New insights into rapid magnetic phrenic nerve stimulation: optimization for transition into the ICU

Abstract

In the intensive-care-unit (ICU), mechanical ventilation is often needed in the event of respiratory failure in a variety of patients with underlying pathophysiologies. Although mechanical ventilation serves to promote gas exchange and ultimately save lives, it is not without its own limitations. One such limitation includes ventilator induced-lung injury (VILI), which is characterized by lung tissue damage and inflammation as a result of the stress and strain induced upon the lung by the unphysiological inspiratory positive pressure from the ventilator. In addition to lung insult, the respiratory musculature, in particular the diaphragm, may develop dysfunction deemed ventilator-induced diaphragmatic dysfunction (VIDD). VIDD is characterized by many insults to the respiratory muscles including myofibrillar damage, lipid accumulation, mitochondrial dysfunction and atrophy. A hallmark of VIDD is a decreased force generating capability of the respiratory muscles, which dramatically impacts the ability of patients to successfully wean off ventilators. Ultimately, weaning may become difficult. Unsuccessful weaning, in turn, increases the time patients spend on mechanical ventilation, which can further promote VILI and VIDD, as well as increase healthcare costs. Inducing ventilation, or at least diaphragm contractions, via phrenic nerve stimulation may serve as a therapeutic tool to combat both VILI and VIDD, simultaneously. When both phrenic nerves are stimulated, each hemi-diaphragm contracts and descends assisting with inspiration by generating negative pleural pressure. This negative pressure generation would ultimately reduce the amount of positive pressure needed from the mechanical ventilator, and thus reduce lung stress and strain (protecting against VILI). In addition, diaphragmatic contraction reduces muscle disuse, and may combat VIDD. Although many methods to perform phrenic nerve stimulation exist, only few are non-invasive, one of them being rapid magnetic stimulation (rMS). However, many questions are still open with respect to the use of rMS in an ICU environment. First, there is a variety of coil shapes and anatomical stimulation locations possible to perform rMS, and the most promising combination for ICU use is unknown. Second, in case of hemi-diaphragm paralysis, it is unclear whether performing rMS on a single phrenic nerve is feasible, and which effect this might have. Third, reliability of rMS over several therapeutic sessions within and between days needs to be quantified. Fourth, the feasibility of performing rMS in different body positions needs to be established given patients’ body positions are often altered during the day, and that certain conditions have more promising clinical outcomes than others depending on position. Finally, whether altering thoracoabdominal compliance extrinsically can be used as a method to increase “stimulation dose” is in need of exploration. Therefore, the aims of the OptiStim study (chapter 4), were to determine the optimal coil design and stimulation location to perform rMS. To accomplish this task, three different commercially available magnetic coils with three distinct shapes (a parabolic shape coil, a D-shape coil, and a butterfly [sometimes referred to as a figure of 8] coil) were tested across a series of stimulation frequencies and stimulator outputs while evaluating inspiratory responses and side-effects. In addition, two possible coil placements for phrenic nerve stimulation were tested, i.e. bilaterally anterolaterally on the neck, and anteriorly on the chest above the sternum. Finally, the OptiStim study aimed to determine if the optimal technique and coil shape could be used to ventilate healthy humans for up to ten minutes. The findings revealed that all three coil shapes were able to induce similar inspiratory responses, however, the butterfly coil needed greater stimulator output, and thus resulted in greater unwanted side-effects. In addition, coils not connected to an active cooling unit, such as the parabolic coil, had a propensity to overheat. In addition, performing rMS on the chest could not induce inspiratory responses in all but one participant. Therefore, rMS-induced ventilation was attempted with the D-shape coil placed bilaterally and anterolaterally on the neck, and which successfully ventilated all participants. The aims of the UniStim study (chapter 5) were to quantify inspiratory responses and side-effects to rMS of a single phrenic nerve, i.e., placing a single coil unilaterally, anterolaterally on the neck during stimulation. The UniStim study was conducted in order to determine if unilateral rMS may produce enough diaphragm activation to potentially attenuate VILI and VIDD. Responses to unilateral and bilateral rMS were also compared. Unilateral rMS was indeed capable of inducing physiological resting breathing activation of the diaphragm. Notably, maximal inspiratory responses were roughly half that of bilateral rMS, however, side-effects did not dramatically differ, excluding an increase in movement with unilateral stimulation. The aim of the ReStim study (chapter 6) was do determine reliability of responses within and between days by applying three sessions of bilateral rMS within a single day, and one session on each of the two subsequent days. rMS was conducted across a multitude of stimulation frequencies and stimulator outputs. Inspiratory responses to rMS displayed good-to-excellent reliability within a single day, but decreased slightly between days, indicating that stimulation parameters such as stimulation frequency and stimulator output may need to be adjusted between days in order to achieve matched responses. The aim of the PosiStim study was to determine if bilateral rMS was feasible in the semirecumbent, supine and sitting position (chapter 7), given the change in prognosis seen in patients during mechanical ventilation in various body positions. rMS was capable of inducing at least physiological resting breathing levels of diaphragm activation in all three body positions. However, activation was the highest in the semirecumbent position, while lowest in the supine position. The changes in activation may serve as a useful tool to alter responses in patients in which simply adjusting the stimulator output may not be feasible. For example, increasing the stimulator output may induce significant increases in discomfort and pain, while decreasing the stimulator output may eradicate any response in some persons. Finally, the sitting position was deemed the least tolerable by the participants. The final study, the WeBiStim study (chapter 8) was a pilot to determine if adding a 10kg-weight belt to the abdomen, or performing elastic thoracoabdominal binding during bilateral rMS could: 1) increase the pressure generation by the diaphragm in response to rMS at matched tidal volumes; and/or 2) increase the stimulator output required to reach a matched target tidal volume and thus to increase the “stimulation dose”. Thoracoabdominal binding had no effect when compared to no binding. Adding the 10kg-weight belt, however, modestly increased the pressure generation of the diaphragm compared to no weight at matched tidal volumes. Thus, whether this modest increase in pressure generation can translate into improved clinical outcomes remains open, but seems unlikely. Collectively, the present work adds essential information to the body of research currently being conducted with the goal of driving the technique of rMS into the ICU. Further innovations in technology are needed to optimize the equipment to perform rMS within a clinical setting. In addition, future research must test the feasibility of performing rMS in mechanically ventilated patients, and to determine the “stimulation dose” that can translate into meaningful clinical outcomes.