© 2002 American Thoracic Society
Phrenic Nerve Pacing in a Tetraplegic Patient via Intramuscular Diaphragm ElectrodesDepartments of Physiology and Biophysics and Biomedical Engineering, Case Western Reserve University and MetroHealth Medical Center; and Department of Surgery, University Hospitals of Cleveland, Cleveland, Ohio Correspondence and requests for reprints should be addressed to Anthony F. DiMarco, M.D., MetroHealth Medical Center, Rammelkamp Center for Education & Research, 2500 MetroHealth Drive, Cleveland, OH 44109-1998. E-mail: afd3{at}po.cwru.edu ABSTRACT In patients with ventilator-dependent tetraplegia, phrenic nerve pacing (PNP) provides significant clinical advantages compared with mechanical ventilation. This technique however generally requires a thoracotomy with its associated risks and in-patient hospital stay and carries some risk of phrenic nerve injury. We have developed a method by which the phrenic nerves can be activated via intramuscular diaphragm electrodes. In one patient with ventilator-dependent tetraplegia, two intramuscular diaphragm electrodes were implanted into each hemidiaphragm near the phrenic nerve motor points via laparoscopic surgery. The motor points were identified employing a previously devised mapping technique. Because inspired volumes were suboptimal on the right, a second laparoscopic procedure was necessary to position electrodes near the anterior and posterior branches of the right phrenic nerve. During bilateral stimulation, inspired volume was 580 ml. After a reconditioning program of progressively increasing diaphragm pacing, maximum inspired volumes on the left and right hemidiaphragms increased significantly. Maximum combined bilateral stimulation was 1120 ml. Importantly, the patient has been able to comfortably tolerate full-time pacing. If confirmed in additional patients, PNP with intramuscular diaphragm electrodes via laparoscopic surgery may provide a less invasive and less costly alternative to conventional PNP.
Key Words: spinal cord injury • diaphragm pacing • laparoscopy Phrenic nerve pacing (PNP) is a clinically useful alternative to mechanical ventilation in patients with ventilator-dependent tetraplegia (1–6). Unfortunately, placement of phrenic nerve electrodes generally requires a thoracotomy and associated hospital stay, with its attendant risks, inconvenience, and high cost. Moreover, this technique involves phrenic nerve dissection and electrode placement that carries some risk of injury to the phrenic nerves. Consequently, PNP represents a formidable undertaking for patients with spinal cord injury. To obviate some of the disadvantages of PNP, we have developed a method of phrenic nerve stimulation via intramuscular diaphragm electrodes, which can be implanted laparascopically. In this report, we describe successful long-term ventilatory support by this technique, in a previously ventilator-dependent tetraplegic patient. METHODS This investigation was approved by the Investigational Review Boards at University Hospitals of Cleveland, Ohio, MetroHealth Medical Center, the Veterans Administration Hospital, and also the Food and Drug Administration. Informed consent was obtained before enrollment in the study. T.C., a 35-year-old male, suffered a cervical spinal cord injury (C2 level) after a diving accident that resulted in tetraplegia and dependence on mechanical ventilatory support. Phrenic nerve conduction studies indicated normal bilateral phrenic nerve function (7). Standard laparoscopic techniques were employed to place intramuscular diaphragm electrodes within the muscular tissue of the diaphragm. After placement of four trocars into the abdominal wall and development of a pneumoperitoneum (Figure 1) , a previously devised mapping procedure (8, 9) (see online data supplement) was performed to determine the phrenic nerve motor points, i.e., the area of the muscle contained within the space defined by the entrance points of the phrenic nerves into the diaphragm (Figure 2) . Initially, several test sites were evaluated in the general region of the motor point with a suction electrode, which could be reversibly applied to the diaphragm. At each test site, a recruitment curve was constructed by determining the magnitude of change in intra-abdominal pressure (Validyne Eng. Corp., Northridge, CA) after a broad range of applied stimulus currents between 0 and 24 mA. During stimulation the patient was off mechanical ventilation. After testing at a single site, the minimum current value for full recruitment (i.e., plateau in pressure development) was used as input to the current–distance model. Mathematic analyses of data from multiple stimulation sites allowed prediction of the approximate location of the motor points within each hemidiaphragm. Subsequently, two stainless steel intramuscular diaphragm electrodes (Peterson Electrodes, Axon Eng. Inc., Cleveland, OH) were inserted via the laparoscope into the region of the motor points of each hemidiaphragm (Figure 1). The electrodes were inserted using a specially designed delivery device (Case Western Reserve University, Cleveland, OH) that allowed insertion of the electrodes in the same plane as the diaphragm (9, 10). On the right hemidiaphragm, the mapping procedure and insertion of electrodes was much more difficult due to the presence of the liver. During diaphragm stimulation, there was no evidence of coincident cardiac stimulation.
Wires from the electrodes were then brought out through the epigastric port with several centimeters of slack such that movement would not dislodge the electrodes. The electrodes were then tunneled subcutaneously to the mid chest region and attached to a connecting circuit. Indifferent electrodes were implanted into this region of the chest wall and also attached to the connecting circuit. From this region, each wire was tunneled individually, subcutaneously to the right subclavicular region where they exited the chest wall. The patient was observed overnight in the hospital and discharged the next morning. No stimulation was applied for a 10- to 14-day period to allow for regression of edema and inflammation at the electrode site. A four-channel electrical stimulator (Axon Eng. Inc.) was used to apply stimulation over a wide range of stimulus parameters (0–24 mA; 0–50 Hz; 0.1-millisecond pulse width). Tidal volume was measured by electrical integration (Model PI-830; CWE, Inc., Ardmore, PA) of the flow signal from a pneumotachograph (Model 3700; Hans Rudolph, Inc., Kansas City, MO) and recorded on an eight-channel recorder (Model DASH8; Astro-Med Inc., Warwick, RI) and also a Wright's Respirometer (Model Mark 14; Ferraris Medical Ltd., Enfield, UK). Maximum inspired volume resulting from stimulation of the left hemidiaphragm (24 mA, 50 Hz) was 380 ml, whereas the right side produced markedly lower inspired volumes of 100 ml. It was surmised therefore that that the electrodes in the right hemidiaphragm were not positioned properly in the motor point region. After approval from both Institutional Review Board and the Food and Drug Administration, the patient underwent a second laparoscopic procedure to reposition electrodes in the right hemidiaphragm. Because no single electrode produced complete diaphragm contraction, it was believed that the motor point region was localized in the region of the central tendon. Therefore, anterior and posterior branches of the phrenic nerve were identified, and one electrode was implanted into each of these regions. Intraoperative stimulation of both electrodes indicated marked contraction of the right hemidiaphragm. The patient was discharged the same day. No stimulation was applied for an additional 2-week period, after which the patient participated in a reconditioning program. RESULTS The initial effects of applied electrical stimulation (after the second laparoscopic procedure) are shown for each electrode alone and in combination in Figure 3 . Because stimulation of both electrodes within each hemidiaphragm provided substantially greater inspired volumes compared with a single electrode, chronic pacing was performed using stimulation of all four electrodes. With maximum current (24 mA), stimulation of one electrode on the right side resulted in mild shoulder discomfort and therefore the level of current was reduced to 12 mA, at which the patient had no discomfort. Bilateral diaphragm pacing was initiated with a stimulus frequency of 20 Hz, which resulted in an inspired volume of 420 ml. Initially, the patient was able to comfortably tolerate 45 minutes periods off mechanical ventilation; this was applied several times per day, 5 to 6 days/week. Oxygen saturation was monitored continuously during each study session using a pulse oximeter (Model N-1000, Nellcor Inc., Hayward, CA). Over the course of the subsequent 20 weeks, pacing time was gradually increased, until he was able to comfortably tolerate full time pacing. Over the course of the reconditioning period, stimulus frequency was gradually reduced to 15 Hz.
Over the course of the reconditioning period, inspired volume production resulting from stimulation of each individual electrode increased substantially, as did that resulting from their stimulation in combination (Figure 3). Maximum stimulation of all four electrodes in combination (24 mA, 50 Hz) resulted in an inspired volume of 1,120 ml. By Week 20 after the second surgical procedure, the patient was able to maintain full-time ventilatory support with diaphragm pacing alone. Moreover, the patient has maintained continuous full-time diaphragm pacing for over 15 months. He reports that his breathing is effortless and there is a marked improvement, in terms of comfort, compared with mechanical ventilation. Repeat analyses of inspired volume production indicate that his inspired volume production has remained stable over the entire period of stimulation. After 40 and 75 weeks of stimulation, his maximum inspired volumes were 1,100 and 1,250 ml, respectively. DISCUSSION The introduction of PNP more than two decades ago by Glenn and associates has provided many ventilator-dependent tetraplegic patients with freedom from mechanical ventilation (4–6). Artificial ventilation by this technique eliminates many of the disadvantages of mechanical ventilatory support and has afforded these patients the ability to breathe more normally (1). The results of this report suggest that diaphragm activation can also be accomplished via a much less invasive, less expensive, and more convenient technique, i.e., laparoscopic surgery. Unlike conventional PNP, which generally requires a thoracotomy, laparoscopy can be performed on an outpatient basis, providing a substantial cost saving. Moreover, placement of electrodes within the diaphragm obviates the need for manipulation of the phrenic nerve and potential risk of nerve injury, an important concern for patients who have already sustained a devastating injury. It should be noted that the success of intramuscular diaphragm pacing depends on activation of the phrenic nerve rootlets near their entrance point into the diaphragm. Consequently, the success of this technique, as with direct PNP, requires intact bilateral phrenic nerve function. Therefore, this technique will not be successful in patients with lower motor neuron disease or primary muscle disorders. The possibility of intramuscular diaphragm pacing was first suggested in previous experiments performed in our laboratory in anesthetized dogs (11–14). In these studies, electrodes were implanted on the ventral surface of the diaphragm through a midline abdominal incision. Results of diaphragm stimulation by this method were compared with that produced by direct phrenic nerve stimulation with conventional cuff electrodes placed directly on the nerve. A single intramuscular electrode positioned within 1 to 2 cm of the site of phrenic nerve entry into each hemidiaphragm produced inspired volumes that were virtually the same as that resulting from phrenic nerve stimulation. Subsequent studies with a variety of intramuscular electrode designs led to the development of the Peterson electrode that is fabricated from a pair of tandem wound wire cables; each cable has seven 316 L stainless steel wires insulated within fluoropolymer (12–15). In chronic animal studies, there was evidence of tissue ingrowth between the coils of the electrode lead along the entire length of the electrode tract (16). Fibrotic tissue appeared on the epimysium of the diaphragm without damage to the diaphragm itself (16). In earlier animal trials, laparoscopic placement of intramuscular diaphragm electrodes often resulted in the inadvertent positioning of the electrodes through the diaphragm and into the thoracic cavity. This occurred consequent to delivery of the electrode through the laparoscope with a virtually perpendicular approach to the diaphragm. For that reason, an insertion instrument that allowed electrode placement in the same plane as the diaphragm was developed (10). In subsequent animal studies with this instrument, there were no instances in which the electrode traversed the diaphragm and entered the thorax. Unlike the prior animal studies (11, 14), it was necessary to activate two electrodes in each hemidiaphragm to achieve adequate inspired volumes. Although the electrodes in the right hemidiaphragm are located some distance apart and most likely are activating different portions of the muscle, the electrodes within the left hemidiaphragm are located only 1 cm apart. The reason for the discrepancy, in terms of the number of electrodes required to activate the diaphragm, between prior animal studies and the subject of this report are unclear. Further studies are necessary to determine optimal electrode number and parameters of stimulation. Electrode wires exiting the skin carry a small risk of infection. Further development of this technique is necessary therefore, to allow connection of electrode wires to a radiofrequency receiver that can be internalized subcutaneously, eliminating the need for wires exiting the skin. External radiofrequency stimulation can then be applied across the skin, as is currently performed with conventional phrenic nerve stimulation devices (1, 6) and combined intercostal and diaphragm stimulation (5). In summary, the results from this subject suggest that intramuscular diaphragm electrodes can be placed via laparoscopic surgery performed on an outpatient basis. If further investigations show this method to be safe and effective, laparoscopic implantation of intramuscular diaphragm electrodes may provide a less invasive and less costly alternative to conventional PNP. FOOTNOTES Supported by Food and Drug Administration—FD-R-001839 and by the Department of Veterans Affairs. This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form March 5, 2002; accepted in final form September 24, 2002 REFERENCES
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
