Suitability of control materials for determination of alpha-amylase activity.

The suitability of control materials for determination of alpha-amylase activity was assessed in comparison with reference groups of authentic human serum specimens containing alpha-amylase of either pancreatic or salivary origin, specimens from patients with no pancreatic pathology, and normal specimens to which porcine pancreatic alpha-amylase was added. After determination of alpha-amylase activity by 11 commonly used techniques (five different principles), the results were processed by both classical (linear representation, regression) and multivariate (correspondence analysis, principal-components analysis) statistical techniques. Specimens containing porcine pancreatic alpha-amylase did not behave like any of the other groups. We conclude that porcine enzyme should not be used for interlaboratory quality-control surveys or intermethod comparison studies. Determination of human salivary and pancreatic alpha-amylase showed intermethod biases similar to those for authentic patients' specimens. Human salivary alpha-amylase, both because of its behavior and its commercial availability, is a satisfactory source for alpha-amylase activity of quality-control specimens. The nature of the matrix (polyvinylpyrrolidone, albumin, delipidated serum, bovine serum, or human serum) little influenced the behavior of the specimens for any of the methods studied.