Investigation of liver dysfunction: who should we test for hepatitis E?

AIM: Hepatitis E virus (HEV) is endemic in developed countries, but unrecognized infection is common. Many national guidelines now recommend HEV testing in patients with acute hepatitis irrespective of travel history. The biochemical definition of 'hepatitis' that best predicts HEV infection has not been established. This study aimed to determine parameters of liver biochemistry that should prompt testing for acute HEV. METHODS: This was a retrospective study of serial liver function tests (LFTs) in cases of acute HEV (n=74) and three comparator groups: common bile duct stones (CBD, n=87), drug-induced liver injury (DILI, n=69) and patients testing negative for HEV (n=530). To identify the most discriminating parameters, LFTs from HEV cases, CBD and DILI were compared. Optimal LFT cutoffs for HEV testing were determined from HEV true positives and HEV true negatives using receiver operating characteristic curve analysis. RESULTS: Compared with CBD and DILI, HEV cases had a significantly higher maximum alanine aminotransferase (ALT) (P<0.001) and ALT/alkaline phosphatase (ALKP) ratio (P<0.001). For HEV cases/patients testing negative for HEV, area under receiver operating characteristic curve was 0.805 for ALT (P<0.001) and 0.749 for the ALT/ALKP ratio (P<0.001). Using an ALT of at least 300 IU/l to prompt HEV testing has a sensitivity of 98.6% and a specificity of 30.3% compared with an ALT/ALKP ratio higher than or equal to 2 (sensitivity 100%, specificity 9.4%). CONCLUSION: Patients with ALT higher than or equal to 300 IU/l should be tested for HEV. This is simple, detects nearly all cases and requires fewer samples to be tested than an ALT/ALKP ratio higher than or equal to 2. Where clinically indicated, patients with an ALT less than 300 IU/l should also be tested, particularly if HEV-associated neurological injury is suspected.

The natural history of autoimmune hepatitis presenting with jaundice.

BACKGROUND: Forty percent of patients with autoimmune hepatitis (AIH) present with acute jaundice/hepatitis. Such patients, when treated promptly, are thought to have a good prognosis. OBJECTIVES: The objective of this study was to describe the natural history of AIH in patients presenting with jaundice/hepatitis and to determine whether the diagnosis could have been made earlier, before presentation. METHODS: This study is a retrospective review of 2249 consecutive patients who presented with jaundice to the Jaundice Hotline clinic, Truro, Cornwall, UK, over 15 years (1998-2013) and includes a review of the laboratory data over a 23-year period (1990-2013). RESULTS: Of the 955 patients with hepatocellular jaundice, 47 (5%) had criterion-referenced AIH: 35 female and 12 male, the median age was 65 years (range 15-91 years); the bilirubin concentration was 139 µmol/l (range 23-634 µmol/l) and the alanine transaminase level was 687 IU/l (range 22-2519 IU/l). Among the patients, 23/46 (50%) were cirrhotic on biopsy; 11/47 (23%) died: median time from diagnosis to death, 5 months (range 1-59); median age, 72 years (range 59-91 years). All 8/11 patients who died of liver-related causes were cirrhotic. Weight loss (P=0.04) and presence of cirrhosis (P=0.004) and varices (P=0.015) were more common among those who died. Among patients who died from liver-related causes, 6/8 (75%) died less than 6 months from diagnosis. Cirrhosis at presentation and oesophageal varices were associated with early liver-related deaths (P=0.011, 0.002 respectively). Liver function test results were available in 33/47 (70%) patients before presentation. Among these patients, 16 (49%) had abnormal alanine transaminase levels previously, and eight (50%) were cirrhotic at presentation. CONCLUSION: AIH presenting as jaundice/hepatitis was mainly observed in older women: 50% of the patients were cirrhotic, and liver-related mortality was high. Some of these deaths were potentially preventable by earlier diagnosis, as the patients had abnormal liver function test results previously, which had not been investigated.