Doreen Engelberts

Permissive hypercapnia, involving tolerance to elevated Pa(CO(2)), is associated with reduced acute lung injury (ALI), thought to result from reduced mechanical stretch, and improved outcome in ARDS. However, deliberately elevating inspired CO(2) concentration alone (therapeutic hypercapnia, TH) protects against ALI in ex vivo models. We investigated whether TH would protect against ALI in an in vivo model of lung ischemia-reperfusion (IR). Anesthetized open chest rabbits were ventilated (standard eucapnic settings), and were randomized to TH (FI(CO(2)) 0.12) versus control (FI(CO(2)) 0.00). Pa(CO(2)) and arterial pH values achieved in the TH versus CON groups were 101 +/- 3 versus 44.4 +/- 4 mm Hg and 7.10 +/- 0.03 versus 7.37 +/- 0.03, respectively. Following left lung ischemia and reperfusion, TH versus control was associated with preservation of lung mechanics, attenuation of protein leakage, reduction in pulmonary edema, and improved oxygenation. Indices of systemic protection included improved acid-base and lactate profile, in the absence of systemic hypoxemia. In the TH group, mean BALF TNF-alpha levels were 3.5% of CON levels (p < 0.01), and mean 8-isoprostane levels were 30% of CON levels (p = 0.02). Western blot analysis demonstrated reduced lung tissue nitrotyrosine in TH, indicating attenuation of tissue nitration. Finally, preliminary data suggest that TH may attenuate apoptosis following lung IR. We conclude that in the current model TH is protective versus IR lung injury and mechanisms of protection include preservation of lung mechanics, attenuation of pulmonary inflammation, and reduction of free radical mediated injury. If these findings are confirmed in additional models, TH may become a candidate for clinical testing in critical care.

RATIONALE: Many authors have suggested that the mechanism by which atelectasis contributes to injury is through the repetitive opening and closing of distal airways in lung regions that are atelectatic. However, neither the topographic nor mechanistic relationships between atelectasis and distribution of lung injury are known. OBJECTIVES: To investigate how atelectasis contributes to ventilator-induced lung injury. METHODS: Surfactant depletion was performed in anesthetized rats that were then allocated to noninjurious or injurious ventilation for 90 min. MEASUREMENTS: Lung injury was quantified by gas exchange, compliance, histology, wet-to-dry weight, and cytokine expression, and its distribution by histology, stereology, cytokine mRNA expression, in situ hybridization, and immunohistochemistry. Functional residual capacity, percent atelectasis, and injury-induced lung water accumulation were measured using gravimetric and volumetric techniques. MAIN RESULTS: Atelectasis occurred in the dependent lung regions. Injurious ventilation was associated with alveolar and distal airway injury, while noninjurious ventilation was not. With injurious ventilation, alveolar injury (i.e., histology, myeloperoxidase protein expression, quantification, and localization of cytokine mRNA expression) was maximal in nondependent regions, whereas distal airway injury was equivalent in atelectatic and nonatelectatic regions. CONCLUSIONS: These data support the notion that lung injury associated with atelectasis involves trauma to the distal airways. We provide topographic and biochemical evidence that such distal airway injury is not localized solely to atelectatic areas, but is instead generalized in both atelectatic and nonatelectatic lung regions. In contrast, alveolar injury associated with atelectasis does not occur in those areas that are atelectatic but occurs instead in remote nonatelectatic alveoli.

Hypoventilation, associated with hypercapnic acidosis (HCA), may improve outcome in acute lung injury (ALI). We have recently reported that HCA per se protects against ALI. The current study explored whether the mechanisms of protection with HCA were related to acidosis versus hypercapnia. Because CO(2) equilibrates rapidly across cell membranes, we hypothesized that (1) HCA would afford greater protection than metabolic acidosis. We further hypothesized that (2) buffering HCA would attenuate its protection. Forty isolated perfused rabbit lung preparations were randomized to: control (normal pH, PCO(2)); HCA; metabolic acidosis; or buffered hypercapnia. After ischemia-reperfusion (IR) injury wet:dry ratio was greatest with control and buffered hypercapnia, and rank order of capillary filtration coefficient was: control approximately buffered hypercapnia > metabolic acidosis > HCA. Isogravimetric pressure reduction was greatest with buffered hypercapnia. Despite comparable injury, pulmonary artery pressure elevation was less with buffered hypercapnia versus control. In vitro xanthine oxidase (XO) activity depended on pH, not PCO(2). We conclude that: (1) HCA and metabolic acidosis are protective, but HCA is the most protective; (2) buffering HCA attenuates its protection; (3) buffering HCA causes pulmonary vasodilation; (4) because metabolic acidosis and HCA similarly inhibit in vitro XO activity, the differential effects cannot be explained solely on the basis of extracellular XO activity.

Relative hypoventilation, involving passively-or "permissively"-generated hypercapnic acidosis (HCA), may improve outcome by reducing ventilator-induced lung injury. However, the effects of HCA per se on pulmonary microvascular permeability (Kf,c) in noninjured or injured lungs are unknown. We investigated the effects of HCA in the isolated buffer-perfused rabbit lung, under conditions of: (1) no injury; (2) injury induced by warm ischemia-reperfusion; and (3) injury induced by addition of purine and xanthine oxidase. HCA (fraction of inspired carbon dioxide [FICO2] 12%, 25% versus 5%) had no adverse microvascular effects in uninjured lungs, and prevented (FICO2 25% versus 5%) the increase in Kf,c following warm ischemia-reperfusion. HCA (FICO2 25% versus 5%) reduced the elevation in Kf,c, capillary (Pcap), and pulmonary artery (Ppa) pressures in lung injury induced by exogenous purine/xanthine oxidase; inhibition of endogenous NO synthase in the presence of 25% FICO2 had no effect on Kf,c, but attenuated the reduction of Pcap and Ppa. HCA inhibited the in vitro generation of uric acid from addition of xanthine oxidase to purine. We conclude that in the current models, HCA is not harmful in uninjured lungs, and attenuates injury in free-radical-mediated lung injury, possibly via inhibition of endogenous xanthine oxidase.

During mechanical ventilation, lung recruitment attenuates injury caused by high VT, improves oxygenation, and may optimize pulmonary vascular resistance (PVR). We hypothesized that ventilation without recruitment would induce injury in otherwise healthy lungs. Anesthetized rats were ventilated with conventional mechanical ventilation (VT 8 ml/kg; respiratory frequency 40 per minute) and 21% inspired oxygen, with or without a recruitment strategy consisting of recruitment maneuvers plus positive end-expiratory pressure, in the presence or absence of a laparotomy. Additional experiments examined the impact of atelectasis on right ventricular function using echocardiography, as well as functional residual capacity and PVR. Lack of recruitment resulted in reduced overall survival (59% nonrecruited vs. 100% recruited, p < 0.05), increased microvascular leak, greater impairment of oxygenation and lung compliance, increased PVR, and elevated plasma lactate. Echocardiography demonstrated that right ventricular dysfunction occurred in the absence of recruitment. Finally, samples from nonrecruited lungs demonstrated ultrastructural evidence of microvascular endothelial disruption. Although such effects clearly do not occur with comparable magnitude in the clinical context, the current data suggest novel mechanisms (microvascular leak, right ventricular dysfunction) whereby derecruitment may contribute to development of lung injury and adverse systemic outcome.