Using a commercially available enzyme-linked immunosorbent assay (ELIS
A) and an immunoblot assay (IB), we tested sera from 100 patients with
erythema migrans (EM) seen in 1991 at the Westchester County Medical
Center Lyme Disease Diagnostic Center. Convalescent-phase sera were av
ailable from 59 patients. Fifty-five patients had EM of < 7 days' dura
tion, 31 had EM of 7 to 14 days' duration, and 14 had EM of > 14 days'
duration. During the acute phase of infection, 35 patients had a posi
tive ELISA result and 43 had a positive IB result by the recently publ
ished criteria of Dressler et al. (F. Dressler, J. A. Whalen, B. N. Re
inhardt, and A. C. Steere, J. Infect. Dis. 167:392-400, 1993) for inte
rpretation of IB in patients with Lyme disease. A greater sensitivity
of IB was observed in patients with EM of <7 days' duration, as follow
s: 14 of 55 (25%) for IB versus 7 of 55 (13%) for ELISA (P = 0.144). S
era of all 14 patients with EM of > 14 days' duration were reactive by
both tests, as follows: 13 positive and 1 equivocal by ELISA and 12 p
ositive and 2 indeterminate by the IB. The band reactivity most freque
ntly observed in the IB was to the 41- and 25-kDa antigens, the latter
being the most frequent band observed in immunoglobulin M blots. Sero
conversion was observed in 74 and 64% of evaluable patients by ELISA a
nd IB, respectively, despite the use of antibiotic therapy.