Locomotor training restores walking in a nonambulatory child with chronic, severe, incomplete cervical spinal cord injury

Abstract

Background and purpose: Locomotor training (LT) enhances walking in adult experimental animals and humans with mild-to-moderate spinal cord injuries (SCIs). The animal literature suggests that the effects of LT may be greater on an immature nervous system than on a mature nervous system. The purpose of this study was to evaluate the effects of LT in a child with chronic, incomplete SCI.

Subject: The subject was a nonambulatory 4 1/2-year-old boy with an American Spinal Injury Association Impairment Scale (AIS) C Lower Extremity Motor Score (LEMS) of 4/50 who was deemed permanently wheelchair-dependent and was enrolled in an LT program 16 months after a severe cervical SCI.

Methods: A pretest-posttest design was used in the study. Over 16 weeks, the child received 76 LT sessions using both treadmill and over-ground settings in which graded sensory cues were provided. The outcome measures were ASIA Impairment Scale score, gait speed, walking independence, and number of steps.

Result: One month into LT, voluntary stepping began, and the child progressed from having no ability to use his legs to community ambulation with a rolling walker. By the end of LT, his walking independence score had increased from 0 to 13/20, despite no change in LEMS. The child's final self-selected gait speed was 0.29 m/s, with an average of 2,488 community-based steps per day and a maximum speed of 0.48 m/s. He then attended kindergarten using a walker full-time.

Discussion and conclusion: A simple, context-dependent stepping pattern sufficient for community ambulation was recovered in the absence of substantial voluntary isolated lower-extremity movement in a child with chronic, severe SCI. These novel data suggest that some children with severe, incomplete SCI may recover community ambulation after undergoing LT and that the LEMS cannot identify this subpopulation.

Figure 1.

Consequences of spinal cord injury: (a) Acute T2-weighted sagittal magnetic resonance image (MRI) of spinal cord revealing central hemorrhagic necrosis at C6. Bold white arrow indicates lesion site. (b–e) Acute T2-weighted axial MRIs through the lesion epicenter. The ventral aspect of the spinal cord is at the top of the images, and the dorsal aspect is at the bottom. The fine white arrow in panel b indicates an area of complete tissue interruption in the dorsal lateral aspect of the spinal cord. The fine white arrow in panel c indicates a thin rim of ventral tissue, and the fine dark arrow indicates a thin rim of dorsal tissue. Between these tissue rims is a fluid-filled cyst. (f–i) Drawings of the remaining tissue corresponding to the axial MRIs shown in panels b through e. Asterisk in panels d and h represents the comparable area of lesion cavity. Asterisk in panels e and i is centered over tissue not present rostally. (j) Before training, the child's primary means of mobility outside of a standing frame or wheelchair: child pulling his body across the floor using only his arms.

Figure 2.

Progression from wheelchair dependence to independent ambulation with a walker across the 76 sessions of locomotor training (LT) and at a 1-month retention test following completion of training: the total number of steps over ground per training session relative to the number of independent steps with a rolling walker. After treadmill-based LT sessions 1 through 9, no over-ground steps were attempted. Following treadmill-based LT sessions 10 through 29, the range in number of steps assisted over ground was 10 to 15. From session 23 to session 29, manual cueing of the tibialis anterior and iliopsoas muscles was intensified over ground, complementing the training on the treadmill. At session 29 and over ground, the first steps without direct cueing of the limb occurred (2/14 steps), although each step followed a contralaterally cued step. The decrease in the number of steps taken after session 41 and over ground reflects the period of videotaping rather than the child's overall capacity. His increasing skill after session 41, however, is reflected in the percentage of independently initiated steps exhibited over ground with the rolling walker, which increased rapidly and steadily from 69% at session 34 to 100% at session 76.

Figure 3.

Progression of self-selected (SS) and fastest (FS) gait speed (meters per second) walking over ground with a rolling walker or with manual assistance and ranked by the Walking Index for Spinal Cord Injury II (WISCI-II) across training sessions and at the 1-month retention test following completion of training.

Figure 4.

Independent walking. (a) Using a rolling walker at 1-month follow-up after completion of training. (b, c) Amount of self-selected walking in the home and community for a 24-hour period following 2 different training sessions: (b) 936 steps after session 49 (week 10), (c) 2,488 steps after session 74 (week 16). (d, e) The most productive hour of walking from data sets b and c, respectively, and demonstrating the greatest capacity of the child for walking activity: (d) 280 steps in 1 hour with 8.3% of steps at >20 steps/min after session 49, (e) 956 steps in 1 hour with 43.3% of steps at >20 steps/min after session 74.