High plasma concentration of myeloperoxidase in cirrhosis: A possible marker of hypersplenism

Plasma myeloperoxidase levels in patients with cirrhosis were compared with those in patients with chronic hepatitis and healthy controls by means of a specific radioimmunoassay for myeloperoxidase. The mean concentration of plasma myeloperoxidase in cirrhotic patients (309.1 ± 17.2 ng/ml, n = 41) was markedly higher than that in chronic hepatitis patients (222.6 ± 17.2 ng/ml, n = 21) (p < 0.01) and normal controls (219.5 ± 5.7 ng/ml, n = 50) (p < 0.01). Plasma myeloperoxidase showed good negative correlations with neutrocyte count (r = −0.32, p < 0.01), thrombocyte count (r = −0.40, p < 0.01), red blood cell count (r = −0.32, p < 0.01), serum albumin (r = −0.35, p < 0.01), and cholinesterase (r = −0.32, p < 0.02) and positive correlations with serum alkaline phosphatase (r = 0.49; p < 0.01) and lactate dehydrogenase (r = 0.31, p < 0.01) in patients with cirrhosis or chronic hepatitis. Among lactate dehydrogenase isozymes, a good positive correlation was seen between plasma myeloperoxidase and lactate dehydrogenase-2 (r = 0.40, p < 0.01) and lactate dehydrogenase-1 (r = 0.03, p < 0.02). Plasma myeloperoxidase was significantly higher in the cirrhotic and chronic hepatitis patients with splenomegaly (341.1 ± 19.4 ng/ml, n = 31) than in those without splenomegaly (217.4 ± 12.2 ng/ml, n = 29) (p < 0.01). We also examined the difference between plasma levels of myeloperoxidase in the portal and peripheral blood. In the patients with nonhepatic disease, we found a significant portalsystemic difference (portal myeloperoxidase, 221.5 ± 19.7 ng/ml, vs. systemic myeloperoxidase, 179.1 ± 10.8 ng/ml; n = 15) (p < 0.01), whereas there was no such difference in the cirrhotic patients (portal myeloperoxidase, 281.5 ± 15.2 ng/ml vs. systemic myeloperoxidase, 268.0 ± 28.2 ng/ml; n = 7). A significantly high concentration of plasma myeloperoxidase was evident in the case of esophageal varices, which is a sign of portalsystemic shunting in human beings. The results of our study imply that the release of myeloperoxidase from destroyed neutrocytes, in addition to its decreased hepatic clearance, are responsible for the high concentrations of plasma myeloperoxidase in cirrhotic patients. Therefore measurement of plasma myeloperoxidase may be useful in assessing hypersplenic function in liver disease. (HEPATOLOGY 1993;18:1377-1383.)