| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Departments of 1Obstetrics and Gynecology, 2Microbiology, and 3Infectious Diseases, IRCCS Policlinico S. Matteo, University of Pavia, Pavia, Italy.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT PATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Two hundred sixteen HIV-seropositive patients were followed up by Pap test, colposcopy, and targeted cervical biopsies for a median of 16 months (range 0–94). A diagnosis of low-grade CIN was made on the basis of Pap test and either colposcopy or cervical biopsy. High-grade CIN was diagnosed solely on the basis of cervical biopsy. At each follow-up visit, we measured HIV-1 RNA in plasma, proviral HIV-1 DNA, and cell-associated and cell-free HIV-1 RNA in cervicovaginal secretion by competitive polymerase chain reaction (cRT-PCR) and reverse transcriptase PCR. The univariable and multivariable associations between the occurrence of CIN and the presence of HIV-related nucleic acids in cervicovaginal secretions were evaluated with logistic generalized estimating equations.
RESULTS: Overall, at enrollment and during the follow-up period, a diagnosis of either low-grade or high-grade CIN was made in 14.4% (99/689) and 6.7% (46/689) of the visits, respectively. The presence of measurable levels of plasma HIV-1 RNA was a significant risk factor for the detection of cervicovaginal HIV-1 DNA (odds ratio [OR] 1.86, 95% confidence interval [CI] 1.32–2.61, P < .001), cell-associated (OR 1.69, 95% CI 1.18–2.43, P = .004), and cell-free HIV-1 RNA (OR 1.84, 95% CI 1.28–2.63, P = .001). After the adjustment for the effect of plasma HIV-1 RNA, CD4+ positive cell counts less than 200 mm3, and bacterial vaginosis, the detection of cell-associated (OR 1.75, 95% CI 1.23–2.49, P = .006) and cell-free HIV-1 RNA (OR 2.0, 95% CI 1.39–2.87, P = .001) in cervicovaginal secretions was significantly associated with the diagnosis of CIN.
CONCLUSION: The presence of CIN lesions is a significant risk factor for genital HIV shedding. Given the high prevalence of cervical disease among HIV-positive women, this finding could have important epidemiological implications in both heterosexual and perinatal transmission of HIV.
LEVEL OF EVIDENCE: II-2
|
PATIENTS AND METHODS |
|---|
|
TOP
ABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Each gynecologic examination included a standard Pap test and a collection of vaginal specimens with sterile cotton swabs to diagnose infection by Candida, trichomonas, or bacterial vaginosis. To evaluate Candida infection, vaginal specimens were inoculated on Saboraud’s dextrose agar containing gentamicin. Vaginal specimens for trichomonas isolation were inoculated in a cysteine-peptose-liver maltose medium, whereas the diagnosis of vaginosis was made according to the criteria of Amsel et al.7 A standardized colposcopic examination of the lower genital tract was performed after the result of Pap test was known. Colposcopic lesions were classified into minor and major colposcopic abnormalities. Minor abnormalities included areas of mild acetowhitening without vascular changes. Major abnormalities included areas of dense white epithelium with sharp margins, prominent mosaic, or coarsely punctuated vascular patterns. High-grade lesions on cytologic or colposcopic screening were routinely biopsied. During the period of the study, HPV DNA detection was not routinely performed in these patients. The result of cervical screening was categorized as follows: 1) Negative screening: negative Pap test and absence of abnormalities on colposcopy. 2) Low-grade CIN: low-grade SIL on cytology confirmed either by minor colposcopic abnormalities or cervical biopsy. 3) High grade CIN: moderate-to-severe cervical dysplasia confirmed by cervical biopsy.
Peripheral blood samples for estimating CD4+ lymphocyte cell counts and quantifying HIV-1 RNA in plasma were collected from the subjects at each visit. Cervicovaginal secretions for HIV-related nucleic acid identification were obtained by gently rotating a Dacron swab within the posterior fornix and by lavage after the insertion of 10 mL of RPMI-1640 medium into the vagina, followed by aspiration of the suspension after allowing 1 minute for pooling. Swabs were used to detect cell-free HIV-1 RNA, whereas proviral HIV-DNA and intracellular HIV-RNA transcripts were detected from lavage samples. Upon arrival in the laboratory, and again after centrifugation, samples were examined under the microscope to confirm the absence of red blood cells and spermatozoa. The possible presence of blood contamination was further checked by using a routine screening test for hemoglobin detection (reactive strips, Bayer, multistic-10 visual). A detailed description of methods used to detect and quantify HIV-related nucleic acids from blood and cervicovaginal secretions has been reported elsewhere.8 Swabs were incubated in 1 mL of RPMI 1640 medium, vortexed, and centrifuged. The supernatant was used to extract cell-free HIV-1 RNA. Human immunovirus-1 DNA was extracted from the nuclei of cervicovaginal cells to minimize contamination with unintegrated DNA. Ribonucleic acid was extracted from cervicovaginal and plasma samples by using the guanidinium thiocyanate method.9 The following substrates were analyzed by quantitative polymerase chain reaction (c-PCR) and reverse transcription polymerase chain reaction (RT-PCR): 1) genomic HIV-1 RNA from plasma and cell-free cervicovaginal secretions, 2) virus-specific unspliced HIV-1 RNA transcripts from cervicovaginal cells, and 3) proviral HIV-1 DNA from cervicovaginal nuclei.
Qualitative analysis of specific RNA and DNA sequences was first performed by PCR using the SK 426/431 pair of primers.10 Ribonucleic acid samples were reverse-transcribed by using 100 U of Moloney murine leukemia virus reverse transcriptase (RT, Gibco Life Technologies, Paisley, UK), 20 pmol of the SK 431 primer, 0.2 mmol/L deoxynucleotide triphosphate (dNTP), and 20 U of Rnasin (Gibco Life Technologies). Deoxyribonucleic acid was subsequently amplified using 50 pmol of primers SK 462 and SK 431 and 1.5 U of Taq-DNA polymerase (Perkin-Elmer Cetus, Emeryville, CA). Quantification of HIV-1 DNA and RNA was performed by using a competitive PCR and RT-PCR as described by Menzo et al.10
For cervicovaginal samples, quantitative results are expressed as HIV-1 RNA transcripts and HIV-1 DNA copy number per 105 cells and as HIV-1 RNA cell-free copy number per milliliter of cervicovaginal secretions. The dilution factor of the cervicovaginal sample with the known amount of transport medium (1 mL/swab sample) was taken into account for the final calculation of the cell-free HIV-1 RNA copy numbers. Each swab sample contained 200–300 µL of cervicovaginal lavage. In our experimental conditions, the lower limit of detection of the assay was 2 DNA or RNA copies/105 cells, 20 RNA copies/mL of cervicovaginal secretions, and 200 RNA copies/mL of plasma. To evaluate the relationship between increasing concentrations of cervicovaginal HIV and explanatory variables, the loads of HIV were categorized into 3 groups (negative, 
50th percentile, and > 50th percentile).
Spearman rank correlation coefficient was used to analyze the association between the severity of SIL and HIV-1 related loads in cervicovaginal secretions. Univariable and multivariable associations between CIN and HIV-related nucleic acids in cervicovaginal secretions were evaluated with logistic generalized estimating equations.11 This procedure provides estimates of odds ratios via methods that compensated for within-subject correlation across multiple observations.12 The generalized estimating equation method estimates the regression parameters assuming that the observations are independent, uses the residuals from this model to estimate the correlations among observations from the same subjects, and then uses the correlation estimates to obtain new estimates of the regression parameters. This process is repeated until the change between 2 successive estimates is very small. Computations were performed with STATA/SE 8.0 (Stata Corporation, College Station, TX). Because this was not a randomized study, we did not compute a preliminary sample size. However, assuming a within-group correlation of 0.4 and a median of 4 visits per case, a sample size of 142 subjects would be sufficient to demonstrate an increase in the rate of HIV-related nucleic acid detection in cervicovaginal secretion from 30% to 50%, at the usual values of 
= 0.05 and ß = 0.1.13
|
RESULTS |
|---|
|
TOP
ABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Overall, at enrollment and during the follow-up period, cervical biopsies were taken in 74 patients (10.7%). A diagnosis of low-grade CIN was made in 14.4% of visits (99/689), including 88 (12 biopsies) cases with cytologic and colposcopic screening suggestive for low-grade lesion, 8 cases with negative or atypical squamous cells of undetermined significance (ASC-US) cytology and low-grade lesion at biopsy, and 3 high-grade lesions on cytology that turned out to be low-grade lesions at biopsy. A high-grade CIN on biopsy was diagnosed in 46 of 689 visits (6.7%), including 38 cases with cytology and colposcopy suggestive of high-grade lesion, 4 cases with ASC-US cytology, and 4 cases of low-grade SIL on cytology. Finally, in 5 visits with positive cytology (either ASC-US or low-grade SIL) but negative biopsy, the screening of cervical lesions was considered negative. All but 2 cases of high-grade CIN lesions were treated either by loop electrosurgical excision procedure (LEEP) or cold-knife conization. Two cases of high-grade CIN with multiple recurrences were treated by vaginal 5-fluorouracil. Low-grade lesions were followed up every 3–6 months without treatment in 73.7% (73/99) of cases. The remaining cases of low-grade lesion (26/99) were treated by either LEEP or diathermy.
The median number of HIV-1 RNA genomes during the study was 1,400 copies/mL (range 50–400,000) in plasma samples and 110 copies/mL (range 5–5,800) in cell-free cervicovaginal samples. The median number of HIV-1 DNA copies and HIV-1 RNA transcripts per 105 cells was 50 (range 5–1,500) and 60 (range 2–2,000), respectively. The presence of measurable levels of plasma HIV-1 RNA was a significant risk factor for the detection of cervicovaginal HIV-1 DNA (127/227 compared with 183/462, OR 1.86, 95% CI 1.32–2.61, P < .001 by logistic generalized estimating equation model), cell-associated (102/187 compared with 208/502, OR 1.69, 95% CI 1.18–2.43, P = .004), and cell-free HIV-1 RNA (113/199 compared with 197/490, OR 1.84, 95% CI 1.28–2.63, P = .001). The HIV-1 RNA viral load in plasma correlated with the amounts of HIV-1 DNA (Spearman rho 0.17 ± 0.04, P < .001), cell-associated (Spearman rho 0.16 ± 0.04, P < .001), and cell-free HIV-1 RNA (Spearman rho 0.15 ± 0.04, P < .001). In generalized estimating equation models, antiretroviral treatment was associated with a significant reduction in the shedding of HIV-DNA (OR 0.41, 95% CI 0.27–0.64, P < .001), cell-associated (OR 0.62, 95% CI 0.40–0.98, P = .04), and cell-free HIV-1 RNA (OR 0.66, 95% CI 0.44–0.99, P = .05)
Table 2 reports the results of univariable analysis using logistic generalized estimating equation models. After adjusting for within-subject correlation across multiple observations, the presence of cervical neoplasia was significantly associated with detectable blood HIV-1 RNA, bacterial vaginosis, and a CD4+ cell count less than 200 mm.3 These variables, together with the presence and loads of HIV-1 related nucleic acids in cervicovaginal secretions, were inserted as explanatory variables in generalized estimating equation models including the diagnosis of CIN as independent variable. The results of this modeling, together with univariable analysis, are reported in Table 3. After the adjustment for the potential confounders reported above, high loads of proviral HIV-1 DNA and cell-associated and cell-free HIV-1 RNA in cervicovaginal secretions were significantly correlated with the diagnosis of cervical neoplasia. Overall, the presence of detectable cell-associated (OR 1.75, 95% CI 1.23–2.49, P = .006) or cell-free HIV-1 RNA (OR 2.0, 95% CI 1.39–2.87, P = .001) in cervicovaginal secretions was significantly more frequent among patients with cervical neoplasia than in negative controls.
|
|
Table 4 reports the results of multivariable analysis of the association between the clinical category of CIN (first-time diagnosis versus persistence/recurrence) and the presence of HIV-related nucleic acids in cervicovaginal secretions. Although we could not find a statistically significant difference in the concentrations of HIV-related nucleic acids between the 2 groups of patients, first-time diagnosis of cervical neoplasia was more uniformly correlated with the presence of HIV in cervicovaginal secretions.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The high rates both of prevalence of CIN at the beginning of the study and of persistences/recurrences during the follow-up in this series are consistent with the results of several other studies.3,4,15,16 In addition, the high incidence (13.7%) of CIN during a 16-month follow-up is comparable with the figure (20%) reported by Ellerbrock et al16 during a 30-month follow-up. Although HIV-related immunodeficiency is a well-known risk factor for cervical neoplasia, the relationship between HIV genital shedding and CIN has been poorly studied. Factors associated with HIV-1 genital shedding in the female genital tract in previous studies included, among others, CD4 cell count, concentration of plasma HIV-RNA, genital ulcerations or other sexually transmitted diseases, and use of hormonal contraceptives.2 In addition to these variables, the presence of cervical ectopy has been associated with an increased risk for heterosexual transmission of HIV, and cervical inflammation has been associated with increased HIV cervical shedding.17
In our study we found, in both univariable and multivariable models, a significant association between all the measures of HIV shedding and a diagnosis of cervical neoplasia. Contrary to the findings of a previous cross-sectional study,5 this association remained significant even after we corrected for potential confounders such as the concentration of HIV in plasma, CD4+ cell counts, and vaginal infection. Differences in the study populations, in the design of the study, and in the sensitivities of the methods used could account for the differences between the 2 investigations. The simultaneous measure of both cell-associated and cell-free HIV-1 RNA and the detection of HIV-1 DNA in our study is a more complete assessment of the genital HIV-shedding than the exclusive measure of cell-free HIV-1 RNA reported by Kovacs et al.5
Biological data support the association between HIV genital shedding and cervical neoplasia. Studies of CIN biopsies of HIV-positive women have shown that these patients have increased lymphocyte and macrophage infiltrates compared with either normal cervix or CIN biopsies of HIV-negative controls.18 Wright et al18 found that cervical inflammation and ulceration associated with treatment of cervical neoplasia was associated with a 10-fold to more than 10,000-fold increase of HIV-1 in cervicovaginal secretions, and this increase was the result of local viral replication in the genital tract. All these data suggest that the association between HIV genital shedding and cervical neoplasia is biologically plausible and possibly mediated by the immune upregulation at the site of the cervical lesion. This could lead to both an increased number of intraepithelial lymphocytes and macrophages in the squamous epithelium of the cervix and to an increased intralesional concentration of proinflammatory cytokines modulating HIV replication.2,15,18 From our data, recurrence/persistence of CIN did not seem to influence heavily the shedding of HIV. This finding is consistent with recent data correlating plasma HIV-1 RNA to SIL. In fact, Strickler et al15, in a large cohort of HIV-positive women, found that the association between plasma HIV-1 RNA and cervical lesions was stronger for incident than for persistent SIL.
It is possible that the association between CIN and HIV genital shedding could be mediated by HPV infection. Genital HPV infection, which is strongly correlated with cervical neoplasia, has been shown to be more frequent and severe among HIV-infected women than negative controls.3,15 In addition, Luque et al20 found that the prevalence of HPV infections was strongly correlated with HIV-1 plasma levels. Studies of cervical biopsies of HIV-positive women suggest that HPV infection is associated with increased intralesional cytokine levels modulating HIV expression.21 In our study HPV detection was not carried out routinely, and we did not have precise data on the relative role of HPV infection and CIN on HIV genital shedding. Although, in a small cohort,22 we found that HPV cervical infection was more frequent among patients with high HIV-1 RNA, but not proviral DNA, genital shedding, factors regulating the simultaneous presence and replication of HIV and HPV in cervical tissue need to be clarified in larger studies. The small number of high-grade cervical lesions in our investigation did not permit a reliable analysis of the effects of the severity of CIN on the presence and load of HIV-related nucleic acids in cervicovaginal secretions. At present, on the basis of our data, the presence of CIN lesions should be considered a significant risk factor for genital HIV shedding. Given the high prevalence of cervical disease among HIV-positive women, this finding could have important epidemiologic implications. Future studies should clarify whether early treatment of CIN lesions in these patients could reduce, at least in the long-term, HIV genital shedding and, possibly, sexual transmission.
|
Footnotes |
|---|
Corresponding author: Arsenio Spinillo, MD, Department of Obstetrics and Gynecology, IRCCS Policlinico S. Matteo P.le Golgi, 27100 Pavia, Italy; e-mail: spinillo{at}smatteo.pv.it.
doi:10.1097/01.AOG.0000196723.76228.ea
|
REFERENCES |
|---|
|
TOP
ABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Coombs RW, Reichelderfer PS, Landay AL. Recent observations on HIV type-1 infection in the genital tract of men and women AIDS 2003;17:455–80.[Medline]
3. Heard I, Potard V, Foulot H, Chapron C, Costagliola D, Kazatchkine MD. High rate of recurrence of cervical intraepithelial neoplasia after surgery in HIV-positive women. J Acquir Immune Defic Syndr 2005;39:412–8.[Medline]
4. Davis AT, Chakraborty H, Flowers L, Mosunjac MB. Cervical dysplasia in women infected with the human immunodeficiency virus (HIV): a correlation with HIV viral load and CD4+ count. Gynecol Oncol 2001;80:350–4.[Medline]
5. Kovacs A, Wasserman SS, Burns D, Wright DJ, Cohn J, Landay A, et al. Determinants of HIV-1 shedding in the genital tract of women. Lancet 2001;358:1593–601.[Medline]
6. Andreoletti L, Chomont N, Gresenguet G, Matta M, de Dieu Longo J, Carreno MP, et al. Independent levels of cell-free and cell-associated human immunodeficiencyvirus-1 in genital-tract secretions of clinically asymptomatic, treatment-naive African women. J Infect Dis 2003;188:549–54.[Medline]
7. Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach D, Holmes KK. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983;74:14–22.[Medline]
8. Spinillo A, Zara F, De Santolo A, Brerra R, Maserati R, Romero E, et al. Quantitative assessment of cell-associated and cell-free virus in cervicovaginal samples of HIV-1-infected women. Clin Microbiol Infect 1999;5:605–11.[Medline]
9. Ehrlich GD. PCR-based methods for the detection of the human Retroviridae and Hepadnaviridae. In: Ehrlich GD, Greenberg SJ, editors. PCR-based diagnostics in infectious disease. Boston (MA): Blackwell Scientific Publications; 1994. p. 415–46.
10. Menzo S, Bagnarelli P, Giacca M, Manzin A, Varaldo PE, Clementi M. Absolute quantitation of viremia in human immunodeficiency virus infection by competitive reverse transcription and polymerase chain reaction. J Clin Microbiol 1992;30:1752–7.
11. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics 1988;44:1049–60.[Medline]
12. Lipsitz SR, Kim K, Zhao L. Analysis of repeated categorical data using generalized estimating equations. Stat Med 1994;13:1149–63.[Medline]
13. Pan W. Sample size and power calculations with correlated binary data. Control Clin Trials 2001:22;211–27.[Medline]
14. Hart CE, Lennox JL, Pratt-Plamore M, Wright TC, Schinaz RF, Evans-Strickfaden T. Correlation of human immunodeficiency virus type 1 RNA levels in blood and the female genital tract. J Infect Dis 1999;179:871–82.[Medline]
15. Strickler HD, Burk RD, Fazzari M, Anastos K, Minkoff H, Massad LS, et al. Natural history and possible reactivation of human papillomavirus in human immunodeficiency virus-positive women. J Natl Cancer Inst 2005;97:577–86.
16. Ellerbrock TV, Chiasson MA, Bush TJ, Sun XW, Sawo D, Brudney K, et al. Incidence of cervical squamous intraepithelial lesions in HIV-infected women. JAMA 2000;283:1031–7.
17. Plourde PJ, Pepin J, Agoki E, Ronald AR, Ombette J, Tyndall M, et al. Human immunodeficiency virus type 1 seroconversion in women with genital ulcers. J Infect Dis 1994;170:313–7.[Medline]
18. Wright TC Jr., Subbarao S, Ellerbrock TV, Lennox JL, Evans-Strickfaden T, Smith DG, et al. Human immunodeficiency virus 1 expression in the female genital tract in association with cervical inflammation and ulceration. Am J Obstet Gynecol 2001;184:279–85.[Medline]
19. Bell MC, Schmidt-Grimminger D, Turbat-Herrera E, Tucker A, Harkins L, Prentice N, et al. HIV+ patients have increased lymphocyte infiltrates in CIN lesions. Gynecol Oncol 2000;76:315–9.[Medline]
20. Luque AE, Demeter LM, Reichman RC. Association of human papillomavirus infection and disease with magnitude of human immunodeficiency virus type 1 (HIV-1) RNA plasma level among women with HIV-1 infection. J Infect Dis 1999;179:1405–9.[Medline]
21. Nicol AF, Fernandes AT, Grinsztejn B, Russomano F, E Silva JR, Tristao A, et al. Distribution of immune cell subsets and cytokine-producing cells in the uterine cervix of human papillomavirus (HPV)-infected women: influence of HIV-1 coinfection. Diagn Mol Pathol 2005;14:39–47.[Medline]
22. Spinillo A, Debiaggi M, Zara F, De Santolo A, Polatti F, Filice G. Human immunodeficiency virus type 1-related nucleic acids and papillomavirus DNA in cervicovaginal secretions of immunodeficiency virus-infected women. Obstet Gynecol 2001;97:999–1004.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
