Adaptation of the Mayo primary biliary cirrhosis natural history model for application in liver transplant candidates

Abstract

The Mayo natural history model has been used widely as a tool to estimate prognosis in patients with primary biliary cirrhosis (PBC), particularly liver transplant candidates. We present an abbreviated model in which a tabular method is used to approximate the risk score, which may be incorporated in the minimal listing criteria for liver transplant candidates. Data used in the development and validation of the original Mayo model were derived from 418 patients with well-characterized PBC. To construct an abbreviated risk score in a format similar to that of Child-Turcotte-Pugh score, 1 to 3 cut-off criteria were determined for each variable, namely age (0 point for <38, 1 for 38 to 62 and 2 for >/=63 years), bilirubin (0 point for <1, 1 for 1 to 1.7, 2 for 1.7 to 6.4, and 3 for >6.4 mg/dL), albumin (0 point for >4.1, 1 for 2.8 to 4.1, and 2 for <2.8 g/dL), prothrombin time (1 point for normal and 2 for prolonged) and edema (0 point for absent and 1 for present). The intervals between these criteria were chosen in a way to enable a meaningful classification of patients according to their risk for death. This score is highly correlated with the original risk score (r = 0.93; P <.01). The Kaplan-Meier estimate at 1 year was 90.6% in patients with a score of 6. The abbreviated risk score is a convenient method to quickly estimate the risk score in patients with PBC. An abbreviated score of 6 may be consistent with the current minimal listing criteria in liver transplant candidates.