Publication Year: 2023
DOI:
https://doi.org/10.1016/S0168-8278(23)00638-4
Authors:
Remih K,
Hufnagel F,
Durkalski-Mauldin V,
Lee WM,
Su Z,
Rule J,
Krieg L,
Karkossa I,
Schubert K,
von Bergen M,
Fontana RJ,
Strnad P
Request IDs:
23373
Studies:
Acute Liver Failure Study Group: Adult Acute Liver Failure Study
Acute liver failure (ALF) indicates sudden hepatocellular dysfunction with coagulopathy and encephalopathy in patients without known liver disease. Liver transplantation represents the only effective treatment, but the decision for/against transplantation remains challenging. We used mass spectrometry to characterize day 1 proteomic profiles in adult ALF patients of varying aetiology to identify diagnostic and prognostic biomarkers. Method: Serum proteomic patterns were compared in 200 and 119 (discovery/validation cohort) adult patients with ALF, ~50% of them acetaminophen (APAP)-related, as well as in 30 healthy controls without liver disease. The former were randomly selected from admission samples (<48h) of the US ALF study group database. Non-survivors were defined as patients who passed away or required liver transplantation within 21 days. Ingenuity pathway analysis (Qiagen) was used to obtain mechanistic insights. Results: In the discovery cohort, 187 proteins were detected in ?70% of subjects and most differed between ALF cases and controls. The key altered pathways were IL-6 signalling, acute phase response and prothrombin activation. 158 proteins differed between 95 APAP and 105 non-APAP cases. Three proteins reproducibly discriminated between the groups (AUROCs > 0.9 in both cohorts) and were superior to other available markers. In the discovery cohort, 46 proteins significantly differed between 21-day survivors and non-survivors. The most significantly enriched pathways were activated immune and acute-phase response which coincided with a better outcome. In particular, higher alpha1-antitrypsin (SERPINA1) and leucine-rich alpha-2 glycoprotein 1 (LRG1) levels were associated with better prognosis in the overall group (see figure, panel A). In both cohorts, they constituted the best discriminators (AUROCs > 0.7) and were comparable to MELD score (see figure, panel B). Conclusion: Unbiased proteomics may help identify a panel of new diagnostic and prognostic biomarkers with biological plausibility in ALF