The diagnosis of Lyme borreliosis should be made only after careful evaluation of the patient’s clinical history, physical findings, laboratory evidence and exposure risk evaluation. Exposure to ticks prior to disease manifestations is essential for the diagnosis of Lyme borreliosis, although recollection of a tick-bite may not always be present. Suitable case definition criteria, which can form the basis of standardised diagnoses are specified in the EUCALB case definitions (see case definitions). As a rule, microbiological findings are a major criterion for the clinical diagnosis of Lyme borreliosis but diagnosis should not be based on laboratory testing alone.
Serology
Detection of antibodies to B. burgdorferi is currently the laboratory method of choice in a routine clinical setting as the direct detection of the pathogen in clinical samples has a lower sensitivity. Serological examination is indicated in all cases of clinically suspected Lyme disease, i.e. when symptoms possibly associated with Lyme borreliosis are present, with the exception of erythema migrans (EM). Patients with a typical EM should be diagnosed clinically and treated promptly without serological testing, which is insensitive at this stage of disease. Negative serology does not exclude the diagnosis of early manifestations of Lyme borreliosis with short duration of symptoms and so repeat testing may be indicated.
Enzyme Immunoassay (EIA)
EIA is the most commonly used serodiagnostic screening method for Lyme borreliosis. Its major advantages are the relative ease of use in large-scale testing, coupled with the avoidance of the need for subjective interpretation. First generation tests were based on sonicate antigens. Second or third generation tests, with improved specificity, use either: semi-purified antigens; one or several purified antigens; recombinant antigens or synthetic peptides.
Immunoblot
Immunoblot is important for the characterisation of immune responses to a range of specific Borrelia burgdorferi s.l. proteins and is generally used in the second step of a two-tier testing procedure, the first test of which is a sensitiveEIA. The use of an immunoblot as a first line or stand alone test is not recommended.
Historically western blots, produced from whole cell lysates, were utilized. The interpretation criteria of Engstrom et al., (J. Clin. Microbiol. 1995, 33 419) was recommended for IgM blots and those of Dressler et al., (J. Infect. Dis. 1993, 167 392) for IgG but these rules were not universally applicable to immunoblots across Europe. Interpretation criteria depended on the strain or species used as source of antigen (Hauser et al., J. Clin. Microbiol 1997 35 1433).
Standardisation of criteria for interpretation of immunoblot in Europe was the subject of a study by EUCALB (Robertson et al.,2000, J. Clin. Microbiol, 38, 2097-2102). This multicentre study, involving six European laboratories using different immunoblot protocols, identified eight bands that were discriminatory in all the laboratories, though with variations in significance. From these bands, five closely related European rules were formulated giving acceptable sensitivity and specificity though there was no single rule that could be applied in all laboratories.
Commercial immunoblots, which utilize defined recombinant or synthetic antigens are now universally used. These are standardized, more reproducible, easier and less subjective to read than traditional western blots utilizing whole cell lysates. Typically, reactivity against at least two or more antigens is required for a positive interpretation of an IgG immunoblot whereas for IgM, OspC alone may be sufficient. The characteristic spectrum of bands, particularly in the IgG immunoblot, can also potentially provide evidence to distinguish the immune response of an early from a later stage infection. A narrow spectrum of bands with antibodies against early-phase antigens (e.g. VlsE, OspC) is typically compatible with an early manifestation. In contrast, a wide spectrum of bands, including late-phase antigens (e.g. p100, p17/18) fits well with late manifestations.
Two-tier testing strategies
Some Scandinavian countries recommend diagnosis by EIA alone, with an assay specificity of at least 95%, but preferably 98%. In other European countries (and also the USA) a two test procedure is recommended with the more sensitive EIA as a preliminary screening test followed by the more specific Immunoblot for confirmation (Standard Two-Tier Testing, STTT). Alternatively, a second EIA may be used (Modified Two-Tier Testing, MTTT) (CDC 2019).
A minimum standard of at least 90% specificity for the screening tests and 95% specificity for the immunoblot should be established in the population where the assay is to be used. The two-step approach may eventually be superseded by a single assay that has both sensitivity and specificity equal to or in excess of these values, but it is likely that immunoblots will still have a role in follow-up investigations.
Antibody detection
As a general rule, B. burgdorferi-specific antibodies are expected to develop in almost all patients within 6-8 weeks. IgM antibodies are relevant for detecting early infection, but their detection does not contribute to the serodiagnosis of late Lyme borreliosis (LB). In some individuals, an IgM antibody response can persist for months or even years after treatment or past infection, although this phenomenon is not associated with a (persistent) infection with B. burgdorferi. In patients with disease duration >6 weeks, a specific IgG response is a prerequisite, but an isolated IgM response is of no diagnostic relevance.
Diagnostic use of stand alone IgG assays may be sufficient if highly sensitive screening tests including VlsE antigen are used. In this instance, IgM detection appears to have no significant advantage over IgG testing in the recognition of early LB and may actually reduce the specificity of diagnostic testing. As a result, some countries have now stopped using IgM tests.
Detection of antibodies to B. burgdorferi cannot discriminate between active, latent or past infection and cannot be used to monitor treatment response.
CSF serology
When Lyme neuroborreliosis is suspected, the CSF should be examined for signs of inflammation and B. burgdorferi-specific antibody production. Specific antibodies in CSF are found earlier than in serum. For the diagnosis of early neuroborreliosis, it is necessary to demonstrate intrathecal antibody production, which requires simultaneously drawn blood and CSF samples.
Methods that take into account dysfunction of the blood-CSF barrier are suitable for the detection of intrathecal antibody production, for example determination of the CSF/serum index (specific antibody index [AI]). This method requires measurement of specific antibodies in the blood and in the CSF by a quantitative EIA and then determination of a quotient from the antibody units and the total IgG concentrations in CSF and serum.
The AI expresses the proportion of pathogen-specific IgG antibodies of the total IgG content in the CSF compared with the serum. This method can be also adapted for determination of intrathecal IgM antibody production. Determination of the AI is more specific than detection of antibody reactivity in either serum or CSF alone. However, an elevated AI may persist for several years post-infection even after successful treatment. Hence, confirmation of active Lyme neuroborreliosis requires the presence of additional signs of CSF inflammation such as pleocytosis.
CXCL13 is a B lymphocyte attractant chemokine that is detectable in high concentrations in the CSF of patients with definite Lyme neuroborreliosis, but rapidly declines during antibiotic therapy. However, CXCL13 may also be detected in some other diseases, such as neurosyphilis, human immunodeficiency virus (HIV) infection and lymphoma.
CXCL13 can be of use in very early cases of suspected Lyme neuroborreliosis (<2 weeks) if the AI is not yet positive, and to differentate an active from resolving disease.
Use of serology in different clinical manifestations of Lyme borreliosis
Erythema migrans (EM)
Specific IgG and/or IgM are found in only 40-60% of untreated cases, and particularly in patients with signs of haematogenous spread, e.g. multiple EM and extensive general symptoms. Early treatment and a superficial lesion can result in absence of a detectable specific antibody response. Serology for EM diagnosis is therefore not recommended unless there is clinical uncertainty.
Early neuroborreliosis
In early neuroborreliosis the finding of a moderate pleocytosis, almost invariably lymphocytic is important for the diagnosis, though some patients only have elevated CSF protein. Lymphocytic pleocytosis may be absent in the rare patients with isolated peripheral neuropathy. Intrathecal IgM and IgG production and oligoclonal IgG bands are common findings and are supportive of the diagnosis. CSF protein is often elevated while sugar levels are usually normal, although a few cases with low sugar have been described.
The failure to demonstrate intrathecal antibody production does not exclude acute neuroborreliosis in the case of short duration of symptoms. A significant change in titre of specific IgG and/or IgM between paired blood samples, one taken at the first visit and a second 3-6 weeks later are supportive for the diagnosis, but serum antibodies can be absent in early cases.
Lyme arthritis
For the diagnosis of Lyme arthritis, it is essential to demonstrate the presence of specific IgG antibodies (usually high levels). A positive IgM test in the absence of IgG antibodies argues against the diagnosis of Lyme arthritis. Follow-up is recommended only in cases with short duration of symptoms. Note that arthralgias may occur early during disease and should not be mistaken for Lyme arthritis.
Acrodermatitis chronica atrophicans (ACA)
For the diagnosis of acrodermatitis chronica atrophicans, it is essential to demonstrate the presence of specific IgG antibodies (usually high levels). IgM antibodies may or may not be present. A positive IgM test in the absence of IgG antibodies argues against the diagnosis of ACA.
Limitations of serology
Sensitivity
Antibody response in early Lyme borreliosis may be weak or absent, especially in erythema migrans and antibiotic treatment may abrogate antibody production. Serology may also be negative in acute neuroborreliosis with short duration of the disease. For patients with a short duration of symptoms a serological follow-up is recommended. Almost all patients with late manifestations have a positive IgG test. The diagnosis of so called “seronegative chronic Lyme disease” is highly speculative and requires further investigation.
Specificity
Specificity of serological assays may be affected by:
- Cut-off levels too low
- Cross-reacting antibodies, e.g. syphilis
- Oligoclonal stimulation, seen in some infections, e.g. EBV, CMV and Mycoplasma pneumoniae, can result in an increase of antibodies to a large number of antigens
- False positive reactions, predominantly IgM and mainly in sonicate antigen ELISA, caused by rheumatoid factor, some autoimmune diseases or disease of unknown origin, e.g. MS, SLE
- Leakage of serum antibodies caused by a damaged serum/CSF barrier, leading to antibodies in the CSF, not due to intrathecal antibody production
High seroprevalence of specific antibodies in the general population in high-endemic areas will cause the problem of relevance to clinical disease. Clinicians must take local seroprevalence into account when interpreting the clinical relevance of positive serology in patients. In addition, the less specific the symptoms, the weaker the probability of Lyme borreliosis and the lower the predictive value.
Minimum Standards
Serological diagnosis is always a balance between sensitivity and specificity of the assays, which depend on the choice of the cut-off level. Irrespective of methods used, a high level of specificity, i.e. a limited number of false positives, is always more important than a high level of sensitivity, since a lack of sensitivity can usually be compensated by rising antibodies in later samples. A minimum specificity of at least 90% is recommended for screening tests (EIA). The cut-off level giving such a specificity should be established in at least 100 serum samples from healthy donors. The same approach should be adopted for qualitative assays such as immunoblot.
A complete evaluation of a quantitative assay, e.g. a commercial kit, should include validation of the recommended cut-off level in the population where the assay is to be used, as described above. The performance of the assay should also be investigated with samples of patients with diseases known to cross-react with Lyme borreliosis, e.g. syphilis, or other infectious diseases (e.g. EBV and CMV) or disorders (e.g. rheumatoid factor) that are known to cause false positive reactions in IgM assays. The sensitivity of the assay should be established in samples from clinically confirmed cases of Lyme borreliosis at different stages.
Other laboratory diagnostic methods
Culture
Culture of spirochetes from patient material is still the gold standard for laboratory diagnosis of LB. However, due to the low numbers of viable B. burgdorferi spirochaetes often present in patient biopsies and the fastidious nature of many B. burgdorferi strains, negative results in culture do not always exclude active infection. The highest recovery rates of borrelia spirochaetes were reported for culture from skin biopsies of EM patients (up to 80%) and biopsies from ACA patients (60%). For cerebrospinal fluid a success rate of 17% has been reported. Borrelia spirochetes have been isolated from blood of patients with early LB, and from synovial fluid of patients with Lyme arthritis. However, blood, synovial fluid, and cardiac tissue are unrewarding sources for culture.
Molecular detection
Polymerase chain reaction (PCR) technology emerged in 1988 and has greatly assisted in the detection and identification of a wide range of fastidious pathogens. PCR methods are now available that are species-specific and can detect low copy numbers of B. burgdorferi s.l. Since PCR (in contrast to culture) only detects spirochaetal DNA but not viable organisms a positive PCR, although suggestive cannot unequivocally establish whether infections are active or not. Nevertheless, it is a particularly valuable tool in both the diagnosis and clinical management of patients with arthritis, since isolation of borreliae from synovial fluids is almost always unsuccessful and serology cannot distinguish between chronic active and past infections. In contrast PCR can detect borrelia DNA in >50% of synovial fluid samples and even higher levels of detection of DNA in synovial membranes can be achieved. In EM and ACA patients, borrelia DNA has been detected in 50-70% of the skin biopsies and rarely in serum. In acute neuroborreliosis patients, borrelia DNA was detected in 20% to 30% of the cerebrospinal fluid samples tested though in a few cases higher detection rates were found. The utility of urine-PCR has been investigated by several groups, but results are rather contradictory and urine PCR is therefore not recommended for routine diagnosis.
Molecular methods are now available that detect and identify individual genospecies of B. burgdorferi sensu lato. These methods either use species-specific primers for PCR, or analyse the PCR product by DNA sequencing, hybridisation to species-specific probes, or restriction endonuclease polymorphism. Mixed infections of 2 or 3 different species have been detected in European neuroborreliosis patients and patients with an EM or ACA. Despite the occurrence of mixed infections, the association of B. afzelii with ACA remains unchallenged. In neuroborreliosis about 60-70% of the strains belong to the species B. garinii and so far the majority of isolates from European Lyme arthritis cases have proved to be B. burgdorferi s.s., though several studies have reported genospecies heterogeneity in Lyme arthritis cases
Histology
In the diagnosis of cutaneous manifestations of Lyme borreliosis, histopathology can be a valuable complement:
Erythema migrans
The histological picture in erythema migrans (EM) is generally non-specific with some perivascular infiltration mainly of lymphocytes and sometimes plasma cells. Histopathological investigation is seldom required in EM but can be of help in atypical cases by differentiating EM from disorders with more characteristic features.
Borrelia lymphocytoma
Borrelia lymphocytoma is dominated by a lymphocytic infiltration in the dermis, but plasma cells, macrophages and eosinophils are also present. Histology may be supportive in the diagnosis of borrelia lymphocytoma but the features are not unique and may be difficult to differentiate from malignant lymphomas.
Acrodermatitis chronica atrophicans
The histopathological picture of acrodermatitis chronica atrophicans (ACA) is characterised by the presence of telangiectases and by a lymphocytic infiltration, which is mixed with plasma cell, in the dermis and often also in the subcutis. Histology in ACA may be supportive of diagnosis but it is not typical enough to be exclusive.
Peripheral blood
Lyme borreliosis is not characterised by any specific findings in peripheral blood. Most patients have normal or slightly elevated sedimentation rates and normal to slightly elevated white blood cell counts. Elevated cryoglobulin levels are occasionally found in active stages of the disease. Unlike relapsing fever spirochaetes, B. burgdorferi cannot be reliably detected in the blood by microscopy or PCR.
Lymphocyte transformation tests
Cellular tests such as lymphocyte transformation or activation tests should not be used until they have been adequately assessed. The available data do not support a high diagnostic accuracy. A recent study performed in the Netherlands evaluated cellular tests for LB and concluded that they had a low specificity compared with serological tests, leading to a high number of false positive tests results, and thus are unfit for clinical use at this stage (Baarsma et al 2022. Lancet Infect Dis 22(9):1388-1396).
Dessau RB, van Dam AP, Fingerle V, Gray J, Hovius JW, Hunfeld KP, Jaulhac B, Kahl O, Kristoferitsch W, Lindgren PE, Markowicz M, Mavin S, Ornstein K, Rupprecht T, Stanek G, Strle F. To test or not to test? Laboratory support for the diagnosis of Lyme borreliosis: a position paper of ESGBOR, the ESCMID study group for Lyme borreliosis. Clin Microbiol Infect. 2018 Feb;24(2):118-124. doi: 10.1016/j.cmi.2017.08.025. Epub 2017 Sep 5. PMID: 28887186.