Performance of a Modified Two-Tiered Testing Enzyme Immunoassay Algorithm for Serologic Diagnosis of Lyme Disease in Nova Scotia

Abstract

Compared to the standard two-tiered testing (STTT) algorithm for Lyme disease serology using an enzyme immunoassay (EIA) followed by Western blotting, data from the United States suggest that a modified two-tiered testing (MTTT) algorithm employing two EIAs has improved sensitivity to detect early localized Borrelia burgdorferi infections without compromising specificity. From 2011 to 2014, in the Canadian province of Nova Scotia, where Lyme disease is hyperendemic, sera submitted for Lyme disease testing were subjected to a whole-cell EIA, followed by C6 EIA and subsequently IgM and/or IgG immunoblots on sera with EIA-positive or equivocal results. Here, we evaluate the effectiveness of the MTTT algorithm compared to the STTT approach in a Nova Scotian population. Retrospective chart reviews were performed on patients testing positive with the whole-cell and C6 EIAs (i.e., the MTTT algorithm). Patients were classified as having Lyme disease if they had a positive STTT result, a negative STTT result but symptoms consistent with Lyme disease, or evidence of seroconversion on paired specimens. Of the 10,253 specimens tested for Lyme disease serology, 9,806 (95.6%) were negative. Of 447 patients who tested positive, 271 charts were available for review, and 227 were classified as patients with Lyme disease. The MTTT algorithm detected 25% more early infections with a specificity of 99.56% (99.41 to 99.68%) compared to the STTT. These are the first Canadian data to show that serology using a whole-cell sonicate EIA followed by a C6 EIA (MTTT) had improved sensitivity for detecting early B. burgdorferi infection with specificity similar to that of two-tiered testing using Western blots.

Keywords: Lyme disease; modified two-tiered test (MTTT); serology; specificity.

FIG 1

Breakdown of LD serological testing and chart review results. PTLDS, posttreatment Lyme disease syndrome. *, three charts had no IB results. **, of the six individuals who had a positive IgM immunoblot but a negative IgG immunoblot, for two symptoms were ongoing for >30 days and one was diagnosed with optic neuritis and ultimately multiple sclerosis, all consistent with false-positive IgM results; of the remaining three, one patient had 3 days of chest pain and no other symptoms (not a typical presentation for LD), one patient had no clinical data associated with the sample, and one patient had 1 week of subjective fever without documentation of temperature but with no other symptoms. ***, two patients with negative IgG IBs did not have IgM tested, as symptoms were ongoing for >30 days.