Leader: Laurent Dubuquoy
Alcoholic hepatitis is the most severe form of alcoholic liver disease that is associated with a high risk of premature death. Symptoms are not specific with hepatomegaly, moderate fever and recent jaundice. Thus, the diagnosis requires a histological confirmation, which is a clinical challenge for the patient and the clinician since it requires an invasive and painful sampling of a liver biopsy in a compromised patient (transjugular biopsy). This diagnosis is important since the symptoms are similar to those of decompensated cirrhosis while the patient management is significantly different between these two conditions.
In addition, corticosteroids are the only treatment that has shown efficacy in AH. It can improve the survival in about 60% of patients. The response to the corticosteroids is evaluated, after 7 days of treatment, by the Lille score, which combines several clinical and biological variables, including age, presence of renal insufficiency, bilirubin level and its evolution during the first days of treatment. In case of a Lille score <0.45, patients are responders to medical treatment with a survival at 6 months of about 85%. In case of non-response to corticosteroids (Lille score ≥0.45), the 6-month survival is only about 25%. Since AH patients are highly susceptible to infections it is important to stop the corticosteroids as early as possible if they are inefficient. It is therefore essential to find non-invasive biomarkers able to predict the response to corticosteroids and therefore improve significantly the patient management.
In this task, we will use the cohort “targetOH”, which we have developed in the hepatology department with more than 400 sera of patients with alcoholic cirrhosis without AH, patients with AH responding to treatment and patients with AH non-responding to treatment. The proteomic profile of these 3 groups of patients will be compared to control patient sera using a new high throughput MALDI-TOF-MS developed by E Heuson (Platform REALCAT, U Lille, France), in collaboration with Ch Flahaut (Institut Charles Violette, Lille, France). A preliminary experiment with 10 patients in each group already identify a modified spectrum for serum albumin specific for patient with ALD and other differentially expressed biomarkers, which remain to be fully characterized. Bioinformatic analysis are performed in collaboration with the Bilille Platform (https://wikis.univ-lille1.fr/bilille/).