作者: G Castaldo , G Oriani , M M Lofrano , L Cimino , M Topa
DOI: 10.1093/CLINCHEM/42.8.1263
关键词:
摘要: We applied a multivariate analysis to large series of serum biochemical tests in an attempt identify function that could efficiently discriminate cirrhosis from hepatocellular carcinoma (HC). analyzed two successive temporal cohorts (1987-90; 1991-94) HC and cirrhotic patients, all histologically classified (first cohort: 69 39 HC; second 66 38 HC). Using data the first cohort we obtained discriminant based on seven analytes: alpha-fetoprotein, hepatic isoenzyme alkaline phosphatase, lactate dehydrogenase 5, total gamma-glutamyltransferase (GGT), GGT isoforms complexed with low-density lipoprotein, aspartate aminotransferase, copper. The same panel analytes emerged when was tested also both were together. In (total, 212 patients) prevalence vs approximately 2:1, correctly 93% cases, highest percentage correct classification diseases so far by laboratory approaches. Validation jackknife reallocation statistical algorithm confirmed these results. addition, six patients liver for whom had opportunity following up observing evolution HC, five as at diagnosis analysis; i.e., provided diagnostic lead time 6-12 months over histology. This function, easy-to-perform tests, may help solve fundamental problem differential chronic HC.