Microcirculatory reactivity in patients with decompensated liver cirrhosis

INTRODUCTION. Cirrhosis is tipically associated with a hyperdynamic pattern, similar to sepsis. Although microcirculatory disturbances are common in sepsis, they have not been well studied in cirrhosis.OBJECTIVES. To study the relationship between muscular microvascular function and the levels of mean arterial pressure (MAP) in critically ill patients with decompensated liver cirrhosis.METHODS. Thenar muscle oxygen saturation (StO2) and muscle tissue hemoglobin index (THI) were continuously measured by a tissue spectrometer (InSpectraTM Model 650, Hutchinson Technology Inc, MN) in decompensated cirrhotic patients. In 35 patients, 27 (77%) male, 56 [49-62] years, median APACHE II score on ICU admission 17 [16-20], and SOFA score on the day of enrolment 10 [8-13]; 23 patients (65%) were Child Pugh C. Vaso-occlusive tests (VOT) (upper limb ischemia induced by a rapid pneumatic cuff inflation around the upper arm) were performed for all of them within 24 hours from ICU admission. The following variables were recorded: StO2, THI, the slope of the decrease in StO2 during the occlusion (desc slope; %/sec), the slope of the increase in StO2 following the ischemic period (asc slope; %/sec), and muscle O2 consumption (nirVO2) expressing in arbitrary units using the following formula: - (Descending Slope) * [(Basal THI + Min THI/2)]. We also pre-defined patients' categories as those receiving or not vasopressor agents (VP) or in those having a mean arterial pressure (MAP) = or > 70 mmHg. Data are presented as median [ranges] and count (percentage).RESULTS. Of the 35 patients, 21 (60%) were in severe sepsis/septic shock and 22 (63%) patients were treated with vasopressors (noradrenaline or terlipressin). ICU mortality was 49%. Median MAP was 76 [68-81] mmHg. NIRS values showed high StO2 (84.3 [80-88]%, and normal THI values (10.7 [8.7-13]), altered desc slope (-0.19 [-0.25-(-0.14)] and asc slope 1.27 ([0.79-2.03]) values. There was no correlation between MAP (range: 50-107 mmHg) and NIRS variables in the global cohort of patients. Nevertheless, we found a weak although significant correlation between asc slope and MAP among patients without VP (r2 = 0.07; p = 0.04). In this subgroup of patients, StO2 was significantly lower in those (n=4) with MAP = 70 [50-70] mmHg when compared to the others (n=9) [72-107 mmHg] (78 [62-81] vs. 86 [81-90], p=0.02), with a trend also for lower asc slope (p=0.07). On the other hand, among patients with MAP > 70 mmHg, we found lower desc slope (-0.16 [-0.19/-0.14] vs. -0.25 [-0.28/-0.14], p=0.01) and nirVO2 (90 [71-104) vs. 143 [115-189], p =0.02) among those treated with VP when compared to those without VP.CONCLUSIONS. In decompensated cirrhotic patients, tissue oxygen saturation is related to MAP, unless they are treated with VP. The presence of shock and/or the use of vasopressors may influence microvascular reactivity and muscle oxygen consumption, especially when MAP is higher than 70 mmHg.