Therapeutic plasma exchange in amatoxin associated acute liver failure-results from the multi-center Amanita-PEX study

Abstract

Background: Amatoxin-related acute liver failure (AT-ALF) carries high mortality without liver transplantation (LTX). While therapeutic plasma exchange (PEX) might improve LTX-free survival in other ALF cases, its role in AT-ALF is unclear. Clinical practice varies, and, given the rarity of this ALF entity, the feasibility of conducting a randomized controlled trial to investigate PEX in AT-ALF is more or less impossible.

Methods: The Amanita-PEX study is a multi-center, international, retrospective study analyzing patients with AT-ALF from 2013 to 2024. The primary outcome was 28-day LTX-free survival (composite endpoint: death or LTX) after ALF diagnosis.

Results: The study included 111 patients from 25 centers: 82 received standard-of-care (SOC), and 29 received at least one PEX-session. PEX and SOC-groups were comparable at baseline, but 76% of PEX- vs. 58% of SOC-patients developed hepatic-encephalopathy (HE) grade ≥ 2 (p = 0.021). While the primary outcome of 28-day LTX-free survival in all patients was not different between the SOC and PEX-groups, in the subgroup of patients with maximal HE grade ≥ 2, LTX-free survival was 19.1% (n = 8/42) in the SOC group, while it was 36.4% (n = 8/22) in patients receiving adjunctive PEX (Gehan-Breslow-Wilcoxon-p = 0.041, Log-Rank-p = 0.060). PEX was independently associated with reduced risk of the combined endpoint death or liver transplantation within 28 days from inclusion in patients with HE grade ≥ 2 (HR 0.37, 95%-CI 0.19-0.73, p = 0.004). After propensity-score-matching, LTX-free survival was 28% in the SOC- and 52% in the PEX group (Gehan-Breslow-p = 0.036; Log-Rank-p = 0.035).

Conclusions: In this real-world study, adjunctive use of PEX was associated with increased LTX-free-survival in patients with AT-ALF and HE grade ≥ 2.

Keywords: Amanita; Liver failure; Liver transplantation; Mushroom poisoning; Plasma exchange.

Fig. 1

Flow chart of study participants. Shown are pre-screening, screening, inclusion and analysis of patients. Included were patients with the diagnosis of acute liver failure following laboratory confirmed Amatoxin ingestion. The diagnosis of acute liver failure was made according to the European Association for the Study of the Liver (EASL) criteria in patients with severe acute liver injury, indicated by the presence of both coagulopathy (international normalized ratio [INR] > 1.5) and manifest hepatic encephalopathy, without pre-existing chronic advanced liver disease. The study compared standard of care (SOC) with SOC + therapeutic plasma exchange (PEX). This was a retrospective, multicenter, multi-national, real-world study

Fig. 2

Liver-transplant-free survival in the entire patient cohort as well as stratified for grade of hepatic encephalopathy. Liver-transplant (LTX)-free survival following fulfillment of acute liver failure (ALF) criteria is shown as Kaplan-Meier graphs for all patients (A), and stratified by the grade of maximum hepatic encephalopathy (HE) (B) for a 28-day follow-up. p-values < 0.05 were considered significant

Fig. 3

Influence of therapeutic plasma exchange on liver-transplant-free survival in the entire patient cohort. Liver-transplant (LTX)-free survival following fulfillment of acute liver failure (ALF) criteria is shown as Kaplan-Meier graphs for patients receiving standard of care (SOC) only and those receiving additional therapeutic plasma exchange (PEX) (A). Cumulative incidence function for LTX and death stratified by chosen therapeutic strategy (SOC vs. PEX) (B) and a multistate comparison of the cumulative incidence of LTX and death (C), are shown demonstrating comparable incidences of LTX and death in patients receiving SOC and PEX. Cumulative incidence function and multistate comparisons visualize the first occurring competing event of LTX or death, respectively. A corresponding multivariate cox-regression with hazard ratios (HR) and 95% confidence intervals (CIs) is shown as a forest plot with corresponding table (D). p-values < 0.05 were considered significant

Fig. 4

Influence of therapeutic plasma exchange on liver-transplant-free survival in patients with maximal hepatic encephalopathy grade ≥ 2. Liver-transplant (LTX)-free survival following fulfillment of acute liver failure (ALF) criteria is shown as Kaplan-Meier graphs for patients with maximal hepatic encephalopathy (HE) grade ≥ 2 receiving standard of care (SOC) only and those receiving additional therapeutic plasma exchange (PEX) (A). Cumulative incidence function for LTX and death stratified by chosen therapeutic strategy (SOC vs. PEX) (B) and a multistate comparison of the cumulative incidence of LTX and death (C), are shown demonstrating a lower combined incidence of LTX and death in patients with HE grade ≥ 2 receiving PEX. Cumulative incidence function and multistate comparisons visualize the first occurring competing event of LTX or death, respectively. A corresponding multivariate cox-regression with hazard ratios (HR) and 95% confidence intervals (CIs) is shown as a forest plot with corresponding table (D). p-values < 0.05 were considered significant

Fig. 5

Influence of therapeutic plasma exchange on liver-transplant-free survival in a propensity-score-matching analysis. A love plot showing the variables used for the propensity score matching (PSM) and their balance before and after PSM. All the variables, except for age, achieved proper balance (standardized mean difference < 0.1) after matching (A). Kaplan–Meier curve with Log-rank test for LTX-free survival of the cohort after PSM (B)

Fig. 6

Overall survival in the entire patient cohort and in patients with hepatic encephalopathy grade ≥ 2. Overall survival (OS) for 28-days following fulfillment of acute liver failure (ALF) criteria is shown as Kaplan-Meier graphs for all patients (A) and for patients additionally stratified by therapeutic strategy (standard of care (SOC) vs. adjunctive therapeutic plasma exchange (PEX) (B) in the entire cohort. Further, OS is shown for the subgroup of patients with maximal hepatic encephalopathy (HE) grade ≥ 2 (C) and for these patients additionally stratified by therapeutic strategy (SOC) vs. PEX) (D). p-values < 0.05 were considered significant