Serum alpha-1 proteinase inhibitor in advanced cancer: mass variants and functionally inert forms.

In 1984, we reported that while immunoreactive levels of serum alpha-1 proteinase inhibitor (API) increased significantly in nine patients with advanced solid tumors, the functional activity of the inhibitor, as measured by the serum trypsin inhibitory capacity, did not increase proportionately. This suggested that a portion of the circulating API was functionally inert. We have now assayed immunoreactive titers and trypsin inhibitory capacity of serum API of 49 patients with advanced carcinomas and 27 healthy controls. Immunoreactive levels of API (expressed as percentage of normal pooled serum which was taken as 100%) in cancer subjects were significantly elevated as compared to normals (mean +/- SE: 233 +/- 9.0% versus 102 +/- 2.0%, P less than 0.05). Although the trypsin inhibitory capacity of the cancer group (16.0 +/- 0.9 units/ml) was significantly elevated (P less than 0.05) as compared to normals (9.9 +/- 0.1 units/ml), this increase was less than that in the immunoreactive titer of API, suggesting the existence of functionally inert API in serum. The fraction of API which was functionally active in this group of cancer patients was 71.0 +/- 3.0% which was significantly less than the normal 98.0 +/- 2.0% (P less than 0.05). In 12 patients followed serially, both immunoreactive levels of API and the trypsin inhibitory capacity increased significantly at the time of clinical progression of disease. There was a significant correlation between increasing absolute granulocyte count and increasing trypsin inhibitory capacity (correlation coefficient 0.66; P less than 0.001). Neither disease progression nor increasing granulocyte count, however, was associated with increasing proportion of functionally inactive API. The inactive form of API had the same molecular weight as the native molecule as shown by gel permeation chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis/Western blot analysis of cancer sera. Therefore, the inactive form was not due to a complex between API and a tumor-derived protease or to proteolytic fragmentation of the native API. Elastase inhibitory capacity of cancer sera with subactive API was essentially identical with trypsin inhibitory capacity indicating that the active site methionine was not oxidized in the inert API. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis/Western blot analysis showed that both normal and cancer serum API existed as two mass variants, at Mr 58,000 and 56,000. Both variants formed complexes with elastase and were functionally active.