A Ten Year Immunopersistence Study of Hepatitis E Antibodies in Rural Bangladesh

Abstract

Hepatitis E virus (HEV) is a major cause of acute, viral hepatitis in Southeast Asia. Several studies have suggested that antibody persistence after HEV infection may be transient, possibly increasing the risk of re-infection and contributing to the frequency of outbreaks in HEV endemic regions. The specific conditions under which antibodies to HEV are lost, or "sero-reversion" occurs, are poorly understood. Here, one hundred participants from population-based studies in rural Bangladesh were revisited in 2015, ten years after a documented HEV infection to examine long-term antibody persistence. Twenty percent (95% confidence interval: 12.0, 28.0) no longer had detectable antibodies at follow-up, suggesting that antibodies generally persist for at least a decade after infection in rural Bangladesh. Those who were seronegative at follow-up were generally younger at infection than those who remained positive (14.4 years versus 33.6 years, P > 0.0001). This age-dependent antibody loss could partially explain cross-sectional sero-prevalence data from South East Asia where children have reportedly low antibody prevalence. The results of this study provide new insight into the immunological persistence of HEV infection in a micronutrient deficient rural population of South Asia, highlighting the importance of age at infection in the ability to produce long-lasting antibodies against HEV.

Figure 1.

Cohort selection and follow-up of participants in a study of persistence of hepatitis E virus (HEV) antibodies, Matlab, Bangladesh, 2003–2015. HLI, hepatitis-like illness.

Figure 2.

Loss of antibodies to hepatitis E virus (HEV) among persons with a previous HEV infection (n = 100), by age group, Matlab, Bangladesh, 2015. All 100 participants were positive for HEV antibodies at baseline (2003–2005). Sample sizes were as follows: age <10 years, n = 14; age 10–19 years, n = 21; age 20–29 years, n = 17; age 30–39 years, n = 16; age 40–49 years, n = 20; age 50–59 years, n = 8; age ≥60 years, n = 4. Bars, 95% confidence intervals.

Figure 3.

Observed and expected prevalences of negativity for hepatitis E virus (HEV) antibodies (n = 1,009), by age group, accounting for different risks (risk ratios) of antibody loss, Matlab, Bangladesh, 2015. Observed antibody prevalence obtained from a 2003 study in Matlab (36) is indicated with a solid line. Expected antibody prevalence (dashed lines) was calculated from univariate Poisson regression with age at infection from the current follow-up study of antibody persistence (2015), assuming varying levels of population prevalence of HEV infection. All age groups were assumed to be equally likely to be infected with HEV. ■, 25% prevalence of HEV infection; ▲, 50% prevalence of HEV infection; X, 75% prevalence of HEV infection; ♦, 100% prevalence of HEV infection. Bars, 95% confidence intervals.