Rapid (0.5°C/min) minimally invasive induction of hypothermia using cold perfluorochemical lung lavage in dogs

Objective: Demonstrate minimally invasive rapid body core and brain cooling in a large animal model. Design: Prospective controlled animal trial. Setting: Private research laboratory. Subjects: Adult dogs, anesthetized, mechanically ventilated. Interventions: Cyclic lung lavage with FC-75 perfluorochemical (PFC) was administered through a dual-lumen endotracheal system in the new technique of 'gas/liquid ventilation' (GLV). In Trial-I, lavage volume (V-lav) was 19 ml/kg, infused and withdrawn over a cycle period (tc) of 37 s. (effective lavage rate V'-lav = 31 ml/kg/min.) Five dogs received cold (similar to4 degreesC) PFC; two controls received isothermic PFC. In Trial-II, five dogs received GLV at V-lav = 8.8 ml/kg, tc = 16 s, V'-lav = 36 ml/kg/min. Measurements and main results: Trial-I tympanic temperature change was -3.7 +/-0.6 degreesC (SD) at 7.5 min, reaching -7.3 +/-0.6 degreesC at 18 min. Heat transfer efficiency was 60%. In Trial-II, efficiency fell to 40%, but heat-exchange dead space (VDtherm) remained constant. Lung/blood thermal equilibration half-time was <8 s. Isothermic GLV caused hypercapnia unless gas ventilation was increased. At necropsy after euthanasia (24 h), modest lung injury was seen. Conclusions: GLV cooling times are comparable to those for cardiopulmonary bypass. Heat and CO2 removal can be independently controlled by changing the mix of lavage and gas ventilation. Due to VDtherm of <similar to>6 ml/kg in dogs, efficient V-lav is > 18 ml/kg. GLV cooling power appears more limited by PFC flows than lavage residence times. Concurrent gas ventilation may mitigate heat-diffusion limitations in liquid breathing, perhaps via bubble-induced turbulence. (C) 2001 Elsevier Science Ireland Ltd. All rights reserved.