Purpose
To determine whether routine serologic screening for Borrelia burgdorferi and subsequent aqueous or vitreous humor analysis is useful in patients with uveitis.
Design
Cross-sectional study.
Methods
All patients referred to our tertiary uveitis referral clinic in the period of from January 1, 2004 to October 31, 2014, in whom routine serologic screening for Borrelia burgdorferi (IgG as determined by enzyme-linked immunosorbent assay and confirmed by immunoblot) was performed were retrospectively reviewed. In patients with an unclassified uveitis, aqueous and vitreous humor and cerebrospinal fluid were also analyzed. Local antibody production was determined by Goldmann-Witmer coefficient calculation or polymerase chain reaction for B burgdorferi . The seroprevalence of B burgdorferi among patients with uveitis was compared to the general population.
Results
Borrelia burgdorferi screening was performed in 1126 uveitis patients (44.3% male, mean age 45.9 ± 19.6 years). The seroprevalence of B burgdorferi among uveitis patients was 3.7% (95% confidence interval 2.6%–4.8%) (n = 42) as compared to 5%–10% in the general Dutch population. Of these 42 patients, 14 (1.2% of all uveitis patients) had an unclassified uveitis, 7 of whom underwent aqueous humor (n = 5) or vitreous humor (n = 2) analysis and cerebrospinal fluid analysis (n = 2). None of the patients had local antibody production in either ocular or cerebrospinal fluid.
Conclusion
The prevalence of immunoblot-confirmed B burgdorferi IgG seropositivity in our uveitis patients is only slightly lower as compared to the general Dutch population. Intraocular antibody production and DNA was absent in all tested patients. These findings do not support routine serologic examination for Borrelia in uveitis patients.
Borrelia burgdorferi sensu lato ( B burgdorferi ) is a spirochete that is transmitted to humans through tick bites. Worldwide, 18 Borrelia genospecies have been described so far. In the Netherlands, B afzelii , B garinii , and, to a lesser extent, B burgdorferi sensu stricto are the most common genospecies.
Some genospecies have been correlated with Lyme disease, with a variety of manifestations and symptoms, commonly affecting the skin (erythema migrans, acrodermatitis chronica atrophicans, and lymphocytoma, mostly on the earlobe), the nervous system (meningoradiculoneuritis), joints (arthritis), and, to a lesser extent, the heart (atrioventricular conduction disorders). The pathogenesis of the disease is not entirely clear yet. So far, no B burgdorferi toxins or toxin-like proteins have been identified. The damage caused by an infection with B burgdorferi is thought to be largely the result of the human immune response.
In the last decades, ocular involvement of B burgdorferi is described in several cases with various clinical presentations including uveitis. A few cases were reported in which B burgdorferi DNA is isolated from the vitreous fluid or in which B burgdorferi is cultured from an iris biopsy. For this reason, serologic examinations for B burgdorferi antibodies have become the standard of care in many medical centers. The aim of this study is to determine whether routine serologic screening is useful in the diagnostic phase when a patient presents with a uveitis de novo.
Methods
A retrospective study was performed to evaluate all patients who were diagnosed with uveitis at the University Medical Center in Utrecht, The Netherlands between January 1, 2004 and October 31, 2014 and for whom a serologic B burgdorferi screening was performed. Screening consisted of determining B burgdorferi serum IgG and IgM by enzyme-linked immunosorbent assay (ELISA). Positive and borderline ELISA results were subsequently confirmed by immunoblot to exclude false-positive results. In case of indeterminate immunoblot results, a second serum sample was analyzed similarly 4–6 weeks later. A negative immunoblot or indeterminate immunoblot upon repeat was interpreted as negative. Treponema pallidum serology was determined to exclude cross-reactivity. Demographics including sex, age at testing for Borrelia antibodies, and anatomic classification of uveitis were derived from the medical records. In addition, medical history (tick bite and erythema migrans) and antibiotic treatment were noted. The serology results were derived from the laboratory database. Serologic screening for B burgdorferi was performed using the Enzygnost Borreliosis/IgM and Enzygnost Lyme link VlsE/IgG (Siemens Healthcare Diagnostics, Marburg, Germany) ELISA according to the instructions of the manufacturer. For immunoblotting, the recomLine Borrelia IgG and IgM assays (Mikrogen, Neuried, Germany) were used as recommended by the manufacturer, using the Autoblot 3000 (Bio-Rad Laboratories B.V., Veenendaal, The Netherlands). The intensities of the immunoblot bands are scored relative to the cutoff on a scale from 0 (no reaction) to 8 (very highly reactive). A sum of all scores equal to or larger than 7 is considered positive. Serum testing for T pallidum was performed using the T pallidum hemagglutination assay for IgG detection (Serodia TP.PA; Fujirebio Diagnostics Inc, Malvern, Pennsylvania, USA) and, from January 2010 onward, the Immulite 2000 Syphilis Screen Assay (Siemens Healthcare Diagnostics), and the nontreponemal Venereal Disease Research Laboratory assay (VDRL) for syphilis disease activity (VDRL; Oxoid, Hampshire, UK) and, from February 2012 onward, the RPR Reditest (Biokit; Oxoid, Barcelona, Spain). Borrelia IgG and IgM detection on cerebrospinal fluid was performed using the IDEIA Lyme Neuroborreliosis assay (Oxoid, Barcelona, Spain), according to the instructions of the manufacturer. Intraocular antibody production was determined by Goldmann-Witmer coefficient (GWC) assay similarly as described previously for Toxoplasma gondii using the Enzygnost Lyme link VlsE/IgG (Siemens Healthcare Diagnostics) ELISA. A GWC value exceeding 3 is considered an indication for intraocular antibody production. Borrelia polymerase chain reaction (PCR) analysis on ocular fluids was performed at the Laboratory of Public Health, Friesland, The Netherlands.
For patients who had had multiple serum B burgdorferi IgG examinations over time, the patient was considered positive if IgG seroconversion occurred within 1 year after the onset of uveitis. This time interval was chosen because the isotype switch from IgM to IgG can take several months in B burgdorferi infections. Patients with a diagnosis other than uveitis were excluded from the analyses. Only positive test results in which the result was confirmed by an immunoblot were considered to be positive. All borderline and indeterminate values were considered to be negative.
Results
A total of 1126 consecutive uveitis patients (44% male, mean age 45.9 ± 19.6) were evaluated. The anatomic classification of uveitis was as follows: anterior uveitis 456 (40.5%), intermediate uveitis 145 (12.9%), posterior uveitis 296 (26.3%), and panuveitis 229 (20.3%) ( Table 1 ).
| All Patients (N = 1126) | Seropositive Patients With Unclassified Uveitis (N = 14) | |
|---|---|---|
| Age, y (mean ± SD) | 45.9 ± 19.6 | 46.0 ± 17.4 |
| Sex, n (%) | ||
| Male | 496 (44.3) | 6 (42.9) |
| Female | 624 (55.7) | 8 (57.1) |
| Anatomic classification, n (%) | ||
| Anterior uveitis | 456 (40.5) | 3 (21.4) |
| Intermediate uveitis | 145 (12.9) | 2 (14.3) |
| Posterior uveitis | 296 (26.3) | 2 (14.3) |
| Panuveitis | 229 (20.3) | 7 (50.0) |
| Seropositive patients, n (%) | 42 (3.7) |
Of 1126 patients, 43 (3.8%) tested positive for B burgdorferi IgM (ELISA) and 72 (6.4%) for B burgdorferi IgG (ELISA). An immunoblot confirmation was performed in 67 patients from the IgG-positive group, of which 42 (3.7% of all patients) tested positive, and in 41 patients (3.6% of all patients) from the IgM-positive group, of which 20 (1.8% of all patients) tested positive. Six patients tested positive for both B burgdorferi IgM and IgG. Taken together, 56 of all uveitis patients (5.0%) had IgM or IgG or both against B burgdorferi , as confirmed by immunoblot.
In the 42 IgG-seropositive patients, the etiology of uveitis was unclassified in 14 patients (1.2% of all uveitis patients), of which 1 patient also had B burgdorferi IgM. None of the 14 patients had serum antibodies indicative of T pallidum infection. The remaining 28 patients had other established etiologies, of which the infections prevailed, as confirmed by aqueous humor analysis ( Table 2 ).
| Etiology | N | % |
|---|---|---|
| Unclassified | 14 | 33.3% |
| HSV/VZV | 7 | 16.7% |
| FHUS/rubella virus | 5 | 11.9% |
| Sarcoidosis | 4 | 9.5% |
| Toxoplasmosis | 3 | 7.1% |
| Postoperative uveitis | 2 | 4.8% |
| Syphilis | 2 | 4.8% |
| JIA | 1 | 2.4% |
| HLA B27+ | 1 | 2.4% |
| Cytomegalovirus | 1 | 2.4% |
| Intraocular lymphoma | 1 | 2.4% |
| Streptococcus pneumoniae | 1 | 2.4% |
| Total | 42 | 100% |
Of the 14 Borrelia -seropositive patients with unclassified uveitis ( Table 1 ), 3 remembered a tick bite and 2 reported erythema migrans. Five of the 14 patients were treated with antibiotics. Aqueous humor analysis was performed in 5 patients, vitreous fluid analysis in 2, and cerebrospinal fluid analysis in 2. In addition, PCR analysis on ocular fluid was performed in 4 patients. None of the ocular fluids tested positive, which makes a diagnosis of infectious uveitis caused by B burgdorferi unlikely. Cerebrospinal fluid samples were negative for both B burgdorferi IgG and IgM.
Results
A total of 1126 consecutive uveitis patients (44% male, mean age 45.9 ± 19.6) were evaluated. The anatomic classification of uveitis was as follows: anterior uveitis 456 (40.5%), intermediate uveitis 145 (12.9%), posterior uveitis 296 (26.3%), and panuveitis 229 (20.3%) ( Table 1 ).
