Engaging cervical spinal circuitry with non-invasive spinal stimulation and buspirone to restore hand function in chronic motor complete patients

Abstract

The combined effects of cervical electrical stimulation alone or in combination with the monoaminergic agonist buspirone on upper limb motor function were determined in six subjects with motor complete (AIS B) injury at C5 or above and more than one year from time of injury. Voluntary upper limb function was evaluated through measures of controlled hand contraction, handgrip force production, dexterity measures, and validated clinical assessment batteries. Repeated measure analysis of variance was used to evaluate functional metrics, EMG amplitude, and changes in mean grip strength. In aggregate, mean hand strength increased by greater than 300% with transcutaneous electrical stimulation and buspirone while a corresponding clinically significant improvement was observed in upper extremity motor scores and the action research arm test. Some functional improvements persisted for an extended period after the study interventions were discontinued. We demonstrate that, with these novel interventions, cervical spinal circuitry can be neuromodulated to improve volitional control of hand function in tetraplegic subjects. The potential impact of these findings on individuals with upper limb paralysis could be dramatic functionally, psychologically, and economically.

Figure 1

Improvement in hand function with interventions. Assessments of grip strength were made in 6 paralyzed subjects over 6 weeks Pre-Treatment (Phase 1, Pre Treat), followed by testing over 2 weeks with stimulation (Phase 2, +Stim, −Busp), followed by testing over 2 weeks with stimulation plus oral buspirone (Phase 3, +Stim, +Busp), followed by testing over 2 weeks with stimulation plus placebo (Phase 4, +Stim, −Busp), and followed by at least 4 days of testing over 2 weeks Post-Treatment after a 3–6 month delay (Phase 5, Post Treat) (a–f) (also see Fig. 4e for the number of test sessions and the number of contractions in each Phase). Mean (±SEM) grip strength for all 6 subjects before (Pre), with stimulation (+Stim), and after stimulation (Post) for each phase (g), EMG amplitudes as percent increase over baseline (Pre Treat) levels (h), ARAT and UEMS (i) are shown. (a–f) show plots for each individual with pre, during, and post averaged within a period. G represents averages of the group means within each phase further divided by baseline (Pre), stimulation (+Stim), and post-stimulation (Post) conditions. Horizontal lines indicate significant differences at P < 0.05 (dotted), P < 0.01 (dash-dot), or P < 0.001 (solid line). FD, flexor digitorum; ED, extensor digitorum; Brac, brachioradialis; Bicep, biceps brachii; Tricep, triceps brachii.

Figure 2

Patient demographic and functional data and sagittal T2 MRI Imaging demonstrating the location of cervical spinal cord injury of subjects (a–f) at approximately the C2 (c), C5 (b,d,e) or C6 (a,f) spinal levels. The injury location displays a high intensity T2 signal corresponding to a glial scar formation. Spinal cord tissue distal and proximal to the injury locus has a normal appearance without evidence of post-traumatic syrinx formation. The extent of cord atrophy ranges from 62% loss of cord diameter (c) to 26% loss of cord diameter (b) when compared to adjacent uninjured levels. The glial scar spans approximately one spinal cord level in all subjects with the exception of subject a where the glial scar spans two vertebral levels.

Figure 3

Plot of maximal hand force across contractions during each of the five phases of testing for all six subjects. The slopes in Phase 1 are relatively neutral for all subjects signifying a lack of improvement in force during Phase 1. There is a total of at least 10 sessions (90 hand contractions) with data recorded during at least 3 sessions at the beginning, middle, and end of Phase 1. In general, there were more sessions in Phase 2 due to the fine-tuning of the simulation parameters for each subject. Piecewise linear regression lines are drawn within each phase. Solid diamonds represent stimulation on, unfilled diamonds represent stimulation off; b = slope; r = correlation coefficient; * denotes p < 0.01 for slope (rate) compared to zero. N = Newton, X-axis numerical values represent the phases of the study. Shaded area in e highlights the forces above 15 N.

Figure 4

Baseline testing (Phase 1) demonstrates stable hand function by handgrip device (a), ARAT (b), and UEMS (c) scores prior to initiation of any intervention. Upper extremity modified Ashworth scores were assessed during each session before (squares) and after (diamonds) testing (d). The total number of sessions and hand contractions and the specific UEMS are shown in (e,f), respectively. Maximal handgrip force on the last day of testing was not different from the first day of testing in the 6 subjects (a). Testing during this phase spanned a period of 6 weeks with bi-weekly testing sessions. Each data point represents an average of 3 maximal handgrip contractions. Clinical testing by ARAT (b) and UEMS (c) scores demonstrate no evidence of improvement during this initial baseline-testing period. There was a general decrease in upper extremity spasms within each testing session and as the study progressed (d). During baseline Phase 1 testing, there were more sessions and hand contractions than for all of Phases 2–4 combined (e). Note that the variability in the number of sessions and contractions among subjects and across different Phases was inevitable as a result of complications of co-morbidities associated with SCI, i.e., urinary tract infections, pressure ulcers, etc. During Phase 1, EMG was not recorded during all sessions due to the relative stability of the response: consequently, there are a higher number of contractions during Phase 1 compared to that shown in Fig. 3.

Figure 5

EMG profiles during hand contractions in a normal subject and Subject D. Representative photo of the handgrip device (a). Different spring(s) (arrows) were used to adjust the resistance level of the handgrip. The normal subject was asked to contract his hand in response to perceived percent effort (10, 25, 50, 75, 100%) and normal EMG responses are shown (b). The average EMG throughout the sessions during Phase 1–5 were shown for Subject D (c). Raw EMG tracings of Subject D during a maximal hand contraction in Phase 3 with and without stimulation (d) are shown. In normal subjects, there is a predominant activation of distal arm muscles (FD, ED, Brac) and a lesser activation of proximal arm muscles (Bicep, Tricep, and Delt) during a hand contraction (b). There also is an increase in EMG amplitude as the percentage of effort increases from 10% to 100%. In Subject D, there is a predominant activation of the Tricep during a hand contraction in the early phases of testing, particularly in Phases 1 and 2 (c). During the later phases, particularly Phase 5, there is an increase in the activation of FD, ED, and Brac during hand contraction (c), closely mimicking the pattern observed in the normal subject (b). The handgrip force and EMG during a maximum handgrip performed with (11 N) and without (7 N) stimulation show increased activation in the FD and ED and relative relaxation of the Tricep and Delt to impart a greater force (d). There is no evidence of spastic activity as the hand contraction is performed. FD, flexor digitorum; ED, extensor digitorum; Brac, brachioradialis; Bicep, biceps brachii; Tricep, triceps brachii; Delt, deltoid. Ph, Phase.