Modified interpretation criteria significantly improve performance of commercially available confirmatory assays for the serodiagnosis of Lyme borreliosis: a case-control study with clinically defined serum samples

Case-control study for the evaluation of innovative test formats for second-tier testing for the serodiagnosis of Lyme borreliosis (LB). A head-to-head comparison was performed with the test systems ViraStripe, SeraSpot, ViraChip, and recomBead. Serum samples from 62 patients (21 erythema migrans, 33 Lyme neuroborreliosis, 8 late LB) and 91 controls (including 29 potentially cross-reacting sera) were tested. For ViraChip and recomBead, optimised interpretation criteria were developed for both IgG and IgM. The most important modification for the proposed interpretation criteria for ViraChip is the interpretation of strong (> 2.5-fold above cutoff) singular IgG reactions against VlsE as positive. This significantly improves sensitivity (32 to 85%, p < 0.0001) without significant changes in specificity (borderline reactions interpreted as negative). By application of our modified rules, specificity of ViraChip IgM is significantly increased (89 to 97%, p < 0.05; borderline results included to negatives), and sensitivities of recomBead IgG and IgM are also significantly improved (69 to 87%, p < 0.01, and 57 to 74%, p < 0.01, respectively; borderline results included to positives). Further improvement of sensitivity by the rating of strong singular IgG reactions against VlsE as positive can also be shown for recomBead. IgG/IgM result combinations must be interpreted as a function of the assumed disease stage, and the best combinations differ for the various assays. Application of our proposed interpretation criteria significantly improve the discriminatory abilities of two assays; however, this must be confirmed with other data sets. Recommendations from Scientific Societies should be updated as may be necessary.