Elicited by Stimulation of the Infraorbital Nerve: Head Retraction Reflex *†Cumhur Ertekin, *Nes¸e Çelebisoy, and *†Burhanettin Uludag˘, Departments of *Neurology and †Clinical Neurophysiology, Medical School Hospital, Ege University, Bornova, ˙Izmir, Turkey.

In the current study, the effects of stimulation of the infraorbital nerve (ION) on the trigeminocervical reflexes (TCRs), recorded from the posterior neck muscles, was investigated and the results were compared with the results recorded by stimulation of the supraorbital nerve (SON). TCRs obtained by stimulation of the ION was evaluated as the electrophysiologic counterpart of the head retraction reflex. Twenty normal control subjects, 10 men and 10 women, were enrolled in the study. The SON and the ION were stimulated by using a bipolar surface electrode. Results were recorded by using either concentric needle electrodes inserted into the semispinalis capitis muscle at the level of the third or fourth cervical vertebra or by surface electrodes placed at the C3 and C7 vertebrae on the midline. It was found that stimulation of the supraorbital and infraorbital branches of the trigeminal nerve had different reflexive effects on the posterior neck muscles. A stable positive (or negative-positive) wave, with a very early latency and high amplitude was always recorded after maximal stimulation of the ION, which could never be detected by stimulation of the SON. The C3 response of the TCR, evoked by SON stimulation was always evoked, by stimulation of the ION, at a low threshold. These findings suggest that the head retraction reflex is composed of two phases: inhibitory and excitatory. The early, fixed positive wave represents the general inhibition of the cranial and neck muscles, just before withdrawal of the face and head, from unexpected stimuli, which precedes the dense C3 response, demonstrating activation of the posterior neck muscles. Key Words: Head retraction— Trigeminocervical reflex—Infraorbital nerve—Posterior neck muscles.

The trigeminocervical reflex (TCR) and the head retraction reflex (HRR) have been defined as pathologic reflexes that may be found in patients with involvement of the corticobulbar pathways and brainstem nuclei (Haerer, 1992), but these reflexes, especially the HRR, have not been appreciated as physiologic reflexes. We have recently reported that the TCRs could be elicited easily by electrical stimulation of the supraorbital nerve (SON) and glabellar tapping (Ertekin et al., 1996). Early and late reflex responses were recorded from the posterior neck muscles using a concentric needle electrode.

However, these responses, elicited by SON stimulation were not compared with the TCR responses, which were obtained by infraorbital trigeminal nerve (ION) stimulation. The stimulation of the ION has been reported to evoke a very early reflex response from the sternocleidomastoid muscle (Di Lazzaro et al., 1995, 1996). Later reflex responses recorded from the sternocleidomastoid muscle by SON stimulation have also been reported in other studies (Nakashima et al., 1989; Sartucci et al., 1986). However, the possible differential reflexive effects of SON and ION stimulation have never been compared by posterior neck muscle recordings, which is important for the withdrawal of the head from facial nociceptive stimuli. We have found that these two branches of the trigeminal nerve produced different re-

Address correspondence and reprint requests to Dr. Ertekin, Nilhan apt. 1357 sok. No: 1/10, Alsancak, I˙zmir, Turkey.

378

HEAD RETRACTION REFLEX flexive effects on the posterior neck muscles. The aim of this study was to compare the electromyographic (EMG) effects of SON and ION stimulation on the posterior neck muscles and especially to describe the electrophysiologic counterpart of the head retraction movement. SUBJECTS AND METHODS Twenty normal subjects, 10 men and 10 women, who ranged in age from 18 to 57 years were enrolled in the study. The investigation was approved by the ethical committee of our university hospital, and all subjects provided informed consent. The electrophysiologic method has been described previously (Ertekin et al., 1996). In brief, the subjects were seated on a chair. The EMG reflex responses from the posterior neck muscles were obtained by two methods. The first method used was needle recording. Concentric needle electrodes (Medelec DMC-37, Surrey, UK; diameter, 0.46 mm; recording area, 0.07 mm2) were inserted into the orbicularis oculi muscle for blink reflexes and into the semispinalis muscle (SSM) at the level of the third or fourth cervical vertebra, on the same side as the orbicularis oculi muscle being examined. The needle electrode was inserted into the middle of the SSM at a right angle. Eleven subjects were investigated using this method. The second method used was surface recording. Two silver cup electrodes were placed and fixed at the level of the C3 and C7 vertebrae on the midline. Ten to 12 responses were averaged and/or superimposed. Eleven subjects were investigated using the surface recording technique. In three patients both recording methods (needle and surface) were used for comparison. EMG signals were not integrated unless it was necessary. Because the subjects were in an upright position, absolute inactivity of the neck muscles could not be provided. There was always random activity of the motor unit potentials despite the effort to minimize them by using some minute head maneuvers. The ION and the SON were stimulated percutaneously using a bipolar surface electrode (Medelec bipolar stimulator electrode no. 16893). Rectangular electrical shocks were used with two classes of pulse duration (0.1– 0.2 msec and 0.5 to 1.0 msec) to obtain painless and painful sensations. The intensity of the electrical shocks was increased in a stepwise manner until the subjects felt pain or unpleasant sensations. The electrical shocks with a 0.5 to 1.0-msec duration and more than 50 mA were usually accepted as painful, according to the experience of the subjects. For both painless (0.1– 0.2 msec and 50 – 60 mA) and painful (0.5–1.0 msec and ⬎50 mA) stimulation conditions, the EMG responses were re-

379

corded simultaneously from the orbicularis oculi muscle and the SSM on the same side. During the painful stimulation conditions and after giving successive but random stimuli, the baseline activity of the SSM increased unavoidably. Therefore, it was almost impossible to provide a stable and measurable baseline activity in the SSM. At least 12 or 15 responses were recorded, superimposed and occasionally averaged (Medelec Mystro MS 20). The analysis time was 100 msec or sometimes 200 msec. The filter bandpass was usually 50 to 2,000 Hz, but 10 to 200 Hz or 500 Hz was also used. The room temperature was controlled at 20 to 23°C. Data were evaluated statistically using Student’s unpaired t-test. RESULTS When the ION was stimulated maximally and the subject experienced pain, two types of EMG responses were recorded from the SSM at the C3/C4 level by the needle electrode and at the C3/C7 level by surface electrodes. The first EMG response was labeled C1 and it appeared approximately 18 to 20 msec after the stimulus. The later response, called C3, was recorded at approximately 50 msec. At first glance they looked similar to the C1 and C3 responses of the SSM obtained by stimulation of the SON (Fig. 1, bottom trace). However, when the intensity of the stimulus applied to the ION was increased, it was realized that C1 harbored two different components. With stimuli of lower intensities an unstable EMG response appeared (Fig. 1, first and second traces), whereas with stimuli of higher intensities a stable, fixed-latency positive or negative-positive wave was recorded (Fig. 1, third trace). The first, unstable (C1) EMG response with a lower threshold could not be recorded all the time when compared with the fixed, stable component and with the components obtained by SON stimulation. It was recorded in 14 of 20 subjects investigated, mainly with the needle-recording method. However, the stable, fixed-latency positive component of C1 was demonstrated more easily when the frequency limits were changed to 10 to 200 Hz or 500 Hz, and by stimulating the ION with higher intensities. It was recorded using both the needle and the surface electrodes. The onset of the positive wave was approximately 18 msec, ending before or at 40 msec after ION stimulation. The positive peak was frequently recorded at 23 msec. The onset and the peak latencies of this positive wave were very stable in a given patient when the stimulation and the recording conditions were kept constant. The wave labeled C3 in Fig. 1 was the reflex EMG response recorded directly from the posterior neck musJ Clin Neurophysiol, Vol. 18, No. 4, 2001

380

C. ERTEKIN ET AL.

FIG. 1. Posterior neck muscle responses (semispinalis capitis muscle [SSM]) recorded by a needle electrode at the C4 level after stimulation of the infraorbital nerve (ION) and the supraorbital nerve (SON). Ten superimposed traces. Stimulus parameters can be seen on the left side for each recording condition.

cles around and at the tip of the recording electrode, and it was the same response obtained by stimulation of the SON (Ertekin et al., 1996). C3 was evoked invariably by both painless and painful stimuli in all subjects investigated. C3 latency was almost the same as that obtained by SON stimulation (P ⬎ 0.05). Minimal latency of C3 varied at a range of 40 to 85 msec, but it seemed quite stable in its onset latency in a given patient. The mean latency and amplitude values for C1 and C3, recorded by ION and SON stimulation, are presented in Table 1. In the C3/C7 surface recording, painful stimulation of J Clin Neurophysiol, Vol. 18, No. 4, 2001

the ION invariably produced the neck wave with a prominent positivity, similar to that obtained by the needle recording of the SSM at the C3/C4 level (Fig. 2). Similarly, C3 responses were obtained from the SSM by using both surface and needle recordings. When the filters were used properly, the configuration of the positive wave was found to be diphasic in 10 of 11 subjects. The first deflexion was negative and it was followed by a prominent positive wave of high amplitude. The onset latency of the positive wave was approximately 21 msec, which was slightly later than that obtained by the needle

HEAD RETRACTION REFLEX

381

TABLE 1. The quantitative results of TCRs recorded at the C3/C4 level (needle recording), elicited by ION and SON stimulation in normal subjects ION stimulation*, mean ⫾ SEM

SON stimulation*, mean ⫹ SEM

Electrical stimulation

C1

C3

C1

C3

Latency, msec†

17.23 ⫾ 0.74 SD: 2.68 0.71 ⫾ 0.11 SD: 0.33 17.77 ⫾ 1.29 SD: 4.66

53.35 ⫾ 3.41 SD: 14.07 1.11 ⫾ 0.18 SD: 0.73 36.43 ⫾ 1.93 SD: 8.83

18.2 ⫾ 1.13 SD: 3.58 0.4 ⫾ 0.05 SD: 0.16 12.8 ⫾ 1.33 SD: 4.21

51.57 ⫾ 2.82 SD: 12.93 0.88 ⫾ 0.17 SD: 0.6 35.31 ⫾ 3.06 SD: 12.24

‡

Amplitude, mV Duration, mV

* Rectangular electrical shocks with a duration of 0.5 to 1 msec and an intensity of approximately 50 mA or higher, depending on the subject’s tolerance. † Minimal latency obtained from the superimposed responses. ‡ Maximal peak-to-peak amplitude obtained from the superimposed responses. ION, intraorbital nerve; SON, supraorbital nerve; SEM, standard error of the mean; SD, standard deviation.

recording (17 msec). The maximal positive peak was approximately 27 msec, which was also slightly delayed when compared with the needle recording (23 msec). The end of the positive wave was again approximately 5 msec later in the surface recording (37.5 msec versus 33 msec). The amplitude of the positive wave was higher in the needle recording than in the surface bipolar registration (517 ␮V versus 188 ␮V on average). The mean latency and amplitude values of the positive wave, recorded by needle and surface recordings, are presented in Table 2. During the appearance of the positive wave, the posterior neck muscles became electrically silent, as seen in the rectified EMG (see Fig. 2), especially the rising phase and the maximal positive peak used to develop during the silent period of the EMG activity. As mentioned, there was always random EMG activity in the neck muscles. Therefore, the silent period could be demonstrated by rectified EMG. Fig. 3 shows the relationship between the C1 response and the stimulus intensity. Painless stimulation of the ION first produced an unstable EMG response, then an increase in the stimulus intensity produced the fixed-latency positive response even in painless stimulation conditions (Fig. 3, arrows). The positive wave became more clear when the

stimulus intensity was increased and the subjects reported painful sensations. The head of the subject often moved backward with painful stimuli but it was observed that restraining the head from movement did not produce any change in the C1 and C3 responses of the SSM. When the effects of the other nerves were investigated for the development of the positive wave, ION stimulation was found to play an important role. The positive wave could be obtained only by stimulation of the ION, but could never be elicited by the painful stimulation of the SON or the mental–trigeminal nerve (Fig. 4A), or by the stimulation of the median nerve at the wrist (Fig. 4B). The other cranial muscles such as the orbicularis oculi, masseter, and frontalis muscles did not produce such a positive wave despite the same ION stimulation and recording conditions. The positive wave obtained from the posterior neck muscles was observed during the silent period of the masseter muscle or it was elicited within the time interval of R1 and R2 responses of the blink reflex (Fig. 5). DISCUSSION Electrical and mechanical stimulation of the supraorbital and infraorbital branches of the trigeminal nerve

TABLE 2. Statistical findings of a prominent positive wave of the posterior neck in normal subjects (frequency limits set at 10 –200 H2reset or 500 Hz) Variable Needle recording, msec C3/C7 surface recording, msec†

Onset, mV

Positive peak, mV

End, mV

Amplitude, mV

17.18 ⫾ 0.64 SD: 2.99 18.58 ⫾ 1.08 SD: 3.75

23.82 ⫾ 0.97 SD: 3.22 27.08 ⫾ ⫺1.66 SD: 5.76

33.0 ⫾ 1.26 SD: 4.20 37.58 ⫾ 1.86 SD: 6.46

0.51 ⫾ 0.14 SD: 0.48 0.18 ⫾ 0.4 SD: 0.13

* Subjects with ages ranging from 35 to 50 years. Subjects with ages ranging from 18 to 57 years. SD, standard deviation. †

J Clin Neurophysiol, Vol. 18, No. 4, 2001

382

C. ERTEKIN ET AL.

FIG. 2. The positive wave, especially the rising phase and the maximal positive peak recorded by the C3/C7 surface electrodes after stimulation of the infraorbital nerve, appears within the time span of the C1 and C3 responses of the trigeminocervical reflexes, when the posterior neck muscles become electrically silent, as can be seen in the rectified electromyograph.

were found to have different reflexive effects on the posterior neck muscles, as summarized in Table 3. The C1 response obtained by ION stimulation seemed to have two different components with similar latencies. An unstable EMG response appeared with stimuli of low threshold and a stable, fixed-latency positive wave was recorded with stimuli of a high threshold. The unstable C1 component recorded by ION stimulation is similar to the one obtained by SON stimulation, except that its appearance was comparably rare in the case of ION

stimulation. Because it is obtained by the needle recording of the SSM, it can be suggested to represent a local discharge, like the C1 response of SON stimulation (Ertekin et al., 1996). The major finding in the current study is the posterior neck muscle recording of a positive (or negative-positive) wave, with a very early latency and a high amplitude after maximal stimulation of the ION. This positive wave could never be detected by SON stimulation. It was first thought that this clear positive wave with a fixed latency could be related to the move-

FIG. 3. The threshold of the fixed positive wave depends on the stimulus intensity. Painless stimulation of the infraorbital nerve does not produce a positive wave until an intensity of 100 V is reached and the amplitude of the wave increases linearly with the increase in the stimulus intensity, without any change in latency.

J Clin Neurophysiol, Vol. 18, No. 4, 2001

HEAD RETRACTION REFLEX

383

FIG. 4. (A, B) The fixed positive wave can be obtained only by painful stimulation of the infraorbital nerve. Stimulation of the supraorbital nerve (SON), the mental nerve (A), or the median nerve at the wrist (B) does not produce such a response. Note the slower sweep time in B.

ment of head retraction. However, a contradictory observation was that the electrophysiologic appearance of the early, fixed positive wave was not associated with any kind of head movement, unless the stimulation of the ION was often painful or unpleasant. We propose that this fixed positive wave is composed of the activities picked up from many posterior neck muscles, excited by stimulation of the ION. Our other complementary proposal is that the early, fixed positive wave represents the general inhibition of the cranial and neck muscles, just before withdrawal of the face and head from unexpected and/or nociceptive stimuli. These kind of stimuli, applied especially to the facial innervation area of the maxillary branch of the trigeminal nerve, seem to be very important. The period of inhibition should be the first part of the HRR. This is followed by the second, excitatory phase, demonstrated as withdrawal of the head and face. The C3 component of the TCR, which is also elicited by SON stimulation as a more generalized reflex withdrawal, should be the electrophysiologic counterpart of the excitatory phase of the HRR. Our conclusion can be supported by some important experimental and human

studies. In cats, light mechanical stimulation of the facial skin, particularly the skin around the nose, produces a head withdrawal response, and stimulation of the branches of the trigeminal nerve or mechanical stimulation of the facial skin readily excites neck motoneurons (Abrahams et al., 1992; Alstermark et al., 1992; Manni et al., 1975). It is also reported that the reflex responses of most neck muscles of the cat are dominated by an early inhibition (Richmond and Loeb, 1992; Richmond et al., 1992). Later investigators (Richmond and Loeb, 1992) stressed inhibitory rather than excitatory effects in many neck muscles, and inhibitory and excitatory motor control of the head involves more than 20 muscle pairs of the posterior neck (Abrahams et al., 1993). The inhibitory and excitatory reflex responses evoked by maxillary nerve stimulation in the cat are similar to our EMG responses recorded from the human posterior neck muscles, which were named C1 (composed of an unstable EMG activity and a fixed positive wave) and C3. There are some human studies that support our results. Shortlatency trigeminal neck reflexes or TCRs have been demonstrated previously in humans (Browne et al., 1993; J Clin Neurophysiol, Vol. 18, No. 4, 2001

384

C. ERTEKIN ET AL.

FIG. 5. The fixed positive wave develops during the silent period of the masseter muscle and within the time interval of the R1 and R2 responses of the blink reflex. OOM, orbicularis oculi muscle.

Di Lazzaro et al., 1995, 1996; Ertekin et al., 1996; Nakashima et al., 1989; Sartucci et al., 1986). Di Lazzaro et al. (1995, 1996) observed a fixed positive/negative wave, with a latency of approximately 13 msec, which corresponded to a period of inhibition of the underlying motor unit activity recorded from the sternocleidomastoid muscle after electrical stimulation of the ION. The latency of this simple response recorded from the sternocleidomastoid muscle is similar to the fixed-latency

positive wave obtained from the posterior neck muscles. In addition, Di Lazzaro et al. (1995) found that the stimuli administered to other branches of the trigeminal nerve (supraorbital and mental) did not produce such clear effects. They suggested that the infraorbital response must be a part of the head withdrawal reflex involving an oligosynaptic trigeminocervical system similar to that described in the cat. Thus, our conclusion is that the first part of the HRR

TABLE 3. Reflexive effects of the SON and the ION Supraorbital nerve stimulation

Infraorbital nerve stimulation

C1 was quite frequent (Ertekin et al., 1996) C1 with low threshold by electrical stimulation C3 with high threshold by electrical stimulation No obvious early positive wave with painful stimuli

C1 was rare (for the low-threshold, variable component) C1 with high threshold (for the high-threshold, stable component), by electrical stimulation C3 with low threshold by electrical stimulation Clear-cut early positive wave with painful stimuli

J Clin Neurophysiol, Vol. 18, No. 4, 2001

HEAD RETRACTION REFLEX is the inhibition of the neck and the cranial muscles. This conclusion can be discussed in the light of some newly published literature: It has been proposed previously that an initial positive peak in unrectified, averaged EMG recordings should reflect inhibition of the underlying motor unit activity during tonic voluntary contraction (Colebatch and Rothwell, 1993; Poliakov and Miles, 1992). When an infraorbital stimulus is administered, some motor units are inhibited at a short latency, and the cancellation of the positive and the negative phases of asynchronous motor unit action potentials produce the surface positive/negative wave (Colebatch and Rothwell, 1993; Di Lazzaro et al., 1995). It seems that the random EMG activity can be recorded from close and distant motor units by using a needle electrode, as used in the current study, and similar electrophysiologic changes may occur even during needle recording, without averaging, although these findings are hard to explain electrophysiologically and they remain to be clarified. The fixed positive wave, recorded both by needle and surface electrodes, had similar latencies, although the amplitudes were different. Electrophysiologically, inhibition in the cranial muscles is presented during the silent period as observed in the masseter, whereas inhibition in the neck muscles is presented with the fixed positive wave as observed in the sternocleidomastoid muscle (Di Lazzaro et al., 1995, 1996), and in the posterior neck muscles, which were investigated in the current study. This is probably specific for ION stimulation. Additional studies to elucidate the electrogenetic nature of such a fixed positive wave are needed. The following excitatory phase of the HRR should be very dense electrophysiologically, which is the C3 response of the TCR. It must be associated with intense efferent discharges from the cervical motoneurons to the extensor neck muscles. The HRR and its electrophysiologic components were demonstrated readily by stimulation of the ION. The stimulation of the median nerve failed to produce these reflexes electrophysiologically, whereas the C3 response was obtained clearly by SON stimulation. This may indicate that the HRR and its electrophysiologic counterparts recorded from the neck muscles are segmental and specific for the organism, to protect the face, mouth, and head from unexpected nociceptive stimuli. The absence of a fixed-latency early response from neck muscles by stimulating the SON cannot be explained at this stage of our study. However, in some early human

385

studies of the TCR, SON stimulation at or above the pain threshold did not produce earlier responses of any kind before 50 msec (Sartucci et al., 1986) or 35 msec (Nakashima et al., 1989). The absence of an early fixedlatency response after stimulating the SON may depend on the anatomic and functional differences between the two branches of the trigeminal nerve. We also do not know the meaning of this short latency response in relation to C3, which is obtained by both ION and SON stimulation, and seems to be more related to withdrawal of the head.

REFERENCES Abrahams VC, Downey ED, Kori A. The superior colliculus and head movement in the cat. In: Berthoz A, Vidal PP, Graf W, eds. The head–neck sensory motor system. New York: Oxford University Press, 1992:289 –91. Abrahams VC, Kori AA, Loeb GE, Richmond FJR, Rose PK, Keirstead SA. Facial input to neck motoneurons: trigemino-cervical reflexes in the conscious and anesthetized cat. Exp Brain Res 1993;92: 183–93. Alstermark B, Pinter MJ, Sasaki S, Tantisira B. Trigeminal excitation of dorsal neck motoneurons in the cat. Exp Brain Res 1992;92: 183–93. Browne PA, Clark GT, Yang Q, Nakano M. Sternocleidomastoid muscle inhibition induced by trigeminal stimulation. J Dent Res 1993;72:1503– 8. Colebatch JG, Rothwell JC. Vestibular evoked EMG responses in human neck muscles [abstract] J Physiol (Lond) 1993;473:18P. Di Lazzaro V, Quartarone A, Higuchi K, Rothwell JC. Short-latency trigemino-cervical reflexes in man. Exp Brain Res 1995;102:474 – 82. Di Lazzaro V, Restuccia D, Nardano R, et al. Preliminary clinical observations on a new trigeminal reflex: the trigemino-cervical reflex. Neurology 1996;46:479 – 85. Ertekin C, Çelebisoy N, Uludag˘ B. Trigemino-cervical reflexes in normal subjects. J Neurol Sci 1996;143:84 –90. Haerer AF. De Jong’s the neurological examination. 5th ed. Philadelphia: JB Lippincott, 1992:165–79. Manni E, Palmieri G, Marini R, Petterossi VE. Trigeminal influences on extensor muscles of the neck. Exp Neurol 1975;47:330 – 42. Nakashima K, Thompson PD, Rothwell JC, Day BL, Stell R, Marsden CD. An exteroceptive reflex in the sternocleidomastoid muscle produced by electrical stimulation of the supraorbital nerve in normal subjects and patients with spasmodic torticollis. Neurology 1989;39:1354 – 8. Poliakov AV, Miles TS. Quantitative analysis of reflex responses in the averaged surface electromyogram. J Neurosci Methods 1992;43: 195–200. Richmond FJR, Loeb GE. Electromyographic studies of neck muscles in the intact cat. II—reflexes evoked by muscle nerve stimulation. Exp Brain Res 1992;88:59 – 66. Richmond FJR, Thomson DB, Loeb GE. Electromyographic studies of neck muscles in the intact cat. Exp Brain Res 1992;88:41–58. Sartucci F, Rossi A, Rossi B. Trigemino-cervical reflex in man. Electromyogr Clin Neurophysiol 1986;26:123–9.

J Clin Neurophysiol, Vol. 18, No. 4, 2001