Performance of commercially 1 available serological screening tests for human T-cell 2 lymphotropic virus infection in Brazil

Abstract

ABSTRACT Serological screening for 24 HTLV-1 is usually performed using enzyme-linked immunosorbent assay, particle agglutination or chemiluminescence assay kits. Due to antigen matrix improvement entailing the use of new HTLV-antigens and changes in the format of HTLV screening tests, as well as newly introduced CLIAs, a systematic evaluation of the accuracy of currently available commercial tests is warranted. We aimed to assess the performance of commercially available screening tests for HTLV diagnosis. A diagnostic accuracy study was conducted on a panel of 397 plasma samples: 200 HTLV-negative, 170 HTLV-positive and 27 indeterminate under Western blotting analysis. WB-indeterminate samples (i.e. those yielding no specific bands for HTLV-1 and/or HTLV-2) were assessed by PCR and results were used to compare agreement among the commercially available ELISA screening tests. For performance analysis, WB-indeterminate samples were excluded, resulting in a final study panel of 370 samples. Three ELISA kits (Murex HTLV-1/2, anti- HTLV-1/2 SYM Solution and Gold ELISA HTLV-1/2) and one CLIA kit (Architect r- HTLV-1/2) were evaluated. All screening tests demonstrated 100% sensitivity. Concerning the HTLV-negative samples, SYM Solution and Gold ELISA kits had specificity values >99.5%, while the Architect r-HTLV-1/2 test presented 98.1% specificity, followed by Murex (92.0%). Regarding the 27 samples with WB-indeterminate results, after PCR confirmation, all ELISA kits showed 100% sensitivity, but low specificity. Accuracy findings were corroborated by Cohen’s Kappa, which evidenced slight and fair agreement between PCR analysis and ELISA tests for HTLV diagnosis. Based on the data, we believe that all evaluated tests can be safely used for HTLV-infection screening. Human T-cell lymphotropic virus 47 type 1 (HTLV-1) and type 2 (HTLV-2) were identified in 1980 and 1982, respectively (1, 2). Subsequently, HTLV-3 and HTLV-4 were discovered in 2005 (3, 4). It has been estimated that at least 10 million people harbor HTLV-1 worldwide (5). Large foci of this infection exist in Japan, Africa, the Caribbean Islands, Melanesia, Australia, the Mashhad area of northeastern Iran and South America (5–7). HTLV-1 is associated with, or causes, a broad range of inflammatory conditions and a severe proliferative disease (5, 8–14). HTLV-2 infection is endemic in native Amerindian populations in both North and South America, certain tribes of Pygmies in Africa and in intravenous drug users (IDUs) in urban areas around the world (15, 16). In contrast to HTLV-1, this type rarely is associated with neurological or lymphoproliferative disorders (17). HTLV-3 and HTLV-4 are restricted to Western Africa and have not yet been associated with any diseases (3, 4). Brazil, a country of 200 million inhabitants, has a population of 800,000 who potentially harbor HTLV-1, representing one of the largest endemic areas for the virus and its associated diseases anywhere in the world (5). The virus is disseminated throughout the country, with higher rates found in the Northeast and Northern regions compared with the South and Southeast (18, 19). HTLV-2 is present mainly in the North, among indigenous populations and in IDUs in urban centers (17). Achieving an accurate diagnosis of HTLV infection is a complex task. Serological screening for HTLV-1 is usually performed using enzyme-linked immunosorbent assay (ELISA), particle agglutination testing or chemiluminescence assay (CLIA) kits. The Brazilian Ministry of Health recommends the use of ELISA or particle agglutination tests as a screening protocol. Western blotting (WB) or immunoblot is used for confirmation, and polymerase chain reaction (PCR) is employed in the case of inconclusive confirmatory test results (20). Among the screening 71 options, ELISA is used most extensively due to an elevated level of automation, simplicity and low cost. ELISA performance depends on antigen composition and assay format (21–24). Tests providing low accuracy present a public health problem, as false-positive results can have a negative impact, not only economically due to the need for confirmation by WB, but also on individuals’ quality of life. In light of this scenario, we endeavored to conduct a systematic evaluation of the commercial screening test kits for HTLV diagnosis. Statistical tools were used to obtain a robust assessment of the performance of each molecule by determining the following diagnostic test parameters: sensitivity (probability of test being positive in the presence of infection) and specificity (probability of test being negative in the absence of infection).