A new paradigm for lung-conservative total liquid ventilation

Abstract

Background: Total liquid ventilation (TLV) of the lungs could provide radically new benefits in critically ill patients requiring lung lavage or ultra-fast cooling after cardiac arrest. It consists in an initial filling of the lungs with perfluorocarbons and subsequent tidal ventilation using a dedicated liquid ventilator. Here, we propose a new paradigm for a lung-conservative TLV using pulmonary volumes of perfluorocarbons below functional residual capacity (FRC).

Methods and findings: Using a dedicated technology, we showed that perfluorocarbon end-expiratory volumes could be maintained below expected FRC and lead to better respiratory recovery, preserved lung structure and accelerated evaporation of liquid residues as compared to complete lung filling in piglets. Such TLV below FRC prevented volutrauma through preservation of alveolar recruitment reserve. When used with temperature-controlled perfluorocarbons, this lung-conservative approach provided neuroprotective ultra-fast cooling in a model of hypoxic-ischemic encephalopathy. The scale-up and automating of the technology confirmed that incomplete initial lung filling during TLV was beneficial in human adult-sized pigs, despite larger size and maturity of the lungs. Our results were confirmed in aged non-human primates, confirming the safety of this lung-conservative approach.

Interpretation: This study demonstrated that TLV with an accurate control of perfluorocarbon volume below FRC could provide the full potential of TLV in an innovative and safe manner. This constitutes a new paradigm through the tidal liquid ventilation of incompletely filled lungs, which strongly differs from the previously known TLV approach, opening promising perspectives for a safer clinical translation. FUND: ANR (COOLIVENT), FRM (DBS20140930781), SATT IdfInnov (project 273).

Keywords: Biomedical engineering; Critical care; Liquid ventilation; Therapeutic hypothermia.

Fig. 1

Evaluation of lung volumes in normal conditions in anesthetized piglets submitted to gas ventilation. A- Picture of a thoracic computerized tomography (CT)-scan in one anesthetized piglet during mechanical ventilation. Images were obtained during prolonged expiratory pause. B- 3D reconstruction of the lung with an acquisition during a prolonged expiratory pause with positive end-expiratory pressure (PEEP) set at 0 cmH₂O. C- 3D reconstruction of the lung with an acquisition during a prolonged expiratory pause with positive end-expiratory pressure (PEEP) set at 5 cmH₂O. D- Measured lung volumes in 6 piglets during a prolonged expiratory pause at PEEP = 0 or 5 cmH₂O. Circles represent individual values and bold line mean values in each condition, respectively.

Fig. 2

Evaluation of different ventilation strategy for total liquid ventilation in piglets. A- Schematic representation of the liquid ventilators for TLV, including piston pumps and valves to drive the liquid into and from the lung, thermal exchanger, condenser for perfluorocarbon condensing and oxygenator. The liquid ventilator also includes sensors, graphic user interface and electronic algorithm to control the entire process [19]. B- Experimental protocol including five groups of piglets submitted to 30 min of TLV with different tidal volumes (TV of 8 or 16 ml/kg) and end-expiratory volumes (EV of 15 or 30 ml/kg), as compared to Sham animals with conventional mechanical ventilation only. The four corresponding groups are so-called TV₈-EV₁₅, TV₁₆-EV₁₅, TV₈-EV₃₀ and TV₁₆-EV₃₀, respectively. C- Volumes of perfluorocarbons within the lungs at the end of expiration during TLV and static pulmonary pressures measured during end-expiratory and end-inspiratory pauses, respectively. D- Blood pH and carbon dioxide and oxygen partial pressure (pCO₂ and pO₂, respectively). *, p < 0.05 vs Sham; †, p < 0.05 vs TV₈-EV₁₅; ‡, p < 0.05 vs TV₁₆-EV₁₅.

Fig. 3

Morphological alterations and perfluorocarbon (PFC) residues in the different groups of piglets submitted to total liquid ventilation (TLV). A- Typical appearance of lung parenchyma on magnetic resonance imaging using T1W sequence, apparent transverse relaxation rate (R2*) map and ¹⁹F dual-nuclei imaging in piglets from the different groups. The T1W allows visualizing lung parenchyma and anatomy. Reduced R2* suggests enhanced hemorrhage or edema in the TV16-EV30 group [32]. ¹⁹F signal shows PFC residues. B- Average value of apparent transverse relaxation rate R2* in the entire lungs in the different groups. Low values indicate morphological alterations including hemorrhage and pulmonary edema. C- Average volume of PFC residues in the different groups (as percentage of entire lung volume). D- 3-D reconstruction of entire lungs after ¹⁹F dual-nuclei imaging. Blue and green areas represent lung parenchyma and PFC-filled spots, respectively. A tube filled with 100% PFC is located under the sample for absolute quantification of ¹⁹F images. E- Normal pulmonary histological appearance in a Sham piglet. F- Abnormal pulmonary histological appearance in a Sham piglet demonstrating interstitial inflammation and foci of bronchiolitis. G- Abnormal pulmonary histological appearance in a piglet from the TV₁₆-EV₃₀ group including hyaline membranes, alveolitis, hemorrhage and serous edema. H- Abnormal pulmonary histological appearance in a piglet from the TV₁₆-EV₃₀ group demonstrating alveolar and bronchiolar distension. *, vs TV₈-EV₁₅; See legend in Fig. 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Evaluation of perfluorocarbons (PFC) pulmonary repartition during total liquid ventilation (TLV) in piglets. A- Transverse images of thoracic computerized tomography (CT-scan) during prolonged pauses at end-inspiration or end-expiration in one anesthetized piglet positioned in dorsal recumbency and submitted to TLV with various end-expiratory or tidal volumes (EV and TV set at 8/16 and 15/30 ml/kg, respectively). B- Lung regions of interest (ROI) on a typical thoracic CT-scan picture. The amount of perfluoctylbromide was calculated in each ROI, taking into account its own attenuation as compared to lung normal attenuation (+2300 and −600 Houndsfield Units, respectively). The so-called zones U, I and L corresponds to upper (sternal), intermediate and lower (dorsal) ROI. C- Amount of PFOB in each ROI, along with the total volume of PFC in the lungs. A pooled analysis was done among inspiration and expiration measurement. Five points were analyzed with the following expected volumes: 1) 15 ml/kg (expiration in TV₈-EV₁₅ and TV₁₆-EV₁₅ conditions), 2) 24 ml/kg (inspiration in TV₈-EV₁₅ conditions), 3) 30–31 ml/kg (inspiration in TV₁₆-EV₁₅ and expiration in TV₈-EV₃₀ and TV₁₆-EV₃₀ conditions), 4) 38 ml/kg (inspiration in TV₈-EV₃₀ conditions) and 5) 46 ml/kg (inspiration in TV₁₆-EV₃₀). Mean values were calculated in case of replication, at 15 and 30–31 ml/kg. D- Relationship between pulmonary pressure and PFC volume during a slow instillation into the trachea in a piglet. Dashed red lines emphasizes inflexion points. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Evaluation of lung-conservative total liquid ventilation (TLV) in pathophysiological condition of hypoxic-ischemic encephalopathy (HIE) in piglets. A- Experimental protocol describing the hypoxic-ischemic injury. After resuscitation, animals were either treated by conventional mechanical ventilation under normothermia (Control group) or hypothermic TLV. A third group was submitted to a Sham procedure with no HIE induction. B- Neurological dysfunction scores after HIE induction or Sham procedure (0% = lack of dysfunction; 100% = death). Open circles and bold lines represent individual and median values, respectively. C- Blood concentration of the S100B protein as a marker of brain injury. D- Kaplan-Meyer survival curves in all experimental groups. E- Typical histological appearance of the hippocampus after hemalun-eosin and fluorojade-C staining (left and right column rows, respectively). The latter staining show no or very few degenerating neurons in Sham and TLV groups, as compared frequent degenerating neurons in Control group. F- Number of positive fluoro-Jade C cells, expressed ad mean number per analyzed field, in each areas of interest of the central nervous system. *, p < 0.05 vs Sham; †, p < 0.05 vs Control.

Fig. 6

Evaluation of total liquid ventilation with a new dedicated technology for large pigs. A- Schematic representation of the new specifically designed liquid ventilator. B- Typical perfluocarbon flow (upper raw), pressure at mouth and pulmonary volume of perfluocarbon during the first 5 min of total liquid ventilation (TLV) in a 63 kg pig. C- Schematic representation of experimental protocol in large pigs submitted to 30 min of hypothermic TLV followed by conventional gaseous ventilation and rewarming, before awakening. Animals were followed during 10 days before euthanasia for post-mortem analyses. D- Body temperatures in the different compartments during the TLV episode, showing a rapid decrease of target temperature (32–33 °C) within 20 min in all compartments. E- Blood pH and carbon dioxide and oxygen partial pressure (pCO₂ and pO₂, respectively). F- Thoracic computerized tomography (CT-scan) of an explanted lung in a pig at the end of the follow-up. No macroscopic foci of perfluorocarbons can be observed, suggesting complete elimination. G- Morphological appearance of the lung upon histological examinations. The left panel shows normal appearance. The right panel show an area with dilation of bronchioles (arrow) and alveolae, as typically observed after mechanical ventilation.

Fig. 7

Evaluation of the long-term tolerance of lung conservative total liquid ventilation (TLV) in two aged non-human primates (Macaca fascicularis). A- Transverse images of thoracic computerized tomography (CT-scan) at the initial stage (i.e., two weeks before the TLV episode). B- 3D reconstruction of the lung in the same animal at the initial stage. C- Rectal temperature of the two primates submitted to an episode of hypothermic TLV, followed by rewarming during conventional mechanical ventilation before awakening. D- Pulmonary compliance of the two primates before and after the episode of hypothermic TLV. E- Arterial blood pH, partial pressure of CO₂ and hemoglobin oxygen saturation in the two primates at baseline, during the episode of TLV and 3 weeks later. F- CT-scan four weeks after the TLV episode. G- 3D reconstruction of the lung in the same animal four weeks after the TLV episode.