P27 Deaths due to extrahepatic disease in patients with autoimmune hepatitis (AIH): is there an excess?

<h3>Introduction</h3> Excess mortality in patients with treated AIH has usually been ascribed to liver disease; however, excess cancer-related deaths have recently been reported. We therefore aimed to characterise this excess mortality further. <h3>Methods</h3> Single-centre study of patients with AIH (1999 IAIHG criteria) presenting since 01/01/1987 and followed until death, liver transplant, moving away or 31/12/2023. We consider case capture essentially complete. Cause of death was classified using clinical records and death certificates as: liver-related (including transplants), extrahepatic, or unclassified. Total expected deaths were calculated using age- and sex-specific regional population mortality data, adjusted for calendar year (1991 [for 1987–2002], 2013 [for 2003–2012] and 2022 [for 2013–2023]). Expected extrahepatic deaths were derived by subtracting rates due to liver disease (data available for 2013 and 2022, extrapolated to 1991) from total rates. Expected cancer-related deaths were calculated using population rates from 2013. Standardised mortality ratio (SMR) was calculated as observed/expected (O/E) deaths. Life expectancy values from the time of AIH diagnosis were obtained from age- and sex-specific regional life tables (also adjusted for calendar year). <h3>Results</h3> 496 patients with AIH were followed up from diagnosis for (median [range]) 9.5 (0–36) years; 91% received prednisolone and 86% azathioprine. 179 patients died or were transplanted: 56 from liver disease (11 transplanted), 121 from extrahepatic disease, 2 uncategorised. There were 45 cancer-related deaths (7 hepatoma, 1 cholangiocarcinoma, 37 extrahepatic). Independent predictors of extrahepatic death were older age, diabetes, and failure to normalise serum ALT; but not sex, ethnicity, cirrhosis, decompensation, deprivation index, smoking, alcohol excess, body mass index or cumulative prednisolone dose. In 107 of the 121 extrahepatic deaths, survival time following AIH diagnosis was shorter than corresponding life expectancy values (mean ± SD difference: 10 ± 11 years). SMR values significantly exceeded 1.0 (indicating excess mortality compared to population) for both all-cause death/transplantation and extrahepatic death (see table). Although ‘extrahepatic’ SMR was higher in patients with cirrhosis and with diabetes, it remained elevated even in those without. It still exceeded 1.0 even after reclassifying infection-related deaths in cirrhotic patients as liver-related. SMR also exceeded 1.0 for all (but not for extrahepatic) cancer deaths. <h3>Conclusions</h3> Patients with AIH have excess mortality due to extrahepatic disease.