Five-year native liver survival analysis in biliary atresia from a single large Chinese center: The death/liver transplantation hazard change and the importance of rapid early clearance of jaundice.

PURPOSE

This study aimed to analyze the change of death/liver transplantation hazard and biochemical indexes over time after Kasai procedure (KP) based on a retrospective biliary atresia (BA) cohort, and to evaluate the predictive value of early jaundice clearance rate to 5-year native liver survival (NLS).

METHODS

A retrospective cohort with follow-up results of 139 BA patients from January 2009 to December 2012 was established, and the pre- and postoperative data were collected. NLS rates were estimated with Kaplan-Meier curves, and any differences between groups were tested by log-rank test. Hazard curve of death/liver transplantation was fitted with Weibull distribution, and hazards at certain time points were calculated. Trend charts of biochemical indexes were drawn to show any changes over time. Rate of jaundice clearance was indicated as the proportion of decreased total bilirubin level at a certain postoperative time point to preoperative total bilirubin level. In multivariate analysis for prediction of 5-year NLS, COX proportional hazard regression model was used and results were expressed as hazard ratios with 95% confidence intervals (CIs). The predictive value of early jaundice clearance rates for 5-year NLS was analyzed by receiver operating characteristic (ROC) curve, and a cut-off value of 4-week jaundice clearance rate was determined.

RESULTS

The estimated 5-year NLS rate of the 139 patients was 58.0%. The patients had a high hazard of death/liver transplantation early after KP, which gradually decreased and stabilized at a lower level 1 year later. Most death/liver transplantation events occurred within 1 year after KP. The total bilirubin (TBIL), direct bilirubin (DBIL) and total bile acid (TBA) levels of successful Kasai group decreased continuously after KP, and the biggest decline was seen in the first month. The aspartate aminotransferase (AST), alanine aminotransferase (ALT) and γ-glutamyl transpeptidase (GGT) levels increased during the first month after KP and decreased continuously thereafter. All the biochemical indexes of successful Kasai group tended to stabilize within/close to normal range 1 year after KP. On the contrary, all the biochemical indexes of failed Kasai group fluctuated at obvious abnormal levels after KP. The estimated 5-year NLS rates of successful Kasai group and failed Kasai group were 90.1% and 10.7% (p = 0.000). The most significant clinical protective factor of 5-year NLS was 4-week jaundice clearance rate, revealed by COX proportional hazard regression model, and the HR was 0.089 (95%CI 0.018-0.432, p = 0.003). In predicting 5-year native liver survival, the largest area under ROC (AUROC) curve belonged to 4-week jaundice clearance rate, which was 0.731 (p = 0.000). A cut-off value of 0.457 was determined, with sensitivity 0.827, specificity 0.552, positive predictive value 0.720, and negative predictive value 0.696. NLS rates of patients divided by cut-off value showed significant statistical difference demonstrated by Kaplan-Meier curve and log-rank test (p = 0.000).

CONCLUSIONS

The 5-year death/liver transplantation hazard of biliary atresia patients reduces greatly and stabilizes 1 year after KP. A successful KP enables patients to achieve long-term stable normal biochemical indexes. A rapid clearance of jaundice is of great positive significance to 5-year NLS, and the 4-week jaundice clearance rate is of some predictive value.

TYPE OF STUDY

Prognosis study.

LEVEL OF EVIDENCE

Level II, retrospective study.