Lithium-thermal double indicator dilution: a new method of extravascular lung water measurement in the critically ill?

Abstract

There is evidence to suggest that therapy targeted at normalising extravascular lung water volume (EVLW) can improve outcomes from critical illness. Indocyanine green-thermal double indicator dilution (ICG-thermal) is considered the clinical reference standard of EVLW volume measurement but is no longer commercially available. The accuracy and reliability of the only clinically available technology (single-thermal indicator dilution) has been questioned in several studies. This thesis incorporates two clinical studies and one laboratory study designed to assess the measurement of EVLW and intrathoracic blood volumes using a prototype lithium-thermal double indicator dilution technique. The proof of concept study suggested our hypothesis that intrathoracic blood volume (ITBV), which is required for the calculation of EVLW volume, could be determined using lithium indicator dilution, was valid. Peri-operative trends and absolute values of ITBV were consistent with those obtained using ICG-thermal in a similar patient study group. The median absolute value of ITBV measured at baseline using indocyanine green (1417 [±208] ml) was similar to that obtained using lithium indicator dilution 1542 (±601) ml. EVLW volume measured by three indicator dilution techniques was then compared to postmortem gravimetry in porcine models of acute lung injury. Sepsis and acute lung injury were associated with increased EVLW volume, (9.2 [±3.0] ml kg-1), compared to sham operated animals (6.6 [±0.45] ml kg-1) in keeping with previous studies. The Li-thermal (Bias-1.8 [±13.1] ml kg-1) and ICG-thermal (Bias-1.0 [±6.6] ml kg-1) techniques demonstrated acceptable accuracy, but wide limits of agreement suggested poor reliability. 4 The single-thermal technique systematically over-estimated EVLW, with unacceptably wide limits of agreement (Bias +8.5 [±14.5] ml kg-1). In this laboratory investigation, the double indicator methods appeared more reliable than the single-thermal technique. However none could be considered ideal. Results of the final clinical study suggested EVLW volume measurement in man with the Li-thermal method was clearly erroneous (Bias -7.6 [±7.4] ml kg-1) and compared poorly to simultaneous measurements made using the ICG-thermal method (Bias +13.2 [±14.4] ml kg-1). A considerable over-estimation of mean transit time (MTT) when compared to the ICG-thermal technique (Bias 12.8 [±13] sec) was observed, a likely consequence of using an external lithium ion electrode instead of an intra-arterial catheter. Manual analysis of the dilution curves suggested considerable variability when compared to the automated analysis. The poor accuracy of MTT, and consequently ITBV measurements in the clinical study, may partly be due to software analysis of the lithium dilution curves. Thoracic blood volumes derived from measurement of ICG transit time are reliable. However, EVLW calculations based on the thermal indicator transit time are likely to be inaccurate. The findings of these clinical and laboratory investigations demonstrate poor agreement between both the prototype Li-thermal and the single thermal measurements of EVLW volume and the ICG-thermal method. Trans-pulmonary lithium indicator dilution measurements of ITBV and EVLW volume using an external lithium ion electrode are not sufficiently accurate to safely guide clinical interventions in individual patients. Consequently we decided not to further develop the lithium-thermal technique of EVLW volume measurement.