Diagnostic efficacy of tests for the detection of iron overload in chronic liver disease.

The value of tests for the detection of body iron overload was investigated in 8 patients with clinically manifest primary hemochromatosis, 12 patients with cirrhosis and iron overload and 20 patients with liver disease and low or normal iron stores. Iron overload was defined as the presence of stainable iron in more than 50% of hepatocytes in a liver biopsy specimen. The percentages of patients with a true-positive (abnormal) or true-negative (normal) result were: serum iron concentration 65%, transferin saturation 85%, serum ferritin concentration 78%, serum ferritin:serum glutamic oxaloacetic transaminase (SGOT) index 78%, percent iron absorption 58%, percent iron absorption in relation to serum ferritin concentration 80% and percent iron absorption in relation to serum ferritin:SGOT index 93%. The calculated predictive value of a normal test result for the exclusion of iron overload in patients with liver disease, a group with an assumed prevalence of iron overload of 10%, was 98% to 99% for transferrin saturation and serum ferritin concentration used alone and 100% for these measures used together; the predictive value of an abnormal result for the diagnosis of iron overload was less than 50% for all of the above measures used alone or in combination. Hence, in patients with an increased serum ferritin concentration or transferrin saturation, or both, determination of the hepatocellular iron content of a specimen from a percutaneous liver biopsy is required for the diagnosis of iron overload.