Predictors of survival in autoimmune liver disease overlap syndromes

Survival in patients with autoimmune liver disease overlap syndromes (AILDOS) compared to those with single autoimmune liver disease is unclear.

To investigate the survival of patients with AILDOS and assess the accuracy of non-invasive serum models for predicting liver-related death.

Patients with AILDOS were defined as either autoimmune hepatitis and primary biliary cholangitis overlap (AIH-PBC) or autoimmune hepatitis and primary sclerosing cholangitis overlap (AIH-PSC) and were identified from three tertiary centres for this cohort study. Liver-related death or transplantation (liver-related mortality) was determined using a population-based data linkage system. Prognostic scores for liver-related death were compared for accuracy [including liver outcome score (LOS), Hepascore, Mayo Score, model for end-stage liver disease (MELD) score and MELD incorporated with serum sodium (MELD-Na) score].

Twenty-two AILDOS patients were followed for a median of 3.1 years (range, 0.35-7.7). Fourteen were female, the median age was 46.7 years (range, 17.8 to 82.1) and median Hepascore was 1 (range, 0.07-1). At five years post enrolment, 57% of patients remained free from liver-related mortality (74% AIH-PBC, 27% AIH-PSC). There was no significant difference in survival between AIH-PBC and AIH-PSC. LOS was a significant predictor of liver-related mortality (P < 0.05) in patients with AIH-PBC (n = 14) but not AIH-PSC (n = 8). A LOS cut-point of 6 discriminated liver-related mortality in AIH-PBC patients (P = 0.012, log-rank test, 100% sensitivity, 77.8% specificity) (Harrell's C-statistic 0.867). The MELD score, MELD-Na score and Mayo Score were not predictive of liver-related mortality in any group.

Survival in the rare, AILDOS is unclear. The current study supports the LOS as a predictor of liver-related mortality in AIH-PBC patients. Further trials investigating predictors of survival in AILDOS are required.