Evaluation of an Immunochromatographic Assay for the Rapid and Simultaneous Detection of Rotavirus and Adenovirus in Stool Samples

Kim, Jayoung; Kim, Hyun Soo; Kim, Han-Sung; Kim, Jae-Seok; Song, Wonkeun; Lee, Kyu Man; Lee, Sunhwa; Park, Kyoung Un; Lee, Woochang; Hong, Young Jun

The overall agreement among the 4 methods for adenovirus detection was 85.5%. The overall agreement, positive agreement, and negative agreement of ICA compared with mRT-PCR were 94.0%, 71.4%, and 94.8%, respectively. A possible explanation for discordant results among different adenovirus assays in our study could be differences in the detectability of different adenovirus subtypes in stool samples. Table 4 shows the analytical reactivity of the 4 assays on adenovirus subtypes. ICA and ELISA could detect every adenovirus type tested, whereas the Seeplex DV assay (mRT-PCR) could detect only AdV 40 and 41. The Simplexa Adenovirus real-time PCR assay (real-time PCR) could detect AdV 1, 3, 4, 5, 8, 31, 40, and 41, but not AdV 2, 18, 23, or 37. The different results observed for ICA, ELISA, realtime PCR, and mRT-PCR could be because the antibodies used in the ICA and ELISA reagents could detect all types of adenovirus capsid antigens, whereas specific adenovirus types could be detected by using PCR according to the primers used. Twenty-eight samples were found to be positive for adenovirus when they were tested by using the Simplexa real-time PCR assay, but were found to be negative when tested by using the Seegene mRT-PCR. These samples might have contained adenovirus serotypes other than 40 and 41. In addition, the discordant results observed among the different assays could be explained by the differences in the ability of each assay to detect variable adenovirus burdens in stool samples; for example, PCR is more sensitive than antigen-based detection tests (ICA and ELISA). When we investigated the detection limit of ICA (Table 5) , we found that real-time PCR and mRT-PCR could detect rotavirus and adenovirus in diluted culture supernatants at a dilution of 1:1,024, whereas ICA could only detect rotavirus in 1:16 dilutions and adenovirus in 1:64 dilutions (data is not shown in Table 5 ). However, similar to the rotavirus results, we found that 15 stool samples were positive for adenovirus when tested by using ICA, but negative when tested by using ELISA, real-time PCR, and mRT-PCR ( Table 2 ); 14 of these 15 samples showed weak band intensities when tested by using ICA. This finding also suggests a high probability of false positive results by using ICA, especially when the positive band intensity of ICA is weak.