| |||||||||||||||||||||||
Case Report
A 29-year-old HIV-infected man was admitted to Westchester County Medical Center, Valhalla, New York, on January 2, 1991, because of a six-month history of right-sided weakness, a three-month history of changes in mental status, a seizure one week before admission, and difficulty swallowing for one week. Seven weeks earlier he had had a lumbar puncture at another hospital because of the changes in mental status. Analysis of the cerebrospinal fluid yielded the following findings: 4 white cells per cubic millimeter (all lymphocytes), 3 erythrocytes per cubic millimeter, a protein level of 140 mg per deciliter, a glucose level of 40 mg per deciliter (2.2 mmol per liter), and a negative Venereal Disease Research Laboratory (VDRL) test.
Review of the patient's medical history revealed that in 1984 he had been treated for primary syphilis. In 1988, neurosyphilis was diagnosed on the basis of a reactive VDRL test of cerebrospinal fluid after a lumbar puncture was performed because of delusional behavior. The patient was treated with 12 x 106 units of intravenous penicillin daily for 10 days. His CD4 cell count was 9 cells per cubic millimeter, and a computed tomographic (CT) scan of the brain (without contrast material) revealed no abnormalities at that time.
On admission the patient had a temperature of 36.8 °C, a pulse of 100 beats per minute, and a respiratory rate of 24 breaths per minute. Thrush was noted. His neck was supple, and funduscopic examination was normal. Skin, chest, heart, abdominal, and genital examinations were unremarkable. Neurologic examination revealed a disoriented, tremulous patient with a poor memory. Cranial-nerve examination showed no abnormalities, but a suboptimal examination of the lower extremities revealed weakness. Reflexes were intact and symmetric. Plantar reflexes were normal. The white-cell count was 3500 cells per cubic millimeter, the hematocrit was 31.7 percent, and the platelet count was 73,000 per cubic millimeter. Biochemical analyses of the blood were normal except for a 
-glutamyltransferase level of 249 U per liter (4.15 microkat per liter). Serum toxoplasma antibodies were not detected by enzyme-linked immunosorbent assay. A rapid plasma reagin test demonstrated a titer of 1:8, and the fluorescent treponemal-antibody-absorption test was reactive. A cranial CT scan revealed multiple ring-enhancing lesions (Figure 1A). Analysis of the cerebrospinal fluid revealed 12 white cells per cubic millimeter (11 lymphocytes), no erythrocytes, a glucose level of 50 mg per deciliter (2.78 mmol per liter), a protein level of 178 mg per deciliter, and a reactive VDRL test at a titer of 1:2. No cryptococcal antigen was detected in cerebrospinal fluid.
|
Methods
Detection of Treponemes by Direct Immunofluorescence Staining
The direct fluorescent-antibody test for the detection of T. pallidum was performed as previously described6,7. Deparaffinized sections of tissue fixed to slides were incubated with a fluorescein isothiocyanate-conjugated mouse monoclonal antibody to T. pallidum directed against a 37-kd surface protein7 or with a fluorescein isothiocyanate-conjugated polyclonal antibody to T. pallidum that had been previously absorbed for cross-reactivity with T. phagedenis. Positive control sections obtained from rabbit testicular tissue infected with T. pallidum were prepared and stained in a manner similar to that used for the brain tissue. If organisms were observed with the polyclonal conjugate, additional sections were stained with an anti-Borrelia burgdorferi conjugate.
Detection of T. pallidum DNA by PCR and Hybridization
The following tissues, submitted as coded specimens, were examined for the presence of DNA from T. pallidum subspecies pallidum: tissue from a gummatous brain lesion from the patient, brain tissue from a patient who died of a heart attack and who had no HIV antibodies and a negative VDRL test (a negative control), and orchitic tissue from a rabbit infected intratesticularly two weeks earlier with 107 T. pallidum per milliliter (a positive control). All tissues used for examination were formalin-fixed and embedded in paraffin.
The tissues were prepared for DNA extraction as previously described8. DNA was extracted according to the method of Boom et al., with 50 microl of the specimens or T. pallidum lysate adsorbed to diatoms in the presence of guanidine thiocyanate9. The in vitro amplification of DNA was performed with the bmp gene of T. pallidum10 and primers Tp7 and Tp8 for PCR and Tp3 and Tp4 for nested PCR. The primers were produced at the Wadsworth Center for Laboratories and Research (Albany, N.Y.) with a DNA synthesizer (model 8700, MilliGen, Bedford, Mass.). Thirty cycles of PCR produced fragments of 617 base pairs (bp); these fragments were then diluted 1:10 for use in the nested PCR, in which 39 cycles produced 500-bp fragments. For each run 250, 25, and 2.5 fg of chromosomal T. pallidum DNA and DNA per 5 microl extracted from T. pallidum suspensions (2 x 103 and 2 x 102 per milliliter) were used as positive controls of sensitivity. A variety of reagents without DNA were used as negative controls. The PCR products were analyzed in 2 percent agarose containing ethidium bromide. The reagents and the technique have been described elsewhere in detail11,12.
To confirm the specificity of the PCR product, biotinylated probes specific for T. pallidum (356 bp) were prepared by PCR and sequenced. The probe-target hybrid was detected by a BluGENE nonradioactive system of nucleic acid detection as described in detail in the manufacturer's manual (GIBCO BRL, Grand Island, N.Y.)13.
Results
Pathological Findings
Examination of the brain at autopsy revealed two lesions in the left parieto-occipital lobe measuring 3.0 and 3.5 cm in diameter. Each lesion had a rubbery greenish core surrounded by a darker area (Figure 1B). Similar lesions (measuring 3.0 and 1.5 cm) were noted in the white matter of the left side of the cerebellum. Microscopical examination demonstrated areas of coagulative necrosis with a marked chronic inflammatory exudate consisting of lymphocytes and plasma cells (Figure 1C). Rare multinucleated giant cells were noted. Special stains for bacteria, mycobacteria, and fungi were negative. B-cell markers (Dako, Carpinteria, Calif.) for monoclonality were negative. T. gondii was not identified after a careful review of specimens stained with hematoxylin and eosin. Silver staining with a modified Steiner stain (Sigma Chemical, St. Louis) revealed spirochetal forms (Figure 1D).
Immunofluorescence Staining
Organisms morphologically resembling treponema were observed only when sections were stained with the polyclonal conjugate. Spiral organisms were not observed when the monoclonal antibodies to T. pallidum or B. burgdorferi were used for staining.
PCR
PCR examination of coded specimens from the brain lesion of the patient provided proof (Figure 2A, lanes 3 and 4) that the gummatous lesion was syphilitic. The specificity of the PCR products was confirmed by DNA hybridization (Figure 2B). Controls included paraffin-embedded brain tissue from an age-matched patient who died of a heart attack (a negative control) (Figure 2, lane 2) and orchitic tissue from T. pallidum-infected rabbits (a positive control) (Figure 2, lane 7).
|
The HIV-infected man we describe had a space-occupying gummatous lesion of the brain. Molecular techniques proved that the invasion of the tissue was due to T. pallidum. The use of PCR to detect T. pallidum DNA in experimental and clinical specimens (including cerebrospinal fluid and amniotic fluid) is highly sensitive and specific11,12,14,15.
The deparaffinization of tissue sections does not seem to affect the sensitivity or specificity of PCR. Using diatom-based extraction of DNA and PCR with nested primers, we were able to detect 2.5 fg of T. pallidum DNA, or one organism in samples of 5 microl. Figure 2 shows the results obtained with five sections of tissue. Positive results were also obtained when single sections were used for the extraction of T. pallidum DNA (data not shown), but the results were less reliable.
Direct immunofluorescence staining of paraffin-embedded tissue with a monoclonal conjugate did not detect T. pallidum; however, spiral organisms were demonstrated with a polyclonal conjugate. The discrepancy in results between conjugated monoclonal and polyclonal antibodies is most likely explained by a difference in the degree of antibody activities. It is also possible that mouse monoclonal antibodies cannot penetrate thick tissue sections16.
Silver-staining techniques do not specifically identify T. pallidum; they can only confirm the presence or absence of spiral structures. Tissue artifacts are major sources of misinterpretation. For the evaluation of treponemal tissue invasion, the visualization of many spirochetal forms with silver stains in conjunction with positive serologic results provides strong but not definitive evidence of T. pallidum as the etiologic agent. In cases in which the diagnosis is questionable, PCR is a helpful addition.
Cerebral gummata are a rare complication of late tertiary syphilis in immunocompetent as well as immunocompromised hosts,17 and spiral organisms are rarely found with silver-staining techniques5. It is not clear whether antibiotic treatment had failed in this patient, allowing reactivation of syphilis as his immune system became impaired. It is also possible that he had been reinfected with T. pallidum since 1988, with rapid progression to central nervous system gummata. Rapid spread of T. pallidum to the central nervous system has been reported in patients with HIV infection4,18,19. The formation of gummata may be due to a hypersensitivity reaction caused either by a reactivation of infection with T. pallidum or by reinfection20. Despite the profound alteration of the immune response in AIDS, gummata can still form, as illustrated in our patient. The rarity of such lesions despite the higher frequency of neurosyphilis in HIV infection may relate to the immunologic deficits, particularly in cellular immunity, associated with AIDS.
In contrast to our patient, the two other HIV-infected patients known to have central nervous system gummata had complete responses to high-dose penicillin given intravenously5. However, the failure to respond to penicillin treatment for T. pallidum infection in HIV-infected patients is well known2,21,22.
Syphilitic gummata should be included in the differential diagnosis of space-occupying lesions in HIV-infected patients. The application of PCR to paraffin-embedded tissue offers an exquisitely sensitive and specific method of identifying T. pallidum as the cause of mass lesions of the central nervous system. Moreover, the infrequency with which treponemes were demonstrated in gummata in the past may reflect the lack of methods sensitive enough to detect the organisms.
Supported in part by a grant (A 121833) from the National Institute of Allergy and Infectious Diseases (to Drs. Wicher and Wicher) and by a research grant (6-FY93-0080) from the March of Dimes Birth Defects Foundation (to Dr. V. Wicher).
We are indebted to Timothy Moran for synthesizing the oligonucleotides and sequencing the 300-bp PCR product, to Dr. Ira Schwartz for his critical review of the manuscript, to Martha Fears and Janine Gerson for technical assistance, to Dan Benevento for photographic assistance, and to Dr. Jerome Shevell for assistance in the case preparation.
Source Information
From the Department of Medicine, Division of Infectious Diseases (H.W.H., G.P.W.), and the Department of Pathology (M.P.V.), New York Medical College, Valhalla; Wadsworth Center for Laboratories and Research, New York State Department of Health, Albany (V.W., F.A., K.W.); and the Division of Sexually Transmitted Diseases Laboratory Research, Centers for Disease Control and Prevention, Atlanta (S.A.L.).
Address reprint requests to Dr. Horowitz at Westchester County Medical Center, Division of Infectious Diseases, Rm. 209, Macy Pavilion SE, Valhalla, NY 10595.
References
| |||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
