Clinical assessment of the electrophoretic separation of alkaline phosphatase isoenzymes.

Over a five-month period, using data from patients in whom alkaline phosphatase (ALP) isoenzyme studies were requested routinely, we compared actual clinical diagnoses with the predicted diagnoses based on the results of electrophoretic separation of ALP isoenzymes on cellulose acetate before and after heat treatment and on elevated enzymatic activity of gamma glutamyl transferase (GGT) and alanine aminotransferase (ALT) activity. ALP isoenzymes were interpreted on a qualitative basis (presence or absence of liver, bone, or other isoenzyme) by individual clinical pathologists. Overall, the consistency of agreement in 61 patients was 66% for GGT, 51% for ALP isoenzymes, and 21% for ALT. In 44 patients with definite diagnoses, the sensitivity and specificity, respectively, of each laboratory test for patients with liver disease were 88 +/- 5.7% and 64 +/- 14.5% (ALP isoenzymes); 88 +/- 5.7% and 91 +/- 8.6% (GGT); and 6 +/- 4.1% and 91 +/- 8.6% (ALT). In patients with bone disease, the sensitivity and specificity of ALP isoenzymes was 75 +/- 10.8% and 86 +/- 6.6%, respectively. The results indicate that isoenzymes as currently performed need to be improved through standardization of the interpretation of ALP isoenzyme patterns to establish uniformity of comments.