Risk of transfusion-transmitted malaria: evaluation of commercial ELISA kits for the detection of anti-Plasmodium antibodies in candidate blood donors

BACKGROUND: Transfusion with Plasmodium-infected blood represents a risk for malaria transmission, a rare but severe event. Several non-endemic countries implement a strategy for the screening of candidate blood donors including questionnaire for the identification of at-risk subjects and laboratory testing of blood samples, often serology-based, with temporary deferral from donation for individuals with a positive result. In Italy, the most recent legislation, issued in November 2015, introduced the use of serological tests for the detection of anti-Plasmodium antibodies. METHODS: In the absence of a gold standard for malaria serology, the aim of this work was to evaluate five commercial ELISA kits, and to determine their accuracy (sensitivity and specificity) in comparison to immuno-fluorescence antibody test (IFAT), and their agreement (concordance of results). Serum samples from malaria patients or from subjects with malaria history (N = 64), malaria naïve patients with other parasitic infections (N = 15), malaria naïve blood donors (N = 8) and malaria exposed candidate blood donors (N = 36) were tested. RESULTS: The specificity of all ELISA kits was 100%, while sensitivity ranged between 53 and 64% when compared to IFAT on malaria patients samples. When tested on candidate blood donors’ samples, ELISA kits showed highly variable agreement (42–94%) raising the possibility that the same individual could be included or excluded from donation depending on the test in use by the transfusion centre. CONCLUSIONS: These preliminary results indicate how the lack of a gold standard for malaria serology must be taken into account in the application and future revision of current legislation. There is need of developing more sensitive serological assays. Moreover, the adoption of a unique serological test at national level is recommended, as well as the development of screening algorithms based on multiple laboratory tests, including molecular assays. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12936-019-2650-0) contains supplementary material, which is available to authorized users.