The Influence of Human Immunodeficiency Virus Infection on Langerhans Cell Counts in the Normal Cervical Tissue

To the Editor: In the past few years, several studies have shown that human immunodeficiency virus (HIV)-seropositive women are at a higher risk of cervical intraepithelial neoplasia (CIN) and human papillomavirus (HPV)-related lesions of the anogenital tract (1-3). The severity of cervical lesions correlates well with the degree of HIV-induced immunosuppression (3-5). Although HIV-induced general immunosuppression may, at least in part, explain the increased risk of CIN in HIV-seropositive patients, local immunologic factors could also play a significant role (2,6). Epithelial Langerhans cells of the cervix act as antigen-presenting cells and have a central role in the process of specific T-cell activation (7,8). In a previous study (6) we found that the number of Langerhans cells was significantly lower in CIN lesions of HIV-seropositive patients than in similar lesions of HIV-seronegative controls. In the present study we have evaluated the number of Langerhans cells in biopsy specimens of the transformation zone obtained from HIV-seropositive and -seronegative women with normal cytologic and colposcopic findings. The subjects participating in this study were enrolled from among past or current intravenous drug abusers attending several drug dependency units and referred to our outpatient clinic for periodic screening for gynecologic diseases during a 2-year period (1991-1993). After informed consent, we enrolled 14 known HIV-seropositive women and 14 HIV-seronegative controls with negative Pap smears and colposcopic examinations and no history of cervical intraepithelial neoplasia. All the patients were past or current intravenous heroin users, and their HIV serologic status had been confirmed by serology tests performed in the 3 months before the study. A standardized history was taken from each woman, including details of cigarette smoking, sexual activity, and contraceptive use. Vaginal, endocervical, and urethral specimens for the diagnosis of bacterial vaginosis, C. trachomatis, Neisseria gonorrhoeae, Trichomonas vaginalis, and candida infection were also obtained. After a thorough colposcopic examination, all patients underwent cervical biopsy of the transformation zone by a colposcopist unaware of the patient's HIV status. Paraffin-embedded biopsy specimens were processed for conventional histology. In situ hybridization for detection of HPV-DNA 6/11, 16/18, 31/35/51 was carried out on paraffin sections using biotinylated probes (Patho-Gene; ENZO Diagnostics, Inc., New York, NY, U.S.A.). Two biopsy specimens from the HIV-seropositive groups and one from the HIV-seronegative controls were judged abnormal (histological evidence of HPV infection) by a pathologist unaware of the patients' HIV status and stained positive for HPV-DNA 6/11. These subjects were excluded from the study, thus leaving a final study population made up of 12 patients in the HIV-positive and 13 in the HIV-negative group whose biopsy specimens stained negative for HPV-DNA. Langerhans cell bodies were stained in deparaffinized sections as previously reported (9). Langerhans cell bodies were counted per 100 basal cells at 250 magnifications by a pathologist blind to the patients' HIV status. The number of readings per biopsy specimen and the concentration of Langerhans cells per 100 basal cells are given as medians and interquartile ranges (25th and 75th percentiles). Differences in median Langerhans cell counts per 100 basal cells between subgroups of patients were tested by median test. Fisher's exact test was used to compare proportions. Among HIV-seropositive patients, eight (66.7%) had asymptomatic HIV infection (category A of the CDC classification), three had symptomatic infection (two patients had constitutional symptoms, and one had idiopathic thrombocytopenic purpura), and one had AIDS (Pneumocystis carinii pneumonia). Table 1 reports the characteristics of the subjects under study. Although all the patients were asymptomatic with regard to cervicovaginal diseases, colonization by candida, trichomonas, and C. trachomatis was common in both groups. CD4+ blood cell counts within 3 months of the visit were significantly lower in the cases than in controls. No other significant differences among the variables studied were found between the two groups. The median number of readings per biopsy specimen was eight (range of six to 10) in the HIV-seropositive and seven (range of 5.5 to 11.5) in the HIV-seronegative patients (p = 0.54 by median test). The concentration of Langerhans cells per 100 basal cells was 1.1 (range of 0-1.8) in the HIV-seropositive and 1.7 (range 0.5-2.4) in the HIV-seronegative patients (p = 0.0003 by median test). In addition, no S-100-positive cells were detected in five of 12 HIV-seropositive patients, while all of the HIV-seronegative controls had S-100-positive cells in their biopsies (p = 0.015 by Fisher's exact test). Among HIV-seropositive subjects, the concentration of Langerhans cells per 100 basal cells was lower in patients with symptomatic HIV infection (group B and C of the CDC classification system) than in patients with asymptomatic HIV infection (group A of the CDC classification) (median = 0 [range 0-1.1] as compared with 1.3 [range 0-3.6], respectively; p = 0.0001 by median test). In the HIV-positive group the median number of readings taken for each biopsy specimen was seven (range of six to 12.5) in asymptomatic and 7.5 (range of six to 12.5) in symptomatic patients (p = 0.47 by median test). The results of this study have shown that Langerhans cell counts are significantly lower in the normal transformation zone of HIV-seropositive women. Although it remains possible that low levels of HPV expression not detected by in situ hybridization could have affected final results, cases and controls were homogeneous with regard to both cytologic and histologic findings. We have already shown that Langerhans cell counts are lower in CIN lesions from HIV-seropositive compared with HIV-seronegative women (6). However, Langerhans cell counts in cervical dysplastic tissue are markedly influenced by the severity of CIN and coexistent HPV infection (6-8). These two variables can act as confounders when comparing Langerhans cell counts between HIV-seropositive and HIV-seronegative women. In fact, HIV-seropositive women are more frequently infected by “high risk” HPV type 16 or 18 and more frequently have high-grade CIN than uninfected controls (1-5). For this reason we have replicated our previous study (6) on cervical biopsies from cytologically and colposcopically normal transformation zones free of HPV infection. Several markers (thymocyte antigen, leukocyte-associated antigen, major histocompatibility complex class II antigen) have been used to stain Langerhans cells in cervical tissue (8,10). S-100-positive cells represent ≈35-40% of total Langerhans cells and are selectively reduced in cervical intraepithelial neoplasia and HPV infection (10). HIV can infect Langerhans cells by binding to human leukocyte antigen (D related) or to IgG-antibody-virus complexes to IgG-Fc or complement receptors (11). According to Macatonia et al. (12), Langerhans cells and the related dentritic cells are the cells most readily infected by HIV in the squamous epithelium. At present, there are no data regarding the effect of HIV infection on the different subpopulations of Langerhans cells. However, the finding of a reduced number of Langerhans cells in the otherwise normal cervix of HIV-seropositive women may help explain the relationship between cervical neoplasia and HIV infection. Since Langerhans cells play a central role in immunosurveillance of the squamous cervical epithelium (7,8,12), their depletion may facilitate the acquisition of HPV infection and the development of cervical neoplasia in such women, who are already at high risk of cervical infection or neoplasia because of sexual behavior and drug abuse. On the other hand, the reduction of Langerhans cells associated with HIV-induced immunosuppression could also favor the progression of cervical intraepithelial neoplasia. The depletion of Langerhans cells and the severity of CIN correlate well with the stage of HIV disease (6). These findings suggest a direct relationship of systemic HIV infection, reduced general and local immune response and occurrence, and progression of cervical dysplastic lesions. Future studies aimed at the identification of the effects of HIV infection and antiretroviral therapies on the different subpopulations of Langerhans cells could help us understand the mechanisms leading to the increased prevalence of squamous intraepithelial and invasive cervical lesions among HIV-seropositive women. A. Spinillo; P. Tenti; F. Baltaro; R. Zappatore; P. Migliora; F. Zara Departments of Obstetrics and Gynecology and Pathology; University of Pavia; IRCCS Policlinico S. Matteo; Pavia, Italy