Molecular Diagnosis of Lymphogranuloma Venereum in Patients with Genital Ulcer Disease

Current textbooks divide lymphogranuloma venereum (LGV) into three stages (7, 31). The primary stage is characterized by small, painless herpetiform genital ulcers which are often not recognized and resolve spontaneously. The secondary stage is characterized as lymphadenopathy, mostly without a genital lesion. Inguinal lymphadenitis with or without a history of genital lesions is therefore the characteristic presentation (11, 12, 14). The tertiary or anorectal/elephantiasis stage results from destroyed inguinal lymphoid tissue. Diagnostic methods to diagnose LGV are microimmunofluorescence (MIF), culture, direct immunofluorescence (DIF), and nucleic acid amplification tests (NAATs). The use of NAATs is the preferred method to diagnose herpes, chancroid, and syphilis, but these tests are rarely applied to diagnose LGV. One study used a PCR targeting the cryptic plasmid without further analysis (9), and one used sequence analysis of major outer membrane protein gene amplicons (1). MIF and culture are not widely available and lack sensitivity, while the targets for DIF and commercially available NAATs are present in all Chlamydia trachomatis biovars (4). The lack of sensitive and specific diagnostic tests is likely responsible for the wide variation in prevalence of LGV in patients with genital ulcer disease (GUD) reported from Africa (2, 5, 6). Biovar identification of C. trachomatis in specimens from genital ulcers is important because non-LGV ulcers can be contaminated with the trachoma biovar from concurrent urethritis or cervicitis (13).

A PCR with restriction digestion of the product to distinguish between the C. trachomatis biovars, with the cysteine-rich outer membrane protein (CrP) gene (GenBank AF 304332 ) as the target, has been described previously (15). We were first in applying this combination of PCR and restriction fragment length polymorphism in a treatment efficacy study with patients with GUD (13), resulting in a rise in the prevalence of LGV in patients with GUD in Durban from 2% to 10% (5, 13). Here we report on the validity of this NAAT for the diagnosis of LGV. We also describe the clinical presentation of LGV diagnosed with this methodology in patients presenting with GUD.

The study cohort was the same as that in the treatment efficacy study (13) and included 520 consecutive consenting patients presenting with GUD at the Prince Cyril Zulu Communicable Diseases Clinic in Durban between October 2000 and April 2001.

Epidemiological and clinical characteristics of patients with LGV were compared with those of patients with genital herpes, chancroid, syphilis, and/or granuloma inguinale (Table 1). All infections except granuloma inguinale were diagnosed by PCR (13). The study was approved by the Ethics Committee of the Nelson R. Mandela School of Medicine.

Specimen collection and preparation have been described previously (13). A blood and tissue mini kit (QIAGEN, Frankfurt, Germany) was used to extract DNA. LGV was diagnosed by PCR-restriction fragment length polymorphism as previously described (13, 15). A PCR targeting the 60-kDa cysteine-rich outer membrane protein was followed by digestion with AccI. C. trachomatis L2 (ATCC VR-902B), C. trachomatis biovar trachoma (a laboratory isolate from a cervical specimen), and water were included as controls in each run. This test is further referred to as CrP-PCR-D. Figure 1 shows the results of one run.

The specimens of 66 (13%) patients tested positive by CrP-PCR-D, and these specimens were further analyzed. To confirm the presence of chlamydia, the strand displacement amplification assay (BDProbeTec ET; Becton Dickinson, Cockeysville, Md.) was performed on all samples with positive CrP-PCR-D results. The sensitivity of CrP-PCR-D was measured by the amplification of 10-fold serial dilutions, in phosphate-buffered saline (pH 6.8), of a C. trachomatis culture in McCoy cells. The number of inclusion bodies per ml in the undiluted culture was established by immunofluorescence microscopy (MicroTrak; Trinity Biotech, Wicklow, Ireland). Six randomly selected PCR products with LGV biovar restriction patterns and one with a trachoma biovar restriction pattern were sequenced. PCR products were purified by using the QIA Quick PCR purification kit (QIAGEN). Sequence reactions were done by using the BigDye terminator cycle-sequencing ready reaction kit (Applied Biosystems, Foster City, Calif.). Products were analyzed on an ABI 3100 analyzer.

Human immunodeficiency virus (HIV) infection was diagnosed by the Determine HIV1/2 test (Abbott Laboratories, Chicago, Ill.). Positive results were confirmed by the Capillus HIV1/2 (Trinity Biotech) test.

Groups were compared by use of Wilcoxon nonparametric tests for numeric data and the chi-square test or Fisher's exact test for categorical data.

The CrP-PCR-D was positive with McCoy cell dilutions containing as few as 5 to 10 inclusion bodies. This is in keeping with what has been reported previously (15). All 66 specimens positive by CrP-PCR-D also tested positive with the strand displacement amplification assay. Identified sequences matched those of the CrP gene of C. trachomatis in GenBank and demonstrated the expected numbers and locations of restriction sites for AccI in both biovars (15).

For two specimens (both from male patients), the signal obtained with the CrP-PCR-D was repeatedly faint and the restriction pattern was considered unreliable. Of the remaining 64, 63 were positive for LGV and 1 positive for the trachoma biovar. This suggests that the detection of chlamydiae in genital ulcer specimens by culture, DIF, or non-biovar-specific NAATs has an acceptable specificity for LGV.

The prevalence of LGV in women was higher than that of LGV in men (19% versus 10%; P = 0.006). Characteristics of patients diagnosed with LGV only were compared with those of patients with no LGV after stratification for gender (Table 1). Ulcers were, subjectively, equally as painful in LGV and non-LGV cases on specimen collection. Men, but not women, with LGV had more frequently palpable inguinal lymph nodes than men without LGV. These lymph nodes had no particular features, e.g., groove sign.

Patients included in this study all had primary LGV. Early publications acknowledge the difficulty in differentiating LGV lesions from chancroid lesions (3, 7, 8, 10), while current textbooks compare LGV lesions with genital herpes lesions (11, 12). This change in the description of the lesions over time may have resulted from the application of MIF as the standard diagnostic test. Being an antibody detection test, MIF is likely to produce positive results only with cases of advanced disease.

LGV was not associated with HIV. The clinical presentations of LGV for the 45 HIV-infected patients were similar to those for the 8 non-HIV-infected patients with LGV, as was the case for patients with non-LGV ulcers (data not shown). Although changes in clinical presentation of LGV due to HIV infection cannot be excluded, the similarity between our findings and the early descriptions (3, 7, 8, 10) suggests that this is not the sole explanation.

In conclusion, the presence of C. trachomatis biovar LGV in painful ulcers of appreciable size suggests that the description of the clinical features of LGV in current textbooks is incomplete.

• Open in new tab
• Download powerpoint

FIG. 1.

Gel electrophoresis of PCR products of specimens investigated for the presence of Chlamydia species by CrP-PCR-D. Lane 1, molecular weight marker; lane 10, undigested 552-bp product; lanes 2 and 9, digested products (280 and 272 bp) of C. trachomatis, LGV genotype (seen as one line); lanes 4 through 6, digested products (280, 195, and 77 bp) of C. trachomatis, urogenital genotype.

• Copyright © 2005 American Society for Microbiology