- Email: [email protected]
Unexpected Recovery After Robotic Locomotor Training at Physiologic Stepping Speed: A Single-Case Design
1476
CLINICAL IMPLICATIONS OF BASIC RESEARCH
Unexpected Recovery After Robotic Locomotor Training at Physiologic Stepping Speed: A Single-Case Design Martina R. Spiess, PT, MPTSc, Jeffrey P. Jaramillo, PT, MS, Andrea L. Behrman, PT, PhD, Jeffrey K. Teraoka, MD, Carolynn Patten, PhD, PT ABSTRACT. Spiess MR, Jaramillo JP, Behrman AL, Teraoka JK, Patten C. Unexpected recovery after robotic locomotor training at physiologic stepping speed: a single-case design. Arch Phys Med Rehabil 2012;93:1476-84. Objectives: To investigate the effect of walking speed on the emergence of locomotor electromyogram (EMG) patterns in an individual with chronic incomplete spinal cord injury (SCI), and to determine whether central pattern generator activity during robotic locomotor training (RLT) transfers to volitional EMG activity during overground walking. Design: Single-case (B-A-B; experimental treatment–withdrawal– experimental treatment) design. Setting: Freestanding rehabilitation research center. Participant: A 50-year-old man who was nonambulatory for 16 months after incomplete SCI (sub-T11). Interventions: The participant completed two 6-week blocks of RLT, training 4 times per week for 30 minutes per session at walking speeds up to 5km/h (1.4m/s) over continuous bouts lasting up to 17 minutes. Main Outcome Measures: Surface EMG was recorded weekly during RLT and overground walking. The Walking Index for Spinal Cord Injury (WISCI-II) was assessed daily during training blocks. Results: During week 4, reciprocal, patterned EMG emerged during RLT. EMG amplitude modulation revealed a curvilinear relationship over the range of walking speeds from 1.5 to 5km/h (1.4m/s). Functionally, the participant improved from being nonambulatory (WISCI-II 1/20), to walking overground with reciprocal stepping using knee-ankle-foot orthoses and a walker (WISCI-II 9/20). EMG was also observed during overground walking. These functional gains were maintained greater than 4 years after locomotor training (LT). Conclusions: Here we report an unexpected course of locomotor recovery in an individual with chronic incomplete SCI. Through RLT at physiologic walking speeds, it was possible to
From the Brain Rehabilitation Research Center, Malcom Randall VAMC, Gainesville, FL (Patten, Spiess); the Department of Physical Therapy, University of Florida, Gainesville, FL (Patten, Spiess, Behrman); VA Palo Alto HCS, Palo Alto, CA (Jaramillo, Teraoka); and Division of Physical Medicine and Rehabilitation, Stanford University School of Medicine, Stanford, CA (Teraoka). Presented to the 4th National Spinal Cord Injury Conference, sponsored by Toronto Rehabilitation Hospital, October 28-30, 2010, Niagara Falls, Ontario, Canada; and to the International Society for Posture and Gait Research Conference, July 14-18, 2007, Burlington, VT. Supported by the Department of Veterans Affairs Rehabilitation Research & Development Service (Project no. B540231) and Research Career Scientist Award (F7823S); and by a University of Florida Alumni Fellowship. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Reprint requests to Carolynn Patten, PhD, PT, VA Brain Rehabilitation Research Center, 1601 SW Archer Rd, 151A, Gainesville, FL 32608, e-mail: [email protected]. In-press corrected proof published online on May 8, 2012, at www.archives-pmr.org. 0003-9993/12/9308-01018$36.00/0 http://dx.doi.org/10.1016/j.apmr.2012.02.030
Arch Phys Med Rehabil Vol 93, August 2012
activate the central pattern generator even 16 months postinjury. Further, to a certain degree, improvements from RLT transferred to overground walking. Our results suggest that LT-induced changes affect the central pattern generator and allow supraspinal inputs to engage residual spinal pathways. Key Words: Locomotion; Neurophysiology; Recovery of function; Rehabilitation; Spinal cord injuries. © 2012 by the American Congress of Rehabilitation Medicine OCOMOTOR TRAINING (LT) has been shown to proL mote recovery of reciprocal locomotor electromyogram (EMG) patterns after spinal cord injury (SCI), even in individ-
uals who are unable to produce voluntary movement.1,2 Under attenuation or absence of supraspinal inputs, it is generally agreed that these EMG patterns are elicited by a spinal central pattern generator3-6 and are influenced by changes in somatosensory feedback resulting from alterations in stepping speed, loading, or joint range of motion induced during locomotion. For example, higher treadmill speeds facilitate both the emergence of patterned EMG and increased EMG amplitude.7-9 However, the feasibility of delivering manually assisted LT at higher treadmill/stepping speeds is limited because it is strenuous and very labor intensive.10 Robotic devices were developed as one approach to address these challenges. Robotic locomotor training (RLT) devices, such as the Lokomat,a offer assistance with limb movement over sustained periods, even at higher stepping speeds.11-13 Lunenburger et al14 demonstrated increased EMG activity in complete SCI patients as a function of speed when walking between 1.5 and 2.5km/h in the Lokomat. With technical advances and updates, higher speeds (5km/h), approaching normal overground walking speed, are now possible in the robot. The effect of training at these higher stepping speeds (eg, between 2.5 and 5km/h) on EMG activity has not been investigated systematically in persons with SCI. In acutely injured individuals, the capacity to produce patterned EMG activity during assisted locomotion seems ubiq-
List of Abbreviations AIS ASIA EMG ISNCSCI LT MR MRI RLT SCI WISCI-II
American Spinal Injury Association Impairment Scale American Spinal Injury Association electromyogram International Standards for Neurological Classification of Spinal Cord Injury locomotor training magnetic resonance magnetic resonance imaging robotic locomotor training spinal cord injury Walking Index for Spinal Cord Injury II
1477
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
uitous, even in individuals with no voluntary control of their lower limbs. However, this capacity does not translate to functional gains in all cases. Some individuals, especially those with incomplete SCI, reveal significant gains in walking function after different forms of LT.15-18 However, in other human subjects—mostly those with complete SCI—there is no evidence reporting the capacity to reproduce stepping under conditions of full body weight-bearing.7,8,14 As a result, it is not possible to conclude that the presence of spinally mediated EMG activity leads to functional gains in human subjects with complete SCI. This phenomenon stands in contrast to the animal literature19-22 and suggests that despite preservation of the central pattern generator during the transition from quadrupedalism to bipedalism, there might be a higher dependence of supraspinal input during bipedal gait, especially when walking overground.7,23,24 Overground stepping does not provide the same sensory inputs as walking on a treadmill, even in healthy individuals.25,26 Additionally, in impaired populations, volitional overground stepping typically occurs at much slower speeds than treadmill stepping and, in part because of the use of assistive devices, it is more difficult to achieve an upright posture and adequate hip extension.27 These factors compromise locomotor-relevant sensory feedback and may therefore place higher demand on supraspinal inputs. It remains unclear whether, and to what extent, patterned EMG activity acquired during treadmill/LT translates to overground stepping. Here we explored the effect of RLT on an individual with incomplete SCI who was nonambulatory 16 months postinjury. Our aims in this single-case design were to determine (1) whether initially absent EMG activity could be elicited during robotic-driven locomotion and how it would adapt over the course of LT; (2) how treadmill stepping speed influences the locomotor EMG pattern and amplitude, especially at speeds higher than 2.5km/h; and (3) whether EMG activity would be revealed during voluntary overground walking where the same sensory cues are not present. METHODS Participant A complete history was obtained from a review of the participant’s medical records. Previously in excellent physical condition, over a 6-week period this 50-year-old man (weight, 100kg; height, 183cm) experienced a gradual onset of back pain with a subsequent development of myelopathic features including increased tone, spasticity, clonus, decreased strength, and decreased ambulation status. He consequently underwent elective decompression surgery for a large focal disk protrusion at T10-11. The intraoperative report revealed that during surgery the patient briefly moved, and simultaneously there was a loss of tibial, but not ulnar, evoked potentials. Postsurgery, the patient exhibited dense paraplegia. Immediate postoperative magnetic resonance imaging (MRI) findings revealed relief of preoperative findings including herniated lumbar disk and severe spinal stenosis, and the presence of minimal edema (T10-11 level), hematoma, and mild central canal stenosis. No frank hemorrhage was identified. On postoperative day 8, the participant was transferred to acute inpatient rehabilitation, which included 15 days of intense physical therapy and microcurrent stimulation.28 During this period, trace quadriceps activation and intermittent flicker of motor activity were reported in the right and left hips, right and left ankles, and bilateral toes. However, consistent, voluntary activation of any lower extremity muscle was not revealed during this period. Sensory assessment revealed slight improvement with a zone of partial preservation extending to the
second or third lumbar dermatome. He achieved standing with knee-ankle-foot orthoses between parallel bars and parallel swing-through gait. However, on discharge, the participant stated his preference for a more recovery-based rehabilitation approach and voluntarily elected to use a manual wheelchair for mobility rather than adopting a compensatory gait strategy. Follow-up examination revealed a large, recurrent disk herniation at T10-11; therefore, surgical decompression and diskectomy were performed 6 months later with no attendant functional changes. At 16 months postinjury, the participant was no longer receiving any medically directed rehabilitation therapies when he heard about the driven gait orthosis in the press and contacted the principal investigator. He expressed his specific interest in late rehabilitation for locomotor function with the driven gait orthosis. After providing written informed consent and Health Insurance Portability and Accountability Act of 1996 authorization, he was enrolled in the current research protocol. All procedures were approved by the Stanford University panels on human subjects in medical research. Measurements Physical and neurologic evaluation. At study baseline, the participant had a sensory/motor level of T11, some sacral sparing (perianal sensation to light touch 1/2), some sensory sparing in the lower extremities, right hip flexor and toe extensor motor scores of 1/5, and 0/5 in all other key lower extremity muscles. He was able to pull to stand and stand with support of parallel bars, and step for less than 10m with manual assistance from 2 people, thus scoring 1 on the Walking Index for Spinal Cord Injury II (WISCI-II) scale.29 Clinical assessments are summarized in table 1. Clinical MRI. Clinical magnetic resonance (MR) images were reviewed by an independent neuroradiologist blinded to the study hypothesis and outcomes. Assessment revealed features consistent with posttraumatic injury. While images revealed a near-complete spinal cord transection with tissue damage primarily in the gray matter, definitive evidence of some white matter continuity is noted in the anterior tract (fig 1).
Table 1: CHART and LEMS Scores Over Time Clinical Assessments
CHART Short Form Physical independence Mobility Occupational abilities Social integration LEMS Hip flexion right Hip flexion left Knee extension right Knee extension left Ankle dorsiflexion right Ankle dorsiflexion left Ankle plantarflexion right Ankle plantarflexion left Great toe extension right Great toe extension left
Aug 14, 2006
Sep 22, 2006
Jan 8, 2007
Feb 15, 2007
88 99 34.75 100
DNA DNA DNA DNA
96 100 56.25 100
96 100 72.5 100
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 0 0 0
Abbreviations: CHART, Craig Handicap Assessment and Reporting Technique66; DNA, data not available; LEMS, lower extremity motor score.
Arch Phys Med Rehabil Vol 93, August 2012
1478
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
reaching up to 5km/h and lasting up to 17 minutes. Walking distances ranged between 606 and 2424m per session (fig 3). The training protocol was defined by the principal investigator (C.P.) and 1 of the authors (A.L.B.), based on a body of previous work.15,31 Body weight support was decreased until full extension of the hips and knees could no longer be maintained during stance, and remained relatively constant at 70% over both training blocks. During block A, the participant returned home where he received clinic-based outpatient physical therapy twice a week that included manually assisted overground gait training using the Lite-Gait,b home-based strengthening exercises, swimming, and high-intensity electrical stimulation to the gluteal and quadriceps muscle groups 3 times per week.32-34
Fig 1. MR image (T1) of the spinal cord 6 months postsurgery. The marker identifies the lesion site, revealing near complete cord disruption with a residual rim of white matter continuity mostly on the ventral aspect.
Training Procedures and Study Design In this single-subject B-A-B design (experimental treatment–withdrawal– experimental treatment),30 LT was performed in 2 blocks of 6 weeks each using a robotic-driven gait orthosis, the Lokomat11 (fig 2). The Lokomat used in this study was modified to attain physiologic walking speed (eg, speeds of up to 5km/h [1.4m/ s]). During blocks B1 and B2, the participant trained 4 times weekly. A total walking time of 30 minutes was targeted (range, 19 –35min) per session. Training speed was continuously adjusted between 1.5 and 4.4km/h, with some bouts
Testing During the LT blocks, the participant was instrumented once weekly to observe muscle activation patterns both while stepping in the Lokomat over a range of speeds (eg, “speed experiment” explained below) and walking overground. Overground walking took place either between parallel bars or while using a walker. Surface EMGc was sampled from 7 lower extremity muscles per leg (rectus femoris, vastus medialis, semimembranosus, biceps femoris, tibialis anterior, medial gastrocnemius, soleus). Footswitches were used to provide an event marker for the step cycle. EMG and footswitch data were sampled concurrently at 1000Hz using a Powerlab SP-16 A/D Convertord and written to disk for offline analysis using custom routines developed in Matlab.e Lokomat speed experiment. The speed experiment was conducted once during each week of LT. Each speed experiment was initiated with an interval of approximately 2 minutes of stepping at 1km/h, during which we determined the participant’s physiologic responses to stepping and verified that all EMG electrodes were recording. After this familiarization period, the Lokomat speed was increased and EMG was sampled for 45 seconds. This procedure was repeated at increments of 0.5km/h over a range of speeds from 1.5 to 5.0km/h. To ensure cardiovascular safety, we monitored the participant closely and ensured that blood pressure remained below 195/110mmHg.
Fig 2. Study design. Arrows on top indicate application of the experimental treatment during block B1, withdrawal of the experimental treatment during block A, and reintroduction of the treatment in block B2. Boxes below list testing that occurred during the different phases. Abbreviations: CHART, Craig Handicap Assessment and Reporting Technique; LEMS, lower extremity motor score; PT, physical therapy.
Arch Phys Med Rehabil Vol 93, August 2012
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
1479
Fig 3. LT parameters used over each 24-session training block. Data reflect weighted averages for training session to account for parameter adjustments and the amount of time spent training at each setting. Abbreviation: km/h, kilometers per hour.
Volitional overground walking. Overground walking function was tested each week. Surface EMG was sampled concurrently using the same procedure described above. EMG data analysis. EMG data were demeaned, bandpass filtered (first-order Butterworth, 10 –200Hz), rectified, and smoothed using a first-order Butterworth low-pass filter (10Hz). For data collected during the speed experiment, the rectified integrated linear EMG envelope was calculated, normalized by step time, averaged (8 –10 steps per speed), and expressed relative to EMG amplitude at 1.5km/h. For overground walking trials, a manually supervised detection algorithm was used to identify EMG activity. Mean background EMG activity was measured during standing and used as a reference to define muscle bursts during volitional activity. The EMG was considered “on” if the threshold (ie, mean background activity plus 3.5 SDs) was exceeded for a minimum of 100 milliseconds, and “off” if the signal fell below this threshold for at least 100 milliseconds. Muscle burst number and duration, and total “on” time during each trial were calculated and normalized to the distance walked. Functional testing. WISCI-II scores were assessed daily during LT blocks B1 and B2 and weekly during block A. Clinical assessments including Craig Handicap Assessment and Reporting Technique and American Spinal Injury Association (ASIA) lower extremity motor scores35 were evaluated at the beginning and end of each treatment block.
RESULTS Functional Tests Figure 4 shows improved walking capacity as measured by the WISCI-II score. Initially, the participant was able to pull to stand, and step with assistance of 2 persons for less than 10m (WISCI-II, 1/20). However, during the first block of RLT, he improved gradually to independent walking with a walker and braces over at least 10m (WISCI-II, 9/20). According to follow-up phone calls and video sent by the participant to the principal investigator, this functional improvement has been maintained more than 4 years after completion of RLT. EMG While Walking in the Lokomat: Speed Experiment During week 4 of the first training block, patterned EMG activity was revealed in all muscles bilaterally during RLT, with more robust patterns (ie, amplitude and timing) revealed at higher treadmill speeds. EMG amplitudes at each speed are expressed relative to the EMG amplitude at 1.5km/h. The use of this approach revealed modest EMG modulation (0.9 –2.5 times) between 1.5 and 2.0km/h, corresponding with typical LT speeds. However, at speeds greater than 3.5km/h, EMG modulation was marked (fig 5). Speed-related EMG amplitude modulation of the mean of all muscles for each extremity was modeled using a third-order polynomial function (R2⫽.97– .99). The overall EMG modulation pattern suggests a threshold
Fig 4. Evolution of walking capacity, as measured by the WISCI-II during the first LT block (B1), the 16 weeks of experimental training withdrawal (A), and the second LT block (B2). Please note the different scales on the time axes between the 2 LT blocks B1 and B2 (daily measurements) and the 16 weeks of block A (weekly measurements). Asterisks in the first and second training block indicate days on which LT took place.
Arch Phys Med Rehabil Vol 93, August 2012
1480
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
Fig 5. EMG responses elicited during the Lokomat speed experiment. Representative data from 1 assessment of the first and second training blocks are shown separately for the left and right legs. EMG amplitudes are expressed as percentage of the amplitude measured while walking at 1.5km/h within each experiment. Gray shaded areas >3.2km/h indicate typical walking speeds of individuals without disability. R2 values are presented for the third-order polynomial fit models of the mean of all muscles within 1 extremity. Red-shaded areas represent 95% confidence intervals of the mean per leg. Abbreviation: kph, kilometers per hour.
speed above which afferent inputs produce robust excitatory influences on motor output. EMG From Overground Walking Trials Overground walking was extremely limited during the first 4 weeks; thus, EMG data were not recorded. Beginning with the fifth week, EMG bursts were recorded in many muscles during voluntary (self-initiated) stepping. We analyzed EMG data to determine whether activation became more robust and corresponded with volitional stepping. However, a systematic change in EMG activation was not apparent in any of the overground walking conditions (ie, both braces locked, 1 brace unlocked, brace straps loosened) between week 5 and the end of the study. We qualitatively analyzed whether any walking condition provoked more EMG activity than another. In general, our data suggest that unlocking the brace of either leg produced more EMG activity than having both braces locked or Arch Phys Med Rehabil Vol 93, August 2012
loosening the straps of either brace. Thus, it appears that less restrictive support contributes to increased neural activation even if in the context of requisite bracing support (fig 6). DISCUSSION Individuals post-SCI with some sacral sparing usually have a considerable chance of regaining some degree of walking function.36,37 Thus, intensive LT might activate spared fibers and lead to the more thorough exploitation of residual physiologic capacity than in the absence of sacral sparing. The individual in this study revealed some remaining sensory function (eg, light touch in S4-5); however, he remained nonambulatory 16 months after his initial injury. During the course of LT he demonstrated recovery of locomotor EMG patterns. Recently, Manella et al38 reported functional locomotor recovery to a very similar extent in a participant 2 years postSCI. Both the present case and Manella’s38 stand in contrast to
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
1481
Fig 6. EMG activity during overground stepping. EMG “on time” (milliseconds [ms] of muscle activation) was normalized by distance (feet) to assess differences while stepping with varying levels of external support. For sake of illustration, EMG “on time” was collapsed over 7 muscles measured within each leg and averaged over the 4 assessments in the second LT block. Error bars represent 95% confidence intervals.
previous studies.39,40 Typically, individuals with chronic SCI who demonstrate improved locomotor function in response to LT reveal higher baseline motor function below the lesion,41,42 recordable lower extremity EMG activity43 at baseline, or both. In these studies,39-43 individuals demonstrating functional levels similar to our participant did not improve locomotor function. Furthermore, on the basis of contemporary prediction rules, our participant responded to LT with improvements that would not have been expected before initiation of the training.44,45 Of note, these clinical prediction rules have been developed for a more acute stage post-SCI where even more spontaneous recovery can be expected. Among other factors, we hypothesize that the unexpected outcome in the present case could be attributable to the differences in training speed between the earlier LT studies (between 0.7 and 2.1km/h)39-42 and the training speeds used with our participant (up to 5km/h). Technical advancements implemented in the Lokomat enabled stepping at speeds greater than 3.2km/h (.88m/s or 1.8mph). Comparable manual LT in a participant of his stature and requiring such a high level of guidance would have been very challenging and labor intensive.10 While it may be possible to attain physiologic walking speeds with manual assistance, sustaining this activity over training bouts lasting up to 17 minutes would not be feasible. Physiologic stepping speeds provide both increased intensity and increased amplitude of locomotor-related afference that may contribute to reactivation of residual pathways that were inactive because of nonuse in this participant. It has been argued that slower walking speeds are critical in promoting recovery of corticospinal input because, unlike faster speed, they allow the participant to actively and voluntarily participate in producing stepping movements.43,46 However, it is also possible that integration of spared descending input can be facilitated by appropriate sen-
sory inputs at the spinal level according to the Hebbian principle.47 Our participant not only revealed improved EMG activity with higher walking speeds, but was able to transfer his regained stepping ability to overground stepping where the facilitatory peripheral inputs were no longer present. This result suggests that the higher training speeds7-9 and, potentially, the variability48 of speeds within each session either directly influenced the descending tract function or the manner in which the input from those tracts is integrated at the spinal level. Although we encouraged the participant to “walk with the robot,” the Lokomat in the present study used position control of the limbs, which allows the participant to remain fairly passive.49,50 Evolving robotic control approaches, including patient-cooperative control, are intended to combine the advantages of high stepping speeds with more voluntary participation and enhanced kinematic and temporal variability to promote greater motor learning.48,49,51,52 In this single-case design, we refrained from assigning the participant a severity level (ASIA Impairment Scale [AIS]) according to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI).35 Accurate ISNCSCI testing is challenging and needs continuous training of the assessors.53-55 The data we were able to obtain from the patient’s records did not allow for a reliable classification. However, we believe the actual AIS score is less important than the fact that the participant possessed untapped potential for neurophysiologic recovery. Even in clinically complete SCIs, spared fibers cross the injury site56 and appear to be physiologically intact.57,58 However, motor output relies on activity in many pathways and the appropriate integration of these influences at the spinal level. Even small dysfunction in the gray matter is disruptive to this integrative process and therefore can interfere with purposeful Arch Phys Med Rehabil Vol 93, August 2012
1482
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
movement.58 Alternatively, findings from animal studies reveal that a small number of spared fibers can suffice to recover at least some movement in the lower extremities.59-61 Dimitrijevic et al62 assigned participants with clinically complete, but electrophysiologically incomplete, lesions the term “discomplete” and suggested such individuals may possess greater potential for functional recovery than individuals with complete SCI. Both short- and long-latency EMG responses were recorded in these discomplete individuals, suggesting that both direct corticospinal pathways and collateral pathways through propriospinal networks could be recruited simultaneously. Manella38 speculated that their participant experienced a discomplete injury, leading to the marked functional recovery in the chronic stage. This hypothesis can be further elaborated in our participant by monitoring EMG as an outcome measure during overground walking. Clinical MR images from our participant reveal damage primarily to gray matter at the injury site, while a rim of mostly ventral white matter remained intact. It is possible that LT in this individual led to recovery of a few critical connections that allowed for reintegration of necessary pathways and thus the observed recovery of EMG patterns. Additionally, our participant exhibited signs of central neuronal damage (clonus, weakness) for at least 6 weeks before his surgery. It is possible that these signs and symptoms reveal a “preinjury” that served as a primer for neuroplasticity after his principal spinal cord damage.63 This ostensibly staged injury, combined with the intensive LT, might have contributed to activation of residual, spared pathways. Our hypothesis is supported by the emergence of patterned EMG within a very short time (a few weeks), therefore likely representing transmission in existing synapses that had not been activated, or activated effectively, before LT.64 Finally, according to most prediction rules, this participant would not have been identified as a candidate for LT in a clinical setting. Thus, it would have been unlikely that he would have been given the chance to train with the intensity and for the duration offered in this study. Nevertheless, he made progress that was both quantifiable and subjectively regarded as meaningful. The participant continues to walk/step daily even 4 years posttraining. Study Limitations The valuable insights gained from this single-case design serve as proof of principle of the potential for late recovery and add to the body of literature demonstrating recovery of locomotor function in the chronic state of incomplete SCI. However, we recognize that the results from this study might not be generalizable to a larger population of individuals with incomplete or discomplete SCI and do not reflect the same level of evidence as a randomized controlled trial. In addition, this study lacked inclusion of a clinical gait measurement tool that is more sensitive to change in individuals with greater functional abilities. We were also unable to retrieve complete information on ISNCSCI data, which prevented us from assigning an AIS score. Finally, because of the severe gait deficits, collection of consistent EMG data from overground walking was challenging in this participant, especially early in the study. CONCLUSIONS In this individual with chronic incomplete SCI who was nonambulatory, intensive LT at physiologic walking speeds led to reemergence of locomotor EMG patterns. Volitional EMG activity could also be measured while walking overground, and greater EMG activation could be noted with less restrictive Arch Phys Med Rehabil Vol 93, August 2012
support (eg, braces unlocked). Individuals with chronic incomplete SCI may still benefit from intensive LT, especially at physiologic walking speeds. Cases as described herein are possible and may, in fact, be underreported in the literature.38,65 Acknowledgments: We thank the staff at the Rehabilitation Research & Development Center, VA Palo Alto Health Care System, and the VA Brain Rehabilitation Research Center of Excellence, Gainesville, Florida. We also thank the following persons for their contributions to this work: Aaron Grogan, MS, for his engineering support; and Keith Peters, MD, for independent, blinded assessment of the MR images. References 1. Dietz V, Colombo G, Jensen L, Baumgartner L. Locomotor capacity of spinal cord in paraplegic patients. Ann Neurol 1995;37: 574-82. 2. Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Brooks VB, editor. Handbook of physiology. Sec. 1. The nervous system. Vol. II. Motor control, part 2. Washington (DC): American Physiological Society; 1981. p 1179-236. 3. Dietz V. Central pattern generator. Paraplegia 1995;33:739. 4. Dobkin B, Edgerton V, Fowler E. Training induces rhythmic locomotor EMG patterns in subjects with complete SCI. Neurology 1992;42(Suppl 3):207-8. 5. Edgerton VR, Leon RD, Harkema SJ, et al. Retraining the injured spinal cord. J Physiol 2001;533(Pt 1):15-22. 6. Rossignol S, Drew T, Brustein E, Jiang W. Locomotor performance and adaptation after partial or complete spinal cord lesions in the cat. Prog Brain Res 1999;123:349-65. 7. Dietz V. Role of peripheral afferents and spinal reflexes in normal and impaired human locomotion. Rev Neurol (Paris) 1987;143: 241-54. 8. Dobkin BH, Harkema S, Requejo P, Edgerton VR. Modulation of locomotor-like EMG activity in subjects with complete and incomplete spinal cord injury. J Neurol Rehabil 1995;9:183-90. 9. Beres-Jones JA, Harkema SJ. The human spinal cord interprets velocity-dependent afferent input during stepping. Brain 2004; 127(Pt 10):2232-46. 10. Pohl M, Mehrholz J, Ritschel C, Ruckriem S. Speed-dependent treadmill training in ambulatory hemiparetic stroke patients: a randomized controlled trial. Stroke 2002;33:553-8. 11. Colombo G, Joerg M, Schreier R, Dietz V. Treadmill training of paraplegic patients using a robotic orthosis. J Rehabil Res Dev 2000;37:693-700. 12. Colombo G, Wirz M, Dietz V. Driven gait orthosis for improvement of locomotor training in paraplegic patients. Spinal Cord 2001;39:252-5. 13. Hesse S, Schmidt H, Werner C, Bardeleben A. Upper and lower extremity robotic devices for rehabilitation and for studying motor control. Curr Opin Neurol 2003;16:705-10. 14. Lunenburger L, Bolliger M, Czell D, Muller R, Dietz V. Modulation of locomotor activity in complete spinal cord injury. Exp Brain Res 2006;174:638-46. 15. Behrman AL, Lawless-Dixon AR, Davis SB, et al. Locomotor training progression and outcomes after incomplete spinal cord injury. Phys Ther 2005;85:1356-71. 16. Barbeau H. Locomotor training in neurorehabilitation: emerging rehabilitation concepts. Neurorehabil Neural Repair 2003;17:3-11. 17. Dobkin B, Barbeau H, Deforge D, et al. The evolution of walkingrelated outcomes over the first 12 weeks of rehabilitation for incomplete traumatic spinal cord injury: the multicenter randomized Spinal Cord Injury Locomotor Trial. Neurorehabil Neural Repair 2007;21:25-35. 18. Dietz V. Body weight supported gait training: from laboratory to clinical setting. Brain Res Bull 2008;76:459-63.
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
19. De Leon RD, Hodgson JA, Roy RR, Edgerton VR. Full weightbearing hindlimb standing following stand training in the adult spinal cat. J Neurophysiol 1998;80:83-91. 20. Edgerton VR, Roy RR, Hodgson JA, Prober RJ, Guzman CP, de Leon R. Potential of adult mammalian lumbosacral spinal cord to execute and acquire improved locomotion in the absence of supraspinal input. J Neurotrauma 1992;9(Suppl 1):S119-28. 21. Harkema SJ, Hurley SL, Patel UK, Requejo PS, Dobkin BH, Edgerton VR. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol 1997;77:797-811. 22. Brown G. The intrinsic factors in the act of progression in mammals. Proc R Soc 1911;B(84):308-19. 23. Jahn K, Deutschlander A, Stephan T, et al. Supraspinal locomotor control in quadrupeds and humans. Prog Brain Res 2008;171: 353-62. 24. Pearson K, Gordon J. Locomotion. In: Kandel ER, Schwartz JH, Jessell TM, editors. Principles of neural science. 4th ed. New York: McGraw Hill; 2000. p 737-55. 25. Stolze H, Kuhtz-Buschbeck JP, Mondwurf C, et al. Gait analysis during treadmill and overground locomotion in children and adults. Electroencephalogr Clin Neurophysiol 1997;105:490-7. 26. Van Gheluwe B, Smekens J, Roosen P. Electrodynographic evaluation of the foot during treadmill versus overground locomotion. J Am Podiatr Med Assoc 1994;84:598-606. 27. Day KV. Motor control solutions to dynamic stability deficits during walking after spinal cord injury [dissertation]. 2010, University of Florida: Gainesville. 28. Lee BY, Wendell K, Al-Waili N, Butler G. Ultra-low microcurrent therapy: a novel approach for treatment of chronic resistant wounds. Adv Ther 2007;24:1202-9. 29. Dittuno PL, Dittuno JF Jr. Walking Index for Spinal Cord Injury (WISCI II): scale revision. Spinal Cord 2001;39:654-6. 30. Tawney JW, Gast DL. Single subject research in special education. Columbus: CE Merrill; 1984. 31. Behrman AL, Harkema SJ. Physical rehabilitation as an agent for recovery after spinal cord injury. Phys Med Rehabil Clin N Am 2007;18:183-202, v. 32. Chou LW, Lee SC, Johnston TE, Binder-Macleod SA. The effectiveness of progressively increasing stimulation frequency and intensity to maintain paralyzed muscle force during repetitive activation in persons with spinal cord injury. Arch Phys Med Rehabil 2008;89:856-64. 33. Scott WB, Lee SC, Johnston TE, Binkley J, Binder-Macleod SA. Effect of electrical stimulation pattern on the force responses of paralyzed human quadriceps muscles. Muscle Nerve 2007;35: 471-8. 34. Kebaetse MB, Lee SC, Johnston TE, Binder-Macleod SA. Strategies that improve paralyzed human quadriceps femoris muscle performance during repetitive, nonisometric contractions. Arch Phys Med Rehabil 2005;86:2157-64. 35. ASIA. International standards for neurological classification of spinal cord injury. Chicago: American Spinal Injury Association; 2002. 36. van Middendorp JJ, Hosman AJ, Pouw MH, Van de Meent H. ASIA Impairment Scale conversion in traumatic SCI: is it related with the ability to walk? A descriptive comparison with functional ambulation outcome measures in 273 patients. Spinal Cord 2009; 47:555-60. 37. Ditunno PL, Patrick M, Stineman M, Ditunno JF. Who wants to walk? Preferences for recovery after SCI: a longitudinal and cross-sectional study. Spinal Cord 2008;46:500-6. 38. Manella KJ, Torres J, Field-Fote EC. Restoration of walking function in an individual with chronic complete (AIS A) spinal cord injury. J Rehabil Med 2010;42:795-8. 39. Dietz V, Harkema SJ. Locomotor activity in spinal cord-injured persons. J Appl Physiol 2004;96:1954-60.
1483
40. Müller RM. Lokomotions-, und Reflexaktivität bei kompletter Querschnittlähmung: Einflussfaktoren und zeitliche Entwickling, [dissertation]. 2006, Eidgenössische Technische Hochschule ETH; Zürich. 41. Hicks AL, Adams MM, Martin Ginis K, et al. Long-term bodyweight-supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective well-being. Spinal Cord 2005;43: 291-8. 42. Wernig A, Nanassy A, Muller S. Maintenance of locomotor abilities following Laufband (treadmill) therapy in para- and tetraplegic persons: follow-up studies. Spinal Cord 1998;36:744-9. 43. Gorassini MA, Norton JA, Nevett-Duchcherer J, Roy FD, Yang JF. Changes in locomotor muscle activity after treadmill training in subjects with incomplete spinal cord injury. J Neurophysiol 2009;101:969-79. 44. van Middendorp JJ, Hosman AJ, Donders AR, et al. A clinical prediction rule for ambulation outcomes after traumatic spinal cord injury: a longitudinal cohort study. Lancet 2011;19:1004-10. 45. Yang JF, Norton J, Nevett-Duchcherer J, Roy FD, Gross DP, Gorassini MA. Volitional muscle strength in the legs predicts changes in walking speed following locomotor training in people with chronic spinal cord injury. Phys Ther 2011;91:931-43. 46. Laufband therapy - the manual for spinal cord damage, stroke, brain damage and MS, orthopedic disorders and others [PDF]. Müller S, Wernig A. Bonn, Germany, 2005. Available at: http:// www.meb.uni-bonn.de/wernig/html/manual.html. Accessed April 10, 2012. 47. Hebb D. The organization of behavior: a neuropsychological theory. New York: Wiley; 1949. 48. Ziegler MD, Zhong H, Roy RR, Edgerton VR. Why variability facilitates spinal learning. J Neurosci 2010;30:10720-6. 49. Duschau-Wicke A, Caprez A, Riener R. Patient-cooperative control increases active participation of individuals with SCI during robot-aided gait training. J Neuroeng Rehabil 2010;7:43. 50. Wernig A. Long-term body-weight supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective well-being. Spinal Cord 2006;44:265-6; author reply 267-8. 51. Emken JL, Harkema SJ, Beres-Jones JA, Ferreira CK, Reinkensmeyer DJ. Feasibility of manual teach-and-replay and continuous impedance shaping for robotic locomotor training following spinal cord injury. IEEE Trans Biomed Eng 2008;55:322-34. 52. Mankala KK, Banala SK, Agrawal SK. Novel swing-assist unmotorized exoskeletons for gait training. J Neuroeng Rehabil 2009;6:24. 53. Schuld C, Wiese J, Hug A, et al. Computer implementation of the international standards for neurological classification of spinal cord injury for consistent and efficient derivation of its subscores including handling of data from not testable segments. J Neurotrauma 2011;29:453-61. 54. Spiess MR, Muller RM, Rupp R, Schuld C, van Hedel HJ. Conversion in ASIA Impairment Scale during the first year after traumatic spinal cord injury. J Neurotrauma 2009;26:2027-36. 55. Dietz V, Muller R. Degradation of neuronal function following a spinal cord injury: mechanisms and countermeasures. Brain 2004; 127(Pt 10):2221-31. 56. Kakulas BA, Bedbrook GM. A correlative clinico-pathologic study of spinal cord injury. Proc Aust Assoc Neurol 1969;6: 123-32. 57. Dimitrijevic MR. Residual motor functions in spinal cord injury. Adv Neurol 1988;47:138-55. 58. Sherwood AM, Dimitrijevic MR, McKay WB. Evidence of subclinical brain influence in clinically complete spinal cord injury: discomplete SCI. J Neurol Sci 1992;110:90-8. Arch Phys Med Rehabil Vol 93, August 2012
1484
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
59. Basso DM. Neuroanatomical substrates of functional recovery after experimental spinal cord injury: implications of basic science research for human spinal cord injury. Phys Ther 2000;80:808-17. 60. Basso DM, Beattie MS, Bresnahan JC. Descending systems contributing to locomotor recovery after mild or moderate spinal cord injury in rats: experimental evidence and a review of literature. Restor Neurol Neurosci 2002;20:189-218. 61. Hill RL, Zhang YP, Burke DA, et al. Anatomical and functional outcomes following a precise, graded, dorsal laceration spinal cord injury in C57BL/6 mice. J Neurotrauma 2009;26:1-15. 62. Dimitrijevic MR, Dimitrijevic MM, Faganel J, Sherwood AM. Suprasegmentally induced motor unit activity in paralyzed muscles of patients with established spinal cord injury. Ann Neurol 1984;16:216-21. 63. McPhail LT, Fernandes KJ, Chan CC, Vanderluit JL, Tetzlaff W. Axonal reinjury reveals the survival and re-expression of regeneration-associated genes in chronically axotomized adult mouse motoneurons. Exp Neurol 2004;188:331-40.
Arch Phys Med Rehabil Vol 93, August 2012
64. Nathan PW, Smith MC. Effects of two unilateral cordotomies on the motility of the lower limbs. Brain 1973;96:471-94. 65. Spiess M, Schuld C. Comment on “Restoration of walking function in an individual with chronic complete (AIS A) spinal cord injury.” J Rehabil Med 2011;43:367-8. 66. Whiteneck GG, Charlifue SW, Gerhart KA, Overholser JD, Richardson GN. Quantifying handicap: a new measure of longterm rehabilitation outcomes. Arch Phys Med Rehabil 1992; 73:519-26. Suppliers a. Hocoma AG, Industriestrasse 4, CH-8604 Volketswil, Switzerland. b. Mobility Research, 444 West Geneva Dr, Tempe, AZ 85282-2004. c. MA-300, preamplified electrodes; Motion Lab Systems, Inc, 15045 Old Hammond Hwy, Baton Rouge, LA 70816. d. AD Instruments, 322 North Landings Dr, Grand Junction, CO 81501. e. The MathWorks, 3 Apple Hill Dr, Natick, MA 01760.
CLINICAL IMPLICATIONS OF BASIC RESEARCH
Unexpected Recovery After Robotic Locomotor Training at Physiologic Stepping Speed: A Single-Case Design Martina R. Spiess, PT, MPTSc, Jeffrey P. Jaramillo, PT, MS, Andrea L. Behrman, PT, PhD, Jeffrey K. Teraoka, MD, Carolynn Patten, PhD, PT ABSTRACT. Spiess MR, Jaramillo JP, Behrman AL, Teraoka JK, Patten C. Unexpected recovery after robotic locomotor training at physiologic stepping speed: a single-case design. Arch Phys Med Rehabil 2012;93:1476-84. Objectives: To investigate the effect of walking speed on the emergence of locomotor electromyogram (EMG) patterns in an individual with chronic incomplete spinal cord injury (SCI), and to determine whether central pattern generator activity during robotic locomotor training (RLT) transfers to volitional EMG activity during overground walking. Design: Single-case (B-A-B; experimental treatment–withdrawal– experimental treatment) design. Setting: Freestanding rehabilitation research center. Participant: A 50-year-old man who was nonambulatory for 16 months after incomplete SCI (sub-T11). Interventions: The participant completed two 6-week blocks of RLT, training 4 times per week for 30 minutes per session at walking speeds up to 5km/h (1.4m/s) over continuous bouts lasting up to 17 minutes. Main Outcome Measures: Surface EMG was recorded weekly during RLT and overground walking. The Walking Index for Spinal Cord Injury (WISCI-II) was assessed daily during training blocks. Results: During week 4, reciprocal, patterned EMG emerged during RLT. EMG amplitude modulation revealed a curvilinear relationship over the range of walking speeds from 1.5 to 5km/h (1.4m/s). Functionally, the participant improved from being nonambulatory (WISCI-II 1/20), to walking overground with reciprocal stepping using knee-ankle-foot orthoses and a walker (WISCI-II 9/20). EMG was also observed during overground walking. These functional gains were maintained greater than 4 years after locomotor training (LT). Conclusions: Here we report an unexpected course of locomotor recovery in an individual with chronic incomplete SCI. Through RLT at physiologic walking speeds, it was possible to
From the Brain Rehabilitation Research Center, Malcom Randall VAMC, Gainesville, FL (Patten, Spiess); the Department of Physical Therapy, University of Florida, Gainesville, FL (Patten, Spiess, Behrman); VA Palo Alto HCS, Palo Alto, CA (Jaramillo, Teraoka); and Division of Physical Medicine and Rehabilitation, Stanford University School of Medicine, Stanford, CA (Teraoka). Presented to the 4th National Spinal Cord Injury Conference, sponsored by Toronto Rehabilitation Hospital, October 28-30, 2010, Niagara Falls, Ontario, Canada; and to the International Society for Posture and Gait Research Conference, July 14-18, 2007, Burlington, VT. Supported by the Department of Veterans Affairs Rehabilitation Research & Development Service (Project no. B540231) and Research Career Scientist Award (F7823S); and by a University of Florida Alumni Fellowship. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Reprint requests to Carolynn Patten, PhD, PT, VA Brain Rehabilitation Research Center, 1601 SW Archer Rd, 151A, Gainesville, FL 32608, e-mail: [email protected]. In-press corrected proof published online on May 8, 2012, at www.archives-pmr.org. 0003-9993/12/9308-01018$36.00/0 http://dx.doi.org/10.1016/j.apmr.2012.02.030
Arch Phys Med Rehabil Vol 93, August 2012
activate the central pattern generator even 16 months postinjury. Further, to a certain degree, improvements from RLT transferred to overground walking. Our results suggest that LT-induced changes affect the central pattern generator and allow supraspinal inputs to engage residual spinal pathways. Key Words: Locomotion; Neurophysiology; Recovery of function; Rehabilitation; Spinal cord injuries. © 2012 by the American Congress of Rehabilitation Medicine OCOMOTOR TRAINING (LT) has been shown to proL mote recovery of reciprocal locomotor electromyogram (EMG) patterns after spinal cord injury (SCI), even in individ-
uals who are unable to produce voluntary movement.1,2 Under attenuation or absence of supraspinal inputs, it is generally agreed that these EMG patterns are elicited by a spinal central pattern generator3-6 and are influenced by changes in somatosensory feedback resulting from alterations in stepping speed, loading, or joint range of motion induced during locomotion. For example, higher treadmill speeds facilitate both the emergence of patterned EMG and increased EMG amplitude.7-9 However, the feasibility of delivering manually assisted LT at higher treadmill/stepping speeds is limited because it is strenuous and very labor intensive.10 Robotic devices were developed as one approach to address these challenges. Robotic locomotor training (RLT) devices, such as the Lokomat,a offer assistance with limb movement over sustained periods, even at higher stepping speeds.11-13 Lunenburger et al14 demonstrated increased EMG activity in complete SCI patients as a function of speed when walking between 1.5 and 2.5km/h in the Lokomat. With technical advances and updates, higher speeds (5km/h), approaching normal overground walking speed, are now possible in the robot. The effect of training at these higher stepping speeds (eg, between 2.5 and 5km/h) on EMG activity has not been investigated systematically in persons with SCI. In acutely injured individuals, the capacity to produce patterned EMG activity during assisted locomotion seems ubiq-
List of Abbreviations AIS ASIA EMG ISNCSCI LT MR MRI RLT SCI WISCI-II
American Spinal Injury Association Impairment Scale American Spinal Injury Association electromyogram International Standards for Neurological Classification of Spinal Cord Injury locomotor training magnetic resonance magnetic resonance imaging robotic locomotor training spinal cord injury Walking Index for Spinal Cord Injury II
1477
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
uitous, even in individuals with no voluntary control of their lower limbs. However, this capacity does not translate to functional gains in all cases. Some individuals, especially those with incomplete SCI, reveal significant gains in walking function after different forms of LT.15-18 However, in other human subjects—mostly those with complete SCI—there is no evidence reporting the capacity to reproduce stepping under conditions of full body weight-bearing.7,8,14 As a result, it is not possible to conclude that the presence of spinally mediated EMG activity leads to functional gains in human subjects with complete SCI. This phenomenon stands in contrast to the animal literature19-22 and suggests that despite preservation of the central pattern generator during the transition from quadrupedalism to bipedalism, there might be a higher dependence of supraspinal input during bipedal gait, especially when walking overground.7,23,24 Overground stepping does not provide the same sensory inputs as walking on a treadmill, even in healthy individuals.25,26 Additionally, in impaired populations, volitional overground stepping typically occurs at much slower speeds than treadmill stepping and, in part because of the use of assistive devices, it is more difficult to achieve an upright posture and adequate hip extension.27 These factors compromise locomotor-relevant sensory feedback and may therefore place higher demand on supraspinal inputs. It remains unclear whether, and to what extent, patterned EMG activity acquired during treadmill/LT translates to overground stepping. Here we explored the effect of RLT on an individual with incomplete SCI who was nonambulatory 16 months postinjury. Our aims in this single-case design were to determine (1) whether initially absent EMG activity could be elicited during robotic-driven locomotion and how it would adapt over the course of LT; (2) how treadmill stepping speed influences the locomotor EMG pattern and amplitude, especially at speeds higher than 2.5km/h; and (3) whether EMG activity would be revealed during voluntary overground walking where the same sensory cues are not present. METHODS Participant A complete history was obtained from a review of the participant’s medical records. Previously in excellent physical condition, over a 6-week period this 50-year-old man (weight, 100kg; height, 183cm) experienced a gradual onset of back pain with a subsequent development of myelopathic features including increased tone, spasticity, clonus, decreased strength, and decreased ambulation status. He consequently underwent elective decompression surgery for a large focal disk protrusion at T10-11. The intraoperative report revealed that during surgery the patient briefly moved, and simultaneously there was a loss of tibial, but not ulnar, evoked potentials. Postsurgery, the patient exhibited dense paraplegia. Immediate postoperative magnetic resonance imaging (MRI) findings revealed relief of preoperative findings including herniated lumbar disk and severe spinal stenosis, and the presence of minimal edema (T10-11 level), hematoma, and mild central canal stenosis. No frank hemorrhage was identified. On postoperative day 8, the participant was transferred to acute inpatient rehabilitation, which included 15 days of intense physical therapy and microcurrent stimulation.28 During this period, trace quadriceps activation and intermittent flicker of motor activity were reported in the right and left hips, right and left ankles, and bilateral toes. However, consistent, voluntary activation of any lower extremity muscle was not revealed during this period. Sensory assessment revealed slight improvement with a zone of partial preservation extending to the
second or third lumbar dermatome. He achieved standing with knee-ankle-foot orthoses between parallel bars and parallel swing-through gait. However, on discharge, the participant stated his preference for a more recovery-based rehabilitation approach and voluntarily elected to use a manual wheelchair for mobility rather than adopting a compensatory gait strategy. Follow-up examination revealed a large, recurrent disk herniation at T10-11; therefore, surgical decompression and diskectomy were performed 6 months later with no attendant functional changes. At 16 months postinjury, the participant was no longer receiving any medically directed rehabilitation therapies when he heard about the driven gait orthosis in the press and contacted the principal investigator. He expressed his specific interest in late rehabilitation for locomotor function with the driven gait orthosis. After providing written informed consent and Health Insurance Portability and Accountability Act of 1996 authorization, he was enrolled in the current research protocol. All procedures were approved by the Stanford University panels on human subjects in medical research. Measurements Physical and neurologic evaluation. At study baseline, the participant had a sensory/motor level of T11, some sacral sparing (perianal sensation to light touch 1/2), some sensory sparing in the lower extremities, right hip flexor and toe extensor motor scores of 1/5, and 0/5 in all other key lower extremity muscles. He was able to pull to stand and stand with support of parallel bars, and step for less than 10m with manual assistance from 2 people, thus scoring 1 on the Walking Index for Spinal Cord Injury II (WISCI-II) scale.29 Clinical assessments are summarized in table 1. Clinical MRI. Clinical magnetic resonance (MR) images were reviewed by an independent neuroradiologist blinded to the study hypothesis and outcomes. Assessment revealed features consistent with posttraumatic injury. While images revealed a near-complete spinal cord transection with tissue damage primarily in the gray matter, definitive evidence of some white matter continuity is noted in the anterior tract (fig 1).
Table 1: CHART and LEMS Scores Over Time Clinical Assessments
CHART Short Form Physical independence Mobility Occupational abilities Social integration LEMS Hip flexion right Hip flexion left Knee extension right Knee extension left Ankle dorsiflexion right Ankle dorsiflexion left Ankle plantarflexion right Ankle plantarflexion left Great toe extension right Great toe extension left
Aug 14, 2006
Sep 22, 2006
Jan 8, 2007
Feb 15, 2007
88 99 34.75 100
DNA DNA DNA DNA
96 100 56.25 100
96 100 72.5 100
1 1 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 0 0 0
Abbreviations: CHART, Craig Handicap Assessment and Reporting Technique66; DNA, data not available; LEMS, lower extremity motor score.
Arch Phys Med Rehabil Vol 93, August 2012
1478
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
reaching up to 5km/h and lasting up to 17 minutes. Walking distances ranged between 606 and 2424m per session (fig 3). The training protocol was defined by the principal investigator (C.P.) and 1 of the authors (A.L.B.), based on a body of previous work.15,31 Body weight support was decreased until full extension of the hips and knees could no longer be maintained during stance, and remained relatively constant at 70% over both training blocks. During block A, the participant returned home where he received clinic-based outpatient physical therapy twice a week that included manually assisted overground gait training using the Lite-Gait,b home-based strengthening exercises, swimming, and high-intensity electrical stimulation to the gluteal and quadriceps muscle groups 3 times per week.32-34
Fig 1. MR image (T1) of the spinal cord 6 months postsurgery. The marker identifies the lesion site, revealing near complete cord disruption with a residual rim of white matter continuity mostly on the ventral aspect.
Training Procedures and Study Design In this single-subject B-A-B design (experimental treatment–withdrawal– experimental treatment),30 LT was performed in 2 blocks of 6 weeks each using a robotic-driven gait orthosis, the Lokomat11 (fig 2). The Lokomat used in this study was modified to attain physiologic walking speed (eg, speeds of up to 5km/h [1.4m/ s]). During blocks B1 and B2, the participant trained 4 times weekly. A total walking time of 30 minutes was targeted (range, 19 –35min) per session. Training speed was continuously adjusted between 1.5 and 4.4km/h, with some bouts
Testing During the LT blocks, the participant was instrumented once weekly to observe muscle activation patterns both while stepping in the Lokomat over a range of speeds (eg, “speed experiment” explained below) and walking overground. Overground walking took place either between parallel bars or while using a walker. Surface EMGc was sampled from 7 lower extremity muscles per leg (rectus femoris, vastus medialis, semimembranosus, biceps femoris, tibialis anterior, medial gastrocnemius, soleus). Footswitches were used to provide an event marker for the step cycle. EMG and footswitch data were sampled concurrently at 1000Hz using a Powerlab SP-16 A/D Convertord and written to disk for offline analysis using custom routines developed in Matlab.e Lokomat speed experiment. The speed experiment was conducted once during each week of LT. Each speed experiment was initiated with an interval of approximately 2 minutes of stepping at 1km/h, during which we determined the participant’s physiologic responses to stepping and verified that all EMG electrodes were recording. After this familiarization period, the Lokomat speed was increased and EMG was sampled for 45 seconds. This procedure was repeated at increments of 0.5km/h over a range of speeds from 1.5 to 5.0km/h. To ensure cardiovascular safety, we monitored the participant closely and ensured that blood pressure remained below 195/110mmHg.
Fig 2. Study design. Arrows on top indicate application of the experimental treatment during block B1, withdrawal of the experimental treatment during block A, and reintroduction of the treatment in block B2. Boxes below list testing that occurred during the different phases. Abbreviations: CHART, Craig Handicap Assessment and Reporting Technique; LEMS, lower extremity motor score; PT, physical therapy.
Arch Phys Med Rehabil Vol 93, August 2012
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
1479
Fig 3. LT parameters used over each 24-session training block. Data reflect weighted averages for training session to account for parameter adjustments and the amount of time spent training at each setting. Abbreviation: km/h, kilometers per hour.
Volitional overground walking. Overground walking function was tested each week. Surface EMG was sampled concurrently using the same procedure described above. EMG data analysis. EMG data were demeaned, bandpass filtered (first-order Butterworth, 10 –200Hz), rectified, and smoothed using a first-order Butterworth low-pass filter (10Hz). For data collected during the speed experiment, the rectified integrated linear EMG envelope was calculated, normalized by step time, averaged (8 –10 steps per speed), and expressed relative to EMG amplitude at 1.5km/h. For overground walking trials, a manually supervised detection algorithm was used to identify EMG activity. Mean background EMG activity was measured during standing and used as a reference to define muscle bursts during volitional activity. The EMG was considered “on” if the threshold (ie, mean background activity plus 3.5 SDs) was exceeded for a minimum of 100 milliseconds, and “off” if the signal fell below this threshold for at least 100 milliseconds. Muscle burst number and duration, and total “on” time during each trial were calculated and normalized to the distance walked. Functional testing. WISCI-II scores were assessed daily during LT blocks B1 and B2 and weekly during block A. Clinical assessments including Craig Handicap Assessment and Reporting Technique and American Spinal Injury Association (ASIA) lower extremity motor scores35 were evaluated at the beginning and end of each treatment block.
RESULTS Functional Tests Figure 4 shows improved walking capacity as measured by the WISCI-II score. Initially, the participant was able to pull to stand, and step with assistance of 2 persons for less than 10m (WISCI-II, 1/20). However, during the first block of RLT, he improved gradually to independent walking with a walker and braces over at least 10m (WISCI-II, 9/20). According to follow-up phone calls and video sent by the participant to the principal investigator, this functional improvement has been maintained more than 4 years after completion of RLT. EMG While Walking in the Lokomat: Speed Experiment During week 4 of the first training block, patterned EMG activity was revealed in all muscles bilaterally during RLT, with more robust patterns (ie, amplitude and timing) revealed at higher treadmill speeds. EMG amplitudes at each speed are expressed relative to the EMG amplitude at 1.5km/h. The use of this approach revealed modest EMG modulation (0.9 –2.5 times) between 1.5 and 2.0km/h, corresponding with typical LT speeds. However, at speeds greater than 3.5km/h, EMG modulation was marked (fig 5). Speed-related EMG amplitude modulation of the mean of all muscles for each extremity was modeled using a third-order polynomial function (R2⫽.97– .99). The overall EMG modulation pattern suggests a threshold
Fig 4. Evolution of walking capacity, as measured by the WISCI-II during the first LT block (B1), the 16 weeks of experimental training withdrawal (A), and the second LT block (B2). Please note the different scales on the time axes between the 2 LT blocks B1 and B2 (daily measurements) and the 16 weeks of block A (weekly measurements). Asterisks in the first and second training block indicate days on which LT took place.
Arch Phys Med Rehabil Vol 93, August 2012
1480
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
Fig 5. EMG responses elicited during the Lokomat speed experiment. Representative data from 1 assessment of the first and second training blocks are shown separately for the left and right legs. EMG amplitudes are expressed as percentage of the amplitude measured while walking at 1.5km/h within each experiment. Gray shaded areas >3.2km/h indicate typical walking speeds of individuals without disability. R2 values are presented for the third-order polynomial fit models of the mean of all muscles within 1 extremity. Red-shaded areas represent 95% confidence intervals of the mean per leg. Abbreviation: kph, kilometers per hour.
speed above which afferent inputs produce robust excitatory influences on motor output. EMG From Overground Walking Trials Overground walking was extremely limited during the first 4 weeks; thus, EMG data were not recorded. Beginning with the fifth week, EMG bursts were recorded in many muscles during voluntary (self-initiated) stepping. We analyzed EMG data to determine whether activation became more robust and corresponded with volitional stepping. However, a systematic change in EMG activation was not apparent in any of the overground walking conditions (ie, both braces locked, 1 brace unlocked, brace straps loosened) between week 5 and the end of the study. We qualitatively analyzed whether any walking condition provoked more EMG activity than another. In general, our data suggest that unlocking the brace of either leg produced more EMG activity than having both braces locked or Arch Phys Med Rehabil Vol 93, August 2012
loosening the straps of either brace. Thus, it appears that less restrictive support contributes to increased neural activation even if in the context of requisite bracing support (fig 6). DISCUSSION Individuals post-SCI with some sacral sparing usually have a considerable chance of regaining some degree of walking function.36,37 Thus, intensive LT might activate spared fibers and lead to the more thorough exploitation of residual physiologic capacity than in the absence of sacral sparing. The individual in this study revealed some remaining sensory function (eg, light touch in S4-5); however, he remained nonambulatory 16 months after his initial injury. During the course of LT he demonstrated recovery of locomotor EMG patterns. Recently, Manella et al38 reported functional locomotor recovery to a very similar extent in a participant 2 years postSCI. Both the present case and Manella’s38 stand in contrast to
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
1481
Fig 6. EMG activity during overground stepping. EMG “on time” (milliseconds [ms] of muscle activation) was normalized by distance (feet) to assess differences while stepping with varying levels of external support. For sake of illustration, EMG “on time” was collapsed over 7 muscles measured within each leg and averaged over the 4 assessments in the second LT block. Error bars represent 95% confidence intervals.
previous studies.39,40 Typically, individuals with chronic SCI who demonstrate improved locomotor function in response to LT reveal higher baseline motor function below the lesion,41,42 recordable lower extremity EMG activity43 at baseline, or both. In these studies,39-43 individuals demonstrating functional levels similar to our participant did not improve locomotor function. Furthermore, on the basis of contemporary prediction rules, our participant responded to LT with improvements that would not have been expected before initiation of the training.44,45 Of note, these clinical prediction rules have been developed for a more acute stage post-SCI where even more spontaneous recovery can be expected. Among other factors, we hypothesize that the unexpected outcome in the present case could be attributable to the differences in training speed between the earlier LT studies (between 0.7 and 2.1km/h)39-42 and the training speeds used with our participant (up to 5km/h). Technical advancements implemented in the Lokomat enabled stepping at speeds greater than 3.2km/h (.88m/s or 1.8mph). Comparable manual LT in a participant of his stature and requiring such a high level of guidance would have been very challenging and labor intensive.10 While it may be possible to attain physiologic walking speeds with manual assistance, sustaining this activity over training bouts lasting up to 17 minutes would not be feasible. Physiologic stepping speeds provide both increased intensity and increased amplitude of locomotor-related afference that may contribute to reactivation of residual pathways that were inactive because of nonuse in this participant. It has been argued that slower walking speeds are critical in promoting recovery of corticospinal input because, unlike faster speed, they allow the participant to actively and voluntarily participate in producing stepping movements.43,46 However, it is also possible that integration of spared descending input can be facilitated by appropriate sen-
sory inputs at the spinal level according to the Hebbian principle.47 Our participant not only revealed improved EMG activity with higher walking speeds, but was able to transfer his regained stepping ability to overground stepping where the facilitatory peripheral inputs were no longer present. This result suggests that the higher training speeds7-9 and, potentially, the variability48 of speeds within each session either directly influenced the descending tract function or the manner in which the input from those tracts is integrated at the spinal level. Although we encouraged the participant to “walk with the robot,” the Lokomat in the present study used position control of the limbs, which allows the participant to remain fairly passive.49,50 Evolving robotic control approaches, including patient-cooperative control, are intended to combine the advantages of high stepping speeds with more voluntary participation and enhanced kinematic and temporal variability to promote greater motor learning.48,49,51,52 In this single-case design, we refrained from assigning the participant a severity level (ASIA Impairment Scale [AIS]) according to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI).35 Accurate ISNCSCI testing is challenging and needs continuous training of the assessors.53-55 The data we were able to obtain from the patient’s records did not allow for a reliable classification. However, we believe the actual AIS score is less important than the fact that the participant possessed untapped potential for neurophysiologic recovery. Even in clinically complete SCIs, spared fibers cross the injury site56 and appear to be physiologically intact.57,58 However, motor output relies on activity in many pathways and the appropriate integration of these influences at the spinal level. Even small dysfunction in the gray matter is disruptive to this integrative process and therefore can interfere with purposeful Arch Phys Med Rehabil Vol 93, August 2012
1482
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
movement.58 Alternatively, findings from animal studies reveal that a small number of spared fibers can suffice to recover at least some movement in the lower extremities.59-61 Dimitrijevic et al62 assigned participants with clinically complete, but electrophysiologically incomplete, lesions the term “discomplete” and suggested such individuals may possess greater potential for functional recovery than individuals with complete SCI. Both short- and long-latency EMG responses were recorded in these discomplete individuals, suggesting that both direct corticospinal pathways and collateral pathways through propriospinal networks could be recruited simultaneously. Manella38 speculated that their participant experienced a discomplete injury, leading to the marked functional recovery in the chronic stage. This hypothesis can be further elaborated in our participant by monitoring EMG as an outcome measure during overground walking. Clinical MR images from our participant reveal damage primarily to gray matter at the injury site, while a rim of mostly ventral white matter remained intact. It is possible that LT in this individual led to recovery of a few critical connections that allowed for reintegration of necessary pathways and thus the observed recovery of EMG patterns. Additionally, our participant exhibited signs of central neuronal damage (clonus, weakness) for at least 6 weeks before his surgery. It is possible that these signs and symptoms reveal a “preinjury” that served as a primer for neuroplasticity after his principal spinal cord damage.63 This ostensibly staged injury, combined with the intensive LT, might have contributed to activation of residual, spared pathways. Our hypothesis is supported by the emergence of patterned EMG within a very short time (a few weeks), therefore likely representing transmission in existing synapses that had not been activated, or activated effectively, before LT.64 Finally, according to most prediction rules, this participant would not have been identified as a candidate for LT in a clinical setting. Thus, it would have been unlikely that he would have been given the chance to train with the intensity and for the duration offered in this study. Nevertheless, he made progress that was both quantifiable and subjectively regarded as meaningful. The participant continues to walk/step daily even 4 years posttraining. Study Limitations The valuable insights gained from this single-case design serve as proof of principle of the potential for late recovery and add to the body of literature demonstrating recovery of locomotor function in the chronic state of incomplete SCI. However, we recognize that the results from this study might not be generalizable to a larger population of individuals with incomplete or discomplete SCI and do not reflect the same level of evidence as a randomized controlled trial. In addition, this study lacked inclusion of a clinical gait measurement tool that is more sensitive to change in individuals with greater functional abilities. We were also unable to retrieve complete information on ISNCSCI data, which prevented us from assigning an AIS score. Finally, because of the severe gait deficits, collection of consistent EMG data from overground walking was challenging in this participant, especially early in the study. CONCLUSIONS In this individual with chronic incomplete SCI who was nonambulatory, intensive LT at physiologic walking speeds led to reemergence of locomotor EMG patterns. Volitional EMG activity could also be measured while walking overground, and greater EMG activation could be noted with less restrictive Arch Phys Med Rehabil Vol 93, August 2012
support (eg, braces unlocked). Individuals with chronic incomplete SCI may still benefit from intensive LT, especially at physiologic walking speeds. Cases as described herein are possible and may, in fact, be underreported in the literature.38,65 Acknowledgments: We thank the staff at the Rehabilitation Research & Development Center, VA Palo Alto Health Care System, and the VA Brain Rehabilitation Research Center of Excellence, Gainesville, Florida. We also thank the following persons for their contributions to this work: Aaron Grogan, MS, for his engineering support; and Keith Peters, MD, for independent, blinded assessment of the MR images. References 1. Dietz V, Colombo G, Jensen L, Baumgartner L. Locomotor capacity of spinal cord in paraplegic patients. Ann Neurol 1995;37: 574-82. 2. Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Brooks VB, editor. Handbook of physiology. Sec. 1. The nervous system. Vol. II. Motor control, part 2. Washington (DC): American Physiological Society; 1981. p 1179-236. 3. Dietz V. Central pattern generator. Paraplegia 1995;33:739. 4. Dobkin B, Edgerton V, Fowler E. Training induces rhythmic locomotor EMG patterns in subjects with complete SCI. Neurology 1992;42(Suppl 3):207-8. 5. Edgerton VR, Leon RD, Harkema SJ, et al. Retraining the injured spinal cord. J Physiol 2001;533(Pt 1):15-22. 6. Rossignol S, Drew T, Brustein E, Jiang W. Locomotor performance and adaptation after partial or complete spinal cord lesions in the cat. Prog Brain Res 1999;123:349-65. 7. Dietz V. Role of peripheral afferents and spinal reflexes in normal and impaired human locomotion. Rev Neurol (Paris) 1987;143: 241-54. 8. Dobkin BH, Harkema S, Requejo P, Edgerton VR. Modulation of locomotor-like EMG activity in subjects with complete and incomplete spinal cord injury. J Neurol Rehabil 1995;9:183-90. 9. Beres-Jones JA, Harkema SJ. The human spinal cord interprets velocity-dependent afferent input during stepping. Brain 2004; 127(Pt 10):2232-46. 10. Pohl M, Mehrholz J, Ritschel C, Ruckriem S. Speed-dependent treadmill training in ambulatory hemiparetic stroke patients: a randomized controlled trial. Stroke 2002;33:553-8. 11. Colombo G, Joerg M, Schreier R, Dietz V. Treadmill training of paraplegic patients using a robotic orthosis. J Rehabil Res Dev 2000;37:693-700. 12. Colombo G, Wirz M, Dietz V. Driven gait orthosis for improvement of locomotor training in paraplegic patients. Spinal Cord 2001;39:252-5. 13. Hesse S, Schmidt H, Werner C, Bardeleben A. Upper and lower extremity robotic devices for rehabilitation and for studying motor control. Curr Opin Neurol 2003;16:705-10. 14. Lunenburger L, Bolliger M, Czell D, Muller R, Dietz V. Modulation of locomotor activity in complete spinal cord injury. Exp Brain Res 2006;174:638-46. 15. Behrman AL, Lawless-Dixon AR, Davis SB, et al. Locomotor training progression and outcomes after incomplete spinal cord injury. Phys Ther 2005;85:1356-71. 16. Barbeau H. Locomotor training in neurorehabilitation: emerging rehabilitation concepts. Neurorehabil Neural Repair 2003;17:3-11. 17. Dobkin B, Barbeau H, Deforge D, et al. The evolution of walkingrelated outcomes over the first 12 weeks of rehabilitation for incomplete traumatic spinal cord injury: the multicenter randomized Spinal Cord Injury Locomotor Trial. Neurorehabil Neural Repair 2007;21:25-35. 18. Dietz V. Body weight supported gait training: from laboratory to clinical setting. Brain Res Bull 2008;76:459-63.
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
19. De Leon RD, Hodgson JA, Roy RR, Edgerton VR. Full weightbearing hindlimb standing following stand training in the adult spinal cat. J Neurophysiol 1998;80:83-91. 20. Edgerton VR, Roy RR, Hodgson JA, Prober RJ, Guzman CP, de Leon R. Potential of adult mammalian lumbosacral spinal cord to execute and acquire improved locomotion in the absence of supraspinal input. J Neurotrauma 1992;9(Suppl 1):S119-28. 21. Harkema SJ, Hurley SL, Patel UK, Requejo PS, Dobkin BH, Edgerton VR. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol 1997;77:797-811. 22. Brown G. The intrinsic factors in the act of progression in mammals. Proc R Soc 1911;B(84):308-19. 23. Jahn K, Deutschlander A, Stephan T, et al. Supraspinal locomotor control in quadrupeds and humans. Prog Brain Res 2008;171: 353-62. 24. Pearson K, Gordon J. Locomotion. In: Kandel ER, Schwartz JH, Jessell TM, editors. Principles of neural science. 4th ed. New York: McGraw Hill; 2000. p 737-55. 25. Stolze H, Kuhtz-Buschbeck JP, Mondwurf C, et al. Gait analysis during treadmill and overground locomotion in children and adults. Electroencephalogr Clin Neurophysiol 1997;105:490-7. 26. Van Gheluwe B, Smekens J, Roosen P. Electrodynographic evaluation of the foot during treadmill versus overground locomotion. J Am Podiatr Med Assoc 1994;84:598-606. 27. Day KV. Motor control solutions to dynamic stability deficits during walking after spinal cord injury [dissertation]. 2010, University of Florida: Gainesville. 28. Lee BY, Wendell K, Al-Waili N, Butler G. Ultra-low microcurrent therapy: a novel approach for treatment of chronic resistant wounds. Adv Ther 2007;24:1202-9. 29. Dittuno PL, Dittuno JF Jr. Walking Index for Spinal Cord Injury (WISCI II): scale revision. Spinal Cord 2001;39:654-6. 30. Tawney JW, Gast DL. Single subject research in special education. Columbus: CE Merrill; 1984. 31. Behrman AL, Harkema SJ. Physical rehabilitation as an agent for recovery after spinal cord injury. Phys Med Rehabil Clin N Am 2007;18:183-202, v. 32. Chou LW, Lee SC, Johnston TE, Binder-Macleod SA. The effectiveness of progressively increasing stimulation frequency and intensity to maintain paralyzed muscle force during repetitive activation in persons with spinal cord injury. Arch Phys Med Rehabil 2008;89:856-64. 33. Scott WB, Lee SC, Johnston TE, Binkley J, Binder-Macleod SA. Effect of electrical stimulation pattern on the force responses of paralyzed human quadriceps muscles. Muscle Nerve 2007;35: 471-8. 34. Kebaetse MB, Lee SC, Johnston TE, Binder-Macleod SA. Strategies that improve paralyzed human quadriceps femoris muscle performance during repetitive, nonisometric contractions. Arch Phys Med Rehabil 2005;86:2157-64. 35. ASIA. International standards for neurological classification of spinal cord injury. Chicago: American Spinal Injury Association; 2002. 36. van Middendorp JJ, Hosman AJ, Pouw MH, Van de Meent H. ASIA Impairment Scale conversion in traumatic SCI: is it related with the ability to walk? A descriptive comparison with functional ambulation outcome measures in 273 patients. Spinal Cord 2009; 47:555-60. 37. Ditunno PL, Patrick M, Stineman M, Ditunno JF. Who wants to walk? Preferences for recovery after SCI: a longitudinal and cross-sectional study. Spinal Cord 2008;46:500-6. 38. Manella KJ, Torres J, Field-Fote EC. Restoration of walking function in an individual with chronic complete (AIS A) spinal cord injury. J Rehabil Med 2010;42:795-8. 39. Dietz V, Harkema SJ. Locomotor activity in spinal cord-injured persons. J Appl Physiol 2004;96:1954-60.
1483
40. Müller RM. Lokomotions-, und Reflexaktivität bei kompletter Querschnittlähmung: Einflussfaktoren und zeitliche Entwickling, [dissertation]. 2006, Eidgenössische Technische Hochschule ETH; Zürich. 41. Hicks AL, Adams MM, Martin Ginis K, et al. Long-term bodyweight-supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective well-being. Spinal Cord 2005;43: 291-8. 42. Wernig A, Nanassy A, Muller S. Maintenance of locomotor abilities following Laufband (treadmill) therapy in para- and tetraplegic persons: follow-up studies. Spinal Cord 1998;36:744-9. 43. Gorassini MA, Norton JA, Nevett-Duchcherer J, Roy FD, Yang JF. Changes in locomotor muscle activity after treadmill training in subjects with incomplete spinal cord injury. J Neurophysiol 2009;101:969-79. 44. van Middendorp JJ, Hosman AJ, Donders AR, et al. A clinical prediction rule for ambulation outcomes after traumatic spinal cord injury: a longitudinal cohort study. Lancet 2011;19:1004-10. 45. Yang JF, Norton J, Nevett-Duchcherer J, Roy FD, Gross DP, Gorassini MA. Volitional muscle strength in the legs predicts changes in walking speed following locomotor training in people with chronic spinal cord injury. Phys Ther 2011;91:931-43. 46. Laufband therapy - the manual for spinal cord damage, stroke, brain damage and MS, orthopedic disorders and others [PDF]. Müller S, Wernig A. Bonn, Germany, 2005. Available at: http:// www.meb.uni-bonn.de/wernig/html/manual.html. Accessed April 10, 2012. 47. Hebb D. The organization of behavior: a neuropsychological theory. New York: Wiley; 1949. 48. Ziegler MD, Zhong H, Roy RR, Edgerton VR. Why variability facilitates spinal learning. J Neurosci 2010;30:10720-6. 49. Duschau-Wicke A, Caprez A, Riener R. Patient-cooperative control increases active participation of individuals with SCI during robot-aided gait training. J Neuroeng Rehabil 2010;7:43. 50. Wernig A. Long-term body-weight supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective well-being. Spinal Cord 2006;44:265-6; author reply 267-8. 51. Emken JL, Harkema SJ, Beres-Jones JA, Ferreira CK, Reinkensmeyer DJ. Feasibility of manual teach-and-replay and continuous impedance shaping for robotic locomotor training following spinal cord injury. IEEE Trans Biomed Eng 2008;55:322-34. 52. Mankala KK, Banala SK, Agrawal SK. Novel swing-assist unmotorized exoskeletons for gait training. J Neuroeng Rehabil 2009;6:24. 53. Schuld C, Wiese J, Hug A, et al. Computer implementation of the international standards for neurological classification of spinal cord injury for consistent and efficient derivation of its subscores including handling of data from not testable segments. J Neurotrauma 2011;29:453-61. 54. Spiess MR, Muller RM, Rupp R, Schuld C, van Hedel HJ. Conversion in ASIA Impairment Scale during the first year after traumatic spinal cord injury. J Neurotrauma 2009;26:2027-36. 55. Dietz V, Muller R. Degradation of neuronal function following a spinal cord injury: mechanisms and countermeasures. Brain 2004; 127(Pt 10):2221-31. 56. Kakulas BA, Bedbrook GM. A correlative clinico-pathologic study of spinal cord injury. Proc Aust Assoc Neurol 1969;6: 123-32. 57. Dimitrijevic MR. Residual motor functions in spinal cord injury. Adv Neurol 1988;47:138-55. 58. Sherwood AM, Dimitrijevic MR, McKay WB. Evidence of subclinical brain influence in clinically complete spinal cord injury: discomplete SCI. J Neurol Sci 1992;110:90-8. Arch Phys Med Rehabil Vol 93, August 2012
1484
LOCOMOTOR TRAINING AT PHYSIOLOGIC STEPPING SPEED, Spiess
59. Basso DM. Neuroanatomical substrates of functional recovery after experimental spinal cord injury: implications of basic science research for human spinal cord injury. Phys Ther 2000;80:808-17. 60. Basso DM, Beattie MS, Bresnahan JC. Descending systems contributing to locomotor recovery after mild or moderate spinal cord injury in rats: experimental evidence and a review of literature. Restor Neurol Neurosci 2002;20:189-218. 61. Hill RL, Zhang YP, Burke DA, et al. Anatomical and functional outcomes following a precise, graded, dorsal laceration spinal cord injury in C57BL/6 mice. J Neurotrauma 2009;26:1-15. 62. Dimitrijevic MR, Dimitrijevic MM, Faganel J, Sherwood AM. Suprasegmentally induced motor unit activity in paralyzed muscles of patients with established spinal cord injury. Ann Neurol 1984;16:216-21. 63. McPhail LT, Fernandes KJ, Chan CC, Vanderluit JL, Tetzlaff W. Axonal reinjury reveals the survival and re-expression of regeneration-associated genes in chronically axotomized adult mouse motoneurons. Exp Neurol 2004;188:331-40.
Arch Phys Med Rehabil Vol 93, August 2012
64. Nathan PW, Smith MC. Effects of two unilateral cordotomies on the motility of the lower limbs. Brain 1973;96:471-94. 65. Spiess M, Schuld C. Comment on “Restoration of walking function in an individual with chronic complete (AIS A) spinal cord injury.” J Rehabil Med 2011;43:367-8. 66. Whiteneck GG, Charlifue SW, Gerhart KA, Overholser JD, Richardson GN. Quantifying handicap: a new measure of longterm rehabilitation outcomes. Arch Phys Med Rehabil 1992; 73:519-26. Suppliers a. Hocoma AG, Industriestrasse 4, CH-8604 Volketswil, Switzerland. b. Mobility Research, 444 West Geneva Dr, Tempe, AZ 85282-2004. c. MA-300, preamplified electrodes; Motion Lab Systems, Inc, 15045 Old Hammond Hwy, Baton Rouge, LA 70816. d. AD Instruments, 322 North Landings Dr, Grand Junction, CO 81501. e. The MathWorks, 3 Apple Hill Dr, Natick, MA 01760.