LETTER

Top Letter References

Serological testing for Lyme borreliosis using conventional enzyme-linked immunosorbent and immunoblot assays is hampered by poor correlation between test results and disease status. The high seroprevalence (around 10 to 20%) of Borrelia burgdorferi-specific antibodies in the general population in areas of high endemicity can confound interpretation of positive serology. Additionally, early Lyme disease patients often do not have a detectable antibody response. The recent introduction of a Lyme vaccine is likely to further complicate serological diagnosis. New approaches towards more reliable detection of Lyme borreliosis are urgently needed, as misdiagnosis remains a cause of controversy with both patients and clinicians.

View larger version (11K): [in this window] [in a new window]   FIG. 1.   IFN- production in whole blood from Lyme patients and control subjects stimulated with a B. burgdorferi cell sonicate. The cutoff (mean of control samples + 3 standard deviations) is indicated by a dashed line.

In individuals exposed to B. burgdorferi, an early and sustained specific T-cell response develops that often precedes a measurable antibody response (1). T-cell responses in human Lyme disease are characterized by a pronounced bias towards a Th1-type cytokine profile, especially gamma interferon (IFN-) (3, 4). We investigated whether Borrelia-specific T-cell stimulation in whole blood, and subsequent measurement of specific IFN- responses by enzyme immunoassay (EIA), can be used to discriminate between Lyme disease patients and healthy subjects. In this study, 24 volunteers from Connecticut (11 male and 13 female, 7 to 69 years of age [mean, 48.2]) with clinically diagnosed Lyme disease, all of whom had received antibiotic treatment, and 24 healthy volunteers (12 male and 12 female, 23 to 53 years of age [mean, 35.7]) from Australia, where Lyme disease is not endemic, were recruited into the study. Whole-blood aliquots (1 ml) from patients and control subjects were stimulated for 16 h with either 2.5 µg of B. burgdorferi B31 whole-cell sonicate per ml, saline buffer, or mitogen. Plasma samples were then harvested, and IFN- responses were measured by a rapid two-step simultaneous EIA (2). The mean IFN- response for the control subjects plus 3 standard deviations (0.14) was used as the cutoff for the assay. With this criterion, 16 (67%) of 24 Lyme patients treated for Lyme disease had a positive test result, while 23 (96%) of 24 control subjects were negative (P < 0.001 [Mann-Whitney U test]) (Fig. 1). No spontaneous IFN- production was observed in whole blood from any patient or control subject (mean optical densities [OD] for saline buffer alone were 0.039 and 0.038, respectively), while a response to mitogen was detectable at similar levels in both gorups (mean OD, 1.06 in patients and 1.1 in controls.)

The aim of this pilot study was to prove the concept of the whole-blood IFN- assay for diagnosis of B. burgdorferi infection. For an accurate estimation of sensitivity and specificity a larger study is needed, analyzing both infected and uninfected individuals from an area of endemicity, using B. burgdorferi culture as a "gold standard." However, the data from this pilot study supports the concept of using the whole-blood IFN- assay for diagnosis of B. burgdorferi infection. Future studies will analyze Borrelia-specific IFN- responses in patients with early Lyme disease presenting without erythema migrans and in suspected cases of late Lyme disease where antibody titers are below measurable levels. We propose that this rapid means of measuring the cellular immune response to B. burgdorferi can improve the diagnosis of Lyme disease, either alone or as an adjunct to current serological assays.