365
45
Pacing for Unilateral Diaphragm Paralysis
Raymond P. Onders
surgeries), and trauma. Although sniff tests and phrenic nerve conduction studies with fl uoro- scopic observation can identify intact phrenic nerves, the negative results of these tests do not necessarily determine the extent or location of damage. Localized damage to the myelin sheath or acute edema will result in negative results and loss of central control of the diaphragm. Re- myelinization or resolution of the edema may occur over a number of days to years.1
Diaphragm pacing for ventilatory support has been in use for over 30 years since fi rst reported by Glenn.2 There are several diaphragm pacing systems available including the conventional ones in which phrenic nerve cuff electrodes are placed with staged bilateral thoracotomies. The cervical electrode placement while utilized in the past is discouraged for the following reasons:
there is an accessory branch from the lower segment of the cervical spinal cord that joins the phrenic nerve trunk in the thorax so that neck stimulation may result in incomplete diaphragm activation; brachial plexus nerves are in close proximity and may be activated resulting in pain or undesirable movement; and neck movements can increase mechanical stress on the nerve/elec- trode system which may increase the risk of nerve injury. There have been recent reports of placing the system thorascopically.3,4 Diaphragm pacing with direct phrenic nerve electrodes is underuti- lized because of the scope of the operation, risk of phrenic nerve injury, and theoretical concerns about using it 24h/day. There is a more recent option that involves laparoscopic implantation of intramuscular electrodes at the motor point of Symptoms of unilateral diaphragmatic paralysis
can range from sleep-related symptoms to exert ional dyspnea or orthopnea. At times unilateral diaphragm paralysis is found on routine chest radiograph alone when an elevated hemi- diaphragm is seen. Ventilatory failure will usually only result if there is bilateral diaphragmatic involvement. When diaphragmatic paralysis is suspected, confi rmatory testing is done by inspira- tory fl uoroscopy (sniff test) and electromyography of the phrenic nerve. To determine if the conduc- tion path of the phrenic nerve is intact from the cervical region to the diaphragm, the key test is fl uoroscopic visualization of the diaphragm with transcutaneous stimulation of the phrenic nerve in the neck. If the diaphragm moves during stimu- lation then the phrenic nerve is intact, but there is a disruption of the signal pathway from the respi- ratory center in the brain to the phrenic nerve causing the diaphragm not to function. With the use of fl uoroscopic visualization during stimula- tion, false-positive phrenic nerve conduction studies are virtually eliminated. However, because of diffi culties in locating the phrenic nerve in the cervical region there is a signifi cant potential for false-negative studies, especially in inexperienced hands. The most common causes of an intact phrenic nerve with diaphragm paralysis are high cervical spinal cord injury or central hypoventila- tion syndrome (CHS or Ondine’s Curse). In almost all of these cases the diaphragm paralysis is bilat- eral. Unilateral paralysis of the diaphragm usually involves a nonfunctioning phrenic nerve with the causes in decreasing order of frequency: idio- pathic, postsurgical (cardiac, neck, and thoracic
the diaphragm.5 This has been implanted in 18 spinal cord patients with excellent results. This is an outpatient operation with no risk of phrenic nerve injury and allows 24-h use with the longest patient continuing to pace full-time for over 5 years. In brief, this procedure involves laparo- scopic mapping of the diaphragm to identify the motor point which is the area where a electrical stimulus can cause maximal contraction of the diaphragm.6 Two electrodes are then placed on each hemi-diaphragm with a specially designed laparoscopic implant instrument (Synapse Bio- medical, Oberlin, OH) and tunneled externally to the power source.7
Both the phrenic nerve cuff electrode system and the laparoscopic motor point diaphragm pacing stimulation system require an intact phrenic nerve. The conduction pathway is the phrenic nerve and if that is not intact none of the systems can deliver a stimulus to the target diaphragm muscle. Almost all causes of unilateral diaphragm paralysis in non-spinal-cord-injured patients involve a phrenic nerve that is nonfunctional, at least to some extent, below the cervical region. The medical literature describes nerve transfers to the phrenic nerve and use of a diaphragm pacing system8 as an option in patients with an injured phrenic nerve. This procedure essentially involves the coaption of a proximal foreign nerve to the distal denervated nerve to reinnervate the latter by the donated axons. Cortical plasticity appears to play an important physiological role in the func- tional recovery of the reinnervated muscles. An independent electrical pacing system is necessary because the nerve that is transferred has no con- nection to the central respiratory system so it must be stimulated to cause independent diaphragm contraction to augment respiration.
This chapter will review the extant evidence to assess whether diaphragm pacing is an option for patients with unilateral paralysis of the dia- phragm when there is an intact phrenic nerve and when there is no intact nerve.
45.1. Available Evidence
The initial review will assess the evidence of dia- phragm pacing when there is diaphragmatic dys- function but an intact phrenic nerve. Over the
past 30 years, electrical activation of the phrenic nerves has been used to provide artifi cial ventila- tion in patients with chronic respiratory insuffi - ciency. Despite their clinical effectiveness, their use has been limited to a carefully selected group of patients with bilateral diaphragmatic dysfunc- tion and intact phrenic nerves. The benefi ts of diaphragm pacing have been well described in large series and include: decreased barotrauma with the use of natural negative pressure ventila- tion with their own diaphragm; increased mobil- ity without need for ventilator; improved speech;
improved olfactory sensation; and decreased risk for pulmonary infection.9–14 In some of the early series, diaphragm pacing was considered suc- cessful for ventilatory support in only 50% of patients.11,15,16 These early studies are not refl ec- tive of the modern-day experience with dia- phragm pacing, as the technology and patient selection methods were not well defi ned. There have been few reports of modern-day success rates though several papers describe the use of diaphragm pacing for over 15 years.17,18
Three commercial systems are in current use for trans-thoracic direct phrenic nerve stimula- tion: Avery Biomedical Devices (Commack, NY), Atrotech OY (Tampere, Finland), and Medim- plant Biotechnisches Labor (Vienna, Austria).
These systems differ primarily in the electrode design and stimulus parameters. Phrenic pacers have been implanted in over 1500 patients world- wide. Drawbacks to these systems include the risk of injury to the phrenic nerve either by surgi- cal manipulation or by the electrode itself, system component failure, and the high cost of the systems. Although the risk of injury to the nerve has decreased, it does exist because a section of the nerve must be mobilized for electrode place- ment. The incidence of component failure has declined as the systems have undergone revi- sions. However, all three require some extracor- poreal component. Unlike the cardiac pacemaker, traditional phrenic pacers require an external transmitter and antenna to transmit both the power and control signal to an implanted receiver/
stimulator. Also, at present, none of the systems has any feedback or timing mechanism to make them physiologically responsive, nor are they synchronized with the upper airway. Develop- ment of such a mechanism would be an added
benefi t to phrenic pacers over conventional mechanical ventilators. Cost is perhaps a larger hurdle to overcome. The phrenic nerve electrode pacing systems available today cost nearly
$100,000 (for the system, implant, and rehabilita- tion). Unlike cardiac pacemakers, because of the low number of potential candidates for these systems and the relatively low profi t potential, there is little interest from major manufacturers of medical devices. This may explain the limited effort to develop improved pacing systems.
An alternative to the trans-thoracic phrenic nerve stimulation is the laparoscopic diaphragm pacing stimulation (DPS) system. There have been 22 human subjects implanted with the DPS system [18 spinal cord injury patients and 4 amy- otrophic lateral sclerosis (ALS) patients]. The results of the DPS system for spinal cord patients indicated that the DPS system produced a signifi - cant mean percentage increase in tidal volume relative to the basal required tidal volume.19 The procedure has overcome the learning curve for the operation, with the implantation standard- ized in an outpatient surgical procedure.20 Overall there has been a 94% success rate for the spinal cord injured patients with the only failure being the second patient who had a false-positive inclu- sion criterion. The laparoscopic motor point electrode DPS system is an easy application for diaphragmatic stimulation when the phrenic nerve is intact. It overcomes many of the short- comings of the available phrenic nerve electrode systems. The development of a totally implant- able system is feasible and under way.21 It would be a signifi cant advancement over presently available systems.
For patients with nonfunctioning phrenic nerves, electrical activation of the intercostal muscles is one approach to treat respiratory insuffi ciency. Unlike the diaphragm, these muscles are innervated by a group of nerves (intercostal nerves originating from the ventral rami of T2–T12). However, by placing a single electrode in the epidural surface of the spinal cord through a dorsal laminectomy, this group of nerves/muscles can be activated which can provide up to 40% of vital capacity through the parasternal and external intercostals that are primarily inspiratory. Electrical activation of the intercostal muscles alone has been used in
patients, however, the maximum duration of intercostal pacing (without mechanical ventila- tion or spontaneous breathing activity) remained relatively short (<3h) and is not a viable option on its own.22 Based on this, individuals with only one intact phrenic nerve had a combined inter- costal system with a conventional diaphragm pacing system placed unilaterally. This system was successful in maintaining long-term ventila- tory support in the four patients but presently is not in any further trials.23
In those patients with a nonfunctioning phrenic nerve, diaphragm pacing is not an option unless a nerve is transferred to the phrenic nerve to re-animate the diaphragm. With advances in microsurgical techniques for neural anastomosis and a better understanding of axonal degenera- tion and regeneration, the repair or transfer of a nerve to the phrenic nerve and subsequent rein- nvervation of the diaphragm is a possibility. With a viable nerve, diaphragmatic pacing is then an option. Krieger successfully described transfer- ring a brachial nerve to the phrenic nerve in cats in 1983.8 After a recovery period to allow for growth of axons down the anastomosed phrenic nerve (16–32 weeks), they were able to stimulate the nerve and have adequate diaphragm contrac- tions. Following this initial study, Krieger and colleagues investigated using an intercostals nerve in place of the brachial nerve for the anas- tomosis. The intercostals nerve was a good donor because of its proximity to the phrenic nerve (reducing the time for axonal regeneration), its physiological function (activation of skeletal muscle for respiration), and its size (comparable to the phrenic nerve). The initial article describes a single case and a letter to the editor describes two additional cases.24,25 Subsequently a series of six patients was then described in 2000.26 All of the patients had spinal cord injury with the time from injury to nerve transfer ranging from 6 months to 3 years. In this series of six patients there were a total of 10 nerve transfers. Two patients only had single nerve transfer because the other nerve on direct exploration was found to be intact. Only four patients were available for study. The fi fth patient is on a progressive pacing schedule and the sixth patient was only 1 month postoperative and with accepted growth of regen- erating axons of 1mm per day the distance from
the anastomosis to the diaphragm of 50mm could not have been covered. In this series the average time for diaphragmatic response was 7 months with the shortest 6 months and the longest 13 months, so the true growth rate can be as slow at 1mm every 8 days. Two of the patients are classi- fi ed as capable of pacing but presently are not being paced because of depression in one and one died of unrelated causes. In neither of these cases were the tidal volumes or diaphragmatic move- ment with stimulation given. Two patients (a total of three nerve transfers) are using the system 24h/day, but again no data are given concerning the tidal volumes or diaphragmatic excursion with stimulation. Overall, of the eight nerve transfers that could be studied, all eight showed diaphragm motion with stimulation, which is impressive given the authors’ own description of the operation as diffi cult because of the angles and the fact the anastomosis occurs on the beating heart. There is some concern of the long-term viability of this technique in these patients, though. There was a letter to the editor by a sepa- rate physician stating that one of the patients that was reported as a success is in actuality not using the system at all.27 To date there has been no other reports of this technique in the literature although it is mentioned often in the literature as a pos- sibility both for spinal cord–injured patients and also patients with isolated phrenic nerve injuries.
45.2. Summary of Evidence and Current Recommendations
Overall, none of the available data concerning diaphragm pacing specifi cally identify its use with unilateral diaphragm paralysis. The reason for this is that unilateral paralysis usually involves an injured phrenic nerve and therefore the diaphragm cannot be paced unless the dia- phragm is re-innervated with a nerve transfer. So let us fi rst look at the evidence for diaphragm function with intercostal transfer and diaphragm pacing. The level of evidence for diaphragm pacing using an intercostal nerve transfer is level 4 because it is a case series that only measured end results with tidal volumes with stimulation and measurements of outcomes in less than 80%
of the patients. Without more centers reporting their results or this series re-analyzing their results with a long-term follow-up, the recom- mendation grade is C. With this scarcity of evi- dence patients should not be given the hope of diaphragm pacing for a unilateral paralysis of the diaphragm unless they have an intact nerve.
Patients should not be given the hope of dia- phragm pacing for a unilateral paralysis of the diaphragm unless they have an intact nerve (level of evidence level 4; recommendation grade C).
The results of diaphragm pacing when the phrenic nerve is intact are excellent. The evidence for the ability to pace the diaphragm and provide tidal volumes is level 1 because of the long history of success of pacing in multiple centers and the all or none ability to assess the results.
The patient’s diaphragm either provides a tidal volume for ventilation with stimulation or is non- functional and the patient requires a mechanical ventilator when the device is turned off. The major change in diaphragm pacing is that it can now be done more safely and as an outpatient through the laparoscopic motor point stimula- tion technique with a higher success rate. The recommendation grade is A for bilateral dia- phragm pacing when both phrenic nerves are intact.
When both phrenic nerves are intact, results of bilateral diaphragm pacing are excellent (level of evidence 1; recommendation grade A).
For unilateral diaphragm paralysis pacing is not benefi cial because the phrenic nerve is usually not functional (level of evidence 5;
recommendation grade D).
Unfortunately, for unilateral diaphragm paral- ysis there is no evidence that pacing is done because the phrenic nerve is usually not func- tional. If the nerve is intact but the diaphragm is nonfunctional then the level of evidence for
pacing is 5 and the recommendation grade is D because it is only based on the physiology of the system and has not been reported in the litera- ture. Presently, the discussions of an earlier chapter in this text concerning diaphragm plica- tion may offer the most hope for patients with unilateral dennervated diaphragms.
45.3. Future Research
Future research should involve ways to help a damaged phrenic nerve recover in unilateral paralysis. When diaphragmatic dysfunction is identifi ed after a thoracic or cardiac procedure, instead of waiting to see if recovery occurs we should be proactive in trying to help that recov- ery process. Functional electrical stimulation has been shown to help recovery of injured nerves and, with the intramuscular laparoscopic dia- phragm pacing technique now in clinical use, we may have a way to stimulate the diaphragm so that some afferent affects along the nerve will promote recovery. There is now some prelimi- nary data in a disease where the phrenic nerve is dying at a set rate – amyotrophic lateral sclerosis (Lou Gehrig’s disease; unpublished results). By beginning a process of conditioning the dia- phragm with the DPS system, we have been able to maintain diaphragmatic function in our early patients. This is partly due to the afferent effects of electrical stimulation but also preserving and strengthening the motor units that are left. The continuous decline in forced vital capacity of these initial patients has decreased which will increase their expected lifespan. This technique of using DPS can be expanded into acutely injured phrenic nerves in the hopes of reversing or improving the affects of acute phrenic nerve inju- ries. This technique would not require any nerve transfers and if the nerve recovers it can be easily removed. A prospective trial of using DPS is nec- essary to show if this would help.
There is also a signifi cant number of patients who were told they have a negative phrenic nerve conduction test (a nonfunctioning nerve) when, on repeat evaluation in our laboratory, we were able to show diaphragmatic movement with a nerve conduction study. Phrenic nerve studies are diffi cult to reproduce, especially in patients
that are overweight or have thick necks. We were able to subsequently implant these patients with the laparoscopic motor point electrode system.
With a simple laparoscopic mapping stimulation tool, before giving up on diaphragmatic function or prior to plication, the diaphragm should be surgically studied. If at the time of plication the diaphragm responds to intraoperative stimula- tion, a motor point electrode with the DPS system should be placed and diaphragm function main- tained. This may be a better long-term option than plication. This hopefully will be an option in our armentarium for unilateral diaphragm function in the future.
References
1. Oo T, Watt JW, Soni BM, Sett PK. Delayed dia- phragm recovery in 12 patients after high cervical spinal cord injury. A retrospective review of the diaphragm status of 107 patients ventilated after acute spinal cord injury. Spinal Cord 1999;37:117–
122.
2. Glenn WW, Holcomb WG, Hogan J, et al. Dia- phragm pacing by radiofrequency transmission in the treatment of chronic ventilatory insuffi ciency.
Present status. J Thorac Cardiovasc Surg 1973;66:
505–520.
3. Morgan JA, Ginsburg ME, Sonett JR, et al.
Advanced thoracoscopic procedures are facili- tated by computer-aided robotic technology. Eur J Cardiothorac Surg 2003;23:883–887; discussion 887.
4. Shaul DB, Danielson PD, McComb JG, Keens TG.
Thoracoscopic placement of phrenic nerve elec- trodes for diaphragmatic pacing in children. J Pediatr Surg 2002;37:974–978; discussion 978.
5. DiMarco AF, Onders RP, Kowalski KE, Miller ME, Ferek S, Mortimer JT. Phrenic nerve pacing in a tetraplegic patient via intramuscular diaphragm electrodes. Am J Respir Crit Care Med 2002;166:
1604–1606.
6. Onders RP, Aiyar H, Mortimer JT. Characteriza- tion of the human diaphragm muscle with respect to the phrenic nerve motor points for diaphrag- matic pacing. Am Surg 2004;70:241–247; discus- sion 247.
7. Aiyar H, Stellato TA, Onders RP, Mortimer JT.
Laparoscopic implant instrument for the place- ment of intramuscular electrodes in the dia- phragm. IEEE Trans Rehabil Eng 1999;7:360–371.
8. Krieger AJ, Danetz I, Wu SZ, Spatola M, Sapru HN.
Electrophrenic respiration following anastomosis
of phrenic with branchial nerve in the cat. J Neu- rosurg 1983;59:262–267.
9. Dobelle WH, D’Angelo MS, Goetz BF, et al. 200 cases with a new breathing pacemaker dispel myths about diaphragm pacing. ASAIO J 1994;40:
M244–M252.
10. Elefteriades JA, Quin JA, Hogan JF, et al. Long- term follow-up of pacing of the conditioned dia- phragm in quadriplegia. Pacing Clin Electrophysiol 2002;25:897–906.
11. Tibballs J. Diaphragmatic pacing: an alternative to long-term mechanical ventilation. Anaesth Inten- sive Care 1991;19:597–601.
12. Creasey G, Elefteriades J, DiMarco A, et al. Electri- cal stimulation to restore respiration. J Rehabil Res Dev 1996;33:123–132.
13. DiMarco A. Diaphragm pacing in patients with spinal cord injury. Topics Spinal Cord Rehabil 1999;5:6–20.
14. Glenn WW, Phelps ML, Elefteriades JA, Dentz B, Hogan JF. Twenty years of experience in phrenic nerve stimulation to pace the diaphragm. Pacing Clin Electrophysiol 1986;9:780–784.
15. Carter RE, Donovan WH, Halstead L, Wilkerson MA. Comparative study of electrophrenic nerve stimulation and mechanical ventilatory support in traumatic spinal cord injury. Paraplegia 1987;25:86–91.
16. Weese-Mayer DE, Silvestri JM, Kenny AS, et al.
Diaphragm pacing with a quadripolar phrenic nerve electrode: an international study. Pacing Clin Electrophysiol 1996;19:1311–1319.
17. Elefteriades JA, Quin JA. Diaphragm pacing. Chest Surg Clin North Am 1998;8:331–357.
18. Elefteriades JA, Hogan JF, Handler A, Loke JS.
Long-term follow-up of bilateral pacing of the dia- phragm in quadriplegia. N Engl J Med 1992;326:
1433–1434.
19. DiMarco AF, Onders RP, Ignagni A, Kowalski KE, Mortimer JT. Phrenic nerve pacing via intramus- cular diaphragm electrodes in tetraplegic sub- jects. Chest 2005;127:671–678.
20. Onders RP, Dimarco AF, Ignagni AR, Mortimer JT. The Learning curve for investigational surgery:
lessons learned from laparoscopic diaphragm pacing for chronic ventilator dependence. Surg Endosc 2005;19(5):633–637.
21. Cosendai G, de Balthasar C, Ignagni AR, et al. A preliminary feasibility study of different implant- able pulse generators technologies for diaphragm pacing system. Neuromodulation 2005;8:203–211.
22. DiMarco AF, Supinski GS, Petro JA, Takaoka Y.
Evaluation of intercostal pacing to provide artifi - cial ventilation in quadriplegics. Am J Respir Crit Care Med 1994;150:934–940.
23. DiMarco AF, Takaoka Y, Kowalski KE. Combined intercostal and diaphragm pacing to provide arti- fi cial ventilation in patients with tetraplegia. Arch Phys Med Rehabil 2005;86:1200–1207.
24. Krieger AJ, Gropper MR, Adler RJ. Electrophrenic respiration after intercostal to phrenic nerve anas- tomosis in a patient with anterior spinal artery syndrome: technical case report. Neurosurgery 1994;35:760–763; discussion 763–764.
25. Krieger AJ. Electrophrenic respiration after inter- costal to phrenic nerve anastomosis in a patient with anterior spinal artery syndrome: technical case report [letter]. Neurosurgery 1995;37:553.
26. Krieger LM, Krieger AJ. The intercostal to phrenic nerve transfer: an effective means of reanimating the diaphragm in patients with high cervical spine injury. Plast Reconstr Surg 2000;105:1255–1261.
27. Fodstad H. Electrophrenic respiration after inter- costal to phrenic nerve anastomosis on a patient with anterior spinal artery syndrome: technical case report. Neurosurgery 1996;38:420.
