Near-real-time pulmonary shunt and dead space measurement with micropore membrane inlet mass spectrometry in pigs with induced pulmonary embolism or acute lung failure

Abstract

The multiple inert gas elimination technique (MIGET) using gas chromatography (GC) is an established but time-consuming method of determining ventilation/perfusion (VA/Q) distributions. MIGET-when performed using Micropore Membrane Inlet Mass Spectrometry (MMIMS)-has been proven to correlate well with GC-MIGET and reduces analysis time substantially. We aimed at comparing shunt fractions and dead space derived from MMIMS-MIGET with Riley shunt and Bohr dead space, respectively. Thirty anesthetized pigs were randomly assigned to lavage or pulmonary embolism groups. Inert gas infusion (saline mixture of SF6, krypton, desflurane, enflurane, diethyl ether, acetone) was maintained, and after induction of lung damage, blood and breath samples were taken at 15-min intervals over 4h. The samples were injected into the MMIMS, and resultant retention and excretion data were translated to VA/Q distributions. We compared MMIMS-derived shunt (MM-S) to Riley shunt, and MMIMS-derived dead space (MM-VD) to Bohr dead space in 349 data pairs. MM-S was on average lower than Riley shunt (- 0.05 ± 0.10), with lower and upper limits of agreement of - 0.15 and 0.04, respectively. MM-VD was on average lower than Bohr dead space (- 0.09 ± 0.14), with lower and upper limits of agreement of - 0.24 and 0.05. MM-S and MM-VD correlated and agreed well with Riley shunt and with Bohr dead space. MM-S increased significantly after lung injury only in the lavage group, whereas MM-VD increased significantly in both groups. This is the first work evaluating and demonstrating the feasibility of near real-time VA/Q distribution measurements with the MIGET and the MMIMS methods.