Measurement of liver function capacity in patients at high risk for liver failure is a standard procedure in our institution as a diagnostic tool that can routinely be performed at the bedside when a compromise of liver function is suspected. Since LiMAx can be assessed independently from consciousness it may also be applied for sedated patients. In the three presented cases during isoflurane sedation the measured values were unexpectedly low in the range of 1 to 71 µg/kg/h (Figure 2). We consider several mechanisms for this observation: Interaction of isoflurane with the test procedure itself, interference of isoflurane with CYP1A2 or temporary decrease in liver enzyme activity without liver cell damage (“liver hibernation”).
None of the patients had a history of liver disease (e.g. alcoholic liver disease, hepatitis, non-alcoholic fatty liver or cirrhosis) that could have explained reduced liver enzyme activity. Under certain conditions (i.e. shock of different etiology) the liver enzyme activity may be temporarily reduced. The "standard liver parameters" are quantified from the blood and indicate liver damage, however without reflecting the CYP1A2 enzyme activity. Extremely low LiMAx values were measured in the three patients equally after the start of isoflurane that spontaneously returned to previous values after discontinuation of the inhaled sedative. Meanwhile, neiter clinical nor laboratory findings suggested almost complete liver failure, which would be expected from such low LiMAx values [17].
To rule out toxic effects all administered drugs were scanned for a known interaction with CYP1A2 in the Flockhart Table [19] and for a high likelihood of hepatotoxicity in LiverTox [20]. With the exception of patient B, who received linezolid, none of the patients were treated with a listed drug during the relevant time (Figure 3). The hepatotoxic effect of the oxazolidinone antibiotic linezolid is well known. It is commonly associated with lactic acidosis and leads to an increase in transaminases. Lactic acidosis was not detected and an increase in transaminases was only observed at the time of the last LiMAx measurement, so that a serious impairment of liver function by linezolid at the time of CYP1A2 activity determination is not plausible. Therefore, true liver failure (pre-existing or toxic) as the cause for the significant decrease in LiMAx values seems quite unlikely.
The supply of 13C contained in food or infusions and oxygen have a major influence on the 13CO2:12CO2 ratio measurement [21, 22]. Genetic polymorphisms of CYP1A2 and smoking are supposed only to have a slight influence on the test results [22]. During the investigated period, food, infusions and oxygen supply were almost unchanged suggesting no influence on test results.
The LiMAx instruction manual does not mention a possible influence of isoflurane on test results.
Only few studies have investigated a possible interaction of isotope measurement with isoflurane. After measurement of several pig breath samples that contained isoflurane the quality control drifted further and more quickly from its known value. The authors hypothesized that isoflurane was adsorbed by the gas chromatography column in isotope ratio mass spectrometry (IRMS) [23]. In contrast, LiMAx technology uses laser-based spectrometry and was validated in five patients with total intravenous anesthesia to avoid interference by volatile anesthetics [24]. Ensle et al [25] measured intestinal glucose absorption with a 13C glucose breathing test. Two isoflurane sedated patients were excluded from the analysis because of suspected isoflurane interaction with the 13C measured values, although no excluded values were reported. One could expect decreased LiMAx values if isoflurane prevents the 13C-methacetin pre-hepatic transport from the blood stream to the liver. This, however, is unlikely since lipophilic isoflurane accumulates in the fatty tissue while only low concentrations are available in the blood. We conclude that there is no evidence of a relevant interaction between methacetin or 13CO2 and isoflurane.
Physiologically, isoflurane is almost inert in endogenous metabolism. It is oxidatively metabolized to a small extent (~ 0.2%) in the liver via CYP2E1 [19, 26] while an interaction of isoflurane and CYP1A2 activity has not yet been described [26, 27]. Even so, negative effects of isoflurane on the liver are well known. The liver damage is caused by trifluoroacetylation of liver proteins and a subsequent inflammatory reaction and rapid increase of transaminases termed hepatitis. This effect has been particularly well studied for halothane, but is also known for isoflurane [26]. If in fact isoflurane does compromise CYP1A2 activity genetic polymorphism could explain the differences in enzyme depression, which was less profound in patient C (-91%) as compared to the other patients (-98% and -99%, respectively) [28].
Finally, it remains unclear whether the apparent influence of isoflurane on the LiMAx test is substance-specific or whether other inhaled anesthetics cause similar effects.
To our knowledge, this is the first description of a profound decrease of LiMAx values during simultaneous isoflurane sedation, the reason for which remains unclear. Possible mechanisms are test-specific interactions, an isoflurane induced reduction in CYP1A2 activity or a temporary shut-down of enzymatic liver function without cell damage, suggesting a hibernating liver in the early phase of severe inflammatory disease. A more frequent measurement of Cyp1A2 activity would be desirable, to describe the development of enzyme activity more precisely. Due to the increasing use of both LiMAx and inhaled sedation in the ICU, we consider it important to further investigate this possible interaction in order to prevent erroneous measurements and possible wrong clinical decisions. We propose that, until the breath test on enzymatic liver function has been validated in the critical care setting, LiMAx-test results should be interpreted with caution in patients sedated with isoflurane.