Chen, Cheng-Yen*; Chi, Kai-Hua*; Alexander, Sarah†; Martin, Iona M. C.†; Liu, Hsi*; Ison, Cathy A.†; Ballard, Ronald C.*
From the *Division of STD Prevention, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia; and the †Sexually Transmitted Bacteria Reference Laboratory, Health Protection Agency Centre for Infections, London, U.K.
The authors thank John Papp of the Division of STD Prevention (DSTDP), CDC, for his critical review of the manuscript and Akbar Zaidi of DSTDP for his assistance in statistical analysis.
Correspondence: Cheng-Yen Chen, PhD, Division of STD Prevention, National Center for HIV, STD and TB Prevention, Centers for Disease Control and Prevention, Mail Stop: G-39, 1600 Clifton Road, Atlanta, GA 30333. E-mail: firstname.lastname@example.org.
Received for publication August 24, 2006, and accepted September 5, 2006.
LYMPHOGRANULOMA VENEREUM (LGV) IS A systemic sexually transmitted disease caused by specific biovar types of Chlamydia trachomatis (L1, L2, and L3). The disease is highly prevalent in parts of Africa, Asia, Central and South America but has been rarely seen in the United States and Western Europe for many decades.1 Classically, LGV is a chronic disease that has a variety of acute and late manifestations, including genital ulceration and inguinal/femoral lymphadenopathy (buboes) with suppuration. Subsequent lymphatic spread may result in hemorrhagic proctocolitis and scarring and blockage of the lymphatics resulting in lymphedema.1–3 LGV infection may also increase the risk for acquisition and transmission of other sexually transmitted infections (STIs), including HIV and bloodborne diseases such as hepatitis C virus (HCV).2,4,5
In contrast to the classic presentation, an increase in cases of rectal LGV has been reported in various European cities among men who have sex with men (MSM) since 20032,6–13; unlike classic LGV symptoms with inguinal lymphadenopathy, most cases present initially as a proctitis or proctocolitis, but asymptomatic cases have also been detected. As of September 2004, a total of 92 LGV cases had been confirmed in The Netherlands and were of the L2 genotype.6,14 In the United Kingdom, a total of 344 cases of LGV in MSM have been confirmed between October 2004 and March 2006.15 Most patients are of white ethnicity, HIV-positive, had a concurrent STI and a history of previous STI.
The laboratory diagnosis of LGV infection includes culture and, in some countries, nucleic acid amplification tests (NAATs) followed by genotyping of positives with polymerase chain reaction (PCR)-based restriction fragment length polymorphism (RFLP) analysis and/or sequencing of omp116–18 and serologic testing (indirect microimmunofluorescence, enzyme immunoassay, or complement fixation tests).19 Both culture and NAAT testing lack specificity, whereas the use of serologic methods is also limited owing either to lack of species specificity in the complement fixation test or to broad crossreactivity of antibody responses in cases of LGV in the microimmunofluorescence and enzyme immunoassays. Thus, although serologic test results can support a LGV diagnosis, they often do not provide specific evidence of LGV infection. Current commercial NAATs do not differentiate isolates belonging to the LGV from non-LGV biovars. Genotyping or serotyping methods of C. trachomatis allow reliable and specific discrimination of biovars but are time-consuming and require specially trained personnel in a sophisticated laboratory setting. The current Centers for Disease Control and Prevention (CDC) sexually transmitted disease treatment guidelines recommended regimen is 100 mg doxycycline twice daily for 21 days for LGV infection versus 7 days for other uncomplicated chlamydial genital infections20; thus, a timely laboratory diagnosis to identify early LGV infection is crucial in reducing undertreatment in cases of LGV and overtreatment for non-LGV-related infections.
The aim of this study was to establish a real-time multiplex PCR (M-PCR) test for the diagnosis of chlamydial infection and also to differentiate infection caused by LGV from non-LGV C. trachomatis serovars. Furthermore, we endeavored to validate the performance of the real-time PCR assay with the PCR-based RFLP genotyping method using rectal and urethral swab specimens.
Materials and Methods
Clinical Specimens and DNA Extraction
Rectal and urethral swab specimens obtained from MSM presenting with symptoms suggestive of LGV/non-LGV proctitis were referred to the Sexually Transmitted Bacterial Reference Laboratory, Health Protection Agency (HPA) of the United Kingdom during 2005 for LGV testing. A range of rectal specimens were received, including fresh dry swabs and residues of processed NAAT specimens. Dry swabs were hydrated with 500 μL phosphate-buffered saline and agitated on an orbital shaker at 150 rpm for 1 hour. DNA extractions were performed by one of 2 methods: 1) an automated DNA/RNA extraction system, X-Tractor Gene (Corbett Research, Sydney, Australia), using the Macherey-Nagel NucleoSpin Blood Kit (Düren, Germany); or 2) a manual extraction procedure using the QIAamp Viral RNA Mini Kit (Qiagen, Valencia, CA). Both extractions were performed according to manufacturers' instructions. Only the DNA specimens stored at −20°C for up to 12 months, without patient identifiers, were sent to CDC for real-time M-PCR testing. The requirement of Institutional Review Board approval for this study was waived by the CDC Human Subjects Office.
Detection of Chlamydia trachomatis
The chlamydial status of specimens was determined at the HPA using an inhouse real-time PCR assay. The real-time PCR was run as a duplex containing primers targeting an 88-bp region of the C. trachomatis cryptic plasmid as well as primers targeting the human RNase P gene, which acted as an internal control. All reactions were performed on a real-time PCR instrument (Rotor-Gene 3000; Corbett Research). Each 25-μL reaction contained the final concentrations of the following: 200 nm CTF-008 (GGATTG ACTCCGACAACGTATTC), 300 nm CTR-009 (ATCATTGCCAT TAGAAAGGGC ATT), 200 nm CTP-010 [FAM (6-carboxyfluorescein)-TTACGTG TAGGCGGTTTAG AAAGCGG-BHQ1 (black hole quencher)], 80 nM RNPF-003 (AGATTTGGACCTGCG AGCG), 80 nM RNPR-002 (GAGCGGCTGTCTCCACA AGT), 80 nM RNPP-001 [CY5 (cyanine)-TTCTGACCTGAAGGCTCTGCGCG-BHQ3], 200 μm dATP, 200 μmol/L dCTP, 200 μmol/L dGTP, 400 μmol/L dUTP, 1× PCR buffer, 4 mmol/L MgCl2, 5 units of Ampli Taq Gold polymerase, 0.5 units of uracil-N-glycosylase, and 10 μL of DNA template.
Genotyping by Restriction Fragment Length Polymorphism Analysis
Genotyping of all specimens positive for C. trachomatis was performed at the HPA by the methods of Lan et al16,17 using the chlamydial outer membrane protein 1 gene (omp1) for PCR amplification followed by RFLP analysis of nested PCR products. Briefly, a 1.1-Kb section of omp1 of C. trachomatis was amplified by a nested PCR with positive DNA controls from L1, L2, and L3-type strains. A negative control was included in each PCR run. Amplification was confirmed using a DNA7500 lab chip on an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Palo Alto, CA). PCR products were then digested with restriction endonuclease AluI at 37°C for 2 hours and separated once more on a DNA7500 lab chip. Restriction digest patterns were compared with those obtained for C. trachomatis biovars L1, L2, and L3. Samples with a restriction digest giving an L3 profile were additionally digested using EcoRI and DdeI at 37°C for 2 hours to differentiate between L3 and serovars H and I of C. trachomatis. Samples were assigned as being LGV-associated or non-LGV-associated on the basis of their restriction digest profiles. After genotyping, residual DNA materials from the extraction were shipped to the CDC and tested for LGV and non-LGV C. trachomatis using a real-time M-PCR assay. The genotype results were not revealed to the CDC until the M-PCR testing was completed. The degree of agreement between 2 test formats was measured using the κ test and the McNemar's test.
pmpH Real-Time Polymerase Chain Reaction Amplification and Detection
The polymorphic membrane protein H gene (pmpH) was used as the PCR target for the real-time M-PCR assay based on a unique 36-bp deletion region occurring only among LGV strains.21 A set of PCR primers was selected to amplify a 168-bp DNA fragment that encompasses the deletion region and 2 TaqMan probes were designed: one FAM (6-carboxyfluorescein)-labeled probe hybridizes to the DNA sequences within the deletion region and detects only chlamydial serovars A-K, whereas the second ROX (6-carboxy-X-rhodamine)-labeled probe hybridizes to a conserved sequence outside of the deletion region that is present in all chlamydial serovars (A to L3). The sequence and direction of the primers and fluorescent probes are listed in Table 1.
The Rotor-Gene 3000 real-time PCR instrument (Corbett Research) was used to perform the assay. M-PCR amplifications were performed in 25-μL reaction tubes using 10 μL of DNA samples. A final concentration of 1× PCR buffer (Applied Biosystems); 4 mmol/L MgCl2; 200 nM of each of the primers and probes; 200 μmol/L each of dATP, dGTP, dCTP, and dUTP; 0.5 units of uracil-N-glycosylase (Applied Biosystems); and 1 unit of AmpliTaq Gold DNA polymerase (Applied Biosystems) were used. Appropriate positive and no template controls were included in each run. The following conditions were used for the PCR amplification: the first cycle: 50°C for 2 minutes followed by 95°C for 10 minutes, and in the subsequent 45 to 50 PCR cycles: 95°C for 20 seconds, 60°C for one minute. The analytic sensitivity of the real-time M-PCR assay was determined using serial dilutions of purified C. trachomatis genomic DNA of known infectivity from serovars D and L2, and the specificity tested with a CDC inhouse collection of C. trachomatis serovars from A to L3 and a panel of pathogenic and commensal genital tract microorganisms. Serovars D or L2 strain of C. trachomatis were used as positive controls. Purified L2 genomic DNA was purchased from Advanced Biotechnologies, Inc., Columbia, MD.
Differentiation Between Lymphogranuloma Venereum and Nonlymphogranuloma Venereum Strains by Real-Time Multiplex Polymerase Chain Reaction
Based on the design of 2 TaqMan probes in the real-time M-PCR assay, the FAM-labeled probe detects only chlamydial serovars A-K and the ROX-labeled probe detects all chlamydial serovars (A to L3); thus, LGV and non-LGV strains could be distinguished. Specimens containing DNA from non-LGV strains exhibited positive fluorescent signals in both FAM and ROX channel of the Rotor-Gene 3000, whereas specimens containing LGV DNA had positive fluorescent signals only in the ROX channel. Representative real-time M-PCR amplification curves are illustrated in Figure 1.
Performance of the Real-Time Multiplex Polymerase Chain Reaction Assay
The real-time M-PCR assay allows a broad range of C. trachomatis DNA detection from 10−1 to 108 copies per reaction (Fig. 1). The analytical sensitivity of the assay of each target determined with a series of 10-fold dilutions of purified genomic DNA from serovars L2 and D was 0.1 to 1 copies and 1 to 10 copies per reaction, respectively. The real-time simplex PCR efficiency determined using a 10-fold serial dilution of non-LGV C. trachomatis DNA was 124% with a correlation coefficient (r2) of 0.926. For LGV DNA detection, the PCR efficiency was 124% with r2 = 0.970 (Fig. 1). For determination of specificity of the assay, a CDC inhouse collection of C. trachomatis serovars (A–L3) and a panel of commensal and pathogenic microbes found in the genitourinary tract and skin together with organisms closely related to the intended target species were tested. The primers and the ROX-labeled probe were specific for all C. trachomatis serovars, and the FAM-labeled probe specifically hybridized only to serovars A–K. Nonspecific amplification was not observed using other unrelated bacterial, viral, and fungal DNAs (data not shown).
Comparison of Real-Time Multiplex Polymerase Chain Reaction With Restriction Fragment Length Polymorphism Genotyping
The comparison of test results obtained using the real-time M-PCR assay and the RFLP genotyping method to diagnose LGV and non-LGV infection is shown in Table 2. Overall, 96.5% agreement (111 of 115 specimens) between the 2 assays for the detection of C. trachomatis (both LGV and non-LGV) was noted with a κ value of 0.945 (95% confidence interval = 0.89–0.998, P <0.00001). Both methods identified 53 DNA specimens as LGV-positive, 32 as non-LGV C. trachomatis, and 26 as negative. Eight urethral specimens were identified by both methods as LGV-positive (2 specimens), non-LGV-positive (5), and negative (1). Among the discordant results, 3 were from rectal specimens and one was from an unknown source. A total of 98 rectal specimens were compared; 2 were LGV-positive by the real-time M-PCR but were identified as non-LGV C. trachomatis by the genotyping method. Unfortunately, we were unable to resolve this discrepancy owing to the exhaustion of DNA samples. One LGV and one non-LGV C. trachomatis specimen identified by the genotyping method tested negative by the real-time M-PCR. These discordant results could be explained by the degradation of DNA as a result of freeze–thaw cycles and specimen transport, and not as a result of the presence of low copy number of target DNA below the detection limit of the real-time M-PCR. There was no statistical difference between 2 test formats for either the LGV detection alone or non-LGV infection using the McNemar's test (both P = 0.5)
In the past, serologic assessment has been the primary laboratory method used for diagnosis of LGV; however, the availability of the test and its low specificity has limited its use. In addition, its performance has not been established for rectal LGV infections. In contrast, conventional PCR-based genotyping by direct sequencing and RFLP analysis of omp1 is serovar-specific, but its clinical use is limited owing to long turnaround times (measured in days) and its inherent technical complexity. Given the laboratory tests that are currently available, most clinicians will not be able to make a timely and specific diagnosis of LGV infection other than that based on clinical grounds. The main objective of our study was to develop a relatively rapid (<2.5 hours) and accurate molecular diagnostic method to detect C. trachomatis-specific DNA sequences in rectal specimens using real-time M-PCR and to validate the capability of the assay to differentiate between LGV and non-LGV biovars of C. trachomatis by comparing test results with conventional genotyping using RFLP analysis of omp1.
The design of the real-time M-PCR target was focused on the pmpH gene owing to its DNA sequence showing a pattern of evolution that parallels disease groups, i.e., the nucleotide sequences of pmpH were nearly identical among serovars within a disease group but quite different from serovars of other disease groups.21 There are 6 addition or deletion events in pmpH, but the most notable is a large 36-bp deletion, which occurs only among the LGV serovars; hence, the region surrounding this deletion area became the focus of the real-time M-PCR design for the differentiation of LGV from non-LGV infection. The TaqMan probe designed to hybridize within the deletion region identifies C. trachomatis DNA other than LGV; the second TaqMan probe hybridizes to the conserved nucleotide sequences outside of the deletion region and detects DNA from all C. trachomatis serovars. Thus, the detection of LGV DNA in clinical specimens by our real-time M-PCR assay is an indirect method and based on exclusion. The limitation of the current real-time M-PCR assay lies in its inability to differentiate a coinfection caused by LGV and non-LGV urogenital strains or a mixed infection by different non-LGV strains. LGV would be overlooked if specimens contained a mixture of LGV and non-LGV DNA. The real-time M-PCR assay described here does not include a human DNA control to monitor PCR inhibition, although the potential PCR inhibitors present in the rectal swab specimens might have been removed by the manual or automated DNA extraction. Recently, Morré et al22 published a real-time simplex PCR assay that is capable of diagnosing LGV infection directly. The assay specifically detects only the LGV DNA by targeting the juncture of the 36-bp deletion region of pmpH with a sensitivity of 0.01 inclusion-forming units per reaction; however, the detection of non-LGV DNA requires a second test, and mixed infections of LGV and non-LGV would be overlooked. We have recently combined the real-time PCR assay described here with the test of Morré et al.20 In this quadriplex test (unpublished data), we are able to confirm C. trachomatis infection, to detect mixed infections caused by both LGV and non-LGV, and to incorporate a human DNA control. The RFLP analysis and/or sequencing of omp1 provides the discrimination of C. trachomatis strains and is useful in linking molecular information to contact tracing data for epidemiologic surveillance, whereas the real-time M-PCR assay is not capable of replacing the conventional typing system to reveal the variants of all different serovars or within the same serovar.
Outbreaks of rectal LGV infection in Western Europe suggest there may be an increase in cases in the United States, especially among MSM. In July 2004, the CDC identified an L2 LGV strain on a rectal swab specimen from a patient in the United States who had signs and symptoms similar to those of the patients in The Netherlands.14 Currently, there are no commercially available NAATs for C. trachomatis detection that are cleared for use on the rectal specimens in the United States and all molecular confirmatory tests are still in the process of being validated. In this study, we have shown the performance of the real-time M-PCR assay for the identification of LGV infection is highly comparable with that of the conventional genotyping method using rectal and urethral DNA specimens. Additional advantages of our real-time PCR assay include the relatively rapid turnaround time (<2.5 hours) permitting possible application as a point-of-care test in the future. In addition, because M-PCR is capable of detecting up to 6 molecular targets in one assay (using a Rotor-Gene 6000 instrument), it is possible to include the LGV assay in an M-PCR test for genitourinary disease (syphilis, chancroid, and genital herpes) or to identify LGV with concurrent STIs of interest (e.g., gonorrhea, rectal herpetic infection) in patients to aid outbreak investigation or provide rapid screening.
1. Mabey D, Peeling RW. Lymphogranuloma venereum. Sex Transm Infect 2002; 78:90–92.
2. Nieuwenhuis RF, Ossewaarde JM, van der Meijden WI, Neumann HA. Unusual presentation of early lymphogranuloma venereum in an HIV-1 infected patient: Effective treatment with 1 g azithromycin. Sex Transm Infect 2003; 79:453–455.
3. Spaargaren J, Fennema HSA, Morré SA, de Vries HJC, Coutinho RA. New lymphogranuloma venereum Chlamydia trachomatis variant, Amsterdam. Emerg Infect Dis 2005; 11:1090–1092.
4. Den Hollander JG, Ossewaarde JM, de Marie S. Anorectal ulcer in HIV patients, don't forget lymphogranuloma venereum! AIDS 2004; 18:1484–1485.
5. Van Agtmael MA, Perenboom RM. Tow HIV-positive men with anorectal lymphogranuloma venereum and hepatitis C: Emerging sexually transmitted diseases. Ned Tijdschr Geneeskd 2004; 148:2547–2550.
6. Nieuwenhuis RF, Ossewaarde JM, Götz HM, et al. Resurgence of lymphogranuloma venereum in Western Europe: an outbreak of Chlamydia trachomatis serovar L2 proctitis in The Netherlands among men who have sex with men. Clin Infect Dis 2004; 39:996–1003.
7. French P, Ison CA, Macdonald N. Lymphogranuloma venereum in the United Kingdom. Sex Transm Infect 2005; 81:97–98.
8. van Weel J. Rare sexually transmitted disease hits Europe. Lancet Infect Dis 2004; 4:720.
9. Götz H, Nieuwenhuis R, Ossewaarde T, et al. Preliminary report of an outbreak of lymphogranuloma venereum in homosexual men in The Netherlands, with implications for other countries in Western Europe. Eurosurveillance Weekly 2004; 8:22.
10. Vandenbruaene M. Uitbraak van Lymphogranuloma Venereum in Antwepen en Rotterdam. Epidemiologisch Bulletin van de Vlaamse Gemeenschap 2004; 47:4–6.
11. Robert Koch-Institut. Zum gehäuften Auftreten von Lymphogranuloma Venereum in Hamburg im Jahr 2003. Epidemiologisches Bulletin 2004; 25:18.
12. Institut de Veille Sanitaireqq. Emergence de la Lymphogranulomatose vénérienne rectale en France: cas estimés au 31 mars 2004. Synthése réalisée le 1er Juin 2004.
14. Centers for Disease Control and Prevention. Lymphogranuloma venereum among men who have sex with men—Netherlands, 2003–2004. MMWR Morb Mortal Wkly Rep 2004; 53:985–988.
16. Lan J, Ossewaarde JM, Walboomers JMM, Meijer CJLM, Brule AJC. Improved PCR sensitivity for direct genotyping of Chlamydia trachomatis serovars by using a nested PCR. J Clin Microbiol 1994; 32:528–530.
17. Lan J, Walboomers JMM, Roosendaal R, et al. Direct detection and genotyping of Chlamydia trachomatis in cervical scrapes by using polymerase chain reaction and restriction fragment length polymorphism analysis. J Clin Microbiol 1993; 31:1060–1065.
18. Yuan Y, Zhang Y-X, Watkins NG, Caldwell HD. Nucleotide and deduced amino acid sequences for the four variable domains of the major outer membrane proteins of the 15 Chlamydia trachomatis serovars. Infect Immun 1989; 57:1040–1049.
19. Black CM. Current methods of laboratory diagnosis of Chlamydia trachomatis infections. Clin Microbiol Rev 1997; 10:160–184.
20. Centers for Disease Control and Prevention. Sexually transmitted disease treatment guidelines—2002. MMWR Morb Mortal Wkly Rep 2002; 51:1–80.
21. Stothard DR, Toth GA, Batteiger BE. Polymorphomic membrane protein H evolved in parallel with the three disease-causing groups of Chlamydia trachomatis. Infect Immun 2003; 71:1200–1208.
22. Morrėdot SA, Spaargaren J, Fennema JSA, et al. Real-time polymerase chain reaction to diagnose lymphogranuloma venereum. Emerg Infect Dis 2005; 11:1311–1312.