Acute Liver Injury and Failure
Andrew Junkin, MD
Joy J. Liu, MD
A 47-year-old woman with hypertension presents to the ED with 1 day of nausea, vomiting, anorexia, and right upper quadrant pain. She reports moderate to heavy drinking for many years, with a recent increase due to job-related stress. She is mildly tachycardic with epigastric tenderness and mild scleral icterus. Initial labs show WBC 16 k/µL, AST 110 units/L, ALT 35 units/L, alkaline phosphatase 80 units/L, TBili 4.4 mg/dL, PT 18.5 s, and INR 1.8. She is admitted for presumed alcoholic hepatitis and subsequently rules out for concomitant infection. Despite aggressive nutritional support, she continues to have severe disease based on her persistently elevated Maddrey discriminant function (MDF). You wonder if you should treat her with glucocorticoids.
Which patients with alcoholic hepatitis should be treated with glucocorticoids?
Patients with alcoholic hepatitis should be treated if the MDF score ≥32.
A 2015 network meta-analysis1 assessed the role of glucocorticoids in patients with alcoholic hepatitis. This study included adults with severe alcoholic hepatitis (defined as MDF score ≥ 32 and/or hepatic encephalopathy) and was comprised of 22 randomized controlled trials (2621 patients) with a primary outcome of 4-week mortality. Glucocorticoids, pentoxifylline, and N-acetylcysteine (NAC) administered alone or in
combination for ≥4 weeks were compared to each other or placebo. Compared to placebo, short-term mortality was lower among patients receiving glucocorticoids alone (RR 0.54, 95% CI 0.39-0.73; no P-value reported) or in combination with pentoxifylline (RR 0.53, 95% CI 0.36-0.78; no P-value reported) or NAC (RR 0.15, 95% CI 0.05-0.39; no P-value reported). Study caveats include heterogeneity in studies (I2 = 50% for corticosteroid versus placebo comparisons).
combination for ≥4 weeks were compared to each other or placebo. Compared to placebo, short-term mortality was lower among patients receiving glucocorticoids alone (RR 0.54, 95% CI 0.39-0.73; no P-value reported) or in combination with pentoxifylline (RR 0.53, 95% CI 0.36-0.78; no P-value reported) or NAC (RR 0.15, 95% CI 0.05-0.39; no P-value reported). Study caveats include heterogeneity in studies (I2 = 50% for corticosteroid versus placebo comparisons).
The American Association for the Study of Liver Diseases (AASLD) recommends that patients with MDF ≥32 receive a course of prednisolone (1A recommendation).2
Based on her elevated MDF and the absence of contraindications, you initiate therapy with prednisolone.
Another patient on your service, a 55-year-old woman with diabetes, hypertension, and obesity presented with 5 days of nausea and anorexia and 3 days of constant, progressive right upper quadrant abdominal pain without radiation. She is 1 month into isoniazid treatment for latent tuberculosis. Her initial labs are notable for AST 126 units/L, ALT 183 units/L, alkaline phosphatase 180 units/L, TBili 2.5 mg/dL, PT 16.9 seconds, and INR 1.6.
A broad initial workup for her acute liver injury (ALI) rules out a number of causes including infiltrative, autoimmune, inflammatory, vascular, and many viral diseases (hepatitis A, hepatitis B, hepatitis C, cytomegalovirus, and Epstein-Barr virus). You strongly suspect isoniazid-related drug-induced liver injury (DILI), a diagnosis of exclusion, but wonder whether you should test for hepatitis E virus despite a lack of obvious risk factors.
Is it reasonable to test for hepatitis E virus in ALI?
If initial testing for etiology of ALI is unrevealing, it is reasonable to test for hepatitis E infection even in the absence of obvious risk factors.
Testing for hepatitis E in cases of ALI is rarely positive but may result in changes in diagnosis and management. A prospective study3
of 318 suspected DILI patients evaluated hepatitis E serologies (IgM, IgG) in samples obtained at time of enrollment. Certain patients had reflexive polymerase chain reaction (PCR) testing. Nine (3%) tested positive for hepatitis E IgM, indicating acute hepatitis E infection, and four had hepatitis E viremia at the time of serum sample testing. Patients with hepatitis E IgM positivity were older (P < .001) and male (P < .003), and 2/9 patients had HIV. None had traditional risk factors, such as recent travel to endemic areas, recent contact with farm animals, or consumption of undercooked pork. A positive hepatitis E IgM changed the most likely diagnosis to hepatitis E in 7/9 cases on a repeat causality analysis by three independent reviewers. In 3/7 cases, a “definite” (>95% likelihood) diagnosis of acute hepatitis E infection-induced liver injury was made. This study is limited by the low number of positive cases but demonstrates that patients being evaluated for DILI could have acute hepatitis E as the true cause of ALI.
of 318 suspected DILI patients evaluated hepatitis E serologies (IgM, IgG) in samples obtained at time of enrollment. Certain patients had reflexive polymerase chain reaction (PCR) testing. Nine (3%) tested positive for hepatitis E IgM, indicating acute hepatitis E infection, and four had hepatitis E viremia at the time of serum sample testing. Patients with hepatitis E IgM positivity were older (P < .001) and male (P < .003), and 2/9 patients had HIV. None had traditional risk factors, such as recent travel to endemic areas, recent contact with farm animals, or consumption of undercooked pork. A positive hepatitis E IgM changed the most likely diagnosis to hepatitis E in 7/9 cases on a repeat causality analysis by three independent reviewers. In 3/7 cases, a “definite” (>95% likelihood) diagnosis of acute hepatitis E infection-induced liver injury was made. This study is limited by the low number of positive cases but demonstrates that patients being evaluated for DILI could have acute hepatitis E as the true cause of ALI.
Related posts:
Full access? Get Clinical Tree
