Voluntary supraspinal suppression of spinal reflex activity in paralyzed muscles of spinal cord injury patients

Cioni B, Dimitrijevic MR, McKay WB, Sherwood AM. Voluntary Supraspinal Suppression of Spinal Reflex Activity in Paralyzed Muscles of Spinal Cord Injury Patients. Exp Neurol, 1986 93:574-583. Voluntary Supraspinal Suppression of Spinal Reflex Activity in Paralyzed Muscles of Spinal Cord Injury Patients BEATRICE CIONI, MILAN R. DIMITRIJEVIC, W. BARRY MCKAY, AND ARTHUR M. SHERWOOD Department of Clinical Neurophysiology, The Institute for Rehabilitation and Research, and Department of Rehabilitation, Baylor College of Medicine, Houston, Texas 77030 Received March 19, 1986; revision received April 22, 1986 Having previously demonstrated that residual facilitatory brain influence on seg-mental structures occurs in paralyzed spinal cord injury patients, we sought evidence of suprasegmental suppression in such patients. By recording EMG activity from leg muscles, we studied changes in segmental excitability of the plantar reflex elicited by cutaneous stimulation of the plantar surface. Using surface EMG recordings, 50 paralyzed spinal cord injury patients were examined for their ability to volitionally suppress the plantar reflex on three repeated trials after three baseline trials. The patients, who had no voluntary EMG activity in the monitored muscles, were able to volitionally suppress the plantar reflex responses by 45% in the tibialis anterior, hamstring, and triceps surae muscles and to suppress the quadriceps response by 72%. In this patient group, 73 of 100 tibialis anterior muscle groups showed suppression of more than 20% compared with the control response. On reexamination, these findings were consistent during a period of 2 years in six patients. We conclude that suprasegmental suppression of segmental activity does occur in paralyzed spinal cord injury patients, and that in clinically complete patients, neurological evaluation should include assessment of the degree of preservation of suprasegmental neurocontrol on segmental activity below the lesion. Abbreviations: HS—hamstrings, PR—plantar reflex, QD—quadriceps femoris, SCI—spinal cord injured, TA—tibialis anterior muscle, TS—triceps surae. 1 The authors are in alphabetic order, indicating equal contributions. Support for this work was provided by the Vivian L. Smith Foundation for Restorative Neurology and by National Institute of Handicap Research grants G00830044 and 13-P-59275-6. Dr. Cioni is a visiting postdoctoral fellow from The Institute of Neurosurgery, Catholic University, Rome, Italy. 5740014-4886/86 $3.00 Copyright © 1986 by Academic Press, Inc. All rights of reproduction in any form reserved.

INTRODUCTION Suprasegmental control is normally expressed both as facilitation and suppression of segmental reflexes. We showed elsewhere that the tonic vibratory reflex depended on suprasegmental facilitation, and that it was found even in paralyzed spinal cord injury (SCI) patients (4). We also showed that 40 of 59 SCI patients were able to initiate some motor unit activity in paralyzed muscles through the use of reinforcement maneuvers (5). Thus, because the complete loss of volitional movement did not imply a complete absence of suprasegmental influence, we sought evidence of suprasegmental suppression. It is essential to have a sufficient, consistent level of activity to demonstrate suppression, which in paralyzed patients may be accomplished by inducing segmental reflex activity. Electromyographic (EMG) activity from the plantar reflex (PR) was simultaneously recorded from all major muscle groups in the legs for this purpose. This study quantified changes in the magnitude of EMG responses of paralyzed leg muscles in SCI patients when they attempted to voluntarily suppress the PR, and we here discuss the significance of those changes. MATERIALS AND METHODS Fifty SCI patients drawn from a larger number referred for evaluation of spasticity were selected according to the following criteria: (i) spinal cord injury above T8, (ii) no decubiti, bladder infection, or active diseases, (iii) 9 or more months postinjury, (iv) absence of any movement or motor unit activity recordable with surface electrodes in muscles of the legs when requested to move, and (v) the presence of an obvious and persistent PR in each leg upon repeated trials. Whenever possible, antispastic medications (diazepam, baclofen) were reduced by 50% 3 days prior to the examination, and they were not administered on the day of the examination. The level of injury in this group ranged from C2-3 to T7, with a mean onset time of 47 (10 to 255) months prior to the recording. There were 43 male and 7 female patients, 12 to 55 (mean 26) years old. All 50 patients were completely paralyzed, and 35 were without any evidence of sensory functions in the legs. Neurologic findings for the 15 patients who showed any signs of sensory function are summarized in Table 1. The results reported here are part of a larger study designed for the comprehensive assessment of the clinical neuro-physiology of upper motor neuron dysfunction in SCI patients (6).

The tests were carried out with the patient supine on an examination table. Beckman silver/silver chloride electrodes were placed over the muscle bellies of the quadriceps femoris (QD), hamstrings (HS), tibialis anterior (TA), and triceps surae (TS) muscle groups bilaterally. The EMG activity was recorded with an ink jet recorder (Elema-Schonander), using a bandwidth of 3.2 to 700 Hz, and a display scale of 100 /iV/cm. Paralysis of the patients' legs was verified by the absence of EMG activity when the patient attempted to flex and then extend both hips and knees together, then each leg separately, and finally, to plantar flex and dorsiflex each ankle. The test stimulation to elicit the PR consisted of stroking the plantar surface of the foot with the smooth, blunt end of a reflex hammer handle, beginning at the heel, proceeding along the lateral surface and curving back across the ball of the foot. To verify the consistency of this stimulus, we used a strain-gauge instrumented handle to monitor the applied force during studies in several patients. We found that even with many repetitions of the maneuver, the coefficient of variation offerees imposed by the operator was on the order of 25%. The maneuver, which took about 2 to 3 s to complete, was repeated three times, with an interval of approximately 5 s after cessation of the induced activity between each repetition. During the initial set of three stimuli, the patient was informed as to the nature of the stimulation to be applied. Although the response to such a (mild) stimulation in healthy subjects elicited little or no response, the PR response of the SCI patients was pronounced, particularly so in the more spastic patients, resembling a withdrawal movement. The response was largest in the tibialis anterior muscle, as might be expected. Frequently the PR spread electromyographic activity to the extensor groups and had a tonic characteristic as well. Measurement of the responses averaged for a series of three repetitions of the plantar stimulation revealed that left and right leg values for any of the examined features were not significantly different when considered for the entire group of patients in the study. It was therefore possible to combine the values obtained from the left and right sides into one group

for purposes of analysis. Analysis of the ipsilateral responses revealed that the TA and HS

muscle responses were not significantly different. However, the TS response was smaller (P < 0.001). and the QD was by far the smallest (P < 0.001). As might be expected, the EMG responses recorded from the same muscle group of different patients were moderately variable. After recording the three baseline responses, the patient was instructed to attempt to suppress the induced activity by intentionally relaxing the leg during each of three additional maneuvers, delivered in an identical manner. Verbal prompting from the examiner informed the patient when to attempt the suppression. The patient was not given any advance training in accomplishing the suppression. No feedback of the size of the responses was provided to either the examiner or the patient, and neither had any knowledge of their interpretation. The magnitudes of EMG responses to plantar stimulation were measured by a digitizing tablet to manually trace the area enclosing the envelope of EMG activity recorded from each of the four ipsilateral muscles, measured as microvolt-seconds. Responses for each muscle thus quantified were averaged across the three repetitions of the stimulation for the baseline series and across the three for the suppressed series. The percentage suppression for each muscle was calculated in terms of the ratio of the average of the three suppressed responses to the average of the three baseline responses. Significance of the results was tested using a three-way analysis of variance with repeated measures. RESULTS After the baseline responses were collected (Fig. 1A), volitional ability to suppress reflex activity in paralyzed muscles was tested by asking the patients to relax and thereby consciously suppress activity during a second series of three attempts to elicit the PR. An obviously suppressed response is shown in Fig. 1B. Suppression was present in both flexors and extensors as seen in Fig. 2B, in which the QD response was completely suppressed and the responses in the HS, TA, and TS were reduced. Considered as a whole, all responses were significantly suppressed (P <

0.001). The TS, TA, and HS were suppressed to a similar degree (45%). The largest percentage suppression was in the QD (72%), which was reduced more than any other muscle group (P< 0.001).

Habituation has been shown to be a cause of a diminished segmental response upon long repetition of stimulation in the isolated spinal cord. Possible habituation of the PR was evaluated by comparing the pattern of responses to the three trials within the baseline series as well as the three within the test series. No significant differences in the responses were found due to the sequential order within the baseline or conditioned series. Furthermore, although the right leg was always examined first, there was no significant difference in the responses from the left and right legs. As seen in the histogram in Fig. 3, 73 of 100 TA muscles showed suppression of the PR response by 20% or more. In eight cases, the response in the TA muscle was suppressed completely. A similar distribution of the effects of suppression was observed in the HS and TS muscle groups. An even more pronounced suppression was found in the QD muscle group, in which 49 of 100 muscles were completely suppressed.

Not all patients were able to suppress the PR activity. In 18 of 100 TA muscles, volitional attempts to suppress resulted in no change in the TA response (within ±20% of the baseline value). In 9 TA muscles, the PR response was facilitated. However, in 7 of these 9 muscles, the initial response area was less than 10% of the mean of the group as a whole, and the 3 muscles with facilitation of more than 100% had an extremely small initial response. This small response resulted in the introduction of a large percentage error in their measurement; therefore we excluded those 3 muscle responses when calculating the group mean percentage suppression for the TA.

In six patients who had one to three follow-up examinations during 2 years, we found that the ability to suppress the PR was consistent. The degree of suppression was reproducible in 11 of the 12 TA muscle responses examined, and in the 12th muscle, suppression was consistently absent. To illustrate this reproducibility, values for one of the patients are presented in Table 2, showing that the degree of suppression was constant even though the size of the responses in the baseline series was variable.

DISCUSSION The fact that residual facilitation exists in paralyzed patients raises the possibility that residual suppression also exists. Neuropathologic findings have been reported that support the existence of these residual functions. In the majority of SCI patients, continuity of some fiber tracts, mainly on the periphery of the spinal cord and in the posterior columns, can be found (7-9), even in those patients who are clinically "complete" with total loss of voluntary motor function and sensation below the level of the lesion. We propose that these residual fibers are responsible for the mediation of the suprasegmental influences demonstrated here and in our previous reports (4, 5). Blight examined the relative importance of axonal destruction versus axo-nai dysfunction in the paralyzed SCI cat in a series of morphological (1) and electrophysiological (2) studies. He found that a substantial number of fibers i 10 to 20cc) survived the trauma and functioned in the chronic paralyzed cat. although conduction along such fibers was clearly compromised. He also found a loss of myelin thickness in proportion to the depth of the fiber within the cord and a selective survival of thinner fibers, regardless of depth. We hypothesize that such thin fibers can mediate suppression of the PR. Further support for this hypothesis can be found in the experimental studies of bulbospinal inhibition (11), and Peterson's interpretation of these studies that the inhibition was mediated by fine fiber systems (12). Recent experimental evidence suggested that demyelinated fibers may be able to transmit impulses, either through continuous or saltatory conduction 1 -. 15). Thus residual fibers from a variety of pathways may be available to mediate suprasegmental influence. Four different descending systems from the brain stem have been found to exert a similar inhibitory influence on interneurons involved in segmental reflexes (10). We suggest that the injured portion of the spinal cord can in some way mediate poorly controlled, rudimentary suppression of the test responses in these paralyzed patients. The fibers mediating this suppression are most likely a part of the residual reticulospmal tract (12). After SCI, the residual functioning axons conduct only a small portion of motor commands previously supplied from the brain to the distal segments of the cord. In this study we have shown that it is possible to suppress the PR in paralyzed rnuscles. The suppression was evident in the weakly activated extensors (QD. TS). and in the strongly activated flexors (TA, HS). The relationship of the relative area of the responses during suppression attempts in the HS, TA, and TS muscles did not change; instead, all areas decreased proportionally i Fig. 2C. D). The quadriceps was suppressed to a significantly greater degree. This is consistent with the fact that the PR response in the quadriceps is generalh smaller than in the other measured muscles. The responsiveness within a series of successive stimuli did not change. Thus the reduction seen during volitional attempts to suppress was not due to habituation of the response during repeated trials (3), but to the retention of some volitional ability to suppress segmental activity. The consistency of suppression examinedduring a 2-year period suggests that this motor program did not change in chronic paralysis. Not all patients showed suppression (Fig. 3). Some patients had enhanced responses, further arguing against a systematic

bias. In any particular SCI patient, the pattern of brain influence below the lesion is ultimately determined by the sequelae of the injury. The lesion will result in a "new anatomy" based on different structural relationships of the neurons in the injured portion of the spinal cord, and/or a reduction in number and changed proportion of fibers within the various ascending and descending tracts. These changes have functional consequences as well. In this study we have demonstrated residual suprasegmental suppression of seg-mental activity, which might be a form of residual suprasegmental inhibition. There is experimental evidence in the newborn cat after spinal transec-tion that the presence of inhibition is important in determining the features of motor performance (13). Thus residual inhibition may have both immediate and long-term importance in establishing features of motor performance. Residual suprasegmental suppression of segmental reflex activity can also explain the lack of spasticity in chronic spinal cord injury patients with preserved segmental reflexes. Suprasegmental inhibition activated during different stages of sleep may explain the muscle hypotonia of hypertonic but clinically paralyzed SCI patients during wakefulness. There is anatomic and physiologic evidence of the preservation of brain influence below the level of the lesion in these patients even when there are no such indications on a clinical basis. The neurological evaluation of the SCI patient should, therefore, include an assessment of the degree of preservation of suprasegmental neurocontrol on segmental activity below the lesion.

REFERENCES 1. BLIG HT , A. R. 1983. Cellular morphology of chronic spinal cord injury in the cat: analysis of myelinated axons by line-sampling. Neuroscience 10: 521-543. 2. BLIGHT, A. R. 1983. Axonal physiology of chronic spinal cord injury in the cat: intracellular recording in vitro. Neuroscience 10: 1471-1486. 3. DIMITR IJEVIC, M. R., AND P. W . NATH AN. 1970. Studies of spasticity in man. 4. Changes in flexion reflex with repetitive cutaneous stimulation in spinal man. Brain 93: 743-768. 4. DIM ITRIJEV IC, M . R., W . A. SP EN CE R, J. V . TR ON TE LJ, AND M. M. DIM ITRIJEV IC. 1977 . Reflex effects of vibration in patients with spinal cord lesions. Neurology 27: 1078-1086. 5. DIMITRIJEVIC, M. R., M. M. DIMITRIJEVIC, J. FAGANEL, AND A. M . SHERWO OD. 1984. Suprasegmentally induced m otor unit activity in paralyzed muscles in patients with e stablished spinal cord injury. Ann. Neurol. 16:216-222. 6. DIMITRIJEVIC, M. R. 1984. Neurocontrol of chronic upper motor neurone syndromes. Pages 111-128 in B. SHA HAN I, Ed., Electromyography in CNS Disorders: Central EM G. Butterworths, Boston. 7. KAK ULA S, B. A., AND G. M . BED BRO OK . 1969. A correlative clinico-pathological study of spinal cord injury. Proc. Aust. Assoc. Neurol. 6: 123-132. 8. KAK UL AS. B ., AND G . M. BE DB RO OK .. 197 6. Patholo gy of injuries of the verteb ral spina l cord — with emphasis on the macroscopic aspects. Pages 27-42 in P. J. VINK.EN AND G. \V. BRU Y EN. Eds., Handbook of Clinical Neurology, Vol. 25, Injuries of the Spine and Srn:ii/ Cord. Pan 1. North-Holland, Amsterdam. 9. KAKULAS B. 1985. Pathomorphological evidence for residual spinal cord functions. Pages 163-168 in J. C. ECCLESAND M. R. DIMITRUEV IC, Eds., Recent Achievements in Restorative Neurnlogy I: Upper Motor Neuron Func tions and D ysfunctions. Karger, Basel. 10. L UN DB ER G. V 198 2. Inhib itory control from the b rain stem of transm ission from the p ri-~ jr. arferents to motorneurons, primary afferent terminals and ascending pathways. 197-124 in B. SJOLUND AND A. BjORKLYJND, Eds., Brainstem Control of Spinal Mechanisms. Elsevier. Amsterdam. 11. M AG OO NH W . AND R . RH INE S. 19 46. A n inhibito ry mechanism in the bulbar reticular formation. J Neurophysiol, 9: 165-171. 12. PETERSON BW. 1984. The reticulospinal system and its role in the control of movement. Pages 28-86 in C. D. BA RN ES, Ed., Brainstem C ontro l of Spinal Co rd Fu nction. Acacemc Press. New Y ork/London. 13. RO BIN SON G. A.. AND M. E. G OLD BER GER . 1985. Interfering with inhibition may improve • otor function. Brain Res. 346: 400-403. 14. W AXM AN. S. G. 19 78. Conduc tion through demyelinated plaques in multiple sclerosis: computer simulations of facilitation by short internodes. / Neural. Neurosurg. Psychiatry. 4l:40S-417. 15 Waxman S G. 1982. Membranes, myelin and the pathophysiology of multiple sclerosis. New Eng J Med. 306: 1529-1533.