Abstract
An implanted neuroprosthesis supplying functional neuromuscular stimulation was used to provide grasp and release to tetraplegic individuals. This article describes the results, at a minimum of three years, for the first five patients to have operative implantation of an eight-channel stimulator-receiver. All of the patients had a clinically complete spinal cord injury with motor function remaining at the level of the fifth or sixth cervical nerve root.In addition to implantation of the stimulator system, each patient had augmentative operations on the hand to improve function. The procedures included tendon transfers, side-to-side tendon anastomoses, arthrodesis of the interphalangeal joint of the thumb, and rotational osteotomy of the radius.The neuroprosthesis provides two grasp patterns controlled by voluntary motion of the shoulder or wrist. Functional evaluations included measurement of pinch force, a grasp-release test, evaluation of the level of functional independence, and usage surveys. Pinch force ranged from eight to twenty-five newtons. All five patients demonstrated functional grasp patterns, had increased independence, and were able to use the neuroprosthesis at home on a regular basis. The implanted stimulator has proved to be safe and reliable, with seven years as the longest time in situ at the time of writing.
Functional neuromuscular stimulation has been used to provide grasp and release to tetraplegic individuals7,11,20-23. The neuroprosthesis stimulates the muscles of the forearm and hand in a coordinated fashion to provide functional grasp patterns. Proportional control of opening and closing of the grasp is accomplished by voluntary movement, generally of the shoulder9 although movement of the wrist has also been a command source. The force output of each muscle is controlled by modulation of the stimulus delivered to the muscle. Early neuroprosthetic systems consisted of a few percutaneous electrodes and provided one grasp pattern22. More recently, these percutaneous systems included as many as sixteen active electrodes to provide both palmar and lateral grasp patterns21. A switch, usually mounted on the chest, is used to turn the system on and off and to select the grasp pattern. A mechanism to lock the hand is also provided. These systems have been used successfully by more than fifty patients and have enabled them to perform many activities of daily living26.
More recently, implantable systems have been developed that eliminate percutaneous leads. In these systems, the stimulator-receiver is operatively implanted with leads tunneled subcutaneously to the muscles12,24. In the present study, the first five patients to have received this system were followed for at least three years.
*One or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund or foundation, educational institution, or other non-profit organization with which one or more of the authors are associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were the Department of Veterans Affairs Rehabilitation Research and Development Service Office of Research and Development, the National Institutes of Health Neural Prosthesis Program, the National Institutes of Health Clinical Research Center at MetroHealth Medical Center and Case Western Reserve University Grant M01 RR00080-31, and the Food and Drug Administration Orphan Products Grant FD000832.
†Hamann 601, MetroHealth Medical Center, 2500 MetroHealth Drive, Cleveland, Ohio 44109.
‡NeuroControl Corporation, 1945 East 97th Street, Suite W1-315, Cleveland, Ohio 44106.
Description of the Neuroprosthetic System
The implanted components of the neuroprosthetic system consist of an eight-channel stimulator-receiver, eight epimysial electrodes, leads, and connectors (Fig. 1), as described by Smith et al. Seven of the electrodes are placed on the muscles used to control motion of the fingers and thumb. The eighth electrode is placed in the supraclavicular region and is used for sensory feedback. A radio frequency inductive link provides the communication and power to the stimulator-receiver implant. The external components of the neuroprosthesis are an external control unit, a transmitting coil, and a shoulder control unit. The external control unit processes the signal of the control inputs and generates the output signal (modulated radio frequency) to the stimulator-receiver implant. The external unit is programmed with the appropriate control and stimulus parameters, stored in look-up tables, that relate the position of the shoulder to the stimulus output applied to each muscle. These parameters are customized for each patient15. The external control unit is battery powered and portable. The shoulder control unit consists of a joystick mounted to the chest and a logic switch. The control unit can also be attached to the wrist to allow movement of the wrist to control opening and closing of the grasp. The radio frequency-transmitting coil is taped to the chest directly over the stimulator-receiver implant to make the inductive communication link.
Patients
The stimulator was implanted in five tetraplegic individuals. All of the patients had a clinically complete spinal cord injury with motor function remaining at the level of either the fifth or the sixth cervical nerve root, according to the classification of the American Spinal Injury Association. The classification according to the International Conference on Surgical Rehabilitation of the Upper Limb in Tetraplegia17 was group 0, 1, or 2, with either ocular (O) or cutaneous (Cu) feedback, or both (Table I). One patient (Case 4) had a closed head injury in addition to the spinal cord injury. At the time of implantation, the patients ranged in age from twenty-eight to fifty-seven years (median, thirty-four years). There were three men and two women. The time from the injury to insertion of the implant was two years and two months to nine years (median, three years and seven months). Two patients had previously used a percutaneous system for functional purposes21, and one patient had used a percutaneous system for exercise only. The remaining two patients had muscle-conditioning through surface stimulation before operative placement of the stimulator. The muscle-conditioning program was used to build muscle strength in order to obtain more accurate placement of the electrodes at the time of the operation and to allow planning of the operation. All aspects of this study were supported and evaluated by the Institutional Review Board at MetroHealth Medical Center and the Veterans Affairs Medical Center at Wade Park.
Candidates were required to develop a contractile strength of 3 (full range against gravity) or better, according to the 6-point Medical Research Council scale1 of 0 to 5, during the application of electrical stimulation to the nerve supply of the extensor pollicis longus, extensor digitorum communis, and abductor pollicis brevis and a contractile strength of 4 or better for the adductor pollicis, flexor digitorum superficialis, and flexor digitorum profundus. If at least one of these muscles was weak or had lower motor-neuron damage, augmentative operative procedures were used to achieve both palmar and lateral grasp patterns. Candidates had a nearly normal range of motion of the joints and pharmacologically controlled spasticity such that functional use of the upper extremity was possible. Other medical and psychological factors were used to select patients who were good candidates for rehabilitation.
Operative Procedures
Selection of the Upper Extremity for Implantation
The stronger upper extremity was selected for implantation. If the extremities were of approximately equal strength, muscle excitability, sensory sparing, and dominance were also considered.
Implantation of the Stimulator
The implantation procedure was essentially unchanged from that described by Keith et al.10,12. Epimysial electrodes were sutured on the muscles in an operative procedure. Leads were tunneled from the electrodes up the arm to a connector in the mid-humeral area. The stimulator-receiver unit was implanted in the pectoral region with the leads tunneled subcutaneously to the humeral connector site, where a spring connector was used to connect the proximal implant lead to the distal electrode lead16. The sensory electrode was placed rostral to the clavicle, with the stimulating portion facing the surface of the skin, and sutured in place.
Adjuvant Operative Procedures
Additional operative procedures were performed in conjunction with the placement of the stimulator-receiver and electrodes10. The indications for the procedures included asynchronous motion of the fingers, an intrinsic-minus posture of the hand, damage of the lower motor neurons of key muscles for grasp, and hypermobility of the interphalangeal joint of the thumb. Cross anastomoses of the extrinsic muscles of the fingers compensated for asynchronous motion, providing uniform distribution of force across the digits. The intrinsic-minus posture of the hand was corrected in some patients with the Zancolli lasso procedure8,28, in which the flexor digitorum sublimis tendon is detached from its distal insertion, looped over the first annular pulley, and sutured to itself. This provided an active block for hyperextension of the metacarpophalangeal joint. If there was lower motor-neuron damage to a key muscle, the tendon of another paralyzed muscle was transferred to that muscle to compensate for the missing function10. Arthrodesis of the interphalangeal joint of the thumb was used to treat hypermobility.
Additional operative procedures were also performed to provide the patient with the highest level of function even in the absence of electrical stimulation. Transfers of tendons of muscles under voluntary control were performed to counteract a lack of voluntary extension of the wrist or elbow, or both4-6,19,25. A rotational osteotomy of the radius was performed in patients who had severe supination contracture. This osteotomy consisted of a cut made proximal to the middle third of the radius, rotation of the distal portion, and reattachment with a bone plate.
Postoperative Treatment
An above-the-elbow cast was worn for three weeks postoperatively for stabilization of the electrodes. Elective modifications of the postoperative care were made if the patient had had additional operative procedures on the upper extremity. Muscle-conditioning with use of the neuroprosthesis was done to rebuild the muscle strength after the cast was removed and was continued for at least one month before a functional system was programmed.
Programming Grasp Patterns and Control Movements
Grasp Patterns
Each patient was provided with a lateral and a palmar grasp pattern15,23, typically one to two months after the operation. The grasp patterns were customized for each patient by analysis of the gradation of muscle force with stimulation—that is, the muscle recruitment properties—of each electrode15. The pulse duration of the stimulus delivered to each electrode was used to modulate the muscle force. A series of so-called stimulus maps were generated, which related the single proportional command input signal to the stimulus output to each electrode15. In this way, coordinated control of hand movement and force of either type of grasp was achieved by the single control movement.
Control of Grasp and Release
Proportional control of opening and closing of the grasp was achieved with the remaining voluntary motion of the shoulder9 by four patients. The parameters of the controller were customized to match the range of motion and speed of movement of each shoulder. The proportional control range was chosen as a percentage of the total range of motion. The speed of the movement of the shoulder was used to initiate a lock command. This speed setting was individualized for each patient by measurement of the velocity during normal movements and during small, quick jerks. The patient used a switch, mounted on the chest, to turn the system on and off and to switch grasp patterns. If the switch was depressed momentarily, the neuroprosthesis turned on in the lateral grasp pattern. If it was depressed again, the neuroprosthesis turned on in the palmar grasp pattern. If the switch was depressed for more than two seconds, the system turned off.
Motion of the wrist was used as a proportional control signal by one patient who had active extension of the wrist. The degree of extension corresponded to the relative closing of the grasp, with full extension corresponding to a fully closed grasp. This type of control enhanced the natural tenodesis grasp in these patients. An external switch, typically mounted on the wheelchair, was used to turn the system on and off, switch between grasp patterns, and provide a lock command.
Both shoulder and wrist control allow the patient to perform bimanual tasks while modulating the grasp pattern.
Sensory Feedback
All five patients had absent or abnormal sensation in the hands. Four patients were provided with a single channel of sensory feedback to provide information about the state of the neuroprosthesis. Feedback regarding the level of proportional control was provided with use of five levels of frequency from four to fifty-five hertz. Each level corresponded to 20 per cent of the command range, with the lowest frequency indicating the range of 0 to 20 per cent, when the grasp was open. The sensory feedback was turned off when the system was locked or turned off. In general, the sensory feedback provided patients with information about the state of the system that was difficult to obtain visually.
Functional Training and Evaluation
The patients were trained in the use of the neuroprosthesis as an outpatient or during a three-week stay in the hospital. During this time, the grasp patterns, control parameters, and sensory feedback levels were determined, and a series of functional evaluations were performed. The patient was then discharged to use the device at home. A follow-up evaluation was performed at six months and at yearly intervals thereafter.
The function of the neuroprosthesis was evaluated, beginning with basic grasp, proceeding to the incorporation of grasp into an activity, and ending with the inclusion of this activity into daily life. The grasp force14 was measured; the ability of the patient to grasp, move, and release objects was evaluated with a grasp-release test27; and the ability of the patient to perform activities of daily living was evaluated in the laboratory after extensive training. All evaluations were performed both with and without the neuroprosthesis. The use of the neuroprosthesis in the home was assessed by an interview with the patient and on the basis of information recorded automatically in the external control unit.
Measurements of Grasp Force and Range of Motion
Grasp and pinch force were measured with a modified pinch meter (B and L Engineering, Santa Fe Springs, California) with metal bars added to extend and enlarge the grasping surfaces of the meter. In addition, the individual muscle forces generated by stimulation to each electrode were evaluated with the muscle-grading scale of the Medical Research Council1.
Grasp-Release Test
A six-object grasp-release test was used to evaluate the ability to pick up and release objects while minimizing the influence of other variables, such as strength of the upper arm or postural control27. The patient was instructed to move as many objects as possible (without any mistakes) within a thirty-second trial. Three of the objects required the use of palmar grasp and three required the use of lateral grasp. The objects were of various sizes and weights, were manipulated with one hand only, and included a small juice can (210 grams), a paperweight (264 grams), a videotape (334 grams), and a simulated fork (requiring 440 grams to depress the handle).
Tests of Activities of Daily Living
Each patient was tested with regard to the ability to perform at least four activities of daily living with the neuroprosthesis and without it. The activities included eating, drinking, writing, and brushing teeth. In addition, each patient was tested with regard to the ability to perform activities that were of particular interest to him or her, including applying makeup, brushing hair, removing money from a wallet, rolling dice, and painting. The level of independence for each phase (acquire, release, and so on) of each activity was scored as was the quality of the performance of the task. The patient was asked whether he or she preferred to perform each activity with or without the neuroprosthesis. In order to analyze the results, the patients were divided into three groups: most dependent (no extension of the wrist), mid-dependent (weak tenodesis grasp), and least dependent (strong tenodesis grasp).
Usage Survey
The patients were surveyed to monitor how frequently they used the neuroprosthesis in the home environment. In addition, the external control unit recorded daily use and stored this information in memory. This record could be retrieved at regular intervals through a computer link.
These first five patients were followed for three years and one month to nine years and one month (average, four years and eight months). All five patients gained the ability to actively open and close the hand with use of the implanted neuroprosthesis. Each patient obtained two grasping patterns, lateral and palmar prehension, and developed functional levels of strength. None of the patients had had active grasp before implantation of the neuroprosthesis or had it when the neuroprosthesis was turned off. All of the patients were able to control the degree of grasp closure and the strength of the grip over the full range of force. Each patient demonstrated the ability to use the neuroprosthesis for functional activities and used the neuroprosthesis independently in home activities. All systems continued to function properly, with only minor revisions of, or problems with, the systems or their use.
A total of twenty-two operative procedures were performed to enhance the function of the neuroprosthesis directly; these included twelve cross-tendon anastomoses, involving the flexor or extensor tendons of the digits to achieve more uniform distal motion, and six tendon transfers of paralyzed muscles (Table II). In addition, four patients had tendon transfers of voluntary muscles to provide extension of the elbow alone (two patients) or extension of the elbow and extension of the wrist (two patients) on the side of the neuroprosthesis. Three of the patients also had procedures performed on the contralateral upper extremity.
A total of forty-one epimysial electrodes and leads were implanted in the five patients and stimulated postoperatively, and all but one maintained its electrical and mechanical integrity during the follow-up period. The one mechanical failure occurred at a weld connection at the lead-electrode junction subsequent to axial rotation of the implant. There was no mechanical fatigue failure of an electrode lead or connector. In two patients, we moved an electrode operatively in order to improve muscle recruitment.
The electrode-muscle interface appeared to remain stable over time. There were no changes in response or radiographic findings to indicate that any of the electrodes implanted on muscles had migrated with time. Stimulus thresholds were measured at regular intervals for all eight electrodes in the patient (Case 5) who had had the implant in situ the longest. The results indicated that, although there was some variability in the exact threshold value, there was no long-term increase or decrease in threshold indicative of migration of an electrode. In addition to long-term stability, the epimysial electrodes generally had lower average thresholds than the percutaneous electrodes. The recruitment properties of the epimysial electrodes were as good as or better than the recruitment properties of the percutaneous electrodes14. These factors suggest the advantage of direct visualization during operative placement of the electrode on the muscle.
Electrodes were placed on at least one muscle in each of five functional groups: finger flexors, finger extensors, thumb flexors, thumb extensors, and thumb abductors. With stimulation, twenty-three of the twenty-five muscle groups (five in each patient) generated a force of 4 of 5 (full range against gravity with resistance). Without stimulation, all of these muscle groups were totally paralyzed both before and after implantation of the stimulator. The two muscle groups that did not achieve a grade-4 contraction were both digital extensors and, in both patients, paralyzed muscles were transferred to the digital extensors in an attempt to provide extension of the fingers because of denervation and scarring of the extensor digitorum communis. In the first patient (Case 5), additional tenolysis was necessary to provide weak extension of the fingers with stimulation. In the second patient (Case 2), the transferred muscle (the index flexor digitorum superficialis) was too weak to provide extension of the fingers. Despite weak or absent stimulated extension of the fingers, both patients used the neuroprosthesis functionally either by passive flexion of the wrist to achieve passive opening of the hand or by employing the contralateral hand to push objects into their grasp.
Grasp and Release Patterns
All patients were provided with lateral and palmar grasp patterns. Pinch forces ranged from eight to twenty-five newtons and, in all patients, the strength of the grasp with stimulation was substantially greater than that of the tenodesis grasp alone.
Grasp-Release Test
The neuroprosthesis provided every patient with the ability to manipulate at least three more objects than they could manipulate without the neuroprosthesis. Each patient was able to grasp the two lightest objects with and without the neuroprosthesis. The major difference in performance was exhibited in the grasping of heavier objects. All of the patients could manipulate at least five of the six objects with use of the neuroprosthesis. Each could pick up the paperweight and could depress the simulated fork. One patient was unable to pick up the juice can, and one patient was unable to pick up the videotape. In both cases, this was because of an inability to position the fingers around the object properly.
Tests of Activities of Daily Living
A total of forty-four activities of daily living were evaluated. One patient was tested with respect to four activities, and all of the other patients were tested with respect to at least eight activities.
The two most dependent patients were tested with respect to their ability to perform four (Case 1) or eleven (Case 2) activities, and physical assistance was needed in order to perform twelve of the fifteen activities tested. When the neuroprosthesis was used, physical assistance was not needed for seven of these twelve activities. The neuroprosthesis even allowed the patients to perform four of the twelve activities without any adaptive equipment. The two mid-dependent patients (Cases 3 and 4) were each tested with respect to their ability to perform eight activities. Some form of adaptive equipment was needed to perform ten of the sixteen activities tested, and physical assistance was needed for one other task. The neuroprosthesis eliminated the need for adaptive equipment for nine of the ten activities. The ability of the least dependent patient (Case 5) to perform thirteen activities was tested, and use of the neuroprosthesis eliminated or reduced the need to use adaptive equipment or self-assistance for seven. In summary, all patients demonstrated improvement in independence commensurate with the initial level of injury.
The ability to perform specific activities was also tested. Two tasks, using a fork and writing with a pen, served to demonstrate the type of activities for which the neuroprosthesis was particularly helpful. All five patients were tested with respect to their ability to use a fork for eating. Two patients needed some physical assistance and three patients needed adaptive equipment in order to perform the activity without the neuroprosthesis. Use of the neuroprosthesis allowed all five patients to use a fork for eating with only self-assistance. Similarly, all five patients were tested with respect to their ability to write with a pen. Without the neuroprosthesis, two patients needed physical assistance and two needed adaptive equipment in order to write. Use of the neuroprosthesis increased the independence level of all four patients. The remaining patient needed only self-assistance, and this need was reduced when the neuroprosthesis was used.
The patients preferred to use the neuroprosthesis to perform thirty-three of the forty-four tasks tested. They generally preferred not to use the neuroprosthesis for light tasks, however, such as eating a cookie, for which the tenodesis grasp was sufficient.
Use at Home
All five patients used the device for activities performed at home. Four of the five reported using the device on a regular basis, with usage ranging from 50 to 100 per cent of the days surveyed (days on which the device was used divided by days surveyed). Pain unrelated to the neuroprosthesis developed in the shoulder of the fifth patient and severely limited the use of that upper extremity. The patients generally used the device for eating and personal grooming. Other activities that were performed at home with use of the neuroprosthesis included writing, typing, using a remote control, applying makeup, playing bingo, painting, and putting a cassette into a cassette player (Fig. 2).
Complications
The complications associated with the implanted neuroprosthesis have been few. One device was replaced, two devices rotated in situ, one of the leads fractured, and there was one localized infection. The replacement was done two years after implantation in the first patient (Case 5) because, although it was operational, the stimulator-receiver required increased power to operate. Engineering support staff projected the most likely cause and instituted an appropriate modification, which consisted of a change in the receiving coil12. No additional problems were encountered with the operation of this or any of the other devices. At the time of writing, the longest time for which a device had been operational was seven years and one month and the shortest time was three years and one month. In two patients (Cases 1 and 3), the stimulator-receiver rotated along its long axis after implantation. In both patients, the implant was operatively exposed, unwound, and resutured in place. In one of these patients, the rotation of the implant caused a fracture of the lead at the termination to the sensory electrode. The electrode was replaced uneventfully when the implant leads were unwound. In a third patient, a localized infection developed around an extruded suture near the sensory electrode. To ensure that the infection would not spread along the lead track to the implant, the sensory lead was severed and the electrode was removed to allow the area to heal with the lead remaining in place. The electrode was not replaced because the patient decided to use auditory cues rather than sensory cues as feedback. These complications were minor and did not compromise the result in any patient.
The use of an implanted neuroprosthesis has been shown to provide grasp and release to tetraplegic individuals who have a clinically complete injury of the spinal cord with motor function remaining at the level of the fifth or sixth cervical nerve root. The grasp strength provided by the neuroprosthesis was adequate to perform many activities of daily living, and most patients regularly used the device at home. The neuroprosthesis provides active grasp with versatile gripping motions and requires little or, often, no orthotic support. An orthosis is needed only if the patient has no extension of the wrist and then it is used only to support the wrist and can be worn throughout the day. The patient can perform multiple tasks and move from one task to another without the delays that are caused by the time needed for the application of a splint. Thus, we believe that there are several inherent advantages to the function provided by the neuroprosthesis.
The neuroprosthesis provided increased independence that was consistent with the patient's level of ability, as demonstrated by the tests of activities of daily living. Patients who needed a great deal of physical assistance without the neuroprosthesis did not need it, for most of the activities tested, when they used the device. This may mean a reduction in the time for which these patients need attendants. Patients who needed adaptive equipment when they did not use the neuroprosthesis were able to perform tasks without it when they used the device. Without the need to wear a cuff or another apparatus on the hand, these patients were able to adapt more easily to new environmental situations. In addition, the neuroprosthesis enabled patients to move efficiently from one task to another, which is generally not the case with adaptive equipment.
For patients who have a strong tenodesis grasp, the added function provided by the neuroprosthesis eliminates the need for physical assistance or adaptive equipment for many tasks and reduces the need for self-assistance for most tasks. Patients are not as reliant on use of the mouth or the contralateral hand to assist in the performance of activities. This enables them to perform multiple activities, such as using the contralateral arm for balance when reaching or carrying on a conversation. Just as importantly, the neuroprosthesis provides a more cosmetically acceptable grasp, increased security when objects are grasped, and ease of performance of activities. The importance of these factors is demonstrated by the large number of activities for which the patients preferred to use the neuroprosthesis, even when the level of independence was not greatly improved.
The use of implantation markedly improved the neuroprosthesis compared with the previous-generation system, which used percutaneous electrodes. In addition to offering the utility provided by the percutaneous electrode system, the implanted system was expected to offer improved function, less complex maintenance, improved reliability, and better cosmetic acceptability. Function was improved by better placement of the electrodes and by operative alterations of the anatomy of the upper extremity performed in conjunction with the operative placement of the stimulator10,12. Maintenance was simplified by elimination of the percutaneous interface, which had to be cleaned and maintained by the patient and the attendant18. The reliability of the system was increased through improvement of the designs of the leads and electrodes24. The cosmetic acceptability was improved by the removal of connectors and lead wires from the external surface of the arm. These leads were a source of irritation to patients because they caught on the handle of the wheelchair or broke. These improvements resulted in a neuroprosthesis that is better accepted by and more functional for the user.
The percutaneous electrode system was an essential developmental tool with which the efficacy of the system could be tested with a minimally invasive procedure. However, it had a limited lifetime before the electrodes needed to be replaced because of failure18. Replacement of percutaneous electrodes necessitated that the patient return to the hospital for implantation and for modification of the stimulus levels that establish the grasp patterns. The implanted epimysial electrodes demonstrated recruitment properties that are equivalent to, if not better than, those of percutaneously implanted electrodes14. The epimysial electrodes and leads last longer than the percutaneous electrodes and leads18. For example, Memberg et al. reported that the probability that a single percutaneous electrode will survive for five years is approximately 50 per cent. Assuming that the rate of survival for each electrode is independent, the probability that eight electrodes will survive for five years is 0.4 per cent. In contrast, the first eight epimysial electrodes implanted had not failed at more than nine years. Because the implanted electrodes remain stable, patients typically do not need any modification of the grasp patterns. In fact, at the time of writing, the functional grasp parameters for one patient (Case 5) had not changed in more than six years.
Despite the limitations of the percutaneous system, it was useful as a developmental tool, as mentioned. In addition, the system supplied function to a limited number of patients for many years, while the implant technology was being developed and tested26. Finally, the system was used to demonstrate the feasibility of deployment of an upper extremity neuroprosthesis to collaborating institutions. Through this program, we demonstrated that other centers could implement and deploy the technology but improvements were needed to correct the deficiencies identified as well as to improve the ease with which the system could be programmed by the therapist. These changes have been included in the implanted system.
The hardware, including the implanted stimulator-receiver, leads, connectors, and electrodes have been safe and reliable. There have been no instances of failure of the device, systemic infection, or rejection of the device. The rotation of the device noted in two patients is similar to that reported for cardiac pacemakers2,3, and the problem appears to be solved by suture of the implant to the underlying tissue. Operative installation of this device appears to involve an extension of the skills of the hand surgeon who is familiar with spinal cord injury, and these skills have been transferred successfully to physicians at collaborating institutions13. We anticipate that completion of the research trials in a controlled study at multiple institutions will lead to clinical introduction of this system for more widespread application.
American Spinal Injury Association: Standards for Neurological Classification of Spinal Injury Patients. Atlanta, Georgia, American Spinal Injury Association, 1990.
Bayliss, C. E.; Beanlands, D. S.; and Baird, R. J.: The pacemaker-twiddler's syndrome: a new complication of implantable transvenous pacemakers. Canadian Med. Assn. J.,99: 371-373, 1968.99371
1968
Ellis, G. L.: Pacemaker twiddler's syndrome: a case report. Am. J. Emerg. Med.,8: 48-50, 1990.848
1990
[PubMed]
Freehafer, A. A., and Mast, W. A.: Transfer of the brachioradialis to improve wrist extension in high spinal-cord injury. J. Bone and Joint Surg.,49-A: 648-652, June 1967.49-A648
1967
Freehafer, A. A.; Kelly, C. M.; and Peckham, P. H.: Planning tendon transfers in tetraplegia: "Cleveland technique." In Tendon Surgery in the Hand, pp. 506-515. Edited by J. M. Hunter, L. H. Schneider, and E. J. Mackin. St. Louis, C. V. Mosby, 1987.
Freehafer, A. A.; Peckham, P. H.; and Keith, M. W.: New concepts on treatment of the upper limb in the tetraplegic. Surgical restoration and functional neuromuscular stimulation. Hand Clin.,4: 563-574, 1988.4563
1988
[PubMed]
Handa, Y.; Ohkubo, K.; and Hoshimiya, N.: A portable multi-channel FES system for restoration of motor function of the paralyzed extremities. Automedica,11: 221-231, 1989.11221
1989
House, J. H., and Shannon, M. A.: Restoration of strong grasp and lateral pinch in tetraplegia: a comparison of two methods of thumb control in each patient.. J. Hand Surg.,10A: 22-29, 1985.10A22
1985
Johnson, M. W., and Peckham, P. H.: Evaluation of shoulder movement as a command control source. IEEE Trans. Biomed. Eng.,37: 876-885, 1990.37876
1990
[PubMed]
Keith, M. W.; Kilgore, K. L.; Peckham, P. H.; Wuolle, K. S.; Creasey, G.; and Lemay, M.: Tendon transfers and functional electrical stimulation for restoration of hand function in spinal cord injury. J. Hand Surg.,21A: 89-99, 1996.21A89
1996
Keith, M. W.; Peckham, P. H.; Thrope, G. B.; Buckett, J. R.; Stroh, K. C.; and Menger, V.: Functional neuromuscular stimulation neuroprostheses for the tetraplegic hand. Clin. Orthop.,233: 25-33, 1988.23325
1988
[PubMed]
Keith, M. W.; Peckham, P. H.; Thrope, G. B.; Stroh, K. C.; Smith, B.; Buckett, J. R.; Kilgore, K. L.; and Jatich, J. W.: Implantable functional neuromuscular stimulation in the tetraplegic hand. J. Hand Surg.,14A: 524-530, 1989.14A524
1989
Kilgore, K. L.; Peckham, P. H.; and Keith, M. W.: An implanted upper extremity neuroprosthesis: a twenty patient follow-up. J. Spinal Cord Med.,18: 147, 1995.18147
1995
Kilgore, K. L.; Peckham, P. H.; Keith, M. W.; and Thrope, G. B.: Electrode characterization for functional application to upper extremity FNS. IEEE Trans. Biomed. Eng.,37: 12-21, 1990.3712
1990
[PubMed]
Kilgore, K. L.; Peckham, P. H.; Thrope, G. B.; Keith, M. W.; and Gallaher-Stone, K. A.: Synthesis of hand grasp using functional neuromuscular stimulation. IEEE Trans. Biomed. Eng.,36: 761-770, 1989.36761
1989
[PubMed]
Letechipia, J. E.; Peckham, P. H.; Gazdik, M.; and Smith, B.: In-line lead connector for use with implanted neuroprosthesis. IEEE Trans. Biomed. Eng.,38: 707-709, 1991.38707
1991
[PubMed]
McDowell, C. L.; Moberg, E. A.; and House, J. H.: The Second International Conference on Surgical Rehabilitation of the Upper Limb in Tetraplegia (Quadriplegia). J. Hand Surg.,11A: 604-608, 1986.11A604
1986
Memberg, W. D.; Peckham, P. H.; Thrope, G. B.; Keith, M. W.; and Kicher, T. P.: An analysis of the reliability of percutaneous intramuscular electrodes in upper extremity FNS applications. IEEE Trans. Rehab. Eng.,1: 126-132, 1993.1126
1993
Moberg, E.: Surgical treatment for absent single-hand grip and elbow extension in quadriplegia. Principles and preliminary experience. J. Bone and Joint Surg.,57-A: 196-206, March 1975.57-A196
1975
Peckham, P. H.; Mortimer, J. T.: Restoration of hand function in the quadriplegic through electrical stimulation. In Functional Electrical Stimulation: Applications in Neural Prosthesis, pp. 83-95. Edited by J. B. Reswick, and F. T. Hambrecht. New York, Marcel Dekker, 1977.
Peckham, P. H.; Keith, M. W.; and Freehafer, A. A.: Restoration of functional control by electrical stimulation in the upper extremity of the quadriplegic patient. J. Bone and Joint Surg.,70-A: 144-148, Jan. 1988.70-A144
1988
Peckham, P. H.; Mortimer, J. T.; and Marsolais, E. B.: Controlled prehension and release in the C5 quadriplegic elicited by functional electrical stimulation of the paralyzed forearm musculature. Ann. Biomed. Eng.,8: 369-388, 1980.8369
1980
[PubMed]
Peckham, P. H.; Thrope, G. B.; Buckett, J. R.; Freehafer, A. A.; and Keith, M. W.: Coordinated two mode grasp in the quadriplegic initiated by functional neuromuscular stimulation. In IFAC Control Aspects of Prosthetics and Orthotics, pp. 29-32. Edited by R. M. Cambell. Oxford, Pergamon Press, 1983.
Smith, B.; Peckham, P. H.; Keith, M. W.; and Roscoe, D. D.: An externally powered, multichannel, implantable stimulator for versatile control of paralyzed muscle. IEEE Trans. Biomed. Eng.,34: 499-508, 1987.34499
1987
[PubMed]
Waters, R.; Moore, K. R.; Graboff, S. R.; and Paris, K.: Brachioradialis to flexor pollicis longus tendon transfer for active lateral pinch in the tetraplegic. J. Hand Surg.,10A: 385-391, 1985.10A385
1985
Wijman, C. A.; Stroh, K. C.; Van Doren, C. L.; Thrope, G. B.; Peckham, P. H.; and Keith, M. W.: Functional evaluation of quadriplegic patients using a hand neuroprosthesis. Arch. Phys. Med. and Rehab.,71: 1053-1057, 1990.711053
1990
Wuolle, K. S.; Van Doren, C. L.; Thrope, G. B.; Keith, M. W.; and Peckham, P. H.: Development of a quantitative hand grasp and release test for patients with tetraplegia using a hand neuroprosthesis. J. Hand Surg.,19A: 209-218, 1994.19A209
1994
Zancolli, E.: Structural and Dynamic Bases of Hand Surgery. Ed. 2, p. 177. Philadelphia, J. B. Lippincott, 1979.