We have found a hepatotrophic factor in plasma or sera of patients with fulminant hepatic failure and have purified human hepatocyte growth factor from plasma of these patients. In this study we developed an enzyme-linked immunosorbent assay with high specificity and sensitivity for human hepatocyte growth factor in human serum. This assay for serum human hepatocyte growth factor is a sandwich method consisting of three steps. The standard curve for human hepatocyte growth factor appeared to be linear in the range of 0.20 to 12.50 ng purified human hepatocyte growth factor/ml (2.35 to 147 pmol/L). The assay took about 4 hr. Serum human hepatocyte growth factor values in patients with fulminant hepatic failure measured by enzyme-linked immunosorbent assay showed a strong positive correlation with that by bioassay using rat hepatocytes in primary culture. The mean value of serum human hepatocyte growth factor for 30 normal subjects was 0.24 +/- 0.12 (S.D.) ng/ml; that for 23 patients with fulminant hepatic failure was 8.06 +/- 1.76 (S.E.M.) ng/ml- greater than 30 times greater than the mean value for normal subjects. Serum human hepatocyte growth factor levels in patients with acute hepatitis, chronic hepatitis and cirrhosis were found to be slightly higher than those in normal subjects, but only the increase in serum human hepatocyte growth factor of acute hepatitis patients was statistically significant. The enzyme-linked immunosorbent assay for serum human hepatocyte growth factor should prove useful for serum human hepatocyte growth factor level measurement in patients with various liver diseases.

In normal adult human tissues, sialosyl-Tn antigen, detected by monoclonal antibody TKH2, was uniformly found in the bronchus, uterus, salivary gland, palatine tonsil, testis, stomach, duodenum, and capillary endothelium of several organs. It was also sporadically found in the small intestine, appendix, colorectum, gallbladder, urinary bladder, skin, and esophagus. The antigen was absent in the other organs. Even in the organs showing positive findings, the antigen was observed only in the limited areas. In contrast, sialosyl-Tn antigen was expressed in a large number of adenocarcinomas in many kinds of organs. It was expressed in more than one half the adenocarcinomas of the pancreas, ovary, uterus, stomach, colorectum, and gallbladder, but not in hepatocellular carcinomas, renal cell carcinomas, and papillary carcinomas of the thyroid gland. Sialosyl-Tn antigen expression also was observed in intestinal metaplasia of the stomach and in transitional mucosa adjacent to the colorectal carcinoma, which are considered to be cancer-related lesions. These results indicate that sialosyl-Tn antigen is a useful tumor marker, especially in adenocarcinomas of the mucin-producing organs, and suggest that the regulation of sialosyl-Tn antigen synthesis in adenocarcinomas is different from that in normal tissues.

These findings suggest that the differences observed among ELISA test results may be due principally to differences between the profiles of antigen coated on plates for the assays, rather than to differences between antibodies in serum and urine. The urine-based ELISA (URINELISA H. pylori) developed in this study is very accurate and would be useful for screening H. pylori infection as an alternative to serum ELISAs.