PERIPHERAL NERVOUS SYSTEM INVOLVEMENT IN CHRONIC SPINAL CORD INJURY HATICE TANKISI, MD, PhD,1 KIRSTEN PUGDAHL, MSc, PhD,1 MIKKEL MYLIUS RASMUSSEN, MD,2,3 DORTE CLEMMENSEN, MD,2 YAZAN F. RAWASHDEH, MD, PhD,4 PETER CHRISTENSEN, MD, DMSc,3 KLAUS KROGH, MD, DMSc,5 and ANDERS FUGLSANG-FREDERIKSEN, MD, DMSc1 1
Department of Clinical Neurophysiology, Aarhus University Hospital, Nørrebrogade 44, DK-8000, Aarhus C, Denmark The Spinal Cord Research Centre, Department of Neurosurgery, Aarhus University Hospital, Aarhus, Denmark 3 Pelvic Floor Unit, Department of Surgery, Aarhus University Hospital, Aarhus, Denmark 4 Department of Urology, Aarhus University Hospital, Aarhus, Aarhus, Denmark 5 Neurogastroenterology Unit, Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus, Denmark 2
Accepted 3 March 2015 ABSTRACT: Introduction: Upper motor neuron disorders are believed to leave the peripheral nervous system (PNS) intact. In this study we examined whether there is evidence of PNS involvement in spinal cord injury (SCI). Methods: Twelve subjects with chronic low cervical or thoracic SCI were included prospectively. Needle electromyography was done in 10 different muscles in each subject bilaterally. Nerve conduction studies (NCS) were conducted in the fibular, tibial, and femoral motor and fibular and sural sensory nerves. Results: Half the subjects had widespread abnormal spontaneous activity (SA), and the amount of SA correlated inversely with reflex activity and nerve length. Fibular nerve entrapment across the knee was seen in 6 subjects, and sciatic nerve entrapment was seen in 1. Apart from entrapment neuropathies, NCS changes were found predominantly in motor nerves. Conclusion: The presence of widespread electrophysiologic changes outside entrapment sites indicates that SCI has a significant impact on the entire PNS, affecting the motor part predominantly. Muscle Nerve 52: 1016–1022, 2015
In needle electromyography (EMG), detection of abnormal spontaneous activity (SA) mainly in the form of fibrillation potentials (fibs) and positive sharp waves (PSWs) is the clinical “gold standard” for diagnosing denervated skeletal muscle. Although fibs and PSWs have been described in skeletal muscles in a series of reports in both cerebral1–3 and spinal cord lesions,4,5 it is still widely assumed among neurophysiologists that an upper motor neuron (UMN) lesion should leave the Additional Supporting Information may be found in the online version of this article. Abbreviations: AH, abductor hallucis; ASIA, American Spinal Injury Association; CMAP, compound muscle action potential; EDB, extensor digitorum brevis; EMG, electromyography; Fib, fibrillation potential; GC, gastrocnemius; LMN, lower motor neuron; NCS, nerve conduction studies; PL, peroneus longus; PNS, peripheral nervous system; PSW, positive sharp wave; RF, rectus femoris; SA, spontaneous activity; SCI, spinal cord injury; SNAP, sensory nerve action potential; SSEP, short latency somatosensory evoked potential; TA, tibialis anterior; UMN, upper motor neuron; VL, vastus lateralis; VM, vastus medialis Key words: electromyography; nerve conduction study; peripheral nervous system; spinal cord injury; spontaneous activity This work was supported in part by grants from the Lundbeck Foundation, Copenhagen, Denmark. Correspondence to: H. Tankisi; e-mail: [email protected] C 2015 Wiley Periodicals, Inc. V
Published online 10 March 2015 in Wiley Online Library (wileyonlinelibrary. com). DOI 10.1002/mus.24644
1016
EDx in SCI
lower motor neuron (LMN) anatomically intact, with a normal axon extending to the periphery. Overall, there are inconsistencies in studies of abnormal SA in patients with spinal cord injury (SCI). Some studies of SCI showed a larger number of fibs/PSWs in distal muscles compared with proximal muscles, suggesting a positive correlation between the degree of fibs/PSWs and the length of the peripheral axon,4,6 whereas, in another study, there was no difference in the number of fibs/PSWs between proximal and distal muscles.5 With regard to fibs/PSWs and degree of reflex activity/spasticity, some studies found a negative correlation,4,6 whereas others did not confirm this correlation.5,7 Because new therapies and rehabilitation interventions, that is, functional electrical stimulation, neuroanastomosis,8 and cell- and stem-cell–based therapies,9 for UMN disorders are being developed, confirming normal peripheral nerve function is crucial. This study was done prospectively as part of the preoperative evaluation of the peripheral nervous system (PNS) in subjects with complete or incomplete chronic SCI who were undergoing neuroanastomosis of L5 and S2 ventral nerve roots to obtain bladder and eventually bowel control.8,10–12 The aims of the study were to: (1) examine for LMN involvement other than nerve entrapments in subjects with SCI by determining the pattern of nerve conduction study (NCS) changes and the frequency of abnormal SA; (2) assess whether there is a correlation between the degree of fibs/ PSWs and reflex activity, and (3) determine whether the length of the axon plays a role in the pattern of fibs/PSWs by comparing proximal and distal muscles. METHODS
This study included 12 subjects (11 men and 1 woman) with a mean age of 41.7 years (range 19– 63 years). All subjects had traumatic chronic (>1 year postinjury) cervical or thoracic SCI class A or B on the American Spinal Injury Association (ASIA) scale.13 Mean duration of time since injury MUSCLE & NERVE
December 2015
Table 1. Subjects’ demographics. Patellar reflexes Subject 1 2 3 4 5 6 7 8 9 10 11 12
Years since injury 2 3 5 5 5 1 19 17 13 2 2 2
Gender/age (y) M/55 M/63 M/22 M/36 M/35 W/19 M/62 M/52 M/31 M/34 M/45 M/47
Level Th9 C5 C7 Th3 C4 Th10 Th9 Th4 C8 C5 C7 Th8
AIS score A B B A A A A A B B A A
R – 111 111 – – – 111 111 111 111 111 1
L – 111 111 – – – 111 111 111 111 111 1
Achilles reflexes R
L
– 111 111 111 – – 11 111 111 111 – 11
– 111 111 111 – – 11 111 111 111 – 11
M, man; W, woman; R, right; L, left. Reflexes graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal reflex; and 111 5 hyperreflexia.
was 6.3 (range 1–19) years. Eight subjects were diagnosed with complete injury (ASIA class A), and 4 subjects had incomplete sensory injury (ASIA class B) by clinical examination (Table 1). None of the subjects had a pre-existing neurologic condition, that is, peripheral neuropathy, or were taking medications known to cause peripheral neuropathy, lower limb nerve injury, low back injury, or lower limb fracture before or after the SCI. Peripheral nerve disease, such as root or lumbosacral plexus lesions, was excluded by detailed radiologic examination with MRI and X-ray. All patients had MRI of the cervical, thoracic, and lumbar spine as well as the lumbosacral plexus as a part of the initial investigation protocol. These examinations were re-evaluated by 2 of the authors (M.M.R. and D.C., both neurosurgeons), who, if necessary, also consulted radiologists on the scans after the electrophysiologic examinations in the patients who had widespread severe fibs/PSWs. Informed written consent was obtained from all subjects. In all subjects, a detailed neurologic evaluation was performed, including history, examination of force, deep tendon reflexes, plantar reflexes, and sensory modalities of light touch, pain, vibration, and proprioception. The patellar and ankle reflexes were graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal; and 111 5 hyperreflexia. In all subjects, sensory and motor NCS, EMG, and short-latency somatosensory evoked potentials (SSEPs) were performed using a Keypoint machine (Medtronic, Copenhagen, Denmark). A single electromyographer performed all examinations. Nerve Conduction Studies. Sensory NCS in sural and superficial fibular nerves; motor NCS in femoral, tibial, and fibular nerves; F-waves in tibial and fibular nerves; and H-reflexes in tibial nerve were studied bilaterally in all subjects. Upper extremity EDx in SCI
median and ulnar nerves were only examined in 1 subject, who was suspected of having polyneuropathy. The results were compared with the laboratory reference material. Limb temperatures were maintained at 32 C. For the femoral nerve, the stimulation site was the inguinal area (surface electrode), and recording sites were vastus medialis (VM), rectus femoris (RF), and vastus lateralis (VL) muscles, using concentric needles. For the tibial nerve, the stimulation sites were posterior to the medial malleolus and popliteal fossa, and recording sites were abductor hallucis (AH) muscle, using surface electrodes, and lateral gastrocnemius (GC) muscle using concentric needles. For H-reflexes, the recording site was the soleus muscle, and the stimulation site was the popliteal fossa. For the fibular nerve, motor NCS were performed with recordings from extensor digitorum brevis (EDB) with surface electrodes and from tibialis anterior (TA) and fibularis longus (PL) muscles with concentric needles. The fibular nerve was stimulated at the ankle with surface electrodes, 1–2 cm below the fibular head, and in the popliteal fossa with either surface electrodes or monopolar needle electrodes. Sensory NCS of the superficial fibular nerve were first performed using conventional surface electrode recording.14 For patients in whom a sensory nerve action potential could not be recorded or the suspicion of an entrapment neuropathy could not be ruled out or confirmed, the fibular nerve was examined using the near-nerve needle technique.15 Antidromic sensory NCS of the sural nerve were performed by conventional surface recording.16 In 1 subject, a sensory response could not be recorded by surface electrodes, and the near-nerve needle technique was performed.15,17,18 EMG was conducted bilaterally concentric 30-mm Dantec needle
Electromyography.
using
a
MUSCLE & NERVE
December 2015
1017
electrode. The muscles innervated by the same nerves were evaluated together, that is, TA, PL, and EDB as fibular-; GC and AH as tibial-; and VM, VL, and RF as femoral-nerve–innervated muscles, whereas iliopsoas and rectus abdominis were evaluated individually. The presence of abnormal SA, including fasciculation potentials and complex repetitive discharges, was detected at 10 separate sites. In each muscle, abnormal SA was graded as follows: “normal,” SA in 2 sites; “slightly increased,” SA in 3 sites; “moderately increased,” SA in 4 or 5 sites; and “markedly increased,” SA in 6 sites.19 The length of the nerves was estimated by measuring the distance from the inguinal area to the needle insertion site in each muscle examined. The rectus abdominis was not considered due to the presence of SA in only 1 subject and uncertainty of distance measurement. Short Latency Somatosensory Evoked Potentials. Bilateral tibial and fibular SSEPs were assessed in all patients. The tibial nerve was stimulated posterior to the medial malleolus, and the fibular nerve was stimulated anterior to the medial malleolus at the ankle. Peripheral responses were recorded at the popliteal fossa. Spinal responses were recorded at the T12 level, and cortical potentials (P40/N50 complex) were recorded at Cz0 . Stimulus intensity was adjusted to produce a clear muscle contraction. Two runs of responses of 50–100 stimulations were averaged and superimposed to ensure consistency. The results were compared with laboratory reference material. Analysis. The correlation between the amount of abnormal SA and nerve length was tested by regression analysis. The correlation between the amount of abnormal SA and reflex activity was tested using the Kruskal–Wallis 1-way analysis of variance. The difference between the incidence of fibs/PSWs was analyzed using the non-parametric Wilcoxon signed-rank test. The analysis comprised evaluation of fibs/PSWs in 20 different muscles from 12 subjects. Ten different sites were examined in each muscle. At each site, fibs, PSWs, or both were noted. The Spearman rank correlation test was used to assess the relationship between the amount of abnormal SA and years since injury. P < 0.05 was considered significant for all calculations. Data
RESULTS Nerve Conduction Studies. In all subjects, 1 or more NCS parameters were abnormal in at least 1 nerve. Fibular motor NCS were abnormal in 11 subjects, and 8 also had superficial fibular sensory NCS abnormalities. A localized fibular mononeur1018
EDx in SCI
opathy across the knee was found in 6 subjects, in 4 with the near-nerve technique and in 2 with surface recordings. One subject (subject 5) had bilateral non-localized fibular neuropathies as well as decreased sural sensory nerve action potential (SNAP) amplitudes and abnormal tibial motor NCS, whereas median and ulnar NCS were normal. In this subject, a bilateral sciatic nerve entrapment neuropathy was proposed rather than a polyneuropathy. In another subject, the fibular neuropathy with abnormal motor and sensory NCS could not be localized. In this subject, the near-nerve technique could not be applied due to an absent compound muscle action potential (CMAP). In the remaining 3 subjects, the only abnormal fibular nerve findings were decreased CMAP amplitudes. The tibial nerve was abnormal in 6 subjects; of these, 3 subjects had decreased CMAP amplitudes, and 4 had absent H-reflexes. Decreased amplitudes on 1 or both sides were the most frequent abnormality in the femoral nerve in 5 of the 7 subjects with abnormal NCS. The NCS results are summarized in Table S1 in the Supplementary Material (available online). Increased abnormal SA in at least 1 muscle was found in 9 subjects, whereas 3 subjects had no SA despite motor NCS abnormalities in the fibular, tibial, and femoral nerves. Abnormal SA in fibular-innervated muscles was seen in 8 subjects, and 6 also had SA in tibial- and femoral-nerve–innervated muscles. Two subjects had slightly increased abnormal SA in only fibularinnervated muscles, which could be explained by fibular nerve entrapment neuropathy. In 1 subject, there was slightly increased abnormal SA in only tibial-innervated muscles. Five subjects had abnormal SA in iliopsoas, and 1 subject had markedly increased abnormal SA in rectus abdominis as well as in all other muscles. The EMG data are summarized in the Supplementary Material (Table S1). Fibs and PSWs were the most frequent type of abnormal SA, seen in all 9 subjects with abnormal SA. There were more PSWs per muscle (mean 2.11 6 2.98) than fibs (mean 1.71 6 2.54) (P < 0.001). For each muscle, fibs, PSWs, or both were noted at 10 different sites. There were sites with only PSWs or fibs as well as sites with more PSWs than fibs or vice versa. In addition to fibs and PSWs, complex repetitive discharges were seen in 2 subjects and fasciculation potentials in 1 subject. There was no correlation between presence of fibs/PSWs in a muscle and the approximate length of its nerve when all data were pooled for fibs (r 5 0.05, P 5 0.454) and for PSWs (r 5 0.08, P 5 0.274); however, by exclusion of the muscles without significant abnormal SA, that is, SA in 2 Electromyography.
MUSCLE & NERVE
December 2015
FIGURE 1. Regression analysis between the amount of spontaneous activity in a muscle and the approximate length of that muscle’s nerve showed significant correlation with inclusion of the muscles with significant spontaneous activity for both fibs (–䊊–) (r 5 0.29, *P 5 0.038) and PSWs (---䉮---) (r 5 0.50, **P < 0.001). IP, iliopsoas; RF, rectus femoris; VL, vastus lateralis; VM, vastus medialis; TA, tibalis anterior; PL, peroneus longus; GC, gastrocnemius; EDB, extensor digitorum brevis; AH, abductor hallucis; fibs, fibrillation potentials; PSWs, positive sharp waves. The presence of fibs and PSWs was assessed at 10 separate sites.
sites, a significant negative correlation was found both for fibs (r 5 0.29, P 5 0.038) and PSWs (r 5 0.50, P < 0.001) (Fig. 1). There was a significant difference between the grade of reflex activity and the number of fibs (P < 0.001) and PSWs (P < 0.001), showing less SA with hyperreflexia for both patellar (Fig. 2a) and Achilles (Fig. 2b) reflexes. No significant correlation was found between level of SA and years since injury (P 5 0.080). Short Latency Somatosensory Evoked Potentials. Cortical responses were seen in only 1 subject (subject 3) with incomplete SCI. In this subject, tibial SSEPs showed normal peripheral and spinal latencies, whereas cortical latencies were normal on the right side but prolonged on the left. In another subject, in whom a bilateral sciatic nerve lesion was suspected, all peripheral, spinal, and cortical responses were absent in tibial SSEPs. In the remaining 10 subjects, there were normal peripheral latencies in tibial SSEPs. Fibular SSEPs showed absent or prolonged peripheral latencies on 1 or both sides in 8 subjects, corresponding with abnormal sensory NCS, but spinal and cortical responses were absent. DISCUSSION
A high degree of peripheral nerve dysfunction was shown with NCS and EMG in subjects with SCI. In all 12 subjects, NCS were abnormal in at EDx in SCI
least 1 nerve. Fibular nerve entrapment across the knee was seen in 6 subjects, and sciatic nerve entrapment was seen in 1. NCS changes involved almost entirely motor nerves, except for subjects with entrapment neuropathies. Nine of 12 subjects had SA in at least 1 muscle, and the SA was widespread in 6 subjects. Nerve Conduction Study Findings. The most commonly affected nerve was the fibular nerve (11 of 12 patients). This is consistent with previous reports that showed decreased or absent motor4,5,20 and sensory4,5 potentials. The fibular nerve was not stimulated below and above the knee to localize the pathology in any of those earlier studies. In our study there was localized fibular nerve entrapment across the knee in subjects with decreased or absent superficial fibular SNAPs, probably due to the advantage of using the nearnerve technique. We found normal sural nerves in all subjects, except for the subject in whom we suspected bilateral sciatic nerve lesions. Preserved sural nerves in our study are in contrast to earlier studies showing diminished or absent sural SNAPs in 75%20 and 64%5 of patients with chronic complete SCI, respectively. In 1 recent study, severe axonal sensorimotor peripheral neuropathy was seen in 10 patients in a cohort with pressure sores (n 5 15) and SCI.21 However, the peripheral neuropathy diagnosis was proposed without MUSCLE & NERVE
December 2015
1019
FIGURE 2. Fibs (䊊) and PSWs (䉮) were differed significantly in number in the various grades of reflex activity, showing less spontaneous activity with increased reflexes for (a) patellar reflex (P < 0.001) and (b) Achilles reflex (P < 0.001). Fibs and PSWs were assessed at 10 separate sites in 10 different muscles bilaterally. Reflexes were graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal reflex; and 111 5 hyperreflexia.
examination of upper extremity nerves, although bilateral sciatic neuropathy is equally likely. Bilateral sciatic neuropathy may be caused by pressure 1020
EDx in SCI
due to wheelchair use or because the patients are confined to bed. Our findings almost entirely involved the motor nerves, apart from entrapment MUSCLE & NERVE
December 2015
neuropathies. The changes in fibular, tibial, and femoral nerves were mainly decreased CMAP amplitudes with normal or slight changes in conduction velocities/latencies, indicating axonal loss.22 This is consistent with a possible mechanism where the axons surviving the transition from the early postinjury period to the chronic period conduct at normal rates.4,5,20 Spontaneous Activity in Electromyography. The presence of SA is widely accepted as pathognomonic for motor unit pathology. Previous studies in stroke1,2 and SCI4,5,20 showed abnormal SA. We found abnormal SA in 9 of 12 subjects. Of these, 2 subjects had abnormal SA only in fibularinnervated muscles, and 1 subject had abnormal SA only in tibial-innervated muscles. In the remaining 6 subjects (50%), there was widespread SA in fibular-, tibial-, and femoral-nerve–innervated muscles that could not be explained by entrapment neuropathy alone. The amount of abnormal SA we found is less than previously reported frequencies of 81% in 220 and 72% in 35 of the muscles examined. In those studies, the number of muscles examined was limited to 4 or 5, including the anterior tibial muscle, without taking fibular entrapment neuropathy into account. In some patients, there were motor NCS abnormalities without SA and vice versa. Abnormal SA with normal NCS may be explained by axonal loss with reinnervation, whereas decreased CMAP amplitude without SA may be due to chronicity of the lesion or disuse atrophy. In our study, PSWs were found more frequently than fibs. PSWs are usually associated with fibs and have the same clinical significance. The discrepancy in the occurrence of these 2 abnormal potentials may be related to the temporal order of change in membrane stability after axontomesis, or it may be associated with pathology of different anatomic localizations of the motor injury.23 Although we do not know whether the greater frequency of PSWs over fibs in our subjects is of any clinical significance, it may indicate a different mechanism of abnormal SA in UMN compared with LMN disorders that should be examined further in larger groups of patients. In earlier studies, abnormal SA was described in both the acute and chronic phases of SCI.4,5,24 We included only subjects with chronic SCI. We did not find a significant correlation between the length of time since the SCI and the degree of fibs/PSWs. However, we did find a negative correlation between the grade of fibs/PSWs and reflex activity, consistent with 2 earlier studies,4,6 yet this was disputed in 2 other studies.5,7 The afferent Ia volley and an intact peripheral nerve are necessary EDx in SCI
to elicit the reflex activity/spasticity that cannot be conveyed by an affected nerve. In patients with significant SA, we found a negative correlation between the length of the nerve innervating the muscle and the incidence of fibs/PSWs, in contrast to earlier studies which showed more fibs/PSWs in distal muscles than in proximal muscles.4,6 These studies did not examine foot muscles, where we found the lowest numbers of fibs/PSWs. However, the lower SA in foot muscles may also be because some foot muscles had fewer remaining muscle fibers, and hence lower SA. Somatosensory Evoked Potentials. Eight of 12 patients had complete injury, which was confirmed by absent cortical responses in fibular and tibial SSEPs in addition to clinical examination. In the remaining 4 patients with incomplete injury, cortical responses could be recorded in only 1 subject. This may have been due to uncertainty of the clinical examination or low sensitivity of SSEPs. Peripheral and spinal SSEP recordings contributed further to the evaluation of peripheral sensory function. Normal peripheral responses in tibial SSEPs in 11 subjects confirmed normal peripheral sensory function together with normal sural nerve NCS. However, fibular SSEP peripheral responses were absent or prolonged, which is consistent with the abnormal fibular sensory NCS. Several SSEP studies were conducted specifically in the acute phase of SCI and showed a prognostic value for functional recovery.25–27 To our knowledge, there have been no published studies of SSEPs in chronic SCI. Mechanisms
of
Peripheral
Nervous
System
The mechanism by which SA and NCS changes develop after SCI has not been fully elucidated. Although we could not provide supporting data, “transsynaptic degeneration” is the most often proposed hypothesis to account for peripheral denervation.4,5 When the motoneuron is deprived of trophic inputs and subsequently becomes inexcitable from disconnection due to the SCI, dysfunction of anterior horn cells may produce a disturbance of axonal flow, leading to axonal degeneration. Impaired axonal transport secondary to functional disturbance of anterior horn cells may lead to denervation, followed by SA and decreased CMAP amplitude. Also, there may be extensive SCI when the SCI is severe, and may include anterior horn cells and nerve roots. Our findings may also be explained by myelopathic damage involving anterior horn cells. In fact, the lower abnormal SA we found in association with hyperreflexia for both patellar and Achilles reflexes may favor concomitant anterior horn cell damage rather than transsynaptic degeneration. Involvement in Spinal Cord Injury.
MUSCLE & NERVE
December 2015
1021
In conclusion, the findings of abnormal NCS and/or EMG in all subjects with SCI in this study confirm that SCI has a significant impact on the PNS, in contrast to the classical view of peripheral nerves being spared. We found that 50% of patients had evidence of widespread denervation. Although our sample size was small, large numbers of nerves and muscles were studied in the paralyzed limbs. In addition, our study combined NCS, EMG, and SSEPs for a more comprehensive evaluation. Impaired peripheral nerve function increases the risk of pressure sores and may have crucial implications, particularly for rehabilitation interventions and length of hospital admission. Intact peripheral nerves are required for maintenance of motor function in centrally impaired muscle and for supportive therapies, such as functional electrical stimulation, to be efficient. Moreover, future breakthroughs in management of reestablishing descending cortical input and restoring useful movements will require preserved peripheral nerve function. Our study also highlights the importance of presurgical electrophysiologic evaluation, including neuroanastomosis. The high incidence of fibular entrapment neuropathy across the knee suggests that anastomosis of the fibular nerve/L5 root should be avoided. Further studies should be conducted to ascertain the underlying causes and mechanisms of LMN involvement in SCI. REFERENCES 1. Benecke R, Berthold A, Conrad B. Denervation activity in the EMG of patients with upper motor neuron lesions: time course, local distribution and pathogenetic aspects. J Neurol 1983;230:143–151. 2. Brown WF, Snow R. Denervation in hemiplegic muscles. Stroke 1990; 21:1700–1704. 3. Lukacs M. Electrophysiological signs of changes in motor units after ischaemic stroke. Clin Neurophysiol 2005;116:1566–1570. 4. Campbell JW, Herbison GJ, Chen YT, Jaweed MM, Gussner CG. Spontaneous electromyographic potentials in chronic spinal cord injured patients: relation to spasticity and length of nerve. Arch Phys Med Rehabil 1991;72:23–27. 5. Kirshblum S, Lim S, Garstang S, Millis S. Electrodiagnostic changes of the lower limbs in subjects with chronic complete cervical spinal cord injury. Arch Phys Med Rehabil 2001;82:604–607. 6. Cheng PT, Hong CZ, Liaw MY. Spontaneous electromyographic potentials in cervical cord-injured patients are related to dysesthetic pain. Am J Phys Med Rehabil 1997;76:389–394.
1022
EDx in SCI
7. Nyboer VJ, Johnson HE. Electromyographic findings in lower extremities of patients with traumatic quadriplegia. Arch Phys Med Rehabil 1971;52:256–259. 8. Xiao CG, Godec CJ. A possible new reflex pathway for micturition after spinal cord injury. Paraplegia 1994;32:300–307. 9. Dedeepiya VD, William JB, Parthiban JK, Chidambaram R, Balamurugan M, Kuroda S, et al. The known-unknowns in spinal cord injury, with emphasis on cell-based therapies—a review with suggestive arenas for research. Expert Opin Biol Ther 2014;14:617–634. 10. Xiao CG, Du MX, Dai C, Li B, Nitti VW, de Groat WC. An artificial somatic-central nervous system-autonomic reflex pathway for controllable micturition after spinal cord injury: preliminary results in 15 patients. J Urol 2003;170:1237–1241. 11. Rasmussen MM, Clemmensen D, Rawashdeh YF, Tankisi H, Christensen P, Krogh K. Surgical reinnervation with nerve anastomosis technique for neurogenic bladder and bowel dysfunction [in Danish]. Ugeskr Laeger 2011;173:2412–2415. 12. Rasmussen MM, Rawashdeh YF, Clemmensen D, Tankisi H, FuglsangFrederiksen A, Krogh K, et al. The artificial somato-autonomic reflex arch does not improve lower urinary tract function in patients with spinal cord lesions. J Urol 2015;193:598–604. 13. Kirshblum SC, Burns SP, Biering-Sorensen F, Donovan W, Graves DE, Jha A, et al. International standards for neurological classification of spinal cord injury (revised 2011). J Spinal Cord Med 2011;34: 535–546. 14. Oh SJ, Demirci M, Dajani B, Melo AC, Claussen GC. Distal sensory nerve conduction of the superficial peroneal nerve: new method and its clinical application. Muscle Nerve 2001;24:689–694. 15. Buchthal F, Rosenfalck A. Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 1966;3:1–122. 16. Schuchmann JA. Sural nerve conduction: a standardized technique. Arch Phys Med Rehabil 1977;58:166–168. 17. Tankisi H, Pugdahl K, Otto M, Fuglsang-Frederiksen A. Misinterpretation of sural nerve conduction studies due to anatomical variation. Clin Neurophysiol 2014;125:2115–2121. 18. Horowitz SH, Krarup C. Conduction studies of the normal sural nerve. Muscle Nerve 1992;15:374–383. 19. Tankisi H, Pugdahl K, Johnsen B, Fuglsang-Frederiksen A. Correlations of nerve conduction measures in axonal and demyelinating polyneuropathies. Clin Neurophysiol 2007;118:2383–2392. 20. Riley DA, Burns AS, Carrion-Jones M, Dillingham TR. Electrophysiological dysfunction in the peripheral nervous system following spinal cord injury. PM R 2011;3:419–425. 21. Kamradt T, Rasch C, Schuld C, Bottinger M, Murle B, Hensel C, et al. Spinal cord injury: association with axonal peripheral neuropathy in severely paralysed limbs. Eur J Neurol 2013;20:843–848. 22. Tankisi H, Pugdahl K, Fuglsang-Frederiksen A, Johnsen B, de Carvalho M, Fawcett PR, et al. Pathophysiology inferred from electrodiagnostic nerve tests and classification of polyneuropathies. Suggested guidelines. Clin Neurophysiol 2005;116:1571–1580. 23. Kraft GH. Are fibrillation potentials and positive sharp waves the same? No. Muscle Nerve 1996;19:216–220. 24. Boland RA, Lin CS, Engel S, Kiernan MC. Adaptation of motor function after spinal cord injury: novel insights into spinal shock. Brain 2011;134:495–505. 25. Kovindha A, Mahachai R. Short-latency somatosensory evoked potentials (SSEPs) of the tibial nerves in spinal cord injuries. Paraplegia 1992;30:502–506. 26. Spiess M, Schubert M, Kliesch U, Halder P. Evolution of tibial SSEP after traumatic spinal cord injury: baseline for clinical trials. Clin Neurophysiol 2008;119:1051–1061. 27. Seyal M, Mull B. Mechanisms of signal change during intraoperative somatosensory evoked potential monitoring of the spinal cord. J Clin Neurophysiol 2002;19:409–415.
MUSCLE & NERVE
December 2015
Department of Clinical Neurophysiology, Aarhus University Hospital, Nørrebrogade 44, DK-8000, Aarhus C, Denmark The Spinal Cord Research Centre, Department of Neurosurgery, Aarhus University Hospital, Aarhus, Denmark 3 Pelvic Floor Unit, Department of Surgery, Aarhus University Hospital, Aarhus, Denmark 4 Department of Urology, Aarhus University Hospital, Aarhus, Aarhus, Denmark 5 Neurogastroenterology Unit, Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus, Denmark 2
Accepted 3 March 2015 ABSTRACT: Introduction: Upper motor neuron disorders are believed to leave the peripheral nervous system (PNS) intact. In this study we examined whether there is evidence of PNS involvement in spinal cord injury (SCI). Methods: Twelve subjects with chronic low cervical or thoracic SCI were included prospectively. Needle electromyography was done in 10 different muscles in each subject bilaterally. Nerve conduction studies (NCS) were conducted in the fibular, tibial, and femoral motor and fibular and sural sensory nerves. Results: Half the subjects had widespread abnormal spontaneous activity (SA), and the amount of SA correlated inversely with reflex activity and nerve length. Fibular nerve entrapment across the knee was seen in 6 subjects, and sciatic nerve entrapment was seen in 1. Apart from entrapment neuropathies, NCS changes were found predominantly in motor nerves. Conclusion: The presence of widespread electrophysiologic changes outside entrapment sites indicates that SCI has a significant impact on the entire PNS, affecting the motor part predominantly. Muscle Nerve 52: 1016–1022, 2015
In needle electromyography (EMG), detection of abnormal spontaneous activity (SA) mainly in the form of fibrillation potentials (fibs) and positive sharp waves (PSWs) is the clinical “gold standard” for diagnosing denervated skeletal muscle. Although fibs and PSWs have been described in skeletal muscles in a series of reports in both cerebral1–3 and spinal cord lesions,4,5 it is still widely assumed among neurophysiologists that an upper motor neuron (UMN) lesion should leave the Additional Supporting Information may be found in the online version of this article. Abbreviations: AH, abductor hallucis; ASIA, American Spinal Injury Association; CMAP, compound muscle action potential; EDB, extensor digitorum brevis; EMG, electromyography; Fib, fibrillation potential; GC, gastrocnemius; LMN, lower motor neuron; NCS, nerve conduction studies; PL, peroneus longus; PNS, peripheral nervous system; PSW, positive sharp wave; RF, rectus femoris; SA, spontaneous activity; SCI, spinal cord injury; SNAP, sensory nerve action potential; SSEP, short latency somatosensory evoked potential; TA, tibialis anterior; UMN, upper motor neuron; VL, vastus lateralis; VM, vastus medialis Key words: electromyography; nerve conduction study; peripheral nervous system; spinal cord injury; spontaneous activity This work was supported in part by grants from the Lundbeck Foundation, Copenhagen, Denmark. Correspondence to: H. Tankisi; e-mail: [email protected] C 2015 Wiley Periodicals, Inc. V
Published online 10 March 2015 in Wiley Online Library (wileyonlinelibrary. com). DOI 10.1002/mus.24644
1016
EDx in SCI
lower motor neuron (LMN) anatomically intact, with a normal axon extending to the periphery. Overall, there are inconsistencies in studies of abnormal SA in patients with spinal cord injury (SCI). Some studies of SCI showed a larger number of fibs/PSWs in distal muscles compared with proximal muscles, suggesting a positive correlation between the degree of fibs/PSWs and the length of the peripheral axon,4,6 whereas, in another study, there was no difference in the number of fibs/PSWs between proximal and distal muscles.5 With regard to fibs/PSWs and degree of reflex activity/spasticity, some studies found a negative correlation,4,6 whereas others did not confirm this correlation.5,7 Because new therapies and rehabilitation interventions, that is, functional electrical stimulation, neuroanastomosis,8 and cell- and stem-cell–based therapies,9 for UMN disorders are being developed, confirming normal peripheral nerve function is crucial. This study was done prospectively as part of the preoperative evaluation of the peripheral nervous system (PNS) in subjects with complete or incomplete chronic SCI who were undergoing neuroanastomosis of L5 and S2 ventral nerve roots to obtain bladder and eventually bowel control.8,10–12 The aims of the study were to: (1) examine for LMN involvement other than nerve entrapments in subjects with SCI by determining the pattern of nerve conduction study (NCS) changes and the frequency of abnormal SA; (2) assess whether there is a correlation between the degree of fibs/ PSWs and reflex activity, and (3) determine whether the length of the axon plays a role in the pattern of fibs/PSWs by comparing proximal and distal muscles. METHODS
This study included 12 subjects (11 men and 1 woman) with a mean age of 41.7 years (range 19– 63 years). All subjects had traumatic chronic (>1 year postinjury) cervical or thoracic SCI class A or B on the American Spinal Injury Association (ASIA) scale.13 Mean duration of time since injury MUSCLE & NERVE
December 2015
Table 1. Subjects’ demographics. Patellar reflexes Subject 1 2 3 4 5 6 7 8 9 10 11 12
Years since injury 2 3 5 5 5 1 19 17 13 2 2 2
Gender/age (y) M/55 M/63 M/22 M/36 M/35 W/19 M/62 M/52 M/31 M/34 M/45 M/47
Level Th9 C5 C7 Th3 C4 Th10 Th9 Th4 C8 C5 C7 Th8
AIS score A B B A A A A A B B A A
R – 111 111 – – – 111 111 111 111 111 1
L – 111 111 – – – 111 111 111 111 111 1
Achilles reflexes R
L
– 111 111 111 – – 11 111 111 111 – 11
– 111 111 111 – – 11 111 111 111 – 11
M, man; W, woman; R, right; L, left. Reflexes graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal reflex; and 111 5 hyperreflexia.
was 6.3 (range 1–19) years. Eight subjects were diagnosed with complete injury (ASIA class A), and 4 subjects had incomplete sensory injury (ASIA class B) by clinical examination (Table 1). None of the subjects had a pre-existing neurologic condition, that is, peripheral neuropathy, or were taking medications known to cause peripheral neuropathy, lower limb nerve injury, low back injury, or lower limb fracture before or after the SCI. Peripheral nerve disease, such as root or lumbosacral plexus lesions, was excluded by detailed radiologic examination with MRI and X-ray. All patients had MRI of the cervical, thoracic, and lumbar spine as well as the lumbosacral plexus as a part of the initial investigation protocol. These examinations were re-evaluated by 2 of the authors (M.M.R. and D.C., both neurosurgeons), who, if necessary, also consulted radiologists on the scans after the electrophysiologic examinations in the patients who had widespread severe fibs/PSWs. Informed written consent was obtained from all subjects. In all subjects, a detailed neurologic evaluation was performed, including history, examination of force, deep tendon reflexes, plantar reflexes, and sensory modalities of light touch, pain, vibration, and proprioception. The patellar and ankle reflexes were graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal; and 111 5 hyperreflexia. In all subjects, sensory and motor NCS, EMG, and short-latency somatosensory evoked potentials (SSEPs) were performed using a Keypoint machine (Medtronic, Copenhagen, Denmark). A single electromyographer performed all examinations. Nerve Conduction Studies. Sensory NCS in sural and superficial fibular nerves; motor NCS in femoral, tibial, and fibular nerves; F-waves in tibial and fibular nerves; and H-reflexes in tibial nerve were studied bilaterally in all subjects. Upper extremity EDx in SCI
median and ulnar nerves were only examined in 1 subject, who was suspected of having polyneuropathy. The results were compared with the laboratory reference material. Limb temperatures were maintained at 32 C. For the femoral nerve, the stimulation site was the inguinal area (surface electrode), and recording sites were vastus medialis (VM), rectus femoris (RF), and vastus lateralis (VL) muscles, using concentric needles. For the tibial nerve, the stimulation sites were posterior to the medial malleolus and popliteal fossa, and recording sites were abductor hallucis (AH) muscle, using surface electrodes, and lateral gastrocnemius (GC) muscle using concentric needles. For H-reflexes, the recording site was the soleus muscle, and the stimulation site was the popliteal fossa. For the fibular nerve, motor NCS were performed with recordings from extensor digitorum brevis (EDB) with surface electrodes and from tibialis anterior (TA) and fibularis longus (PL) muscles with concentric needles. The fibular nerve was stimulated at the ankle with surface electrodes, 1–2 cm below the fibular head, and in the popliteal fossa with either surface electrodes or monopolar needle electrodes. Sensory NCS of the superficial fibular nerve were first performed using conventional surface electrode recording.14 For patients in whom a sensory nerve action potential could not be recorded or the suspicion of an entrapment neuropathy could not be ruled out or confirmed, the fibular nerve was examined using the near-nerve needle technique.15 Antidromic sensory NCS of the sural nerve were performed by conventional surface recording.16 In 1 subject, a sensory response could not be recorded by surface electrodes, and the near-nerve needle technique was performed.15,17,18 EMG was conducted bilaterally concentric 30-mm Dantec needle
Electromyography.
using
a
MUSCLE & NERVE
December 2015
1017
electrode. The muscles innervated by the same nerves were evaluated together, that is, TA, PL, and EDB as fibular-; GC and AH as tibial-; and VM, VL, and RF as femoral-nerve–innervated muscles, whereas iliopsoas and rectus abdominis were evaluated individually. The presence of abnormal SA, including fasciculation potentials and complex repetitive discharges, was detected at 10 separate sites. In each muscle, abnormal SA was graded as follows: “normal,” SA in 2 sites; “slightly increased,” SA in 3 sites; “moderately increased,” SA in 4 or 5 sites; and “markedly increased,” SA in 6 sites.19 The length of the nerves was estimated by measuring the distance from the inguinal area to the needle insertion site in each muscle examined. The rectus abdominis was not considered due to the presence of SA in only 1 subject and uncertainty of distance measurement. Short Latency Somatosensory Evoked Potentials. Bilateral tibial and fibular SSEPs were assessed in all patients. The tibial nerve was stimulated posterior to the medial malleolus, and the fibular nerve was stimulated anterior to the medial malleolus at the ankle. Peripheral responses were recorded at the popliteal fossa. Spinal responses were recorded at the T12 level, and cortical potentials (P40/N50 complex) were recorded at Cz0 . Stimulus intensity was adjusted to produce a clear muscle contraction. Two runs of responses of 50–100 stimulations were averaged and superimposed to ensure consistency. The results were compared with laboratory reference material. Analysis. The correlation between the amount of abnormal SA and nerve length was tested by regression analysis. The correlation between the amount of abnormal SA and reflex activity was tested using the Kruskal–Wallis 1-way analysis of variance. The difference between the incidence of fibs/PSWs was analyzed using the non-parametric Wilcoxon signed-rank test. The analysis comprised evaluation of fibs/PSWs in 20 different muscles from 12 subjects. Ten different sites were examined in each muscle. At each site, fibs, PSWs, or both were noted. The Spearman rank correlation test was used to assess the relationship between the amount of abnormal SA and years since injury. P < 0.05 was considered significant for all calculations. Data
RESULTS Nerve Conduction Studies. In all subjects, 1 or more NCS parameters were abnormal in at least 1 nerve. Fibular motor NCS were abnormal in 11 subjects, and 8 also had superficial fibular sensory NCS abnormalities. A localized fibular mononeur1018
EDx in SCI
opathy across the knee was found in 6 subjects, in 4 with the near-nerve technique and in 2 with surface recordings. One subject (subject 5) had bilateral non-localized fibular neuropathies as well as decreased sural sensory nerve action potential (SNAP) amplitudes and abnormal tibial motor NCS, whereas median and ulnar NCS were normal. In this subject, a bilateral sciatic nerve entrapment neuropathy was proposed rather than a polyneuropathy. In another subject, the fibular neuropathy with abnormal motor and sensory NCS could not be localized. In this subject, the near-nerve technique could not be applied due to an absent compound muscle action potential (CMAP). In the remaining 3 subjects, the only abnormal fibular nerve findings were decreased CMAP amplitudes. The tibial nerve was abnormal in 6 subjects; of these, 3 subjects had decreased CMAP amplitudes, and 4 had absent H-reflexes. Decreased amplitudes on 1 or both sides were the most frequent abnormality in the femoral nerve in 5 of the 7 subjects with abnormal NCS. The NCS results are summarized in Table S1 in the Supplementary Material (available online). Increased abnormal SA in at least 1 muscle was found in 9 subjects, whereas 3 subjects had no SA despite motor NCS abnormalities in the fibular, tibial, and femoral nerves. Abnormal SA in fibular-innervated muscles was seen in 8 subjects, and 6 also had SA in tibial- and femoral-nerve–innervated muscles. Two subjects had slightly increased abnormal SA in only fibularinnervated muscles, which could be explained by fibular nerve entrapment neuropathy. In 1 subject, there was slightly increased abnormal SA in only tibial-innervated muscles. Five subjects had abnormal SA in iliopsoas, and 1 subject had markedly increased abnormal SA in rectus abdominis as well as in all other muscles. The EMG data are summarized in the Supplementary Material (Table S1). Fibs and PSWs were the most frequent type of abnormal SA, seen in all 9 subjects with abnormal SA. There were more PSWs per muscle (mean 2.11 6 2.98) than fibs (mean 1.71 6 2.54) (P < 0.001). For each muscle, fibs, PSWs, or both were noted at 10 different sites. There were sites with only PSWs or fibs as well as sites with more PSWs than fibs or vice versa. In addition to fibs and PSWs, complex repetitive discharges were seen in 2 subjects and fasciculation potentials in 1 subject. There was no correlation between presence of fibs/PSWs in a muscle and the approximate length of its nerve when all data were pooled for fibs (r 5 0.05, P 5 0.454) and for PSWs (r 5 0.08, P 5 0.274); however, by exclusion of the muscles without significant abnormal SA, that is, SA in 2 Electromyography.
MUSCLE & NERVE
December 2015
FIGURE 1. Regression analysis between the amount of spontaneous activity in a muscle and the approximate length of that muscle’s nerve showed significant correlation with inclusion of the muscles with significant spontaneous activity for both fibs (–䊊–) (r 5 0.29, *P 5 0.038) and PSWs (---䉮---) (r 5 0.50, **P < 0.001). IP, iliopsoas; RF, rectus femoris; VL, vastus lateralis; VM, vastus medialis; TA, tibalis anterior; PL, peroneus longus; GC, gastrocnemius; EDB, extensor digitorum brevis; AH, abductor hallucis; fibs, fibrillation potentials; PSWs, positive sharp waves. The presence of fibs and PSWs was assessed at 10 separate sites.
sites, a significant negative correlation was found both for fibs (r 5 0.29, P 5 0.038) and PSWs (r 5 0.50, P < 0.001) (Fig. 1). There was a significant difference between the grade of reflex activity and the number of fibs (P < 0.001) and PSWs (P < 0.001), showing less SA with hyperreflexia for both patellar (Fig. 2a) and Achilles (Fig. 2b) reflexes. No significant correlation was found between level of SA and years since injury (P 5 0.080). Short Latency Somatosensory Evoked Potentials. Cortical responses were seen in only 1 subject (subject 3) with incomplete SCI. In this subject, tibial SSEPs showed normal peripheral and spinal latencies, whereas cortical latencies were normal on the right side but prolonged on the left. In another subject, in whom a bilateral sciatic nerve lesion was suspected, all peripheral, spinal, and cortical responses were absent in tibial SSEPs. In the remaining 10 subjects, there were normal peripheral latencies in tibial SSEPs. Fibular SSEPs showed absent or prolonged peripheral latencies on 1 or both sides in 8 subjects, corresponding with abnormal sensory NCS, but spinal and cortical responses were absent. DISCUSSION
A high degree of peripheral nerve dysfunction was shown with NCS and EMG in subjects with SCI. In all 12 subjects, NCS were abnormal in at EDx in SCI
least 1 nerve. Fibular nerve entrapment across the knee was seen in 6 subjects, and sciatic nerve entrapment was seen in 1. NCS changes involved almost entirely motor nerves, except for subjects with entrapment neuropathies. Nine of 12 subjects had SA in at least 1 muscle, and the SA was widespread in 6 subjects. Nerve Conduction Study Findings. The most commonly affected nerve was the fibular nerve (11 of 12 patients). This is consistent with previous reports that showed decreased or absent motor4,5,20 and sensory4,5 potentials. The fibular nerve was not stimulated below and above the knee to localize the pathology in any of those earlier studies. In our study there was localized fibular nerve entrapment across the knee in subjects with decreased or absent superficial fibular SNAPs, probably due to the advantage of using the nearnerve technique. We found normal sural nerves in all subjects, except for the subject in whom we suspected bilateral sciatic nerve lesions. Preserved sural nerves in our study are in contrast to earlier studies showing diminished or absent sural SNAPs in 75%20 and 64%5 of patients with chronic complete SCI, respectively. In 1 recent study, severe axonal sensorimotor peripheral neuropathy was seen in 10 patients in a cohort with pressure sores (n 5 15) and SCI.21 However, the peripheral neuropathy diagnosis was proposed without MUSCLE & NERVE
December 2015
1019
FIGURE 2. Fibs (䊊) and PSWs (䉮) were differed significantly in number in the various grades of reflex activity, showing less spontaneous activity with increased reflexes for (a) patellar reflex (P < 0.001) and (b) Achilles reflex (P < 0.001). Fibs and PSWs were assessed at 10 separate sites in 10 different muscles bilaterally. Reflexes were graded as follows: – 5 areflexia; 1 5 hyporeflexia; 11 5 normal reflex; and 111 5 hyperreflexia.
examination of upper extremity nerves, although bilateral sciatic neuropathy is equally likely. Bilateral sciatic neuropathy may be caused by pressure 1020
EDx in SCI
due to wheelchair use or because the patients are confined to bed. Our findings almost entirely involved the motor nerves, apart from entrapment MUSCLE & NERVE
December 2015
neuropathies. The changes in fibular, tibial, and femoral nerves were mainly decreased CMAP amplitudes with normal or slight changes in conduction velocities/latencies, indicating axonal loss.22 This is consistent with a possible mechanism where the axons surviving the transition from the early postinjury period to the chronic period conduct at normal rates.4,5,20 Spontaneous Activity in Electromyography. The presence of SA is widely accepted as pathognomonic for motor unit pathology. Previous studies in stroke1,2 and SCI4,5,20 showed abnormal SA. We found abnormal SA in 9 of 12 subjects. Of these, 2 subjects had abnormal SA only in fibularinnervated muscles, and 1 subject had abnormal SA only in tibial-innervated muscles. In the remaining 6 subjects (50%), there was widespread SA in fibular-, tibial-, and femoral-nerve–innervated muscles that could not be explained by entrapment neuropathy alone. The amount of abnormal SA we found is less than previously reported frequencies of 81% in 220 and 72% in 35 of the muscles examined. In those studies, the number of muscles examined was limited to 4 or 5, including the anterior tibial muscle, without taking fibular entrapment neuropathy into account. In some patients, there were motor NCS abnormalities without SA and vice versa. Abnormal SA with normal NCS may be explained by axonal loss with reinnervation, whereas decreased CMAP amplitude without SA may be due to chronicity of the lesion or disuse atrophy. In our study, PSWs were found more frequently than fibs. PSWs are usually associated with fibs and have the same clinical significance. The discrepancy in the occurrence of these 2 abnormal potentials may be related to the temporal order of change in membrane stability after axontomesis, or it may be associated with pathology of different anatomic localizations of the motor injury.23 Although we do not know whether the greater frequency of PSWs over fibs in our subjects is of any clinical significance, it may indicate a different mechanism of abnormal SA in UMN compared with LMN disorders that should be examined further in larger groups of patients. In earlier studies, abnormal SA was described in both the acute and chronic phases of SCI.4,5,24 We included only subjects with chronic SCI. We did not find a significant correlation between the length of time since the SCI and the degree of fibs/PSWs. However, we did find a negative correlation between the grade of fibs/PSWs and reflex activity, consistent with 2 earlier studies,4,6 yet this was disputed in 2 other studies.5,7 The afferent Ia volley and an intact peripheral nerve are necessary EDx in SCI
to elicit the reflex activity/spasticity that cannot be conveyed by an affected nerve. In patients with significant SA, we found a negative correlation between the length of the nerve innervating the muscle and the incidence of fibs/PSWs, in contrast to earlier studies which showed more fibs/PSWs in distal muscles than in proximal muscles.4,6 These studies did not examine foot muscles, where we found the lowest numbers of fibs/PSWs. However, the lower SA in foot muscles may also be because some foot muscles had fewer remaining muscle fibers, and hence lower SA. Somatosensory Evoked Potentials. Eight of 12 patients had complete injury, which was confirmed by absent cortical responses in fibular and tibial SSEPs in addition to clinical examination. In the remaining 4 patients with incomplete injury, cortical responses could be recorded in only 1 subject. This may have been due to uncertainty of the clinical examination or low sensitivity of SSEPs. Peripheral and spinal SSEP recordings contributed further to the evaluation of peripheral sensory function. Normal peripheral responses in tibial SSEPs in 11 subjects confirmed normal peripheral sensory function together with normal sural nerve NCS. However, fibular SSEP peripheral responses were absent or prolonged, which is consistent with the abnormal fibular sensory NCS. Several SSEP studies were conducted specifically in the acute phase of SCI and showed a prognostic value for functional recovery.25–27 To our knowledge, there have been no published studies of SSEPs in chronic SCI. Mechanisms
of
Peripheral
Nervous
System
The mechanism by which SA and NCS changes develop after SCI has not been fully elucidated. Although we could not provide supporting data, “transsynaptic degeneration” is the most often proposed hypothesis to account for peripheral denervation.4,5 When the motoneuron is deprived of trophic inputs and subsequently becomes inexcitable from disconnection due to the SCI, dysfunction of anterior horn cells may produce a disturbance of axonal flow, leading to axonal degeneration. Impaired axonal transport secondary to functional disturbance of anterior horn cells may lead to denervation, followed by SA and decreased CMAP amplitude. Also, there may be extensive SCI when the SCI is severe, and may include anterior horn cells and nerve roots. Our findings may also be explained by myelopathic damage involving anterior horn cells. In fact, the lower abnormal SA we found in association with hyperreflexia for both patellar and Achilles reflexes may favor concomitant anterior horn cell damage rather than transsynaptic degeneration. Involvement in Spinal Cord Injury.
MUSCLE & NERVE
December 2015
1021
In conclusion, the findings of abnormal NCS and/or EMG in all subjects with SCI in this study confirm that SCI has a significant impact on the PNS, in contrast to the classical view of peripheral nerves being spared. We found that 50% of patients had evidence of widespread denervation. Although our sample size was small, large numbers of nerves and muscles were studied in the paralyzed limbs. In addition, our study combined NCS, EMG, and SSEPs for a more comprehensive evaluation. Impaired peripheral nerve function increases the risk of pressure sores and may have crucial implications, particularly for rehabilitation interventions and length of hospital admission. Intact peripheral nerves are required for maintenance of motor function in centrally impaired muscle and for supportive therapies, such as functional electrical stimulation, to be efficient. Moreover, future breakthroughs in management of reestablishing descending cortical input and restoring useful movements will require preserved peripheral nerve function. Our study also highlights the importance of presurgical electrophysiologic evaluation, including neuroanastomosis. The high incidence of fibular entrapment neuropathy across the knee suggests that anastomosis of the fibular nerve/L5 root should be avoided. Further studies should be conducted to ascertain the underlying causes and mechanisms of LMN involvement in SCI. REFERENCES 1. Benecke R, Berthold A, Conrad B. Denervation activity in the EMG of patients with upper motor neuron lesions: time course, local distribution and pathogenetic aspects. J Neurol 1983;230:143–151. 2. Brown WF, Snow R. Denervation in hemiplegic muscles. Stroke 1990; 21:1700–1704. 3. Lukacs M. Electrophysiological signs of changes in motor units after ischaemic stroke. Clin Neurophysiol 2005;116:1566–1570. 4. Campbell JW, Herbison GJ, Chen YT, Jaweed MM, Gussner CG. Spontaneous electromyographic potentials in chronic spinal cord injured patients: relation to spasticity and length of nerve. Arch Phys Med Rehabil 1991;72:23–27. 5. Kirshblum S, Lim S, Garstang S, Millis S. Electrodiagnostic changes of the lower limbs in subjects with chronic complete cervical spinal cord injury. Arch Phys Med Rehabil 2001;82:604–607. 6. Cheng PT, Hong CZ, Liaw MY. Spontaneous electromyographic potentials in cervical cord-injured patients are related to dysesthetic pain. Am J Phys Med Rehabil 1997;76:389–394.
1022
EDx in SCI
7. Nyboer VJ, Johnson HE. Electromyographic findings in lower extremities of patients with traumatic quadriplegia. Arch Phys Med Rehabil 1971;52:256–259. 8. Xiao CG, Godec CJ. A possible new reflex pathway for micturition after spinal cord injury. Paraplegia 1994;32:300–307. 9. Dedeepiya VD, William JB, Parthiban JK, Chidambaram R, Balamurugan M, Kuroda S, et al. The known-unknowns in spinal cord injury, with emphasis on cell-based therapies—a review with suggestive arenas for research. Expert Opin Biol Ther 2014;14:617–634. 10. Xiao CG, Du MX, Dai C, Li B, Nitti VW, de Groat WC. An artificial somatic-central nervous system-autonomic reflex pathway for controllable micturition after spinal cord injury: preliminary results in 15 patients. J Urol 2003;170:1237–1241. 11. Rasmussen MM, Clemmensen D, Rawashdeh YF, Tankisi H, Christensen P, Krogh K. Surgical reinnervation with nerve anastomosis technique for neurogenic bladder and bowel dysfunction [in Danish]. Ugeskr Laeger 2011;173:2412–2415. 12. Rasmussen MM, Rawashdeh YF, Clemmensen D, Tankisi H, FuglsangFrederiksen A, Krogh K, et al. The artificial somato-autonomic reflex arch does not improve lower urinary tract function in patients with spinal cord lesions. J Urol 2015;193:598–604. 13. Kirshblum SC, Burns SP, Biering-Sorensen F, Donovan W, Graves DE, Jha A, et al. International standards for neurological classification of spinal cord injury (revised 2011). J Spinal Cord Med 2011;34: 535–546. 14. Oh SJ, Demirci M, Dajani B, Melo AC, Claussen GC. Distal sensory nerve conduction of the superficial peroneal nerve: new method and its clinical application. Muscle Nerve 2001;24:689–694. 15. Buchthal F, Rosenfalck A. Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 1966;3:1–122. 16. Schuchmann JA. Sural nerve conduction: a standardized technique. Arch Phys Med Rehabil 1977;58:166–168. 17. Tankisi H, Pugdahl K, Otto M, Fuglsang-Frederiksen A. Misinterpretation of sural nerve conduction studies due to anatomical variation. Clin Neurophysiol 2014;125:2115–2121. 18. Horowitz SH, Krarup C. Conduction studies of the normal sural nerve. Muscle Nerve 1992;15:374–383. 19. Tankisi H, Pugdahl K, Johnsen B, Fuglsang-Frederiksen A. Correlations of nerve conduction measures in axonal and demyelinating polyneuropathies. Clin Neurophysiol 2007;118:2383–2392. 20. Riley DA, Burns AS, Carrion-Jones M, Dillingham TR. Electrophysiological dysfunction in the peripheral nervous system following spinal cord injury. PM R 2011;3:419–425. 21. Kamradt T, Rasch C, Schuld C, Bottinger M, Murle B, Hensel C, et al. Spinal cord injury: association with axonal peripheral neuropathy in severely paralysed limbs. Eur J Neurol 2013;20:843–848. 22. Tankisi H, Pugdahl K, Fuglsang-Frederiksen A, Johnsen B, de Carvalho M, Fawcett PR, et al. Pathophysiology inferred from electrodiagnostic nerve tests and classification of polyneuropathies. Suggested guidelines. Clin Neurophysiol 2005;116:1571–1580. 23. Kraft GH. Are fibrillation potentials and positive sharp waves the same? No. Muscle Nerve 1996;19:216–220. 24. Boland RA, Lin CS, Engel S, Kiernan MC. Adaptation of motor function after spinal cord injury: novel insights into spinal shock. Brain 2011;134:495–505. 25. Kovindha A, Mahachai R. Short-latency somatosensory evoked potentials (SSEPs) of the tibial nerves in spinal cord injuries. Paraplegia 1992;30:502–506. 26. Spiess M, Schubert M, Kliesch U, Halder P. Evolution of tibial SSEP after traumatic spinal cord injury: baseline for clinical trials. Clin Neurophysiol 2008;119:1051–1061. 27. Seyal M, Mull B. Mechanisms of signal change during intraoperative somatosensory evoked potential monitoring of the spinal cord. J Clin Neurophysiol 2002;19:409–415.
MUSCLE & NERVE
December 2015
