RESUMO
BACKGROUND: Initial symptoms of dengue fever are non-specific, and thus definite diagnosis requires laboratory confirmation. Detection of IgM against dengue virus (DENV) has become widely used for dengue diagnosis. Understanding the persistence of anti-DENV IgM in subjects after acute infection is essential in order to interpret test results correctly. Although the longevity of anti-DENV IgM has been vehemently investigated in symptomatic children, anti-DENV IgM persistence in adults and in asymptomatically infected people have seldom been reported. METHODS: We prospectively investigated 44 adults with detectable anti-DENV IgM in a serosurvey conducted in the 2015 dengue epidemic in Tainan, Taiwan. Among subjects within the cohort, 17 were classified to be symptomatic and 27 were asymptomatic. The enzyme-linked immunosorbent assay (ELISA) from Standard Diagnostic (SD) and Focus Diagnostic were used to detect anti-DENV IgM for specimens collected initially, at 6 and 12 months. Regression analyses were used to estimate the duration of anti-DENV IgM fell below the detectable level. Rapid dengue tests from Standard Diagnostics had been widely adopted to detect anti-DENV IgM in Taiwan during the 2015 dengue outbreak. As such, collected specimens were also evaluated with the SD rapid dengue test in parallel. RESULTS: Anti-DENV IgM was detectable in 70.5 and 46.2% of the 44 subjects at 6 months and 12 months by the SD ELISA, respectively, while 13.6 and 7.7%, respectively, by the Focus ELISA. There was no significant difference in anti-DENV IgM detection for the follow-up specimens between subjects with symptomatic and asymptomatic infections. The regression analysis estimated that anti-DENV IgM persistence fell to the undetectable level at 338.3 days (95% CI 279.7-446.9) by SD ELISA, while at 175.7 days (95% CI 121.9-221.1) by Focus ELISA. The detectable frequency of anti-DENV IgM by rapid tests was 86.4%, 68.2 and 35.9% at initial, 6 and 12 months, respectively. CONCLUSION: Anti-DENV IgM was found to persist much longer than previously thought, suggesting a necessity of re-evaluation of the use of anti-DENV IgM for both the diagnosis of dengue and serological surveillance, especially when large outbreaks have occurred in the preceding year.