Endotoxin-induced liver hypoxia: defective oxygen delivery versus oxygen consumption.

In vivo EPR was used to investigate liver oxygenation in a hemodynamic model of septic shock in mice. Oxygen-sensitive material was introduced either (i) as a slurry of fine particles which localized at the liver sinusoids (pO2 = 44.39 +/- 5.13 mmHg) or (ii) as larger particles implanted directly into liver tissue to measure average pO2 across the lobule (pO2 = 4.56 +/- 1.28 mmHg). Endotoxin caused decreases in pO2 at both sites early (5-15 min) and at late time points (6 h after endotoxin; sinusoid = 11.22 +/- 2.48 mmHg; lobule = 1.16 +/- 0.42 mmHg). The overall pO2 changes observed were similar (74.56% versus 74.72%, respectively). Blood pressures decreased transiently between 5 and 15 min (12.88 +/- 8% decrease) and severely at 6 h (59 +/- 9% decrease) following endotoxin, despite volume replacement with saline. Liver and circulatory nitric oxide was elevated at these times. Liver oxygen extraction decreased from 44% in controls to only 15% following endotoxin, despite severe liver hypoxia. Arterial oxygen saturation, blood flow (hepatic artery), and cardiac output were unaffected. Pretreatment with l-NMMA failed to improve endotoxin-induced oxygen defects at either site, whereas interleukin-13 preserved oxygenation. These site-specific measurements of pO2 provide in vivo evidence that the principal cause of liver hypoxia during hypodynamic sepsis is reduced oxygen supply to the sinusoid and can be alleviated by maintaining sinusoidal perfusion.