This article is more than 5 years old. Information may no longer be current.

Alcoholic hepatitis prevalent in patients with cirrhosis, GI bleeding

Alcoholic hepatitis was prevalent in patients with gastrointestinal bleeding, and alcoholic cirrhosis, but survival was similar with and without gastrointestinal bleeding, according to study data.

“In our center, we systematically plan a liver biopsy in patients with alcoholic cirrhosis admitted for [gastrointestinal bleeding] and who present with a clinical suspicion of severe [alcoholic hepatitis],” the researchers wrote. “In our study, we found that ~80% of patients with alcoholic cirrhosis, presenting with [gastrointestinal bleeding] and [discriminate function greater than or equal to] 32, display histological features of [alcoholic hepatitis] … [though] patients with severe [alcoholic hepatitis] with and without [gastrointestinal bleeding] did not differ in prognosis.”

The researchers sought to determine the prevalence of histologically proven alcoholic hepatitis (AH) in patients with gastrointestinal bleeding (GIB), recent onset of jaundice and discriminate function (DF) greater than or equal to 32. They compared features and outcomes of AH in patients with and without GIB and assessed the application of the Lille model, which is used to determine response to therapy in patients with AH and GIB.

They reviewed data of 161 patients with alcoholic cirrhosis, recent onset of jaundice and discriminate function (DF) ≥ 32 admitted to a hospital in Paris, France between January 2005 and March 2011. All patients underwent liver biopsy to determine AH and GIB occurrence. The primary outcome was 1 month survival.

Of these patients, 86 had GIB and 75 did not. Overall, 105 patients, 58 with GIB and 47 without GIB, were included in the final analysis. Patients began treatment with corticosteroids at a median period of 5 days after a bleeding episode, according to the research.

Overall, the prevalence of AH was similar between patients with GIB and those without (77.3% vs. 81%), respectively, and had similar Child-Pugh and end-stage liver disease scores and DF. Six-month survival rates were also similar among the two groups; 73.9 ± 6.0% in GIB-positive patients vs. 69.9 ± 7% for GIB-negative patients (P = .49). Patients negative for AH and positive for GIB were less prone to infection compared with the GIB-negative patients (24.1% vs. 44.7%; P = .04). 

The area under the curve score for the Lille model predicted 6-month survival to be 0.71 ± 0.06 for all the patients and 0.74 ± 0.06 for patients without AH, but positive for GIB (P > .05).

“Corticosteroids were as effective in patients with GIB as without GIB. Therefore, corticosteroids should be started as soon as possible for patients with concomitant AH and GIB,” the researchers concluded. “The Lille model can be applied in this particular population of patients and can be used to guide corticosteroid therapy.” – by Melinda Stevens

Disclosure: The researchers report no relevant financial disclosures.