Condylomata Acuminata (Anogenital Warts) Contain Accumulations of HIV-1 Target Cells That May Provide Portals for HIV Transmission

Anogenital warts contain significantly higher concentrations of human immunodeficiency virus (HIV)–target cells than site-matched normal skin. These findings provide biologic evidence that anogenital warts may promote HIV sexual transmission and suggest that prevention or treatment of anogenital warts could decrease HIV transmission.

Abstract

Background

Condylomata acuminata (anogenital warts [AGWs]) are prevalent in human immunodeficiency virus (HIV)–infected individuals and sexually active populations at risk for HIV acquisition and have been associated with HIV transmission. We compared AGW specimens to control tissue specimens for abundance, types, and location of HIV target cells and for susceptibility to HIV infection in vitro, to provide biologic evidence that AGWs facilitate HIV transmission.

Methods

We used immunohistologic staining to identify HIV target cells in AGW and control specimens. We also inoculated HIV in vitro into AGW and control specimens from HIV-negative men and assessed infection by means of TZM-bl and p24 assays.

Results

CD1a⁺ dendritic cells, CD4⁺ T cells, and macrophages were significantly more abundant in the epidermis of AGW specimens than control specimens. These HIV target cells also often appeared in large focal accumulations in the dermis of AGW specimens. Two of 8 AGW specimens versus 0 of 8 control specimens showed robust infection with HIV in vitro.

Conclusions

Compared with normal skin, AGWs contain significantly higher concentrations of HIV target cells that may be susceptible to HIV infection. Condylomata may thus promote HIV transmission, especially in the setting of typical lesion vascularity and friability. Prevention or treatment of AGWs may decrease the sexual transmission of HIV.

Human papillomavirus (HPV) is one of the most prevalent sexually transmitted pathogens in humans; most sexually active adults are infected with at least 1 strain of HPV during their lifetime [1–3]. Of >100 known types of HPV, 13 are classified as high risk (ie, cancer causing). Low-risk HPV strains 6 and 11 are associated with the occurrence of condylomata acuminata (anogenital warts [AGWs]), benign epithelial lesions that often develop at sites that are vulnerable to abrasion or injury during sexual intercourse [4, 5]. Anal warts are observed primarily in individuals who engage in receptive anal intercourse, but they may also occur in men and women with no such history [6]. Genital warts are usually found under the foreskin or on the shaft of the penis in men and on the external genitalia/introitus in women [7]. AGWs are often asymptomatic and may not be brought to medical attention until they cause discomfort, pruritus, or bleeding.

AGWs are prevalent in groups at high risk for HIV acquisition and in HIV-infected individuals. In seronegative men who have sex with men, the prevalence of low-risk HPV infection associated with AGWs is about 20% [8, 9]. A recent meta-analysis showed that the prevalence of AGWs in HIV-uninfected populations at high risk for HIV acquisition in sub-Saharan Africa (ie, female sex workers and men and women attending sexually transmitted disease [STD] clinics) ranges from 2.4% to 14% [10]. HIV-infected men and women have a higher prevalence of HPV infection, AGWs, and premalignant and malignant lesions, compared with age-matched uninfected controls [11].

A number of recent studies have shown an increased risk of HIV acquisition in individuals with HPV infection [12–17] and those with AGWs [18, 19], and it has been speculated that HPV may enhance HIV acquisition because of inflammation and an increased numbers of HIV target cells at the infection site [20]. HPV clearance, rather than HPV infection, has also been implicated as a factor in HIV acquisition [17, 21, 22]. AGWs could also be a site of enhanced HIV shedding and transmission from HIV-infected individuals to uninfected partners; however, studies to date have not associated HPV infection or AGWs with HIV transmission.

HIV primarily infects cells that express CD4, the high-affinity HIV receptor, and either CCR5 or CXCR4, which serve as coreceptors that participate in viral fusion events [23]. Cells that express these receptors in anogenital skin, CD4⁺ T cells, macrophages, and CD1a^+high epidermal dendritic cells (DCs; also called Langerhans cells) are primary target cells in HIV sexual transmission [24, 25]. CD1a⁺ DCs are especially important because they reside in the superficial epidermis and are usually the first HIV target cell to encounter HIV [26]. Few studies have described HIV target cell populations in AGWs, and the results are varied. Mild inflammatory infiltrates have been described in AGWs, along with CD4⁺ T cells in the stroma [20]. CD1a⁺ DCs have been described to be either unchanged in number or depleted in AGWs [20, 27, 28]. Regressing AGWs have been described to contain infiltrates of CD4⁺ and CD8⁺ T cells [29, 30] and increased numbers of CD1⁺ DCs [22].

The purpose of this study was to systematically compare the abundance, types, and location of HIV target cells in AGW specimens to those in anatomically matched control skin specimens from a large collection of samples from HIV-infected and HIV-negative individuals. HIV target cells (CD1a⁺ DCs, CD4⁺ lymphocytes, and macrophages) and cells expressing HIV coreceptors (CCR5 and CXCR4) were detected by immunohistochemical staining. Presence and numbers of HIV target cells in AGW specimens were correlated with tissue site, HIV status, and other clinical information, as well as with markers of acute inflammation (CD15⁺ granulocytes) and wart regression (CD8⁺ T cells and macrophages). We also used an anti-HIV p24 antibody to detect HIV-infected cells in AGW and matched control tissue specimens from HIV-infected individuals, and we conducted a pilot study of freshly collected biopsy tissue samples from HIV-seronegative men to determine whether AGWs are more susceptible to HIV infection in vitro as compared to matched normal control tissue specimens.

MATERIALS AND METHODS

Immunohistochemical Study

Tissue Samples

This study was approved by the Institutional Review Board (IRB) of Boston University Medical Center (BUMC; Boston, MA). Archived samples of biopsied warts from 91 subjects with AGW (59 with anal warts and 32 with genital warts) were available for analysis from the BUMC Pathology Department. After medical records were reviewed for relevant information, samples were coded and patient identifiers were eliminated to preserve confidentiality.

Low-grade anal wart specimens from 47 men and 12 women and genital wart specimens from 5 men and 27 women were analyzed. Most specimens from women were from vulvar condylomata; all specimens from men were from penile warts. Normal skin specimens located outside the margins of the AGW lesions (n = 40) and archived normal genital skin specimens from men and women with no signs of AGWs (n = 26) were included as controls.

Clinical Information

Patient age, sex, race/ethnicity, and lesion type/location were obtained from coded pathology reports. Subjects ranged in age from 19 to 70 years (median, 38 years). All AGW specimens used for the study were classified as low-grade dysplasia. Thirty-six percent of subjects were HIV infected (23 of 59 with anal warts and 10 of 32 with genital warts). All HIV-infected subjects had been prescribed antiretroviral drugs, although one third had detectable HIV in blood at their last clinical visit.

Detailed clinical information was available for the subjects with anal warts. Many had received topical therapy for their lesions, including imiquimod (19%), trichloroacetic acid (8%), or podophyllum (13%); 19% had completed treatment with ≥2 agents, while 4% had completed treatment with ≥3. Few patients (8%) had a history of systemic steroid or immunomodulatory therapy in the 6 months prior to anal wart resection. Only 2 subjects had received an HPV vaccine. Ten subjects had a history of an STD in the past 6 months, with chlamydia reported in 6%, gonorrhea in 4%, syphilis in 2%, herpes simplex virus infection in 6%, and trichomoniasis in 2%.

Immunohistochemical Analysis

All tissue specimens were fixed in formalin and embedded in paraffin. Sections were cut at 5 µ, and mounted on glass slides. The immunocytochemistry protocol has been described in detail elsewhere [31]. Primary antibodies were obtained from Dako, (Carpintaria, CA; CD1a, CD68, and p24), Biocare Medical (Concord, CA; CD4, CD8, and CD15), and Novus Biologicals (Littleton, CO; CXCR4 and CCR5). Antibodies were detected with a biotinylated secondary anti-mouse/rabbit antibody and alkaline phosphatase–labeled streptavidin and were visualized with fast red substrate that stains positive cells red (Dako). Since cells were often unevenly distributed and/or located in clusters, the following semiquantitative scoring system was used to assess cell density: +, 1–10 positive cells per 40× high-power field (HPF); ++, 11–50 cells/HPF; +++, 51–100 cells/HPF; and ++++, >100 cells/HPF [32].

HIV Infection of AGWs In Vitro

Nine HIV-uninfected men undergoing surgery for removal of anal warts at Beth Israel Deaconess Medical Center (BIDMC; Boston) were recruited for a prospective HIV infection study. The protocol was approved by the BIDMC IRB, and informed consent was obtained from all subjects.

HIV infection of explant tissue specimens was conducted following an established protocol [33]. The Q23.ENV.17 envelope expression vector (National Institutes of Health AIDS Reagent Program) was transfected into 293T cells, and the supernatant was used to infect PHA-activated peripheral blood mononuclear cells for production of HIV_Q17/23 [34] viral stock. The tissue culture infectious dose (TCID₅₀) was determined by the TZM-bl assay [35]. Biopsy specimens from AGWs and normal perianal tissue (control) were transported on ice to BUMC. Tissue specimens were cut into sections (2 mm × 2 mm × 1 mm), and replicate pieces were transferred to 24-well tissue culture plates. A total of 10³ TCID₅₀ of HIV or tissue culture medium was added, and samples were incubated overnight. On day 1, tissue explants were washed and resuspended in fresh tissue culture medium. On days 3, 7, 11, and 14 of culture, supernatants were harvested and stored at −80°C. Supernatants were tested for the presence and amount of infectious HIV, using TZM-bl cells [36]; HIV infection was assessed by the Galacto-Light Plus kit (Applied Biosystems). In addition, HIV p24 concentrations were measured in supernatants by using the Alliance HIV-1 p24 Antigen enzyme-linked immunosorbent assay kit (PerkinElmer).

Statistical Analysis

For initial analysis, differences in the abundance of HIV target cell populations in AGW versus control specimens were analyzed by 1-way analysis of variance (ANOVA) with post hoc testing by the Fisher protected least significant difference. This was followed by analysis of HIV target cell populations in AGW specimens versus tissue-matched control samples by 2-way ANOVA, with HIV-1 serostatus (a between-subjects factor) and tissue site (a within-subjects factor) as independent variables. When >1 sample was available from an individual participant, the mean of sample values was used. Data were log transformed before ANOVA, to control for nonnormal distribution and heterogeneity of variance. Correlations between numbers of different cell types in AGW specimens were performed using Spearman rank correlation coefficients. Discrete, categorical data were analyzed by the Fisher exact test (or the Freeman-Halton extension) or the McNemar test. Differences or associations were considered to be statistically significant at a P value of <.05. StatView (version 5.0.1; SAS Institute, Cary, NC) statistical software was used for data analysis.

RESULTS

Characterization and Enumeration of Cell Populations by Immunohistochemical Analysis

The focus of this study was the identification and enumeration of principal HIV target cells (CD1a⁺ DCs, CD4⁺ T cells, and macrophages) in the epidermis (ie, the top layer) of AGW specimens and site-matched normal skin specimens. An initial analysis indicated that the number of these cells was similar between anal and genital control specimens and between anal and genital wart specimens and that AGW specimens contained significantly more HIV target cells than normal anogenital skin specimens (P < .01; Figure 1).

Figure 1.

Summary of the concentration of human immunodeficiency virus (HIV) target cells in the epidermis of anogenital warts and normal control skin. Types and numbers of cases: anal (A) warts, n = 51; genital (G) warts, n = 30; A skin, n = 26; G skin, n = 40. Differences in cell distribution between A skin and G skin, and between A warts and G warts were not significant (Fisher Exact test, Freeman-Halton extension). A and G warts had significantly higher concentrations of all three HIV target cell types (CD1a+ DCs, CD4+ lymphocytes and CD68+ macrophages) than did control A and G skin (all P  ’s < 0.01).

We also described HIV target cells in the dermis and used other phenotypic markers (CCR5 and CXCR4 [HIV coreceptors], p24 [HIV-infected cells], CD15 [granulocytes], and CD8 [CD8⁺ T cells]) to further characterize the HIV target cells and identify possible cofactors related to their abundance (described below).

Normal Anogenital Skin Specimens

Because the distribution of HIV target cells was similar in normal genital and anal skin specimens, descriptions of these 2 sites have been combined. Histologic findings in normal anogenital skin specimens are shown in Figure 2A. CD1a⁺ DCs were present in low-to-moderate numbers (score, ++ or less) in the epidermis and were rarely observed in the dermis (Figures 1 and 3A). CD4⁺ T cells and CD68⁺ macrophages were rarely observed in the epidermis (Figures 1 and 4A) but were frequently observed in low-to-moderate numbers in the dermis (Figures 4A and 5A). No CD15⁺ granulocytes and few to no CXCR4⁺ or CCR5⁺ cells were observed in normal skin specimens. Low numbers of CD8⁺ lymphocytes were observed in the stratum basalis of the epidermis and were scattered in the dermis.

Figure 2.

Morphology of normal anogenital skin specimens (A) and anogenital wart (AGW) specimens (B–D) stained by hematoxylin eosin (40× original magnification). The epidermal layers (stratum corneum [SC], stratum granulosum [SG], stratum spinosum [SS], stratum basalis [SB]), dermis, and dermal papillae [DP]) are labeled. AGW specimens can demonstrate abnormal thickening of the SS (acanthosis; B), thickening of the SC (hyperkeratosis; C), and/or thickening of the SG (hypergranulosis; C). AGW specimens also characteristically show parakeratosis (nuclei in the SC) and perinuclear cytoplasmic necrosis (koilocytes; D).

Figure 3.

CD1a⁺ dendritic cells detected by immunohistologic staining (red) in normal anogenital skin specimens (A) and anogenital wart specimens (B and C). A, Normal vulvar skin specimen demonstrating a few CD1a⁺ cells in the epidermis and fewer in the dermis (20× original magnification). B, Vulvar wart specimen displaying hyperkeratosis and acanthosis with numerous CD1a⁺ cells in the epidermis (40× original magnification). C, Clitoral wart specimen with pronounced hyperkeratosis and numerous CD1a⁺ cells in the stratum corneum (SC) and stratum spinosum (40× original magnification). D, Penile wart specimen showing focal accumulations of CD1a⁺ cells in the dermis (40× original magnification).

Figure 4.

CD4⁺ lymphocytes detected by immunohistologic staining (red) in normal anogenital skin specimens (A) and anogenital wart specimens (B–D). A, Normal anal skin specimens demonstrating few CD4⁺ cells, with most located in the dermis (20× original magnification). B, Vulvar wart specimen with CD4⁺ T cells present in a localized area of the epidermis containing abundant koilocytes (40× original magnification). C, Anal wart specimen with large focal accumulations of CD4⁺ T cells in the dermis (20× original magnification). D, Penile wart specimen with focal accumulations of CD4⁺ T cells in the dermis (20× original magnification). SC, stratum corneum.

Figure 5.

CD68⁺ macrophages detected by immunohistologic staining (red) in normal anogenital skin specimens (A) and anogenital wart specimens (B–D). A, Normal vulvar skin specimen with CD68⁺ cells located primarily in the dermis (20× original magnification). B, Vulvar wart specimen showing CD68⁺ cells infiltrating koilocytes in epidermis (40× original magnification). C, Penile wart specimen with numerous CD68⁺ cells associated with accumulations of lymphocytes in the dermis (20× original magnification). D, Vulvar wart specimen showing an area of localized CD68⁺ macrophage infiltration into the epidermis (20× original magnification). SC, stratum corneum.

AGW Specimens

Because the distribution of HIV target cells was similar for genital and anal wart specimens (Figure 1), specimens from these sites have been combined in the descriptions specified below. Characteristic histologic features of AGW specimens are shown in Figure 2B–D.

CD1a⁺ DCs

Approximately half of the AGW cases had abundant (score, +++ or greater) CD1a⁺ DCs in the epidermis (Figure 1). In many samples, these cells were dispersed throughout the width of the stratum spinosum (Figure 3B), whereas in others they were concentrated in the upper epithelial layers (ie, the stratum granulosum and stratum corneum; Figure 3C) or the stratum basalis. CD1a⁺ DCs were also consistently detected in the dermis and/or dermal papillae in AGW samples. In a substantial number of specimens (20% of anal wart specimens and 28% of genital wart specimens), large numbers were observed in the dermis in clusters associated with focal accumulations of lymphocytes (Figure 3D).

CD4⁺ T Lymphocytes

Twenty percent of AGW specimens had moderately high numbers (score, ++ or greater) of CD4⁺ lymphocytes in the epidermis (Figure 1). They were often detected in regions with abundant koilocytes (Figure 4B). CD4⁺ T cells were consistently detected in the dermis of AGW specimens and were often abundant. In some samples, CD4⁺ T cells were organized in a layer just beneath the base of the epidermis; in others, they occurred as distinct focal accumulations in the dermis. These accumulations varied in number from few to numerous and varied in size from small to very large (Figures 4C and 4D). Focal accumulations of CD4⁺ T cells in the dermis usually also contained numerous CD8⁺ T cells, macrophages, and CD1a⁺ DCs.

CD68 ⁺ Macrophages

Twenty percent of AGW specimens contained moderately high numbers (score, ++ or greater) of CD68⁺ macrophages in the epidermis (Figure 1). These cells were found either throughout the epidermal layer or in localized areas often associated with koilocytes (Figure 5B). CD68⁺ macrophages were also consistently present in the dermis and dermal papillae of AGW specimens, ranging in concentration from few to very abundant. For some samples, CD68⁺ macrophages were associated with focal accumulations of lymphocytes in the dermis (Figure 5C) and were sometimes observed infiltrating from dermal aggregates into the epidermis (Figure 5D).

CXCR4⁺ Cells

This HIV coreceptor can be expressed by a variety of cell types found in AGWs, including CD4⁺ T cells, granulocytes, and keratinocytes [37]. In 11 specimens, CXCR4 was expressed on granulocytes or CD8⁺ lymphocytes in the epidermis. Varying concentrations of CXCR4⁺ cells resembling lymphocytes and macrophages were also detected in the dermis. CXCR4⁺ keratinocytes were occasionally detected in both genital and anal wart specimens in basal layers of the epithelium surrounding dermal papillae.

CCR5⁺ Cells

The CCR5 HIV coreceptor can be expressed on a number of cell types found in AGWs, including memory T cells, macrophages, granulocytes, Langerhans cells, and keratinocytes [38]. Only 12% of AGW specimens had large numbers of CCR5⁺ cells in the epidermis; most of these cells were granulocytes and were observed in the stratum spinosum. Many anal wart specimens but few genital wart specimens had a large number of CCR5⁺ lymphocytes and macrophages in the dermis; however, few CCR5⁺ cells were detected in the focal accumulations of HIV target cells present in the dermis.

HIV p24 ⁺ Cells

HIV-infected Jurkat cells were used as a positive control, and all stained positive with the p24 antibody (data not shown). No cells definitively positive for p24 were observed in 52 AGW samples from 31 HIV-infected patients. Most HIV-infected subjects were receiving antiretroviral drugs, which may have suppressed p24 expression by HIV-infected cells.

CD15 ⁺ Granulocytes

CD15⁺ granulocytes were assessed as a marker of acute inflammation [39]. Notable concentrations were observed in 17% of AGW specimens (10 anal specimens and 6 genital specimens). They often appeared in large focal accumulations in the dermis and/or in the stratum corneum and stratum spinosum; they were often observed infiltrating koilocytes.

CD8⁺ Lymphocytes

CD8⁺ lymphocytes were assessed as a potential marker of cell-mediated immunity associated with wart regression [29]. Approximately one third of AGW specimens had high concentrations of CD8⁺ lymphocytes in the epidermis, localized throughout the layer or concentrated in the stratum basalis. Many AGW specimens (54% of genital specimens and 33% of anal specimens) also had high concentrations of CD8⁺ lymphocytes in the dermis, mostly in focal aggregates. There was concurrence between CD8⁺ lymphocyte concentrations in the dermis and epidermis.

Comparison of HIV Target Cell Numbers in AGW Versus Donor-Matched Control Specimens

An in-depth statistical analysis was performed for cell counts in the epidermis of paired wart and control skin samples. Anal wart specimens (from 39 patients) had significantly higher concentrations of all 3 HIV target cell types than did donor-matched control anal tissue specimens (Table 1). Anal wart specimens had elevated concentrations of CD1a⁺ cells (P = .0001), with HIV-infected subjects having the highest concentrations (P = .03 for the interaction between tissue site and HIV infection status). Anal wart specimens also had significantly higher concentrations of CD68⁺ macrophages (P < .0001), CD4⁺ T cells (P = .0006), and CD15⁺ granulocytes (P = .0002) than control anal skin specimens. Anal wart specimens from HIV-uninfected subjects had higher concentrations of CD8⁺ cells than did those from HIV-infected subjects (P = .02 for the interaction between tissue site and HIV infection status). Elevated concentrations and focal accumulations of HIV target cell populations in anal wart specimens were not associated with use of topical therapy or recent history of an STD.

Table 1.

Human Immunodeficiency Virus (HIV) Target Cells, Granulocytes, and CD8⁺ T Cells in the Epidermis of Anogenital Wart Specimens Versus Donor/Site-Matched Control Tissue Specimens (CS) From HIV-Infected and Uninfected Subjects

Variable	Control Specimens		Wart Specimens		Pa
	HIV Negative	HIV Positive	HIV Negative	HIV Positive	HIV Negative vs HIV Positive	Wart vs Control
Anal wart vs control specimens
CD1a+ cells
Mean ± SEb	1.2 ± 0.2	1.1 ± 0.1	1.9 ± 0.3	2.6 ± 0.2	NS	<.0001
Matched pairs (n)	10	10	10	10
CD4+ cells
Mean ± SEb	0.2 ± 0.1	0.3 ± 0.1	0.8 ± 0.1	0.6 ± 0.2	NS	.0006
Matched pairs (n)	14	9	14	9
CD68+ cells
Mean ± SEb	0.04 ± 0.04	0.2 ± 0.1	0.7 ± 0.1	0.6 ± 0.1	NS	<.0001
Matched pairs (n)	12	12	12	12
CD8+ cells
Mean ± SEb	0.9 ± 0.1	0.7 ± 0.1	1.6 ± 0.2	0.6 ± 0.1	.002	NS
Matched pairs (n)	17	12	17	12
CD15+ cells
Mean ± SEb	0.03 ± 0.03	0.0 ± 0.0	0.6 ± 0.2	0.4 ± 0.1	NS	.0002
Matched pairs (n)	16	16	16	16
Genital wart vs control specimens
CD1a+ cells
Mean ± SEb	1.1 ± 0.1	0.9 ± 0.2	2.9 ± 0.4	2.7 ± 0.5	NS	.0006
Matched pairs (n)	7	5	7	5
CD4+ cells
Mean ± SEb	0.1 ± 0.1	0.3 ± 0.1	0.8 ± 0.3	0.8 ± 0.3	NS	.026
Matched pairs (n)	7	6	7	6
CD68+ cells
Mean ± SEb	0.0 ± 0.0	0.0 ± 0.0	1.2 ± 0.4	0.8 ± 0.3	NS	.015
Matched pairs (n)	3	4	3	4
CD8+ cells
Mean ± SEb	0.9 ± 0.2	0.5 ± 0.2	1.7 ± 0.2	1.4 ± 0.2	NS	.002
Matched pairs (n)	9	4	9	4
CD15+ cells
Mean ± SEb	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0	NS	NS
Matched pairs (n)	4	4	4	4

Abbreviations: NS, not significant; SE, standard error.

^aBy two-way analysis of variance (main effects).

^bData are semiquantitative cell density values.

Paired genital wart and control skin samples were available from 13 women. Genital wart specimens contained significantly higher concentrations of CD1a⁺ (P = .0006), CD68⁺ (P = .015), CD4⁺ (P = .026), and CD8⁺ (P = .002) T cells in the epidermis than did donor-matched control tissue specimens; HIV-1 serostatus was not significant for any of the white blood cell counts (Table 1). There were no donor-matched tissue specimen pairs available from men with genital warts, but data were compared for penile wart specimens from 5 subjects and normal penile tissue specimens from 5 unrelated subjects. The epidermal layer of male genital wart specimens had significantly higher concentrations of CD1a⁺ (P = .004) and CD68⁺ (P = .009) cells (both P values were determined by the Mann-Whitney U test).

Correlations Between HIV Target Cells, CD15⁺ Granulocytes, and CD8⁺ T Cells in AGW Specimens

Granulocyte concentrations were positively correlated with CD68⁺ cell numbers in the AGW epidermis (ρ = 0.54; P < .0001) but not with CD1a⁺ DCs or CD4⁺ lymphocytes. CD8⁺ T-cell concentrations were not correlated with numbers of granulocytes or any of the HIV target cells in the epidermis. However, all 3 HIV target cell types were often found associated with CD8⁺ T cells in focal aggregates in the dermis layer.

Infection of AGW and Control Specimens With HIV In Vitro

Donor-matched samples of anal warts and normal anal epithelium were obtained from 9 HIV-negative men (89% were white; age range, 20–29 years). All wart samples were anal intraepithelial neoplasia grade 1–2 on pathologic review. One case was discarded because of contamination. Two of 8 AGW samples versus 0 normal tissue samples showed definitive signs of HIV infection, as evidenced by robust TZM-bl values (>1000 relative light units on >2 days of culture) and p24 assay confirmation (Figure 6); because of the small sample size, these data were not significant (P > .10).

Figure 6.

Human immunodeficiency virus (HIV) infection of anal wart specimens and control anal skin explant tissue specimens from HIV-uninfected men in vitro. Two of 8 anal wart specimens, compared with 0 of 8 control tissue specimens, showed high levels of HIV infection in the TZM-bl assay.

DISCUSSION

HPV is highly infectious and is transmitted between persons via microabrasions during sexual skin-to-skin contact [7, 40]. Most initial HPV infections occur in adolescence, making early prevention critical. After initial HPV infection, patients may have clearance of subclinical infection, development of lesions with subsequent regression, persistence of infection with subsequent regression, or persistence of infection with development of cancer. Most HPV infections are subclinical and are cleared by the immune system via the cell-mediated immune response before any clinical disease is observed [41]. Clinical AGWs may also undergo regression as a result of cell-mediated immunity [29].

Our study demonstrates that the density of principal HIV target cells (CD1a⁺ DCs, CD4⁺ T cells, and CD68⁺ macrophages) is significantly higher in AGW tissue than donor/tissue-matched control skin tissue. Approximately half of AGW specimens had elevated concentrations of DCs in the epidermis. It is generally accepted that DCs play a pivotal role in the initial events of HIV transmission by transferring HIV to CD4⁺ T cells [26]. A subset of AGW specimens also had elevated numbers of macrophages and CD4⁺ T cells in the epidermis. There was a strong association between numbers of macrophages and granulocytes in the epidermis, suggesting the recruitment of macrophages to the superficial layer of AGWs during acute inflammation. Macrophages and CD4⁺ T cells were often observed near koilocytes, classical HPV-infected cells with enlarged nuclei and perinuclear halos, indicating that HIV target cells accumulate in AGWs at sites of active HPV infection.

Previous studies have described immune cells in human AGWs, with varying results [29, 42–45]. Two studies that analyzed naturally regressing warts described increased concentrations of lymphocytes and macrophages [29] and DCs [22]. Although we did not specifically study wart regression, approximately one third of our patients with AGWs had elevated numbers of CD8⁺ T cells (a marker of cell-mediated immunity) in the epidermis and/or massive focal infiltrates in the dermis. Whereas CD8⁺ T cells in the epidermis did not correlate with elevated numbers of HIV target cells at that site, CD8⁺ dermal infiltrates contained numerous CD4⁺ T cells, macrophages, and DCs, providing evidence that AGW regression may be associated with the recruitment of high densities of HIV target cells to the dermis layer.

A recent study demonstrated HIV-1 RNA and episomal DNA in intraanal high-grade HPV lesions in HIV-infected men who have sex with men and speculated that receptive anal intercourse with HIV-infected individuals that have HPV-associated lesions may increase the risk for HIV acquisition [46]. This is an intriguing hypothesis, although the presence of HIV nucleic acid does not necessarily indicate presence of infectious virus. In our study we investigated evidence of assembly of HIV viral particles (p24 protein) in AGW tissues from HIV-infected individuals. We not detect p24-positive cells in AGW from HIV-infected individuals, potentially due to viral suppression by antiretroviral drugs. This research should be extended to individuals who are not receiving antiretroviral therapy, to determine whether AGWs can be foci of HIV infection. In a related experiment, we challenged AGW and normal tissue explants from uninfected individuals with HIV in vitro; 2 AGW specimens and no donor-matched control specimens showed evidence of robust HIV infection, suggesting that some AGWs may be highly susceptible to HIV infection. We propose that future studies be conducted to correlate HIV target cell types and numbers with HIV infectability in vitro.

In summary, our study demonstrates a significant increase in the number of HIV target cells in AGWs and provides evidence of enhanced HIV infection of AGW in vitro. These findings suggest that AGWs may provide portals for HIV transmission and reinforce the importance of prevention and treatment of AGWs to control the HIV epidemic. Potential susceptibility of AGWs to HIV infection provides additional impetus for vaccination of children and adolescents before the age of sexual debut with HPV vaccines that induce immunity to strains associated with AGWs (ie, HPV types 6 and 11) [47]. Prelicensure trial efficacy, modeling, and postvaccination surveillance studies of quadrivalent vaccine programs have already demonstrated the significant short-term impact of this vaccine on the incidence and prevalence of AGWs in the United States [48, 49].

The new nonavalent HPV vaccine is expected to show similar results. Large-scale roll out of HPV vaccines in HIV-endemic areas such as sub-Saharan Africa could significantly impact the HIV epidemic in these regions.

Notes

Acknowledgments. We thank Dr Antonio de las Morenas, Department of Pathology, Boston University School of Medicine, for providing access to samples; Oscar Gonzalez, PhD, Sagar Laboratory, Boston University, for performing p24 assays for the prospective study; Lindsay Bohnert, MS, and Tesfaldet Tecle, PhD, Boston University, for providing assistance with immunohistochemical and infection assays; and Linda Rosen, MSEE, Boston University Clinical Data Warehouse, for her assistance in obtaining clinical information for the immunohistochemical data set.

Financial support. This work was supported by the National Institutes of Health (grant U19 AI096398) and Merck (IISP grant #37966) both to D. Anderson. Supported in part by a research grant from Investigator-Initiated Studies Program of Merck Sharp & Dohme Corp. The opinions expressed in this paper are those of the authors and do not necessarily represent those of Merck Sharp & Dohme Corp.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.