UNLABELLED: Longstanding cirrhosis has been recognized as a risk factor for the development of hepatocellular carcinoma in patients with autoimmune hepatitis (AIH). Thus, the accurate determination of cirrhosis is important for prognostication, decisions regarding treatment and monitoring of disease progression. The aim of this study was to identify independent predictors of cirrhosis and to develop a model for estimating cirrhosis in patients with type 1 AIH. METHODS: Using the training sample, consisting of 121 patients with type 1 AIH, we retrospectively examined independent predictors of cirrhosis and constructed a model for estimating cirrhosis. Validation was prospectively performed in the validation sample, consisting of 35 patients. RESULTS: Using a stepwise multiple linear regression analysis, three predictors of serum immunoglobulin A level, ratio of aspartate aminotransferase to alanine aminotransferase, and platelet count were elicited, and a model for estimating cirrhosis was determined as follows: risk score = -0.113 + 0.0006056 x immunoglobulin A (mg/dL) + 0.155 x ratio of aspartate aminotransferase to alanine aminotransferase - 0.007079 x platelet (x10(4)/mm(3)). In the training sample, the sensitivity and specificity were 90% and 83%, respectively, when patients presenting a risk score >/=0.20 were estimated to be cirrhotic. When this model was applied to the validation sample, the sensitivity, specificity, positive predictive value, negative predictive value and accuracy were 83%, 97%, 83%, 97% and 94%, respectively. CONCLUSION: It is suggested that this model could be useful for the estimation of cirrhosis in patients with type 1 autoimmune hepatitis.
UNLABELLED: Longstanding cirrhosis has been recognized as a risk factor for the development of hepatocellular carcinoma in patients with autoimmune hepatitis (AIH). Thus, the accurate determination of cirrhosis is important for prognostication, decisions regarding treatment and monitoring of disease progression. The aim of this study was to identify independent predictors of cirrhosis and to develop a model for estimating cirrhosis in patients with type 1 AIH. METHODS: Using the training sample, consisting of 121 patients with type 1 AIH, we retrospectively examined independent predictors of cirrhosis and constructed a model for estimating cirrhosis. Validation was prospectively performed in the validation sample, consisting of 35 patients. RESULTS: Using a stepwise multiple linear regression analysis, three predictors of serum immunoglobulin A level, ratio of aspartate aminotransferase to alanine aminotransferase, and platelet count were elicited, and a model for estimating cirrhosis was determined as follows: risk score = -0.113 + 0.0006056 x immunoglobulin A (mg/dL) + 0.155 x ratio of aspartate aminotransferase to alanine aminotransferase - 0.007079 x platelet (x10(4)/mm(3)). In the training sample, the sensitivity and specificity were 90% and 83%, respectively, when patients presenting a risk score >/=0.20 were estimated to be cirrhotic. When this model was applied to the validation sample, the sensitivity, specificity, positive predictive value, negative predictive value and accuracy were 83%, 97%, 83%, 97% and 94%, respectively. CONCLUSION: It is suggested that this model could be useful for the estimation of cirrhosis in patients with type 1 autoimmune hepatitis.