The patient was a 35-year-old African American man without any significant medical history. He presented to our hospital, a 700-bed tertiary referral center, with a 2-week history of progressive numbness of the right thigh and a tingling sensation down his right leg. The numbness began after he scratched the area secondary to a focal pruritus. He denied trauma, lesions, or rashes.
One week before admission, he began to experience numbness and tingling of the left lower extremity. He denied weakness. Three days before admission, he began to have blurry vision in his left eye that became progressively worse. There was no associated discharge, pain, double vision, or photophobia. On the day of admission, he had difficulty with balance while walking and went to the emergency department. He denied fever, chills, weight loss, hearing loss, and bowel or bladder incontinence. About a week before admission, he had a headache that resolved and did not recur. The rest of review of symptoms was negative, although his wife noted new-onset erectile dysfunction. Upon further questioning, he recalled a hospitalization 1 year prior at an outside hospital for pneumonia. He took no medications. Social history was significant for unprotected sex with male and female partners 2 years earlier. He did not know his HIV status.
He was admitted to the neurology service, and by hospital day 3, he developed blurry vision in his right eye. He also noticed adecrease in hearing on the right side without associated tinnitus. Our service was consulted that day and noted on physical examination he was afebrile, pupils were minimally reactive and equal, and he had oral thrush. Neurological examination revealed decreased vision in both eyes, loss of hearing on the right, motor strength of 5/5, ataxic gait, increased deep tendon reflexes in upper extremities bilaterally and normal in lower extremities, diminished sensation on the right side, and an absent Babinski reflex. The rest of the physical examination was normal. A magnetic resonance imaging (MRI) with gadolinium contrast of the brain showed a single white matter lesion in the right temporoparietal region that was nonenhancing and minimal cerebral atrophy (Figure 1). An ophthalmologic evaluation revealed bilateral acute retinal necrosis.
Initial laboratory studies revealed a white blood cell count of4500/μL, a negative rheumatologic workup, and a nonreactive rapid plasma regain (RPR). A cerebrospinal fluid (CSF) analysis showed normal opening pressure, a glucose of 46 mg/dL, protein of 76 mg/dL, and 52 cells/μL white blood cells with 87% lymphocytes. His HIV enzyme-linked immunosorbent assay (ELISA) and Western blot analysis were both positive, and the CD4 count was 50 cells/μL. The CSF VDRL was positive, and a subsequent serum RPR obtained 5 days later was reactive at 1:1. No prozone reaction was present with the second RPR; it was not checked the first time. A microhemagglutination-Treponema pallidum test (MHA-TP) from serum was nonreactive. Fluorescent treponemal antibody absorption (FTA-ABS) was negative; serum Lyme enzyme immunoassay was positive with a negative Western blot analysis. The erythrocyte sedimentation rate was greater than 100 mm/h; IgG anticardiolipin antibody was positive. Toxoplasma serum IgG was positive and IgM was negative. Hep B core Ab was positive, and Hep B surface Ag and Hep C Ab were negative. Antinuclear antibodies were negative.
Administration of intravenous aqueous crystalline penicillin, acyclovir, foscarnet, and high-dose steroids was begun on hospital day 3. Administration of acyclovir was started empirically for the retinal necrosis pending the results of the herpes simplex virus (HSV) DNA polymerase chain reaction (PCR). Foscarnet was given for possible varicella-zoster virus (VZV). High-dose steroids were initially given based on a case report that suggested they may be beneficial.3 He was also given prophylactic trimethoprim-sulfamethoxazole and weekly azithromycin. Highly active antiretroviral therapy was not given because of concerns about precipitating the immune reconstitution inflammatory syndrome, which might have caused a worsening of his neurological status. However, by hospital day 9, the patient was completely blind and deaf. Results of the original CSF studies were returned that day and revealed the HSV-1 and HSV-2 PCR was negative; cytomegalovirus PCR, negative; and CSF Lyme IgG, positive with a negative Western blot analysis. Other serological studies for connective tissue diseases and vasculitis were negative. A subsequent brain MRI with contrast showed no demyelinating plaques, advanced diffuse cerebral atrophy with corresponding ventriculomegaly, and normal brain parenchyma. On hospital day 10, a serum T. pallidum antibody by ELISA returned negative (ARUP Labs, Salt Lake City, Utah). Also that day, a serum FTA returned nonreactive (Quest Labs, Horsham, PA).
On hospital day 20, the patient underwent repeat CSF examination. It showed glucose of 58 mg/dL, protein of 59 mg/dL, 50cells/μL white blood cells with 91% lymphocytes, and a nonreactive VDRL. He completed a 14-day course of intravenous penicillin and 21 days of acyclovir with tapering steroids. The patient was discharged, still blind and deaf, to a rehabilitation facility. Treponema pallidum IgG by immunofluorescent labeled antibody staining from CSF was nonreactive (ARUP labs). After discharge, the CSF VZV PCR was reported positive from the second lumbar puncture, having not been sent from the first one. Over the following 2 months, he recovered most of his hearing and sight.
False-positive CSF VDRL is thought to be a very rare occurrence. The phenomenon was first reported in a patient with lymphosarcoma.1 Griemberg et al2 gave the rate of biological false positives as 0.66%. In our patient, VZV can most likely be the unifying diagnosis that best explains this patient's symptoms and clinical course. His initial focal pruritus was probably subclinical zoster. A case report from Japan identified VZV infection in a patient with acute-onset deafness by virus-specific IgG and IgM serology.3 Therapy involved acyclovir, hyperbaric oxygen, and steroids. However, bilateral hearing loss remained at the time of discharge. Of note, between 8% and 11% of patients with HIV are infected with herpes zoster.4 We hypothesize that the acute immunologic response to the VZV infection gave our patient a false-positive VDRL and a positive anticardiolipin antibody which in turn was identified with the serum RPR and Lyme enzyme immunoassay.
The diagnosis of neurosyphilis can be challenging. The CSF VDRL is insensitive but very specific. A positive CSF VDRL in the right clinical setting establishes the diagnosis of neurosyphilis. Negative treponemal tests from the CSF canrule out neurosyphilis in late disease but not in patients with early disease. Also, 5 or more mononuclear cells per microliter of CSF is also suggestive of active neurosyphilis.5 Muller et al6 used a meta-analysis of external quality control data from German laboratories to determine the impact of test quality on the diagnosis of syphilis. They found that T.pallidum hemagglutination assay, T. pallidum particle agglutination assay, and ELISAs were more reliable than VDRL and FTA-ABS as screening tests. Of concern, 18.3% of laboratories misclassified samples from patients with active syphilis as past infection, and 10.2% reported serological evidence for active infection in patients with past syphilis or in sera from seronegative blood donors. If a false-positive VDRL is suspected, the proper time to repeat syphilis testing is about 10 weeks after the initial serology.7 Thakar et al8 found that VDRL titers of 1:8 and below should be considered inconclusive. After completing therapy for neurosyphilis with IV penicillin, repeat CSF examinations should be performed every 6 months until the cell count is normal. Changes in CSF VDRL or CSF protein occur more slowly than cell counts. Recent data on HIV-infected patients suggest that CSF abnormalities might persist for more extended periods.9
False-positive VDRL reactions (FPRs) are known to occur in patients infected with HIV. In 1 study, a subgroup analysis of HIV patients (n = 1415) showed a 10-fold higher rate of benign false-positive serum VDRL tests (2.1% vs 0.24%), which was statistically significant.10 Another study11 found FPRs in 6 (4%) of 159 HIV-seropositive patients and 34 (0.8%) of 4387 HIV-seronegative patients. Further data from the Zimbabwe AIDS Prevention Project found FPRs in0.5% in a cohort of 580 men, of whom 19.8% were HIV-seropositive.12 Higher false-positive titers occurred in 5 men who were all infected with HIV.
Hepatitis C virus (HCV) infection has been associated with false-positive reactions for syphilis.13 The investigators found that 5 of 50 patients with HCV had a positive VDRL and a negative MHA-TP test. Our patient was negative for HCV antibodies. Spirochete illnesses-like relapsing fever (Borrelia spp), leptospirosis, rat-bite fever, and Lyme disease (Borrelia burgdorferi) can also cause false-positive syphilis tests.5 Our patient had a false-positive Lyme antibody in the CSF (1:4). Systemic lupus erythematosus can produce false-positive nontreponemal tests and positive or borderline positive FTA-ABS reactions.
These cases, in addition to the patient in our report, remind the astute clinician about the pitfalls of false-positive testing in the diagnosis of neurosyphilis. Unfortunately, new molecular tests for syphilis are unlikely to replace serology in the near future because of expense and the need for sophisticated equipment.
1. Wuepper KD, Tuffanelli DL. False-positive reaction to VDRL test with prozone phenomena. Association with lymphosarcoma. JAMA
2. Griemberg G, Ravelli MR, Etcheves PC, et al. Syphilis and pregnancy. Prenatal control, seroprevalence and biological false positives. Medicina
3. Fuse T, Inamura H, Nakamura T, et al. Bilateral hearing loss due to viral infection. J Otorhinolaryngol
4. Gnann JW, Whitley RJ. Natural history and treatment of varicella-zosterin high risk populations. J Hosp Infect
5. Tramont EC. Treponema pallidum
(syphilis). In: Mandell GL, Bennett JE, Dollin R, eds. Principles and Practice of Infectious Diseases
. 5th ed. Philadelphia, PA: Elsevier; 2005:2768-2785.
6. Muller I, Brade V, Hagedorn HJ, et al. Is serological testing a reliable tool in laboratory diagnosis of syphilis? Meta-analysis of eight external quality control surveys performed by the German Infection Serology Proficiency Testing Program. J Clin Micro
7. Wiwanitkit V. Biological false reactive VDRL tests: when to re-test? Southeast Asian J Trop Med Public Health
. 2002;33(suppl 3):131-132.
8. Thakar YS, Chande C, Mahalley AD, et al. Seroprevalence of syphilis by TPHA test. Indian J Pathol Microbiol
9. Marra CM, Maxwell CL, Tantolo L, et al. Normalization of cerebrospinal fluid abnormalities after neurosyphilis therapy: does HIV status matter? Clin Infect Dis
10. Geusau A, Kittler H, Hein U, et al. Biological false-positive tests comprise a high proportion of Venereal Disease Research Laboratory reactions in an analysis of 300,000 sera. Int J STD AIDS
11. Rompalo AM, Cannon RO, Quinn TC, et al. Association of biologic false-positive reactions for syphilis with human immunodeficiency virus infection. J Infect Dis
12. Gwanzura L, Latif A, Bassett M, et al. Syphilis serology and HIV infection in Harare, Zimbabwe. Sex Transm Infect
13. Sonmez E, Ozerol IH, Senol M, et al. False-positive reaction between syphilis and hepatitis C infection. Isr J Med Sci