Evaluation of IgG-ELISA for the diagnosis of Schistosoma japonicum in a high prevalence, low intensity endemic area of China

Abstract

Antibody-based diagnostic methods for detecting infection with Schistosoma japonicum have been developed and integrated into the national control program in China; however, the utility of these methods compared with conventional coprological methods remains unclear. In two consecutive years, we compared the performance characteristics of Kato–Katz with a commercially available enzyme-linked immunosorbent assay (ELISA) that detects anti-egg antigen IgG antibodies in a high prevalence, low intensity village in China (1025 subjects in 2005 and 652 subjects in 2006). In comparison with Kato–Katz based on duplicate stool specimens, each read in triplicate, the sensitivity of IgG-ELISA was high, ranging from 79.3% to 87.4% but with a relatively low specificity of 38.9% to 53.5%. The positive predictive value ranged from 20.8% to 24.6% while the negative predictive value ranged from 93.1% to 94.4%. When analyzed as continuous variables, there was a poor correlation between EPG (eggs per gram feces) and antibody level in both years ( r 2005 = 0.23 and r 2006 = 0.41). We detected a trend toward reduced sensitivity at lower infection intensity as measured by Kato–Katz in 2005 ( P = 0.262) and 2006 ( P = 0.287). We evaluated changes in antibody levels and the prevalence of positive antibody in the cohort of subjects examined in both 2005 and 2006 ( n = 565). The prevalence of positive antibody but not the continuous antibody level, decreased in individuals who were uninfected at both time points or who transitioned from infected to uninfected as assessed by Kato–Katz. In this cohort, the distribution of antibody levels measured in 2006 among individuals who were positive by Kato–Katz in 2006 broadly overlapped with the distribution of antibody levels in individuals who were negative by Kato–Katz in both 2005 and 2006. Our results indicate fairly poor performance characteristics of the anti-egg antigen IgG-ELISA for the detection of active infection with S. japonicum in our community based sample and are in contrast with other reports based on more selected populations. The high prevalence but low intensity of S. japonicum in our study community reflects the evolving epidemiology of schistosomiasis in communities receiving intermittent treatment with praziquantel in China. We suggest marked caution in implementing anti-egg antigen IgG-ELISA based diagnosis for either individual level diagnosis or population-based targeting for national control programs.