The influence of concomitant HIV infection on the serological diagnosis of primary syphilis in southern Africa.
The bi-directional nature of the interaction between conventional sexually transmitted infections (STIs) and HIV has been documented in many studies. (1,2) It has come to be recognised that the presence of STIs can increase the risk of acquisition of HIV infection and also promote its transmission, while HIV infection and associated immunodeficiency can alter the natural history and influence the diagnosis and management of conventional STIs. (3-5) The magnitude of the impact of this interaction is more evident in developing countries where the twin epidemics of HIV infection and conventional STIs coexist and produce a vicious cycle that is less evident in more developed societies. During the 1990s, several studies indicated that the incidence of HIV among patients with STIs was significantly higher than those without. This association was more pronounced among those patients presenting with genital ulcer disease (GUD). (6-9)
As a cause of GUD, syphilis is associated with an increased risk of both acquisition and transmission of HIV. Establishment of a definitive diagnosis in cases of primary syphilis is important in order to provide appropriate therapy as soon as possible to prevent spread of both syphilis and HIV (in the co-infected) and reduce the risk of acquisition of HIV (in those not already infected). Unfortunately, techniques to detect Treponema pallidum in primary lesions (either by darkfield microscopy, direct immunofluorescence or amplified molecular technologies) are usually not available in most settings in developing countries, and clinicians usually have to rely on serological tests to establish a diagnosis.
Several studies have suggested that concomitant HIV infection may change the performance characteristics of serological tests for syphilis. (10-14) In these studies, we have investigated the utility of both a non-treponemal (RPR) test and a treponemal (FTA-ABS) test for the diagnosis of primary syphilis during the emergence of the HIV epidemic in southern Africa.
Patients and methods
Eight hundred and sixty-eight patients with genital ulcerations were enrolled in five aetiological studies conducted in Maseru (Lesotho), Johannesburg, Cape Town, Durban and Carletonville STD clinics during the period 1994-1999. All patients were of black African ethnicity and the majority (97%) were male. Patients were eligible to enter these studies if they presented with visible genital ulceration(s) of at least 2 mm in diameter with or without inguinal and/or femoral lymphadenopathy. However, they were subsequently excluded if they had a history of treatment with antibiotics, or had received antiviral therapy during the past 7 days, or failed to give oral informed consent for completion of a questionnaire and collection of clinical specimens.
In each case, material from the bases of target lesions was collected using a cotton-tipped swab (Medical Wire and Equipment, Corsham, UK). Each swab was then expressed into 0.2 ml sterile distilled water and stored frozen at -70[degrees]C until analysed. Subsequently, a multiplex polymerase chain reaction (M-PCR) assay (Roche Molecular Systems) was used to detect specific target sequences of DNA from T. pallidum, Haemophilus ducreyi, and HSV from the swabs. An aliquot of the processed specimen was also used in a Chlamydia trachomatis PCR test (Amplicor, Roche Molecular Systems). All M-PCR assays were performed at the Centers for Disease Control and Prevention, in Atlanta, Georgia, USA, using methods described previously. (15)
In addition, vacuum collection tubes without anticoagulant were used to collect 10 ml of venous blood from each patient. After collection, blood specimens were allowed to clot at room temperature, centrifuged at 1200 x g for at least 5 minutes and the serum subsequently stored at 4[degrees]C or frozen at -70[degrees]C. Serological testing for syphilis was performed using both a quantitative non-treponemal RPR test (Immunotrep, Omega Diagnostics, Alloa, Scotland, UK) and a treponemal test (FTAABS, Marburg, Germany). HIV testing was performed using a rapid test (Capillus HIV-1/HIV-2, Cambridge Biotech, USA) and all positive samples were confirmed by an ELISA test (AxSYM HIV1/2, Abbott Diagnostics). All indeterminate or low-titre positive specimens were further tested using the Western Blot method (HIV 2.2 Blot, Genelab Diagnostics).
All HIV-seropositive patients were provided with post-test counselling and referred to local HIV clinics for follow-up and appropriate management.
All study protocols were approved by the Committee for Research on Human Subjects, of the University of the Witwatersrand, Johannesburg.
The results of molecular testing to determine the aetiology of the ulcerations seen in the five centres are shown in Table I. Of 868 patients included in these studies, H. ducreyi (HD) was detected by M-PCR in 436 (50.2%), T. pallidum (TP) in 163 (18.8%), and herpes simplex virus (HSV) in 268 (30.9%), while C. trachomatis (CT) was detected by PCR in 29 cases (3.3%). No organism could be demonstrated in 88 cases (10.1%). A single aetiological agent was detected in 672 cases (77.4%), while multiple aetiologies were detected in 108 (12.4%). Among those with a single aetiology, HD was detected in 346 (51%), HSV in 203 (30.2%), TP in 106 (15.8%) and CT in 17 (2.5%) cases. The most common mixed infections involved syphilis, chancroid and genital herpes. Of 163 TP-positive patients, 57 (34.9%) were co-infected with at least one other aetiological agent, namely: 33 (57.9%) with HD, 13 (22.8%) with HSV, and 7 (12.3%) with both. Co-infection with HD and HSV was also common.
Overall, 211 (24.3%) of sera obtained from these patients were reactive in both the RPR and FTA-ABS tests, a further 236 (27.2%) were reactive in the FTA-ABS test alone, and 421 (48.5%) were negative for both. No false-positive RPR reactions were detected. Among RPR-seropositive patients, 137 (64.9%) had a titre of 1:8 or higher (see Fig. 1). The overall geometric mean RPR titre was 1:11.5.
The rates of positive syphilis and HIV serology among patients with different GUD aetiologies are shown in Table II. The highest rate of RPR-seropositivity (69.3%) was found among patients with a positive M-PCR-TP. Of 163 M-PCR-TP-positive patients, the RPR positivity rate was higher in patients with single TP infections (79.2%) than in those with mixed aetiology involving TP and other agents (50.9%) ([chi square] = 14.03, p<0.0001).
The RPR seropositivity rate recorded among patients with a positive M-PCR-TP was significantly higher than that recorded among patients with a negative M-PCR-TP (69.3% v. 13.9% [chi square] = 216.9, p<0.001). The geometric mean RPR titre of sera collected from cases that were solely M-PCR-TP positive was similar to that recorded among cases where ulcerations were M-PCRTP positive but which had another pathogen detected in the same lesion (1:17.1 v. 1:16.8). However, the geometric mean titre was significantly lower in RPR-seropositive patients with non-treponemal ulcerations (1:5.9) (ANOVA, F statistics = 2.9, p<0.05).
Among all patients included in these studies, the rate of FTAABS seropositivity was 51.5%. Of the 657 RPR-seronegative patients, 236 (35.9%) were positive by the FTA-ABS test. As with the RPR test, patients whose ulcerations were only MPCRTP positive had a higher FTA-ABS seropositivity rate than those with other aetiologies recorded (96.2% v. 41.5%, [chi square]=21.4, p<0.001).
The overall HIV seroprevalence among GUD cases was 46.7%. Of 163 patients who were M-PCR-TP positive, 47 (28.8%) were co-infected with HIV. Among patients with a single aetiology, the lowest HIV prevalence was found among M-PCR-TP positive cases (20.8%). However, a higher HIV prevalence rate (43.9%) was recorded in cases that were M-PCR-TP positive but mixed with other aetiological agents (43.9% v. 20.8%, [chi square] = 9.64, p<0.001). The highest HIV prevalence of 329/617 (53.3%) was found in patients infected with non-treponemal causes of ulceration.
When comparing RPR seropositivity rates by HIV serostatus, the RPR test was positive in 27 HIV-positive/M-PCR-TP-positive cases (57.4%) compared with 86 patients (74.1%) in the HIV-negative/M-PCR-TP group (OR 0.47, 95% CI 0.23-0.96, p<0.05). Irrespective of their RPR status, no significant difference in FTA-ABS seropositivity rates was found among HIV-seropositive and -seronegative M-PCR-TP positive patients (83% v. 92.2%, NS). Eight HIV-positive patients (17%) and 9 HIV-negative patients (7.8%) who were M-PCR-TP positive were negative by both serological tests for syphilis. There was no significant difference in the geometric mean titres of the RPR test when comparing HIV-seropositive and -seronegative patients with a positive RPR test (1:13.7 v. 1:18.2, respectively).
Among 163 M-PCR (TP)-positive patients, 106 (65%) were solely infected by TP and 57 (34.9%) had mixed infections with other aetiological agents. Of the 106 patients with a single aetiology, 22 (20.8%) were co-infected with HIV. There was no significant difference in RPR positivity rates between HIV-seropositive and -seronegative groups with a single TP infection (81.8% v. 78.6%, Fisher's exact test NS). However, a significant difference in the RPR seropositivity rates between HIV-positive and -negative groups with mixed GUD aetiologies was detected (36.0% v. 62.5%, [chi square] = 3.9, p<0.05).
Of 705 patients with a negative M-PCR-TP test, 358 (50.8%) were co-infected with HIV, and the overall RPR seropositivity rate was 13.9%. The RPR seropositivity rate in HIV-seropositive patients was 12.3% and in HIV-seronegative patients 15.6% ([chi square]=1.6, p=0.2). The RPR seropositivity rate detected in patients with indeterminate laboratory findings was significantly higher than that seen in patients with known aetiologies other than TP (28.4% v. 11.8%, [chi square]=17.7, p<0.001), indicating that some cases of primary syphilis were not detected by the M-PCR-TP. This significance was independent of HIV serostatus (HIV-seropositives: 24.1% v. 11.2%, [chi square]=4.1, p=0.04; HIV-seronegative group: 30.5% v. 12.5%, [chi square]=12.1, p<0.001).
A comparison of the performance of syphilis serology in HIV-seropositive and -seronegative patients with single and mixed infections is shown in Table III. Eighty-eight patients with unknown or indeterminate aetiology were excluded from this analysis. The overall sensitivity of RPR to detect primary syphilis in our setting was 69.3% and its specificity was 86.1%, while the sensitivity and specificity of the FTA-ABS test were 89.6% and 58.5% respectively. Unfortunately, this test, in common with other treponemal tests, exhibited poor specificity for the detection of primary disease because it measures lifetime exposure to T. pallidum infection and was found to be reactive in many cases of ulceration that were M-PCR-TP negative. The sensitivity of the RPR test was found to be significantly higher in HIV-seronegative than HIV-seropositive patients (74% v. 57.4%, [chi square]= 4.94, p=0.026). The sensitivity of the RPR test was further decreased (to 36%) in HIV-seropositive patients with mixed infections. Among HIV-seropositive patients, there was a significant difference in the sensitivity of the RPR test in patients infected solely by TP when compared to those who had mixed TP infections (81.8% v. 36%, p<0.001). However, this difference was not statistically significant when comparing patients who were HIV-seronegative.
As has been demonstrated previously in southern Africa, the utility of both treponemal and non-treponemal serological tests for the diagnosis of primary syphilis is questionable, (16) since many patients with primary syphilis are seronegative (especially by the RPR test), while others with ulcerations caused by other organisms are seropositive (especially by the FTA-ABS test) as a result of previously treated treponemal infections. In these studies, we have examined the effect of HIV co-infection in patients with M-PCR-confirmed primary syphilis on the performance of treponemal and non-treponemal serological tests. Previous studies have reported atypical treponemal and nontreponemal serological responses in HIV-infected individuals with syphilis, such as false negative tests or a delay in seroconversion (12,14) and falsely positive serological responses. (10,11) In the study reported here, no biological false positive RPR reactions were detected among HIV-seropositive patients. This may be a phenomenon which is more associated with intravenous drug abuse than the direct effect of HIV infection; unlike populations studied in more industrialised countries, intravenous drug abuse is rare among patients co-infected with syphilis and HIV in southern Africa.
Only 69.3% of patients with M-PCR-confirmed TP infection were found to be RPR-seropositive, and the rate of RPR seropositivity among primary syphilis patients co-infected with HIV (57.4%) was significantly lower than that recorded among patients without HIV co-infection (74.1%). These results are consistent with a delay in antibody formation occurring as a result of HIV infection and subsequent immunosuppression. However, the rate of RPR seropositivity was also significantly reduced among those patients presenting with TP mixed with other GUD aetiological agents (50.9%) when compared with those with a single TP infection (79.2%). This effect was also seen among HIV-negative patients--with a significantly reduced rate of RPR seropositivity in those with mixed infections. The utility of RPR tests in detecting primary syphilis may be reduced further in areas where both HIV-coinfection as well as GUD with multiple aetiologies are common.
In southern Africa, the HIV epidemic reached maturity in the late 1990s, with an increasing number of immunocompromised patients having been recorded. Most patients with genital herpes who were included in the studies reported here presented with non-vesicular, purulent ulcerations that would usually be diagnosed clinically as chancroid. These chronic ulcers could subsequently serve as a portal of entry for T. pallidum, resulting in mixed infections and a delay in development of antibody responses to this 'secondary' infection. Regardless of the underlying mechanism of this decrease in sensitivity, it is clear that caution should be exercised in the interpretation of syphilis serological testing in cases of GUD, particularly in those who are co-infected with HIV, because many cases of primary syphilis may be associated with other causes of GUD.
(1.) Wasserheit JN. Epidemiological synergy. Interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Transm Dis 1992; 19: 61-77.
(2.) Clottey C, Dallabetta G. Sexually transmitted diseases and human immunodeficiency virus. Epidemiologic synergy? Infect Dis Clin North Am 1993; 7: 753-770.
(3.) Van de Perre P, Le Polain B, Carael M, Nzaramba D, Zissis G, Butzler JP. HIV antibodies in a remote rural area in Rwanda, Central Africa: an analysis of potential risk factors for HIV seropositivity. AIDS 1987; 1: 213-215.
(4.) Van de Perre P, Carael M, Nzaramba D, Zissis G, Kayihigi J, Butzler JP. Risk factors for HIV seropositivity in selected urban-based Rwandese adults. AIDS 1987; 1: 207-211.
(5.) Nkowane BM. Prevalence and incidence of HIV infection in Africa: a review of data published in 1990. AIDS 1991; 5 Suppl 1: S7-15.
(6.) Latif AS, Katzenstein DA, Bassett MT, Houston S, Emmanuel JC, Marowa E. Genital ulcers and transmission of HIV among couples in Zimbabwe. AIDS 1989; 3: 519-523.
(7.) Simonsen JN, Cameron DW, Gakinya MN, et al. Human immunodeficiency virus infection among men with sexually transmitted diseases. Experience from a center in Africa. N Engl J Med 1988; 319: 274-278.
(8.) Plummer FA, Simonsen JN, Cameron DW, et al. Cofactors in male-female sexual transmission of human immunodeficiency virus type 1. J Infect Dis 1991; 163: 233-239.
(9.) Simonsen JN, Cameron DW, Gakinya MN, et al. Human immunodeficiency virus infection among men with sexually transmitted diseases. Experience from a center in Africa. N Engl J Med 1988; 319: 274-278.
(10.) Joyanes P, Borobio MV, Arquez JM, Perea EJ. The association of false-positive rapid plasma reagin results and HIV infection. Sex Transm Dis 1998; 25: 569-571.
(11.) Yinnon AM, Coury-Doniger P, Polito R, Reichman RC. Serologic response to treatment of syphilis in patients with HIV infection. Arch Intern Med 1996; 156: 321-325.
(12.) Johnson PD, Graves SR, Stewart L, et al. Specific syphilis serological tests may become negative in HIV infection. AIDS 1991; 5: 419-423.
(13.) Rompalo AM, Cannon RO, Quinn TC, et al. Association of biologic false-positive reactions for syphilis with human immunodeficiency virus infection. J Infect Dis 1992;165: 1124-1126.
(14.) Drabick JJ, Tramont EC. Utility of the VDRL test in HIV-seropositive patients. N Engl J Med 1990; 322: 271.
(15.) Morse SA, Trees DL, Htun Y, et al. Comparison of clinical diagnosis and standard laboratory and molecular methods for the diagnosis of genital ulcer disease in Lesotho: association with human immunodeficiency virus infection. J Infect Dis 1997; 175: 583-589.
(16.) Sischy A, Exposto F da L, Dangor Y, et al. Syphilis serology in patients with primary syphilis and non-treponemal sexually transmitted diseases in southern Africa. Genitourin Med 1991; 67: 129-132.
Sexually Transmitted Infections Reference Centre, National Institute for Communicable Diseases, National Health Laboratory Service, Johannesburg
R C Ballard, PhD
H J Koornhof, MB ChB, DSc, FRCPath
F Radebe, MSc, NHDMT (Virology)
H G Fehler, MSc
Ye Htun, MB BS, PhD
Division of STD Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
R C Ballard, PhD
C-Y Chen, PhD
Ye Htun, MB BS, PhD
Corresponding author: R C Ballard (email@example.com)
Table I. Aetiology of GUD among patients in South Africa and Lesotho (1994 - 1999) No. of Aetiology patients % Single aetiology (N=672) Primary syphilis 106 12.2 Chancroid 346 39.9 Genital herpes 203 23.4 LGV 17 2.0 Multiple aetiology (N=108) Primary syphilis & Chancroid 33 3.8 Primary syphilis & g. herpes 13 1.5 Primary syphilis, Chancroid & g. herpes 7 0.8 Primary syphilis & LGV 3 0.3 Primary syphilis, Chancroid & LGV 1 0.1 Chancroid & g. herpes 43 5.0 Chancroid & LGV 6 0.7 G. herpes & LGV 2 0.2 Indeterminate 88 10.1 Total 868 100.0 LGV = lymphogranuloma venereum. Table II. Syphilis and HIV seroprevalence by aetiology of genital ulcer disease Both RPR & FTA-ABS+VE FFA-ABS+ve only Number of Aetiology Patients No. % No. % Single aetiology Primary syphilis 106 84 79.2 18 17.0 Chancroid 346 45 13.0 115 33.2 Genital herpes 203 14 6.9 49 24.1 LGV 17 3 17.6 5 29.4 Multiple aetiology Primary syphilis & chancroid 33 17 51.5 6 18.2 Primary syphilis & g. herpes 13 6 46.2 6 46.2 Primary syphilis, chancroid 7 4 57.1 2 28.6 & g. herpes Primary syphilis & LGV 3 2 66.7 -- Primary' syphilis, chancroid 1 -- 1 100.0 & LGV Chancroid & g. herpes 43 9 2.0 10 23.3 Chancroid & LGV 6 2 33.3 3 50.0 G. herpes & LGV 2 -- 1 50.0 1 Indeterminate 88 25 28.4 20 22.7 Total 868 211 24.3 236 27.2 Both RPR & FTA-ABS-ve HIV-positive Aetiology No. % No. % Single aetiology Primary syphilis 4 3.8 22 20.8 Chancroid 186 53.8 163 47.1 Genital herpes 140 69.0 126 62.1 LGV 9 52.9 10 58.8 Multiple aetiology Primary syphilis & chancroid 10 30.3 14 42.4 Primary syphilis & g. herpes 1 7.7 7 53.8 Primary syphilis, chancroid 1 14.3 3 42.9 & g. herpes Primary syphilis & LGV 1 33.3 1 33.3 Primary' syphilis, chancroid - - & LGV Chancroid & g. herpes 24 55.8 27 62.8 Chancroid & LGV 1 16.7 2 33.3 G. herpes & LGV 50.0 1 5.0 Indeterminate 43 48.9 29 33.0 Total 421 48.5 405 46.7 Table III. Performance of syphilis serology in the diagnosis of primary syphilis RPR Single infection Mixed infection p HIV-seropositive Sensitivity 18/22 (81.8%) 9/25 (36%) 0.001 Specificity 271/299 (90.6%) 21/30 (70%) 0.002 PPV 18/46 (39.1%) 9/18 (50%) 0.4 NPV 271/275 (98.5%) 21/37(56.8%) 0.001 HIV-seronegative Sensitivity 66/84 (78.6%) 20/32 (62.5%) 0.08 Specificity 233/267 (87.3%) 19/21 (90.5%) 0.9 PPV 66/100 (66%) 20/22 (90.9%) 0.02 NPV 233/251 (92.8%) 19/31 (61.3%) 0.001 IFA-ABS Single infection Mixed infection p HIV-seropositive Sensitivity 21/22 (95.5%) 18/25 (72%) 0.08 Specificity 175/299 (58.5%) 14/30 (54.5%) 0.58 PPV 21/145 (14.5%) 18/34 (52.9%) 0.001 NPV 175/176 (99.4%) 14/21 (66.7%) 0.27 HIV-seronegative Sensitivity 81/84 (96.4%) 26/32 (81.3%) 0.02 Specificity 160/267 (59.9%) 12/21 (57.1%) 0.8 PPV 81/188 (43.1%) 26/35 (74.3%) 0.001 NPV 160/163 (98.21%) 6/18 (66.7%) 0.02 Fig. 1. Distribution of RPR titres in 211 GUD patients with a positive RPR test. RPR Titre Number of patients 1 23 2 27 4 24 8 31 16 21 32 45 64 19 128 16 256 4 512 1 Note: Table made from bar graph.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Original Articles; human immunodeficiency virus|
|Author:||Ballard, R.C.; Koornhof, H.J.; Chen, C.-Y.; Radebe, F.; Fehler, H.G.; Htun, Ye|
|Publication:||South African Medical Journal|
|Article Type:||Clinical report|
|Date:||Nov 1, 2007|
|Previous Article:||Antibiotic-resistant gonococci--past, present and future.|
|Next Article:||Changes in the patterns of sexually transmitted infection among South African mineworkers, associated with the emergence of the HIV/AIDS epidemic.|