Claus Niederau1, Stefan Lange, Martin Frühauf, Andreas Thiel. 1. Department of Internal Medicine, Katholische Kliniken Oberhausen gGmbH, St. Josef Hospital Oberhausen, Academic Teaching Hospital of the University of Duisburg-Essen, Oberhausen, Germany. claus.niederau@st-josef.de
Abstract
BACKGROUND/AIMS: Although physicians have looked for cutaneous signs of liver disease for more than a century, their prognostic value has never been evaluated systematically. METHODS: Therefore, cutaneous changes were prospectively recorded in all patients referred for liver biopsy from June 2000 to May 2004. Fibrosis was staged from F0 to F4 according to Desmet and Scheuer. The analysis included 744 patients, 520 of whom had chronic hepatitis C while the remaining had other diseases. RESULTS: By univariate analysis, the frequency of several skin changes was associated with the degree of fibrosis. In general, at fibrosis F0-1 skin changes were infrequent; they became more frequent at F2 and were frequent at F3-4. To analyse the predictive value of skin changes, patients with fibrosis F0-2 were compared with those with F3-4. Final logistic regression included spider naevi, palmar erythema, teleangiectasia, bleeding signs and dry skin as well as age and gender. When routine laboratory values were included in the analysis, prothrombin time, gamma-glutamyltransferase and albumin proved to be significant. Receiver operating characteristic (ROC) showed a good discrimination of fibrosis F0-2 from F3-4 by the modelled score on combining skin changes and laboratory tests: at the cost of 2% of non-diagnosed patients with F3-4, one might have saved 60% of biopsies. ROC was less useful in discriminating fibrosis F0-1 from F2-4. The discriminative power of skin changes was better than the laboratory values and the aspartate aminotransferase/platelet ratio. CONCLUSIONS: The results prove that it is quite useful to look for skin changes in patients with liver disease.
BACKGROUND/AIMS: Although physicians have looked for cutaneous signs of liver disease for more than a century, their prognostic value has never been evaluated systematically. METHODS: Therefore, cutaneous changes were prospectively recorded in all patients referred for liver biopsy from June 2000 to May 2004. Fibrosis was staged from F0 to F4 according to Desmet and Scheuer. The analysis included 744 patients, 520 of whom had chronic hepatitis C while the remaining had other diseases. RESULTS: By univariate analysis, the frequency of several skin changes was associated with the degree of fibrosis. In general, at fibrosis F0-1 skin changes were infrequent; they became more frequent at F2 and were frequent at F3-4. To analyse the predictive value of skin changes, patients with fibrosisF0-2 were compared with those with F3-4. Final logistic regression included spider naevi, palmar erythema, teleangiectasia, bleeding signs and dry skin as well as age and gender. When routine laboratory values were included in the analysis, prothrombin time, gamma-glutamyltransferase and albumin proved to be significant. Receiver operating characteristic (ROC) showed a good discrimination of fibrosisF0-2 from F3-4 by the modelled score on combining skin changes and laboratory tests: at the cost of 2% of non-diagnosed patients with F3-4, one might have saved 60% of biopsies. ROC was less useful in discriminating fibrosis F0-1 from F2-4. The discriminative power of skin changes was better than the laboratory values and the aspartate aminotransferase/platelet ratio. CONCLUSIONS: The results prove that it is quite useful to look for skin changes in patients with liver disease.
Authors: Thomas Göbel; Janine Schadewaldt-Tümmers; Lucas Greiner; Christopher Poremba; Dieter Häussinger; Andreas Erhardt Journal: World J Gastroenterol Date: 2015-01-21 Impact factor: 5.742