Investigating the Mechanics of Positive- versus Negative-Pressure Ventilation

Abstract

Rationale: In response to the COVID-19 pandemic, an increasing number of studies seek to investigate the differences between artificially ventilated lung behavior via positive-pressure ventilation (PPV), and physiological negative-pressure ventilation (NPV) in order to prevent barotrauma and ventilated induced lung injury (VILI). However, these studies have been limited to comparing downstream biochemical inflammatory markers, instead of PPV/NPV energetics and strain mechanics. In this fundamental study, we subject each lung specimen to both PPV and NPV under matched loading conditions and characterize the pressure-volume-strain behavior. Methods: PPV and NPV was applied to ex-vivo porcine lungs using our novel validated custom-designed dual-piston ventilator (Yorkshire farm pigs, n=3, >6months old);PPV pushed air into the lung, and NPV removing air from the sealed tank enclosure housing the tissue, imitating the role of the diaphragm. After preconditioning, a preload/positive end-expiratory pressure (PEEP) of ±5cmH20 was applied followed by various tidal inflation volumes of 675, 900, and 1350mL (corresponding to 6, 8, and 12 mL/kg) at 15 breaths per minute. Each lung was both positively and negatively ventilated, matching the peak pressures and actual displaced lung volume (measured by accounting for air compressibility in real time) to enable direct comparisons. Furthermore, the specimens were speckled to interface global pressure-volume bulk behaviors with local strain measures using new digital image correlation techniques established for lung tissue. Results: The classical pressure-volume loop was analyzed, and the energy loss (measured from the normalized hysteresis response) averaged 23% higher in PPV (0.54±0.06) compared to NPV (0.44±0.06) across all tidal volumes. The bulk material compliancy, measured from the initial inflation slope representing alveolar recruitment and airway resistance, was 2.7-4.5 times greater in NPV (13.18±4.25 mL/cmH2O) than PPV (3.52±1.06 mL/cmH2O), depending on the tidal volume. The strain contours for all inflation volumes of NPV demonstrated uniform stretch topology and histograms revealed reduced strain ranges and means when compared to PPV. Conclusion: Findings highlight loading and deformation-based mechanisms for pulmonary damage and inflammation. The altered response of the same specimen under a positive versus negative pressure gradient demonstrates how artificial ventilation (PPV) underutilizes material elastic recoil, induces a stiffer bulk response, and causes localized regions of high strains compared to natural breathing (NPV). These results have implications for understanding the role of strain-induced damage in VILI and for reevaluating the physiological pressure-volume response in pathologies (e.g. pulmonary emphysema and fibrosis), which have only been based on PPV (not NPV) to date.