Long term diaphragmatic electrical stimulation: a case report of three cases illustrating the positive aspects and pitfalls in high spinal injury

Introduction

Diaphragmatic and phrenic pacing systems, collectively known as diaphragmatic electrical stimulation (DES), have been available and used successfully for over 30 years. Several publications outline the indications, surgical approach and postoperative training required for each system (1-7). There are two broad categories of DES: diaphragmatic motor point pacing (DMPP) and phrenic nerve stimulation (PNS).

In DMPP, impulses are delivered directly to the diaphragmatic muscle via intra-muscular electrodes. Two electrodes are implanted into each hemidiaphragm at the motor points of the phrenic nerves. Motor points are identified as the areas where electrical stimulation causes maximal contraction. Electrode leads are then tunnelled to an exit site in the abdomen where they are connected to an external battery-powered pulse generator (Figure 1A). This procedure is performed laparoscopically.

Figure 1 Diaphragmatic electrical stimulation. (A) DMPP (NeuRx system: Patient 1); 2 electrodes implanted into each hemidiaphragm. (B) PNS. Upper arrows: circumferential phrenic nerve electrodes. (Atrotech left) Patients 2 and 3 (Avery medical right). Lower arrows: implant stimulators. DMPP, direct motor point pacing; PNS, phrenic nerve stimulation.

In PNS, the phrenic nerves are directly stimulated, either via a low cervical approach or within the thorax. A single electrode is placed circumferentially around each phrenic nerve. Leads are then tunnelled and connected to subcutaneously placed implant stimulators. Impulses are transmitted from an external battery-powered pulse generator via antennae using inductive coupling with radiofrequency pulses to the implanted stimulators. The implant stimulator translates the radiofrequency pulses into electrical stimuli to be delivered to each phrenic nerve (Figure 1B).

The phrenic nerves may be paced via a low cervical approach; however, head movement and the proximity of electrodes to the brachial plexus may interfere with impulse transmission. Wound healing may also be compromised by the proximity of cervical incisions to tracheostomies and exposure to respiratory secretions. Additionally, this approach may miss the accessory phrenic nerve, identified in over 60% of patients in one study, with 45% of those joining the phrenic nerve in the thorax (8).

Alternatively, the phrenic nerves may be isolated within the mediastinum where all portions of the nerve are captured. This requires mini-thoracotomies or a thoracoscopic approach.

Connections remain stable with movement, and because the skin is relaxed and away from tracheostomy ties, wound healing is typically uncomplicated. While cervical implants may be done as day-cases, thoracic implants require sequential single lung ventilation and postoperative inpatient care.

DES benefits spinal cord injury (SCI) patients in terms of improved speech, reduced respiratory infections and improved quality of life, however, the limitations of this technology must be acknowledged (9-11). For cardiac pacing, the myocardium needs only a low energy, single-twitch stimulus for full contraction. In contrast, the diaphragm requires a series of rapid, successive pulses. Due to this high energy demand, DES relies on an external power source, with a limited battery life. DES requires an invasive procedure for implementation, with associated risks of general anaesthesia and surgery. Currently, DES systems are not physiologically responsive and require a skilled caregiver to operate.

We describe our experience of DES in 3 patients. We discuss key issues for healthcare professionals considering the use of DES. We present this article in accordance with the CARE reporting checklist (available at https://acr.amegroups.com/article/view/10.21037/acr-2025-72/rc).

Case presentation

Patient 1

This 44-year-old man suffered traumatic bilateral facet joint dislocation at C4/C5 after a fall. He immediately had an American Spinal Injury Association (ASIA) Grade A spinal injury i.e., a complete loss of motor and sensory function below this level. He developed respiratory insufficiency and required mechanical ventilation. He underwent early surgery with cervical fusion. Despite the rapid recognition and correction of his injury, he remained quadriplegic. His post-operative course was complicated by refractory bradycardias requiring placement of a permanent cardiac pacemaker.

Ventilatory weaning followed several conventional strategies; initially, daytime continuous positive airway pressure with pressure support (CPAP-PS) and synchronized intermittent mandatory ventilation (SIMV) overnight. This was followed by prolonged weaning using neurally adjusted ventilatory assist (NAVA). While the patient was comfortable on NAVA with perfect synchronisation, the NAVA level required to generate an appropriate tidal volume was exceptionally high. After several weeks with limited progress on NAVA, we pursued a ventilator-free breathing protocol; this comprised breathing without pressure support for progressively longer periods. While ventilator-free periods of up to 10 minutes were achieved, this strategy led to significant patient anxiety and dyspnoea.

The permanent pacemaker and concerns regarding device-device interaction limited the scope of nerve conduction studies. Needle electromyography with live diaphragmatic fluoroscopy demonstrated recruitment of diaphragmatic motor units on inspiration and confirmed phrenic nerve continuity. A DMPP system was chosen and placed 9 months after his injury to facilitate weaning (NeuRX, Synapse Biomedical Inc., Oberlin, OH, USA).

Following device insertion, we embarked on a regimented conditioning program, with aggressive respiratory clearance using insufflation-exsufflation devices, abdominal binders and multidisciplinary input. Despite a 3-month period of conditioning, spontaneous ventilation was limited to periods of no more than 45 minutes per episode before returning to fully supported mechanical ventilation. Tidal volumes created by the pacing system were approximately 200 mL.

It was our assessment that the use of ventral accessory muscles and activation of a limited number of motor units may have exaggerated the true extent of diaphragmatic descent visible on fluoroscopy. Our patient suffered long-standing diabetes which may have impaired his healing capacity. Finally, the phrenic nerves may have been partially damaged at time of primary injury leading to irreversible diaphragmatic atrophy.

Where the patient had previously failed with leak speech, the use of the diaphragmatic pacing system did allow us to deflate his tracheostomy cuff and enabled communication. This was a major enduring success of placement which significantly enhanced his quality of life. Within 6 months of placement of the pacing system, he succumbed to severe pneumonia and was not resuscitated in accordance with his wishes.

Patient 2

This 31-year-old woman developed a cerebrospinal fluid (CSF) leak following a sports related injury. She subsequently developed meningitis and cerebral venous sinus thrombosis. As a consequence of severe cerebral oedema, she suffered bilateral posterior inferior cerebellar artery occlusion and a large posterior fossa infarction (Figure 2). The patient was mechanically ventilated on a mandatory mode for 6 months prior to phrenic nerve conduction studies. Measured amplitudes were lower than expected and therefore we conducted electrical stimulation of the phrenic nerves with simultaneous diaphragmatic fluoroscopy to measure diaphragmatic descent. Although diaphragmatic contraction was brisk on fluoroscopy, the measured diaphragmatic descent with electrical stimulation was significantly less than normal: 1.2 and 1.7 cm for the left and right sides respectively (normal >4 cm). We proceeded with implantation, attributing the reduced descent to disuse atrophy of the diaphragm rather than denervation (Table 1, Video 1). An Atrotech (Atrotech Oy, Tampere, Finland) phrenic nerve system was implanted. After a period of diaphragmatic conditioning, tidal volumes of 1,400 mL were achieved.

Figure 2 Spinal imaging of patients 1, 2 and 3. Patient 1: bilateral anterior dislocation (anterolisthesis) of C4 on C5; Patient 2: infarct of posterior fossa with pseudomeningocele and infarct of cervical cord; Patient 3: brainstem atrophy with pseudomeningocele of posterior fossa.

The PNS system has allowed this patient to be free of the requisite tethering with tubes mandated by mechanical ventilation. Instead, mobilisation and movement is a much simpler proposition: a heat and moisture exchanger (HME) or Swedish nose is attached to the tracheostomy and the lightweight stimulus controller rests on her lower abdomen or legs. Prior to implantation, she had suffered almost continuous respiratory tract infection and had been on frequent antibiotics. Since establishing respiration with the phrenic pacemaker full-time, these infections have decreased significantly and she has remained well.

While this patient met all criteria for implantation, significant bulbar issues and an absent cough present significant ongoing risks of aspiration. Therefore, we have been unable to progress communication beyond lip reading, letter boards and other augmentative and alternative communication aids. Attempts at “above cuff vocalisation” have not been successful despite varying the model of tracheostomy tube.

Occasional fluctuations in PaCO₂ have led to spells of dizziness where her capture has not been optimal. Weight changes and breast movement have led to loss of capture, particularly on the left side. With the aid of ultrasound, a mark was tattooed on the centre point of the receiver for staff to best place each antenna (Figure 3).

Figure 3 Tattooing for easy identification of implant stimulator receiver.

Patient 3

This 41-year-old man suffered an intracranial haemorrhage and was diagnosed with a brainstem benign cavernous haemangioma. Due to its position, it was deemed inoperable. His neurological status gradually deteriorated and within 18 months, he was confined to a wheelchair. In 2011, following a further haemorrhage, he underwent excision of this lesion (Figure 2). Following this procedure, he had multiple bulbar issues, sialorrhoea, spasticity, dystonia, and respiratory paralysis with quadriplegia. Following a 2-year in-patient stay, he was discharged home and for 5 years was cared for by his family and a homecare team. He was maintained on mandatory ventilation for 7 years.

Neurophysiological studies were supportive of phrenic pacing and, therefore, despite significant bulbar palsies, we proceeded with placement of a phrenic pacing system (Atrotech Oy, Tampere, Finland) (Video 2). Despite 7 years of inactivity, initial diaphragmatic stimulation yielded tidal volumes of 500 mL. After a period of training and with optimum positioning tidal volumes increased to 950 mL (Video 3). His pacemaker has functioned well since implantation.

The patient communicates with the assistance of an eye-gaze device and standard letter number charts. He is not suitable for a speaking valve due to the degree of oromotor dysfunction and the absence of cough and swallow reflexes.

Similar to our second patient, the PNS has facilitated better mobilisation. Concerns about ventilator tubing disconnections, especially during transfers are a constant for all care providers. Tubing may catch on other medical equipment during transfers and may lead to tracheostomy displacement. However, with a PNS, transfer from bed to chair and the tethering effects of the ventilator tubing is not an issue. The assurance that the patient is breathing spontaneously with a PNS allows care providers time and simplifies the attachments a patient needs with movement. Selective ventilation of the lung bases by phrenic stimulation has enabled better mobilisation of secretions and has reduced the frequency of respiratory infections in this man.

Ethical statement

All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patients for the publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Discussion

DES has had a positive impact on the lives of countless patients. It is indicated in congenital or acquired central hypoventilation syndromes (CHS) and high spinal injuries, and there is extensive evidence to support its use in suitable patients. It has proven efficacy: patients can be free of ventilatory support for part, if not all of their day. It liberates them from the tangle of tubes, equipment and accessories, and greatly improves the caregivers’ ability to mobilise patients.

Some patients regain the ability to speak in a quiet environment, without the background noise of modern IPPV and “leak speech”. DES may restore the senses of smell and taste, as well as the ability to eat and drink. In addition to these benefits, reduced respiratory infections and associated hospitalisations have cost savings from a healthcare economics perspective.

When considering DES as a treatment modality, healthcare providers have a broad range of issues to consider including the choice of system, funding, homecare packages and service level agreements. Providers must also be cognisant of the skill mix of carers and the strengths, limitations and contraindications associated with this technology (Figure 4) (12).

Figure 4 Pathway for diaphragmatic pacing in high SCI patients. DES, diaphragmatic electrical stimulation; DMPP, direct motor point pacing; MDT, multidisciplinary team; PNS, phrenic nerve stimulation; SCI, spinal code injury.

Contraindications to DES include raised body mass index, poor lung compliance and mental health issues. Of relevance to patients 2 and 3, significant bulbar dysfunction is also considered to be a relative contra-indication. Prolonged ventilation with a cuffed tracheostomy tube leads to varying degrees of laryngeal and pharyngeal desensitisation. While both patients had bulbar palsies, we proceeded with implantation in the hope that laryngeal sensitivity would improve with time.

Choice of system: DMPP vs. PNS

Our preference, and the institutional preference has been to use Atrotech system as we did in patients 2 and 3. These preferences are based on experience from other local institutions employing this technology, however DMPP may have unique benefits in certain patients.

In this report, our first patient underwent laparoscopic placement of DMPP leads, primarily due to concerns about his existing cardiac pacemaker. DES is an asynchronous technology; there is no sensing arm and therefore concerns regarding triggering of the diaphragmatic pacemaker by a cardiac pacemaker are not applicable. However, the reverse; interference with a sensing cardiac pacemaker by DES is a possibility. A maximum stimulus amplitude of 5 mA is employed for the Atrotech PNS system, 10 mA for the Avery Medical PNS system and up to 25 mA for the NeuRx DMPP system. Despite the increased amplitudes required for DMPP systems, interference with cardiac pacemakers is uncommon, reported in 1 of 20 patients in a one case series. In that patient, pacing was suspended in 1 of the 4 diaphragmatic leads leading to a resolution of the problem without loss of significant tidal volume (13). Post implantation, we monitored patient 1 to examine whether his demand pacemaker was affected by the DMPP system. No interference was detected at any point.

Patients with cardiac pacemakers are more likely to experience interference from PNS implants; however, each case must be evaluated individually, depending on the pacemaker’s complexity, rate modulation, and antiarrhythmic functions. Such interference may be mitigated through device reprogramming (14).

Because our first patient had DMPP, no change in airway was required during the operative procedure to place the leads. While critics of the system question whether all portions of the diaphragm are stimulated by placement of leads at the motor points of the phrenic nerves, it has enjoyed enduring success with numerous large case series (13). In our experience the connection of the diaphragmatic leads with the external pulse generator is very stable and safe.

Our second and third patients had PNS leads implanted via mini-thoracotomies. Following induction of anaesthesia, the tracheostomies were changed to left-sided double-lumen tubes. Neither required a chest drain as each incision avoided damage to the visceral or pulmonary pleura. While postoperative pain management would be a significant consideration for most patients, due to the nature of their injuries and loss of sensation, none of our three patients experienced difficulties in this regard.

Diaphragmatic conditioning

Histological evidence shows that diaphragmatic muscle degradation begins within 18 hours of losing neural input. Prolonged disuse leads to diaphragmatic atrophy, thinning, and loss of contractile force. Reversing this atrophy and increasing the muscle bulk of the diaphragm are the objectives of the diaphragmatic conditioning period. This period may begin immediately in the case of DMPP, however for PNS, a period of healing of about 1 month is recommended (15).

With appropriate stimulation, there is a gradual conversion of Type II diaphragmatic fast-twitch, fatigue-susceptible muscle fibres, to Type I slow-twitch, fatigue-resistant muscle fibres. Diaphragmatic conditioning relies on this adaptation mechanism. In addition to this change in fibre type, increases in tidal volumes over weeks to months represent an increase in the muscle bulk of the stimulated diaphragm. Periodic recalibration of the currents used, tidal volume measurements, and the determination a fatigue threshold are constituent parts of the conditioning period.

The duration of this conditioning period has been reported to take between 2 weeks to 5 months and depends on several factors, including the duration of diaphragmatic inactivity, the frequency and intensity of training sessions, and numerous patient specific factors such as motivation, intercurrent infection and body habitus. Finally, the pinnacle point is the availability of skilled healthcare personnel to conduct training and ensure a seamless transition between the ventilator and stimulator.

Tidal volumes tend to remain stable over time, however, intercurrent illness, pneumonias or interruptions in stimulation of more than a few days may require a cautious re-titration of stimulation to previously tolerated settings.

With the Atrotech PNS system, patients may receive stimulation at three programmed amplitudes: low, normal and high, each producing a distinct tidal volume. Tidal volumes delivered for each amplitude setting vary with position. A manual change in the stimulation setting and therefore tidal volume is required following a change in the body position. For example, moving from a recumbent to a sitting position may require a change from low setting to the high setting to achieve the same tidal volume.

Complications

Modern mechanical ventilators provide a full spectrum of alarms and back-up mechanisms to alert and protect the user and caregiver from issues such as disconnection, leak, pressure changes and power failure (Table 2). Most mechanical ventilators will also measure and display continuous pressure, flow and volume data allowing for real-time assessment of ventilation (16-18). Even the quietest modern ventilators also inherently produce some sound, providing an auditory cue if abruptly discontinued.

In PNS-DES, displacement of the antennae is a constant risk which may lead to loss of conduction with consequently hypoventilation or even apnoea (19-22). Loss of conduction may occur by loosening of securing tapes or slippage of the elasticated fitting belt. The loss of capture on one side due to displacement of an antenna may lead to reduced minute ventilation. To mitigate these risks, we have employed portable pulse oximetry and end-tidal carbon dioxide (ETCO₂) monitoring.

Proper positioning of the implant stimulator at the time of operation is of the utmost importance. In our experience, placement of one implant too close to the axillary tail of the breast has led to issues where conduction has been unreliable at times (Figure 3). Additionally, changes in body habitus may make locating subcutaneously placed implant stimulators difficult. Tattooing may be helpful to identify the centre point to allow best positioning of the external antennae.

Proponents of DES highlight the lack of risk of disconnection as a major advantage for this technology over mechanical ventilation, however, battery or pacer failure present a similarly significant hazard. Manufacturers have attempted to address these risks through the development of bilateral antennae, where independent and parallel circuits generate separate signals for each antenna (19). Irrespective of redundancy within the system, there is a lack of information about patient ventilation inherent to DES and users must be cognisant of this fact. This is particularly true in patients with brainstem injuries whose ability to communicate is compromised.

Future directions

At present, none of the available DES systems are physiologically responsive, nor do they synchronise with upper airway or accessory respiratory muscles. All DES systems require a large power supply, which at present must be extra-corporeal. This has hampered the development of a fully implanted system (6). The ultimate solution would include a fully implanted system with position sensing and rate responsiveness (2,4).

Conclusions

DES is an advanced treatment modality aimed at managing one aspect of the complex clinical syndrome associated with high SCI. While it is recommended by NICE, there are significant limitations in current technology which make DES challenging to employ. A lack of physiological responsiveness, limited monitoring of the adequacy of ventilation, the requirement for a skilled caregiver and the potential for unrecognised hypopnea and apnoea are significant issues. Due to the rarity of conditions which benefit from phrenic pacing, technological advancement has been slower than in other similar areas such as cardiac pacemakers or diabetes closed loop systems (10,21).

Licensing for DES systems whether by the Food and Drug Administration (USA) or under Medical Device Regulation (Europe) should strive to develop new and improved standards for DES technology to address current shortcomings. If phrenic pacing is to become a strategy for the future, a fully implanted, intelligent, rate responsive, position sensitive system must be the objective.

Acknowledgments

We sincerely thank our patients and their families for their kind support in the preparation of this manuscript.

Footnote

Reporting Checklist: The authors have completed the CARE reporting checklist. Available at https://acr.amegroups.com/article/view/10.21037/acr-2025-72/rc

Peer Review File: Available at https://acr.amegroups.com/article/view/10.21037/acr-2025-72/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://acr.amegroups.com/article/view/10.21037/acr-2025-72/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patients for the publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chervin RD, Guilleminault C. Diaphragm pacing: review and reassessment. Sleep 1994;17:176-87. [Crossref] [PubMed]
2. Le Pimpec-Barthes F, Legras A, Arame A, et al. Diaphragm pacing: the state of the art. J Thorac Dis 2016;8:S376-86. [Crossref] [PubMed]
3. Arango-Cortes ML, Giraldo-Cadavid LF, Latorre Quintana M, et al. Diaphragm pacing compared with mechanical ventilation in patients with chronic respiratory failure caused by diaphragmatic dysfunction: a systematic review and meta-analysis. Expert Rev Respir Med 2024;18:1101-11. [Crossref] [PubMed]
4. DiMarco AF. Phrenic nerve stimulation in patients with spinal cord injury. Respir Physiol Neurobiol 2009;169:200-9. [Crossref] [PubMed]
5. Jarosz R, Littlepage MM, Creasey G, et al. Functional electrical stimulation in spinal cord injury respiratory care. Top Spinal Cord Inj Rehabil 2012;18:315-21. [Crossref] [PubMed]
6. Aiyar H, Mortimer J. Intelligent Systems and Technologies in Rehabilitation Engineering. Teodorescu H, Jain L. editors. CRC Press LLC, Boca Raton; 2001.
7. NICE. IPG 792: Phrenic nerve pacing for ventilator-dependent high cervical spinal cord injury: Interventional procedures guidance. 2024.
8. Loukas M, Kinsella CR Jr, Louis RG Jr, et al. Surgical anatomy of the accessory phrenic nerve. Ann Thorac Surg 2006;82:1870-5. [Crossref] [PubMed]
9. Hirschfeld S, Exner G, Luukkaala T, et al. Mechanical ventilation or phrenic nerve stimulation for treatment of spinal cord injury-induced respiratory insufficiency. Spinal Cord 2008;46:738-42. [Crossref] [PubMed]
10. Hirschfeld S, Huhtala H, Thietje R, et al. Phrenic nerve stimulation experiences. A single centre, controlled, prospective study. J Clin Neurosci 2022;101:26-31. [Crossref] [PubMed]
11. Romero FJ, Gambarrutta C, Garcia-Forcada A, et al. Long-term evaluation of phrenic nerve pacing for respiratory failure due to high cervical spinal cord injury. Spinal Cord 2012;50:895-8. [Crossref] [PubMed]
12. The NHS Commissiong Board Clinical Reference Group (2013). Clinical Commissioning Policy: Phrenic Nerve Pacing Following Spinal Cord Injury. Accessed 24 November 2023.
13. Onders RP, Khansarinia S, Weiser T, et al. Multicenter analysis of diaphragm pacing in tetraplegics with cardiac pacemakers: positive implications for ventilator weaning in intensive care units. Surgery 2010;148:893-7; discussion 897-8. [Crossref] [PubMed]
14. Kolb C, Eicken A, Zrenner B, et al. Cardiac pacing in a patient with diaphragm pacing for congenital central hypoventilation syndrome (Ondine's curse). J Cardiovasc Electrophysiol 2006;17:789-91. [Crossref] [PubMed]
15. Tobin MJ, Laghi F, Jubran A. Narrative review: ventilator-induced respiratory muscle weakness. Ann Intern Med 2010;153:240-5. [Crossref] [PubMed]
16. Gregoretti C, Navalesi P, Ghannadian S, et al. Choosing a ventilator for home mechanical ventilation. Breathe 2013;9:394-409.
17. Park S, Suh ES. Home mechanical ventilation: back to basics. Acute Crit Care 2020;35:131-41. [Crossref] [PubMed]
18. Janssens JP, Michel F, Schwarz EI, et al. Long-Term Mechanical Ventilation: Recommendations of the Swiss Society of Pulmonology. Respiration 2020; Epub ahead of print. [Crossref]
19. Giberson CE, Cheshier SH, Poree LR, et al. Diaphragm Pacing: A Safety, Appropriateness, Financial Neutrality, and Efficacy Analysis of Treating Chronic Respiratory Insufficiency. Neuromodulation 2023;26:490-7. [Crossref] [PubMed]
20. Günter C, Delbeke J, Ortiz-Catalan M. Safety of long-term electrical peripheral nerve stimulation: review of the state of the art. J Neuroeng Rehabil 2019;16:13. [Crossref] [PubMed]
21. Wolfe LF. Point: Should phrenic nerve stimulation be the treatment of choice for spinal cord injury? Yes. Chest 2013;143:1201-3. [Crossref] [PubMed]
22. Gay PC. Counterpoint: Should phrenic nerve stimulation be the treatment of choice for spinal cord injury? No. Chest 2013;143:1203-6. [Crossref] [PubMed]