The terminal segment of the human phrenic nerve as a novel implantation site for diaphragm pacing electrodes: Anatomical and clinical description

Highlights

• The human phrenic nerve consistently ends with a free terminal pre-branching segment.

• Running from the pericardium to the diaphragm, this segment measures 60–75 mm.

• This free terminal pre-branching segment contains the totality of phrenic fibers.

• It appears highly suitable for the dissection free implantation of diaphragm pacing electrodes.

Abstract

Background

Diaphragm pacing allows certain ventilator-dependent patients to achieve weaning from mechanical ventilation. The reference method consists in implanting intrathoracic contact electrodes around the phrenic nerve during video-assisted thoracic surgery, which involves time-consuming phrenic nerve dissection with a risk of nerve damage. Identifying a phrenic segment suitable for dissection-free implantation of electrodes would constitute progress.

Study design

This study characterizes a free terminal phrenic segment never fully described before. We conducted a cadaver study (n = 14) and a clinical observational study during thoracic procedures (n = 54).

Results

A free terminal phrenic segment was observed on both sides in 100% of cases, "jumping" from the pericardium to the diaphragm and measuring 60 mm [95% confidence interval; 48–63] and 72.5 mm [65–82] (right left, respectively; p = 0.0038; cadaver study). This segment rolled up on itself at end-expiration and became unravelled and elongated with diaphragm descent (clinical study). Three categories of fat pads were defined (type 1: pericardiophrenic bundle free of surrounding fat; type 2: single fatty fringe leaving the phrenic nerve visible until diaphragmatic entry; type 3: multiple fatty fringes masking the site of penetration of the phrenic nerve) that depended on body mass index (p = 0.001, clinical study). Hematoxylin-eosin and toluidine blue staining (cadaver study) showed that all of the phrenic fibers in the distal, pre-branching part of the terminal segment were contained within a single epineurium containing a variable number of fascicles (right: 1 [95%CI 0.65–4.01]; left 5 [3.37–7.63]; p = 0.03).

Conclusion

Diaphragm pacing through periphrenic electrodes positioned on the terminal phrenic segment should be tested.

Introduction

Lifelong diaphragm pacing allows ventilator-dependent patients with central respiratory paralysis (e.g., congenital central alveolar hypoventilation or high cervical spinal cord lesions) to achieve weaning from external ventilatory assistance (Le Pimpec-Barthes et al., 2016). The reference method for diaphragm pacing in this setting involves surgical implantation of stimulation electrodes in contact with the phrenic nerve trunk —periphrenic electrodes— (Glenn et al., 1973, Talonen et al., 1983).

Phrenic nerve electrodes were historically implanted at the base of the neck for reasons of simplicity. This approach is imperfect because of the risk of damage to the electrodes during neck movements and the risk of incomplete diaphragm recruitment when the accessory phrenic nerve carries a significant proportion of phrenic nerve fibers (Rajanna, 1947). To circumvent these limitations, electrodes are currently implanted during video-assisted thoracic surgery, at the level of the superior vena cava on the right side and the hilum of the lung on the left side (Le Pimpec-Barthes et al., 2011). This method ensures access to all phrenic nerve fibers, but it requires careful dissection of the phrenic nerve to separate it from the underlying anatomical plane and to create electrode accesses (Le Pimpec-Barthes et al., 2011). It is therefore a time-consuming procedure, associated with a risk of nerve damage. Implantation of phrenic nerve electrodes over the intact phrenic nerve without dissection would therefore represent significant progress.

Anatomical descriptions of the human phrenic nerve have abunded over time (Kelley, 1950, Luschka, 1853, Pretterklieber et al., 2020, Rajanna, 1947, Whitley et al., 2021). A depiction of the human phrenic nerve termination published at the end of the 1950s (Muller Botha, 1957) reported that, before branching out to different diaphragm zones, the phrenic nerve leaves the anterolateral angle of the pericardial base in the direction of the diaphragm. This segment of the phrenic nerve has not been precisely described (Muller Botha, 1957), but it is free of any anatomical contact with the pericardium. It can therefore be hypothesized that this segment would be suitable for dissection-free implantation of phrenic nerve electrodes. It can also be hypothesized that implantation at this site would allow full diaphragm recruitment, because the corresponding nerve segment, proximal to any nerve branches, is bound to contain all phrenic nerve fibers. We designed the present study to investigate these hypotheses by providing a detailed anatomical, histological and surgical description of the pre-branching terminal segment of the human phrenic nerve.