Changes in Spinal Cord Excitability in a Patient with Rhythmic Segmental Myoclonus

Home , Cauda equina

Journal ofNeurology, Neurosurgery, and Psychiatry 1996;61:641-644 641 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.6.641 on 1 December 1996. Downloaded from SHORT REPORT

Changes in spinal cord excitability in a patient with rhythmic segmental myoclonus

V Di Lazzaro, D Restuccia, R Nardone, A Oliviero, P Profice, A Insola, P Tonali, J C Rothwell

Abstract by abnormal activity in spinal circuits which Paired stimulation of the common per- lack input from the missing interneurons. oneal and posterior tibial nerve was used To test this hypothesis we evaluated dorsal to study the recovery cycle oflumbosacral horn function in one patient with spinal somatosensory evoked potentials in 10 myoclonus by recording the spinal evoked control subjects and in one patient with potentials produced by stimulation of the rhythmic segmental myoclonus of the leg common peroneal and posterior tibial nerves. involving the L2-L4 myotomes. In normal Spinal responses in humans are similar to subjects the peripheral nerve volley in the those described in monkeys, and are thought cauda equina had recovered at an inter- to reflect the postsynaptic neuronal response stimulus interval of 3 ms whereas the to inputs conveyed by group I and II periph- postsynaptic dorsal horn potential was eral afferent fibres in Rexed layers IV and V.4 reduced to about 60% of its control size. These responses are generated by static trans- Similar results were found in the patient verse dipolar sources in the grey matter of the after posterior tibial nerve but not com- spinal cord.? We recorded the response to sin- mon peroneal nerve stimulation. The sec- gle stimuli and also to pairs of stimuli to study ond, which evokes afferent input to the the recovery cycle of excitability changes affected lumbar segments, produced within the dorsal horn. facilitation ofthe postsynaptic response at 3 ms. This finding suggests that the physiological suppression of dorsal horn Patient Istituto di Neurologia, interneurons which usually takes place A previously healthy woman aged 39 pre- Universita Cattolica, occur L go A Gemelli 8, after paired stimulation fails to in sented with a two month history of lower limb 00168 Rome, Italy segmental myoclonus. This may indicate jerking. She had bilateral, rhythmic, involun- V Di Lazzaro that dorsal horn interneurons are abnor- tary jerks of proximal lower limb muscles. D Restuccia R Nardone mally hyperactive and are involved in the These jerks were intensified by emotional http://jnnp.bmj.com/ A Oliviero pathophysiology of spinal myoclonus. stress and ceased during sleep. Strength, tone, P Profice and tendon reflexes were normal and plantar A Insola responses were bilaterally flexor. Neurological P Tonali (J Neurol Neurosurg Psychiatry 1996;61:641-644) J C Rothwell examination was otherwise normal. Full blood Servizio di examination was normal. Magnetic resonance Neurofisiopatologia, imaging of the CNS was normal, as were CTO, Via S Nemesio Keywords: myoclonus; spinal cord; evoked potentials motor and sensory conduction studies. Needle 21, 00145 Rome, Italy on September 28, 2021 by guest. Protected copyright. A Insola EMG recording from leg muscles disclosed IRCCS "Casa Sollievo Myoclonus may be the result of abnormal repetitive discharges of motor units at a rate of della Sofferenza" San activity in many different parts of the CNS. about 04 Hz in the iliopsoas, adductor, and Giovanni Rotondo, That of spinal origin is distinguished by the rectus femoris muscles on either side. No signs Italy P Tonali fact that (a) the jerking is usually rhythmic, (b) of denervation were found. The EEG, the the muscles involved are innervated by a blink reflex, and the recovery cycle of the blink MRC Human Movement and restricted segment of the cord, and (c) there is reflex were normal. Motor evoked potentials Balance Unit, The no sign that the jerks are preceded by EEG after cortical and paravertebral magnetic stim- National Hospital for activity in the contralateral sensorimotor cor- ulation were of normal latency. Standard Neurology and Neurosurgery, Queen tex. somatosensory evoked potentials (SEPs) after Square, London The pathophysiology of spinal myoclonus stimulation of the median, common peroneal, WC1N 3BG, UK remains speculative. Histological studies have and tibial nerve were of normal amplitude and J C Rothwell shown a striking reduction in the number of latency. Correspondence to: Dr Vincenzo Di Lazzaro, small and medium sized neurons in the poste- Istituto di Neurologia, rior horns of the lumbar cord with relative Universita Cattolica, Lgo A Gemelli 8, 00168 Rome, sparing of large neurons in the anterior horns. Methods Italy. Taken together with EMG studies that have For SEP recording, the patient lay on a couch Received 8 March 1996 demonstrated the absence of denervation in a warm and semidarkened room. Stimuli and in revised form 2 July 1996 potentials,2 this suggests that relatively nor- (02 ms duration, 2 Hz) were delivered Accepted 8 August 1996 mal a motor neurons are driven to discharge through skin electrodes at the popliteal fossa 642 Di Lazzaro, Restuccia, Nardone, Oliviero, Profice, Insola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.6.641 on 1 December 1996. Downloaded from Control subject Patient Common peroneal nerve (knee) A c~~~~~ L4-L2

Ti 2-Ant ::_

Posterior tibial nerve (ankle) B D

L4-L2

Ti 2-Ant

1 IIV 5 ms

Figure 1 Paired stimulation in a control subject and in the patient. (A). Paired stimulation of common peroneal nerve with an interstimulus interval of 3 ms in a normal subject. The conditioned cauda equina response is of comparable amplitude to the control response (thicker line). The conditioned N14 is about 60% of the responses obtained with a single stimulus (thicker line). (B) Paired stimulation ofposterior tibial nerve with an interstimulus interval of 3 ms in a normal subject. The conditioned cauda equina response is of comparable amplitude to the control response (thicker line). The conditioned N24 is about 60% of the responses obtained with a single stimulus (thicker line). (C) Paired stimulation of common peroneal nerve with an interstimulus interval of 3 ms in the patient. The conditioned cauda equina response is of comparable amplitude to the control response (thicker line). The conditioned N14 is about 120% of the responses obtained with a single stimulus (thicker line). (D) Paired stimulation ofposterior tibial nerve with an interstimulus interval of 3 ms in the patient. The conditioned cauda equina response is of comparable amplitude to the control response (thicker line). The conditioned N24 is about 70% of the responses obtained with a single stimulus (thicker line).

for the common peroneal nerve and at the peak. http://jnnp.bmj.com/ ankle for the posterior tibial nerve; stimulus In a preliminary study on three normal sub- intensity was adjusted to be slightly above jects we determined the recovery curves of motor threshold. The filter bandpass was cauda equina and spinal responses by delivering 30-3000 Hz. Responses were averaged with pairs of stimuli of equal intensity to the com- an analysis time of 50 ms. Samples with exces- mon peroneal nerve or posterior tibial nerve sive interference were automatically edited out with interstimulus intervals of 2, 3, 4, and 5 of the average. Two averages of 2048 trials ms. As peripheral nerve excitability, judged by on September 28, 2021 by guest. Protected copyright. each were obtained. The recording electrodes the cauda equina potential, had recovered by 3 (impedance below 5 kohm) were placed over ms (fig 1), the patient and 10 control subjects the spinal processes of L4 and T12. The L4 (mean age 38-1 (SD 4 7) years) were studied electrode was referred to L2 to record the with this interstimulus interval. To measure response generated by the ascending volley of the recovery of the conditioned response, the impulses in the cauda equina.67 For recording test response recorded using a single stimulus the spinal potential, which we labelled as N14 was subtracted off line from the responses for the common peroneal nerve and as N24 recorded using paired stimuli. The test in con- for the posterior tibial nerve, we connected trols and in the patient was performed after grid 1 of the amplifier to the T12 electrode stimulation of nerves of the right side. The and grid 2 to an electrode placed over the upper limit of the conditioned response was anterior abdomen. The rationale for this mon- defined as the mean + 2-5 SDs of the normal tage has been discussed in detail in a previous values. study.8 Briefly, it permits the selective recording of the activity generated by the transverse dipolar source located in the lumbosacral Results spinal cord9; moreover, it can cancel noise TIME COURSE OF SEP RECOVERY IN THREE from the ECG activity that is picked up by NORMAL SUBJECTS both T12 and anterior electrodes. The ampli- The recovery cycle of the cauda equina and tudes of responses were measured peak to spinal responses in three normal subjects were Changes in spinal cord excitability in a patient with rhythmic segmental myoclonus 643 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.6.641 on 1 December 1996. Downloaded from Figure 2 (A) Recovery A Common peroneal nerve In the patient we found a normal inhibition cycle of the cauda equina 140 of conditioned response after posterior tibial response (*) and N14 response (0) to paired nerve stimulation with a conditioned response stimulation of common of 69-6% of the test response and a cauda peroneal nerve. Each trace equina response of 98% of the test (fig 1). is the mean of responses recorded in three different Paired stimuli to the common peroneal nerve subjects; SD are indicated. produced no inhibition of the conditioned The conditioned response response and on the contrary resulted in an amplitude is expressed as a the N14 that was percentage of response increase of the amplitude of obtained with a single 121% of the test response (fig 1). The cauda stimulus. Cauda equina C equina response was 100% of the test. The refractory period ends at 40 0a amplitude and rhythm of the jerks were not 3 ms interstimulus interval with a full recovery ofthe " 20 affected by the nerve stimulation. response. At this interval cn there is stiU a pronounced depression ofconditioned 2 3 4 5 Discussion N14. (B) Recovery cycle QD ofthe cauda equina co Interstimulus interval (ms) Our patient had typical rhythmic segmental response (*) and N24 U was no of corti- 0. myoclonus.10 There evidence response (0) to paired o1 B Posterior tibial nerve stimulation ofthe posterior T2 1407- cal or brainstem hyperexcitability; cortical tibial nerve. Each trace is were and the recov- ._ SEPs of normal amplitude the mean of responses CDI ery cycle of the blink reflex was normal. The recorded in three different 120 subjects; SD are indicated. segmental distribution of jerks and the absence The conditioned response of any excitability changes in cerebral cortex or amplitude is expressed as a 0 brainstem strongly suggest that the jerks had a percentage ofthe response obtained using a single 80 spinal origin. stimulus. Cauda equina The pathophysiology of spinal myoclonus refractory period ends at 60 remains speculative. As outlined in the intro- 3 ms interstimulus interval duction, several authors have suggested that with a full recovery ofthe response. At this interval 40 the underlying pathophysiology involves spinal there is still a pronounced interneurons rather than a motor neurons and depression ofconditioned 20 this has been supported by histological evi- N24. dence in one case.' To date, however, there o 0 1 2 3 4 5 has been little evidence of a physiological Interstimulus interval (ms) abnormality which might parallel the anatomi- cal changes. Davis et al I showed in one patient that stimulation of the common peroneal constructed from pairs of equal stimuli given nerve on one side could evoke abnormal short to the common peroneal nerve or posterior latency responses in the opposite leg. It is tibial nerve with an interstimulus interval likely that such responses are caused by hyper- ranging from 2 to 5 ms. Figure 1 shows typical excitability of interneuronal connections examples of normal responses to both a single between the two sides of the cord which under stimulus, and to the second stimulus of a pair normal circumstances are relatively sup-

with an interstimulus interval of 3 ms. The pressed. The present data provide further evi- http://jnnp.bmj.com/ response to the second stimulus of a pair has dence for physiological hyperexcitability in the been constructed by subtracting the response spinal cord. to a single shock from the total response to a In normal subjects, the spinal potential of pair of shocks. Figure 2 shows the mean the dorsal horn produced by the second stimu- results from the three normal subjects. Cauda lus of a pair to either the common peroneal equina potential recovered completely at 3 ms. nerve or the posterior tibial nerve was sup- At this interstimulus interval spinal responses pressed to 60% of its control value when the evoked by the second stimulus of the pair were interval between the shocks was 3 ms. At the on September 28, 2021 by guest. Protected copyright. still suppressed. Spinal responses fully recov- same interstimulus interval, cauda equina ered at an interstimulus interval of 5s. responses had completely recovered to the control level, suggesting that spinal rather than COMPARISON OF SEP RECOVERY AT 3 ms IN THE peripheral mechanisms were responsible for PATIENT VERSUS 10 NORMAL SUBJECTS the inhibition. This finding is in agreement In control subjects, using an interstimulus with previous studies on the recovery cycle of interval of 3 ms the mean amplitude of the spinal SEPs.l1 Our patient had segmental cauda equina response was 95 5% (19%) of myoclonus involving iliopsoas, quadriceps, the response evoked with a single stimuli for adductor muscles, and thus the L2-L4 the posterior tibial nerve and 97% (14%) for myotomes.12 Spinal responses to a single nerve the common peroneal nerve. The mean ampli- stimulus were clear from both the posterior tude of N24 after stimulation of the posterior tibial nerve and the common peroneal nerve. tibial nerve was 61% (4-6%) of the response The main changes occurred with paired stim- evoked with a single stimuli, and the N14 after uli. After common peroneal nerve stimulation, common peroneal nerve stimulation was the response to the second stimulus of the pair 60-6% (5-9%). Thus the upper limit for a con- was enhanced, whereas it was suppressed to ditioned spinal response was 72-9% for the normal after posterior tibial nerve stimulation. posterior tibial nerve and 75-4% for the com- The dorsal root innervation of the common mon peroneal nerve. peroneal nerve is from lumbar myelomeres" 644 Di Lazzaro, Restuccia, Nardone, Oliviero, Profice, Insola, et al J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.6.641 on 1 December 1996. Downloaded from from the same spinal segments as those monkey. 7 Neurophysl 1 1977,40: 199-211. 5 Jeanmonod D, Sindou M, Mauguiere F. Three transverse involved in the genesis of the myoclonus dipolar generators in the human cervical and lumbo- whereas the sacral myelomeres mainly con- sacral dorsal horn: evidence from direct intraoperative recordings on the spinal cord surface. Electronencephalogr tribute to the posterior tibial nerve effect.'" C/in Neurophysiol 1989;74:236-40. We conclude that there was local hyperex- 6 Magladery JW, Porter WE, Park AM, Teasdall RD. Electrophysiological studies of nerve and reflex activity in citability of the mechanisms responsible for normal man, IV. The two-neurone reflex and identifica- the dorsal horn potential after common per- tion of certain action potentials from spinal roots and cord. Bulletin of_ohns Hopkins Hospital 1951 ;88:499-519. oneal nerve stimulation. 7 Cracco JB, Cracco RQ, Graziani LJ. The spinal evoked The enhancement of the conditioned com- response in infants and children. Neurology 1975;25: 31-6. mon peroneal nerve spinal response in our 8 Restuccia D, Di Lazzaro V, Valeriani M, Colosimo C, patient suggests that in segmental spinal Tonali P. N24 spinal response to tibial nerve stimulation and magnetic resonance imaging in lesions of the lum- myoclonus dorsal horn interneurons are bosacral spinal cord. Neurology 1993;43:2269-75. abnormally hyperactive. This finding, in asso- 9 Desmedt JE, Cheron G. Spinal and far-field components of human somatosensory evoked potentials to posterior tibial ciation with an EMG study in our patient and nerve stimulation analysed with oesophageal derivations in previous studies2' that showed no denerva- and non-cephalic reference recording. Electronencephalogr Clin, Neurophysiol 1983;56:635-51. tion in involved muscles, strongly suggests that 10 Bressman S, Fahn S. Essential myoclonus. In: Fahn S, in spinal myoclonus the underlying pathophys- Marsden CD, Van Woert MH, eds. Myoclonuis. Advances I.n neurology. Vol 43. New York: Raven Press, 1986: iology involves dorsal horn intemeurons. 287-94. From a clinical point of view, the data in 11 Saito T, Yamada T, Hasegawa A, et al. Recovery functions of common peroneal, posterior tibial and sural nerve our patient suggest that study of spinal recovery somatosensory evoked potentials. Electroenicephalogr Clin curves is capable of confirming the spinal ori- Neurosphysiol 1 992;85:337-44. 12 Wilbourn AJ. The value and limitations of electromyo- gin of excitability changes in patients with seg- graphic examination in the diagnosis of lumbosacral mental myoclonus just as the recovery cycle of radiculopathy. In: Hardy RW, ed. Lumbar disc disease. New! York: Raven Press, 1982:65-109. cortical SEPs can for forms of myoclonus orig- 13 LH Bannister, MM Berry, P Collins, et al. Nervous system. inating in the cerebral cortex. In: PL Williams, LH Bannister, MM Berry, P Collins, M Dyson, JE Dussek, MWJ Ferguson, eds. Gray's anatomny. 1 Davis SM, Murray NMF, Diengdoh JV, Galea-Debono A, 38th ed. Edinburgh: Churchill Livingstone, 1995:1286-7. Kocen RS. Stimulus sensitive spinal myoclonus. _7 Neuri-ol 14 Foester, 0. Spezielle Anatomie und Physiologie del periph- Newiosiurg Psychliany 1981 ;44:884-8. eren Nerven. In: Lewandowsky M, ed. Handubuch del 2 Hopkins AP, Michael F. Spinal myoclonus. I7 Neiu ol Neurolosgic, Erganzuingsband, Zweiter Teil. Berlin: Springer, NeUrosurg Psychiatyl1 974;37: 1112-5. 1929:785-974. 3 Nohl M, Doose H, Gross-Selbeck G, Jensen HP. Spinal 15 Chiappa KH. Short-latency somatosensory evoked poten- myoclonus. Eur Neu',ol 1978;17:129-35. tials: methodology. In: Chiappa KH, ed. Evokedpotentials 4 Beall JE, Applebaum AE, Foreman RD, Willis WD. Spinal I.n clinical medicine. New York: Raven Press, 1985: cord potentials evoked by cutaneous afferents in the 203 312. http://jnnp.bmj.com/ on September 28, 2021 by guest. Protected copyright.