Continuous renal replacement therapy and its impact on hyperammonaemia in acute liver failure

Objective: Hyperammonaemia contributes to complications in acute liver failure (ALF) and may be treated with continuous renal replacement therapy (CRRT), but current practice is poorly understood. Design: We retrospectively analysed data for baseline characteristics, ammonia concentration, CRRT use, and outcomes in a cohort of Australian and New Zealand patients with ALF. Setting: All liver transplant ICUs across Australia and New Zealand. Participants: Sixty-two patients with ALF. Main outcome measures: Impact of CRRT on hyperammonaemia and patient outcomes. Results: We studied 62 patients with ALF. The median initial (first 24 h) peak ammonia was 132 μmol/L (interquartile range [IQR], 91–172), median creatinine was 165 μmol/L (IQR, 92–263) and median urea was 6.9 mmol/L (IQR, 3.1–12.0). Most patients (43/62, 69%) received CRRT within a median of 6 hours (IQR, 2–12) of ICU admission. At CRRT commencement, three-quarters of such patients did not have Stage 3 acute kidney injury (AKI): ten patients (23%) had no KDIGO creatinine criteria for AKI, 12 (28%) only had Stage 1, and ten patients (23%) had Stage 2 AKI. Compared with non-CRRT patients, those treated with CRRT had higher ammonia concentrations (median, 141 μmol/L [IQR, 102–198] v 91 μmol/L [IQR, 54–115]; P = 0.02), but a nadir Day 1 pH of only 7.25 (standard deviation, 0.16). Prevention of extreme hyperammonaemia (> 140 μmol/L) after Day 1 was achieved in 36 of CRRT-treated patients (84%) and was associated with transplant-free survival (55% v 13%; P = 0.05). Conclusion: In Australian and New Zealand patients with ALF, CRRT is typically started early, before Stage 3 AKI or severe acidaemia, and in the presence hyperammonaemia. In these more severely ill patients, CRRT use was associated with prevention of extreme hyperammonaemia, which in turn, was associated with increased transplant-free survival.