We set out to investigate not only whether the presence of underlying cirrhosis affects the survival rate of patients with acute hepatitis E, but also to elucidate the impact of ACLF development in HEV-infected patients with cirrhosis and CLD. In this observational cohort study of 74 Korean patients with acute HEV infection, mortality at 180 days was 9.5% (3.8% for patients without cirrhosis vs. 22.7% with cirrhosis, P = 0.013). The age- and sex-adjusted proportional-hazard model revealed that patients with underlying cirrhosis had a significantly increased risk of mortality compared to patients without cirrhosis (HR = 8.111). The development of ACLF (according to EASL and APASL criteria) had a significant effect on the 180-day mortality rate in patients with HEV infection.
A peculiar result of this study was the approximately 8-fold increase in risk of 180-day mortality in patients with cirrhosis compared to patients without cirrhosis. Previous studies conducted in Asia found a high 12-month mortality rate approaching 70% in HEV-infected patients with underlying CLD [9, 20]. To date, the impact of HEV infection on CLD/cirrhosis in Europe has not been studied as well as in Asian countries. Data from small European studies revealed that autochthonous HEV-infected patients (3 patients in the UK and 7 in France) with pre-existing CLD had an approximate mortality rate of 70% [21, 22]. However, recent data from 11 patients with decompensated CLD in UK/France showed three (27%) died within 180 days of presentation [13]. All studies investigating the mortality of HEV-infected patients with underlying CLD/cirrhosis in Europe and Asia had a very small sample size. In two Chinese studies, hepatitis E viral superinfection in patients with chronic hepatitis B resulted in more severe clinical outcomes [23, 24]. The epidemiology of HEV is changing in China, where the previously dominant genotype 1 has become much less common while the zoonotic genotypes 3 and 4 are now the most commonly observed in middle-aged Chinese men [25]. A very recent study with a large sample size of Chinese patients, including 56 patients with cirrhotic CLD, 47 with non-cirrhotic CLD, and 124 with no CLD, showed that superinfection with HEV in patients with cirrhotic CLD had a poorer outcome than HEV-infected patients with non-cirrhotic CLD or without CLD [26]. However, this study did not provide the mortality rate of patients with HEV infection. In our study of 22 and 35 patients with underlying cirrhosis and CLD, the 180-day mortality rate for hepatitis E was 22.7% and 14.3%, respectively. Patients with cirrhosis, but not those with CLD, had a significantly higher mortality rate than those without underlying liver disease.
The mortality of HEV-related ACLF varies widely in different Asian studies. Studies of HEV genotype 1 infection reported short-term mortality rates between 0–67% for cases of ACLF (most studies used the APASL-ACLF criteria) [11]. All case series on HEV-related ACLF patients in Europe also had a very small sample size (maximum sample size of 11 patients) [27]; moreover, these studies did not apply the ACLF criteria of EASL or APASL. Our study, investigating HEV-related ACLF by applying the EASL and APASL criteria, has a larger sample size than previous European studies. Our study revealed a higher 180-day mortality rate in ACLF patients diagnosed by EASL (57.1%) and APASL (28.6%) criteria than those without ACLF and without underlying liver disease. These results are comparable to the high mortality rate of 15–25% reported in HEV-infected pregnant women [28]. Our findings suggest that both worsened pre-existing liver status and the development of organ failure play an important role in the prognosis of patients with acute HEV infection. The PREDICT study, a European, multicenter, prospective, observational study, showed that established bacterial infection and severe alcoholic hepatitis accounted for almost all cases of AD and ACLF [29]. In our study, ACLF was related to bacterial infection in 2 patients (28.6%) and alcoholic hepatitis in 3 patients (42.9%).
Although acute hepatitis E has been considered a rare cause of acute viral hepatitis in Korea, the reported IgG anti-HEV seroprevalence data (17–27%) suggest that the prevalence of HEV infection may be underestimated [30]. In a previous study conducted in the southeastern region of Korea, we reported that acute hepatitis E is no longer a rare cause of acute viral hepatitis. However, there have been no reports on mortality of patients with acute hepatitis E in Korea. We found that the mortality rate of acute hepatitis E was low in the general population without underlying liver disease, but high in patients with cirrhosis, and particularly in those with ACLF.
Our study was not without its limitations. First, clinical symptoms of acute hepatitis and the presence of IgM anti-HEV (not HEV-PCR) were used to diagnose acute hepatitis E in most cases. Another limitation of our study is the retrospective nature of the data collection and associated bias. However, this is a relatively large-scale study that systematically reported the association between mortality and the development of ACLF using both EASL and APASL criteria in patients with cirrhosis/CLD.