Spinal Cord Microstimulation: Long-term Stability of Implanted Wires and Evoked Responses

In previously conducted acute experiments, it was shown that spinal cord microstimulation (SCµstim) can generate functional limb movements in cats: SCµstim can selectively activate individual muscle groups, produce smooth and graded muscle contractions and reduce muscle fatigue [1-3]. Since March 1998 we have been developing a chronically implantable SCµstim system. The goal of the work is to determine the long-term stability of muscle contractions produced by stimulating motoneurons and interneurons in the mammalian lumbo-sacral spinal cord. The intended clinical goal is to improve motor function in individuals with spinal cord injuries and hemiplegia through SCµstim. Arrays of 6-20 microwires of varying diameter (17-50 µm) and stiffness were implanted in the lumbo-sacral spinal cord of 12 adult, intact cats for up to 6 months. The loose wires in each array were pre-bent (3.0–4.5) mm lengths, insulated except for 30-100 µm at the tip) and temporarily held in a silastic gig (spaced 1-3 mm apart), which allowed each to be inserted separately. This had the advantage of causing minimal dimpling to the cord and allowing adequate flexibility in positioning. After insertion, the connecting portions of the microwires lay in an orderly array flush with the dural surface. Plastic film glued over the wires prevented regrowth of muscle, avoiding dislodgement of wires over time [4]. Post-mortem dissections indicated that electrode tips were usually close to their ventral horn target, though their localization could still be improved [4]. Stimulation through single wires (10-300 µA, 300 µs, 2-50 pps) generated three types of responses: movements about one joint (60% of wires), whole-limb synergistic movements involving the hip, knee and ankle (30%), and cocontraction of mutually antagonistic muscles causing stiffening of one or more joints (10%). Torques generated in the first two types of responses where large enough to lift the animals' hindquarters (i.e., load-bearing) [4]. Throughout the duration of testing stimulus thresholds and evoked responses remained constant [4]. Stimulus threshold for all microwires doubled after the first 10 days of implantation but remained stable thereafter. This initial increase in threshold was attributed to electrode encapsulation, a product of the natural immunological reaction to implanted foreign objects [4]. The stability of stimulus thresholds following the initial increase indicates that the microwires were remaining securely in place and were not inducing an ongoing inflammatory reaction. At least 67% (67 – 100%) of the microwires implanted in each animal continued to activate the same muscle(s) throughout the duration …