Objective: To determine the background levels and specificity of antibody to Borrelia burgdorferi by Western blot (immunoblot) in an area nonendemic for Lyme disease, and to correlate antibody specificity with clinical or serologic findings.
Methods: In a prospective survey by consecutive sampling, serum was obtained from patients referred to a tertiary care referral center in a rural area of Pennsylvania not endemic for Lyme disease. A total of 207 consecutive referrals to a rheumatology clinic over a 3-month period from September, 1991 were divided into 3 groups. Those referred because of a positive Lyme serology (Group 1) were compared with patients having positive antinuclear antibodies or rheumatoid factor (Group 2) and with controls having no rheumatic complaints (Group 3).
Results: Antibody to at least one protein of B. burgdorferi was seen in over 40% of patients. Reactivities to the heat shock proteins and the 41 kDa flagellar antigen accounted for the majority of positive bands. There were no differences observed between the 3 groups, and no significant correlation between Western blot and ELISA findings in the absence of Lyme disease.
Conclusion: We conclude that significant levels of antibody to B. burgdorferi may be seen on Western blotting in patients who have not been exposed to this organism by clinical or epidemiologic criteria. The antibodies detected may be natural antibodies, or may result from exposure to homologous antigenic epitopes on other organisms. The definition of a positive Western blot in the diagnosis of Lyme disease should incorporate the background levels of reactivity seen in nonexposed populations. Criteria for positivity should focus on the presence of antibody to the more specific proteins of B. burgdorferi.