Abstract Background Lyme borreliosis is the fastest growing vector-borne illness in North America, with recent estimates as high as 476 000 new cases annually in the U.S. alone. Measurement of Borrelia-specific antibodies via serologic testing remains an important diagnostic aid. Since 1994, serologic testing for Lyme disease in the United States has consisted of a two-stage algorithm, termed standard two-tiered testing (STTT). In the STTT algorithm, specimens equivocal/positive by first tier screening methodologies are subsequently tested via western blots. In recent years, the FDA has cleared several alternatives for replacing western blots with more sensitive second-tier tests, termed modified two-tiered testing (MTTT). We recently developed a VlsE1/pepC10 IgG/IgM chemiluminescent immunoassay (CLIA), to be used in conjunction with a new automated analyzer. The goal of these studies was to conduct a multi-site validation for this new assay and analyzer as a first-tier screening test within an STTT algorithm. Methods Retrospective cohort: 280 clinically characterized samples were provided from the U.S. Centers for Disease Control and Prevention (90 cases, 190 controls). The CDC provided VlsE1/pepC10 IgG/IgM ELISA data derived from their own testing. VlsE1/pepC10 IgG/IgM CLIA and western blotting data were also generated for this panel at ZEUS Scientific. Sensitivity and specificity values were calculated for the retrospective cohort as follows:VlsE1/pepC10-CLIA vs existing VlsE1/pepC10 IgG/IgM-ELISA, and VlsE1/pepC10-CLIA + Western blots, vs VlsE1/pepC10 IgG/IgM-ELISA + Western blots, interpreted as STTT algorithms. All testing of the retrospective cohort was performed at ZEUS Scientific. Prospective Cohort: 600 routine Lyme serology specimens were prospectively collected at three locations (1800 total): Mayo Clinic (MOC), Rochester MN.; Marshfield Clinic (MC), Stevens Point, WI; Health Network Laboratories (HNL), Allentown PA. The samples were assayed and analyzed in the same combinations as described above for the retrospective cohort; however, instead of sensitivity and specificity, the results were reported as positive percent agreement (PPA) and negative percent agreement (NPA) since the prospective samples were not clinically characterized. Testing of the prospective cohort was evenly split across three different laboratories, one of which was ZEUS Scientific. Results Retrospective cohort: [VlsE1/pepC10 CLIA, Sensitivity = (77/90) = 85.56%, Specificity = (177/190) = 93.16%; VlsE1/pepC10 ELISA, Sensitivity = (80/90) = 88.89%, Specificity = (179/190) = 94.21%], [STTT-VlsE1/pepC10-CLIA + Western blots, Sensitivity = (72/90) = 80.00%, Specificity = (190/190) = 100.00%; STTT-VlsE1/pepC10-ELISA + Western blots, Sensitivity = (70/90) = 77.78%, Specificity = (190/190) = 100.00%]. Prospective Cohort (1 sample out of the 1800 collected was omitted due to an invalid western blot result):[VlsE1/pepC10-CLIA vs VlsE1/pepC10-ELISA, PPA = (196/235) = 86.51%, NPA = (1452/1564) = 92.84%; STTT-VlsE1/pepC10-ELISA vs STTT-VlsE1/pepC10-ELISA, PPA = (160/167) = 95.81%, NPA = (1607/11 632) = 98.47%]. Conclusion These studies demonstrate that the newly developed VlsE1/pepC10 IgG/IgM CLIA, in conjunction with a new automated analyzer, yielded accurate results as a first-tier screening assay for both the retrospective and prospective multi-site validation cohorts. Future studies, utilizing the same sample cohorts, will be aimed at validating additional second tier CLIA tests as part of an MTTT algorithm.
Read full abstract