In total liquid ventilation (TLV), the lungs are filled with a breathable liquid perfluorocarbon (PFC) while a liquid ventilator ensures proper gas exchange by renewal of a tidal volume of oxygenated and temperature-controlled PFC. Given the rapid changes in core body temperature generated by TLV using the lung has a heat exchanger, it is crucial to have accurate and reliable core body temperature monitoring and control. This study presents the design of a virtual lung temperature sensor to control core temperature. In the first step, the virtual sensor, using expired PFC to estimate lung temperature noninvasively, was validated both in vitro and in vivo. The virtual lung temperature was then used to rapidly and automatically control core temperature. Experimentations were performed using the Inolivent-5.0 liquid ventilator with a feedback controller to modulate inspired PFC temperature thereby controlling lung temperature. The in vivo experimental protocol was conducted on seven newborn lambs instrumented with temperature sensors at the femoral artery, pulmonary artery, oesophagus, right ear drum, and rectum. After stabilization in conventional mechanical ventilation, TLV was initiated with fast hypothermia induction, followed by slow posthypothermic rewarming for 1 h, then by fast rewarming to normothermia and finally a second fast hypothermia induction phase. Results showed that the virtual lung temperature was able to provide an accurate estimation of systemic arterial temperature. Results also demonstrate that TLV can precisely control core body temperature and can be favorably compared to extracorporeal circulation in terms of speed.

BackgroundFilling the lung with dense liquid perfluorocarbons during total liquid ventilation (TLV) might compress the myocardium, a plausible explanation for the instability occasionally reported with this technique. Our objective is to assess the impacts of TLV on the cardiovascular system, particularly left ventricular diastolic function, in an ovine model of neonatal respiratory distress syndrome.MethodEight newborns lambs, 3.0 ± 0.4 days (3.2 ± 0.3kg) were used in this crossover experimental study. Animals were intubated, anesthetized and paralyzed. Catheters were inserted in the femoral and pulmonary arteries. A high-fidelity pressure catheter was inserted into the left ventricle. Surfactant deficiency was induced by repeated lung lavages with normal saline. TLV was then conducted for 2 hours using a liquid ventilator prototype. Thoracic echocardiography and cardiac output assessment by thermodilution were performed before and during TLV.ResultsLeft ventricular end diastolic pressure (LVEDP) (9.3 ± 2.1 vs. 9.2 ± 2.4mmHg, p = 0.89) and dimension (1.90 ± 0.09 vs. 1.86 ± 0.12cm, p = 0.72), negative dP/dt (-2589 ± 691 vs. -3115 ± 866mmHg/s, p = 0.50) and cardiac output (436 ± 28 vs. 481 ± 59ml/kg/min, p = 0.26) were not affected by TLV initiation. Left ventricular relaxation time constant (tau) slightly increased from 21.5 ± 3.3 to 24.9 ± 3.7ms (p = 0.03). Mean arterial systemic (48 ± 6 vs. 53 ± 7mmHg, p = 0.38) and pulmonary pressures (31.3 ± 2.5 vs. 30.4 ± 2.3mmHg, p = 0.61) were stable. As expected, the inspiratory phase of liquid cycling exhibited a small but significant effect on most variables (i.e. central venous pressure +2.6 ± 0.5mmHg, p = 0.001; LVEDP +1.18 ± 0.12mmHg, p<0.001).ConclusionsTLV was well tolerated in our neonatal lamb model of severe respiratory distress syndrome and had limited impact on left ventricle diastolic function when compared to conventional mechanical ventilation.

This study is the first to simulate ultrafast cooling by TLV in a human model and is a strong motivation to translate TLV to humans to improve the quality of life of postcardiac arrest patients.

This comprehensive thermal model of the lungs and body has the advantage of closely modeling the rapid thermal dynamics in TLV. The model can explain how the time to achieve mild hypothermia between newborn and juvenile lambs remained similar despite of highly different physiological and ventilatory parameters. The strength of the model is its strong relationship with the physiological parameters of the subjects, which suggests its suitability for projection to humans.

No abstract

Total liquid ventilation (TLV) is an emerging and promising mechanical ventilation method in which the lungs are filled with a breathable liquid. Perfluorocarbon (PFC) is the predominant liquid of choice due to its high O2 and CO2 solubility. In TLV, a dedicated liquid ventilator ensures gas exchange by renewing a tidal volume of PFC, which is temperature-controlled, oxygenated and free of CO2. A fundamental difference between TLV and conventional mechanical ventilation relates to the fact that PFCs are approximately 1500 times denser than air. This high density provides PFCs with a large heat capacity, turning the lungs into an efficient heat exchanger with circulating blood. The originality of this study is the development of a lumped thermal model of the body as a heat exchanger coupled to a liquid ventilator. The model was validated with an animal experimentation on a newborn lamb with the Inolivent-5.0 liquid ventilator prototype. TLV was initiated with a fast hypothermia induction, followed successively by a slow posthypothermic rewarming, a fast rewarming and finally a second fast hypothermia induction. Results demonstrate that the model was able to aptly predict, in every phase, the temperature of the lungs, the eardrum, the rectum as well as the various compartments of the liquid ventilator.

Total liquid ventilation is an innovative experimental method of mechanical assisted ventilation in which lungs are totally filled and then ventilated with a tidal volume of perfluorochemical liquid (PFC) by using a dedicated liquid ventilator. The positive end-inspiratory and end-expiratory pressures (PEIP and PEEP) are static pressure measurements that are critic to the safe and efficient control of the ventilation. However, their measurement is impeded by large oscillations of pressure caused by the propagation of pressure waves along the flexible tubes carrying the PFC to the patient. The aim of this paper is to describe a method to accurately estimate the PEEP and the PEIP from noisy data hindered by flexible tubing resonance during short respiratory pauses. The method developped makes use of the least squares technique to estimate the steady state pressure. Preliminary in vivo validation of the algorithm shows that the method gives accurate estimations with respiratory pauses as short as 0.3 second. I. INTRODUCTIONIn the last decade, a vast array of preclinical studies have shown the efficacy and safety of total liquid ventilation (TLV) in pediatric and adult animal models [1]. In TLV, the lungs are totally filled with a breathable perfluorocarbon (PFC) liquid and then ventilated with a tidal volume of PFC controlled by a dedicated liquid ventilator [2]. TLV offers many advantages over conventional mechanical ventilation (CMV). By eliminating the air-liquid interface, it allows the recruitment of collapsed lung regions ensuring more homogeneous alveolar ventilation [1], an efficient lung lavaging effect [3] and the capability to induce ultra-fast mild therapeutic hypothermia [4].The liquid ventilator Inolivent-5 ( Figure 1) includes different control modes found on conventional ventilators [5], [6]. The use of a volume controlled ventilation during inspiration and of a pressure controlled ventilation during expiration to avoid airway collapsus [7] greatly improves the efficiency and the safety of the ventilator for clinical use [8]. The positive end-inspiratory pressure (PEIP) and the positive endexpiratory pressure (PEEP) are critical parameters for the control of the ventilation. They are static pressure measurements reflecting the alveolar pressure and are related to the endinspiratory (EI) and end-expiratory (EE) lung volume. Usually,

Total liquid ventilation (TLV) is an experimental mechanical ventilation technique where the lungs are completely filled with a perfluorocarbon liquid (PFC). It can be used to implement moderate therapeutic hypothermia (MTH) and treat severe respiratory problems. During TLV, the airway pressure must be monitored adequately to avoid overpressure and airway collapses. On the thermodynamic level, rectal, esophageal or tympanic temperature measurements are not suitable (long time constant) to avoid lowering the heart below 30°C. The objective was to design a Y connector positioned at the mouth which integrates the virtual sensors, used by controllers. The first estimates the airway pressure and the second provides the core body temperature. Pressure and RTD sensors were installed in the connector to implement the virtual measurements. In-vitro experiments were done to validate the virtual sensors. In-vivo experiments (on newborn lambs) confirm the accuracy of the airway pressure estimation and of the systemic arterial temperature.