Introduction

About 1% of sexually active adults in the United States have genital warts [1] and HIV-infected persons have higher prevalence and rate of recurrence of genital warts than HIV-uninfected persons [2,3]. Anal warts are a major public health problem for several reasons. They have never been studied separately from genital warts in prior research, and yet anal warts are found to be more common than cervical warts in women [4]. Additionally, a person infected with HPV is more likely to have multiple HPV types (i.e., high-risk and low-risk HPV), and there is evidence that the presence of anal warts may predict risk of anal intraepithelial neoplasia (AIN) or anal carcinoma [5,6], most likely as an indicator that unmeasured high-risk HPV types are present concurrently. Recently, a large Danish cohort study of approximately 50,000 patients found that there was a strong association between the presence of genital warts and anal cancer in men and women (standardized incidence ratio: 12.5 and 7.8 for men and women, respectively) [7]. And finally, a person with anal warts may experience serious financial and psychological burdens [8,9]. A recent study [10] showed that the annual cost to treat anogenital warts in the US is approximately $200 million (US). Patients with anogenital warts were reported to have decreased self-esteem and increased distress, embarrassment, shame, anger, negative self-perception and anxiety [9].

Previous research has indicated that the anal transformation zone (from anal squamous epithelium to rectal columnar epithelium) is morphologically similar to the cervical transformation zone, which is known to be susceptible to oncogenic HPV infection [11]. Also, anal cancer and cervical cancer have been found to share the same route of HPV infection [12]. Subsequent studies have shown that women with cervical dysplasia or cervical cancer are also at increased risk for anal cancer [13-15] and that concurrent HPV cervical infection has been identified as a risk factor for anal HPV infection [16,17]. Accordingly, a woman with cervical HPV infection is 20.5 times (95% CI: 16.3-25.7) more likely to have subsequent anal HPV infection with the same genotype compared to a woman without a previous cervical HPV infection [16].

Cervical Papanicolaou testing has long been used as a screening test to detect cervical precancerous/cancerous lesions. However, research on the use of this test to predict the presence of warts has been limited to genital warts only and no previous study has examined its association with anal warts. For example, in the early 1990s Rowen et al. [18] reported that women with warts (or sexual partners with genital warts) were more likely to have borderline or dyskaryotic smears than women without such a history. Recently, Kanno et al. [19] showed that women with genital warts were 7.03 times (95% confidence interval - CI: 2.82-17.53) more likely to have abnormal Papanicolaou test results than women without genital warts. From a clinical standpoint, it is interesting to ask whether the cervical Pap test might predict the development of anal wart size, given the widespread clinical use of this test.

The purpose of the current study was to determine the association between abnormal cervical Papanicolaou test results and the size of the largest anal wart over time in HIV-infected women in an on-going cohort study in the US.

Methods

Study population

Data for this analysis were obtained from a public dataset (release P09) of the WIHS, an on-going cohort study of HIV-infected and HIV-uninfected women in 6 locations in the US: Bronx/Manhattan, NY; Brooklyn, NY; Washington DC; Los Angeles/Southern California/Hawaii; San Francisco/Bay Area, CA; and Chicago; IL. Details on the WIHS were reported previously [20,21]. Briefly, WIHS had 2 major enrollment periods: October 1994-November 1995, when 2,059 HIV-infected and 569 HIV-uninfected women were recruited from clinic-based and population-based sources; and October 2001-September 2002, when 1,144 women were recruited. All participants provided written informed consent to participate in the WIHS study, and the study was approved by the Institutional Review Boards of all participating institutions. Interviews, medical exams, gynecologic exams, specimen collections, and medical record abstraction were performed at the baseline visit and follow-up visits at 6-month intervals.

The WIHS public dataset has 3,766 HIV-infected and HIV-uninfected women. A participant was included in the current analysis if she had at least one anal wart during follow-up. Participants were not included if they were HIV-uninfected (n = 958), seroconverted during the study follow-up (n = 16), had an unknown HIV status (n = 1), did not have at least one anal wart during the study period (n = 1,777), or received treatment for an anal wart during follow-up (n = 38). After these exclusions, a sub-sample of 976 participants was available for the current analysis (Figure 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Flowchart of inclusion and exclusion of participants in current analysis.

https://doi.org/10.1371/journal.pone.0081751.g001

Results

At the baseline visit, more than 68% of participants in the current analysis had normal cervical Papanicolaou test results, 29% had either ASCUS or LSIL and only 3% had either HSIL or cervical carcinoma. While the proportion of Caucasian and the other race/ethnicity groups (i.e., Hispanic, Asian/Pacific Islanders, and Native America/Alaskan Native) were similar (19.42% and 19.94%, respectively), approximately 66% of HIV-infected women were African-American (Table 1).

Table 2 presents the results on the association between size of anal wart and cervical Papanicolaou test results over time, both in unadjusted and adjusted models. In the unadjusted model, the growth rate of the largest anal wart after each visit in a woman with cervical Papanicolaou test results of HSIL/carcinoma was 19.50 faster than that of a woman with a normal cervical Papanicolaou test result (P = 0.008). However, this finding was insignificant in the adjusted model (P = 0.56), possibly due to the small sample size in the HSIL/carcinoma group. In the adjusted model, a woman with Papanicolaou test results of ASCUS/LSIL had an anal wart, on average, 12.81 mm² larger than a woman with normal cervical Papanicolaou test results. The growth rate of the largest anal wart after each visit in a woman with results of ASCUS/LSIL was 1.56 mm² slower than that of a woman with a normal test result. These two results, however, were not statistically significant (P = 0.54 and P = 0.82, respectively).

Discussion

In this study, we examined the association between cervical Papapicolaou smear results and the size of the largest anal wart over time in HIV-infected women using a public dataset from the largest on-going prospective cohort study in HIV-infected and HIV-uninfected women in the US. We found that women with abnormal cervical Papanicolaou test results (ASCUS/LSIL) had larger anal warts at baseline and slower growth rate of largest anal warts after each visit than those with normal results, although our results were not statistically significant.

To our knowledge, this is the first study to use a linear mixed model to determine whether cervical Papanicolaou test results could predict the size of anal warts over time in HIV-infected women. We, therefore, cannot compare our findings directly to any other study. There are, however, a few studies reporting the relationship between the presence of genital warts and abnormal Papanicolaou tests in the general population [18,19]. For example, Kano et al.[19] reported that genital warts were significantly associated with abnormal Papanicolaou tests among HIV-uninfected women seen in a public STD clinic in Baltimore, MD (Odds Ratio - OR = 7.03; 95% CI: 2.82-17.53; P<0.001). We noted that there are two major differences between our findings and those reported by Kano et al. [19]. First, the outcome variable in our study was the size of the largest anal wart, while Kano et al. [19] examined abnormal Pap smear results. Second, in our study, the size of anal warts was examined over time, whereas Kano et al. [19] examined the presence of a genital wart regardless of the size.

Although we did not find a significant association between cervical Papanicolaou test results and the size of anal warts over time among HIV-infected women, this study provides some clinical implications for future screening programs of HIV-infected women. The cervical Papanicolaou test has contributed to the reduction of cervical cancer prevalence and incidence in developed countries. With regard to screening for anal HPV infections, pre-cancerous/cancerous lesions and warts, a reliable test with high accuracy is still needed. In a small subsample from WIHS, Holly et al. [35] did not find concurrent cervical cytology to be a useful overall indicator of anal cytology (Relative risk-RR = 1.5; 95% CI: 0.97-2.2). Also, using anal Papanicolaou tests to evaluate a subset of the WIHS cohort (223 HIV-infected and 57-uninfected women), Palefsky et al. [36] reported that 76% and 42%, respectively, had anal HPV infection, and 26% and 8%, respectively, had abnormal anal Papanicolaou test results. These findings, together with our own, demonstrate that the cervical Papanicolaou test might not be predictive of the behavior of anal warts, which are very common among HIV-infected women. Instead, anal Papanicolaou testing may be a relevant approach in this population.

Previous studies have indicated that HIV-infected women are more likely to have abnormal anal cytology/histology than HIV-uninfected women [37,38]. Recent local guidelines in New York recommended anal Papanicolaou screening be conducted in men who have sex with men (MSM), females with abnormal vaginal and/or vulvar histology, and any patients with a history of anogential condylomas [39]. This guideline, however, did not recommend anal Papanicolaou screening in HIV-infected women with no other pathological abnormalities. Our findings, along with the prior evidence, support a recommendation for the use of anal Papanicolaou test screening for HIV-infected women, given the fact that anal Papanicolaou testing has proven to be cost-effective in the HIV-infected MSM population [40] and has fairly comparable accuracy to cervical cytology [41].

The current analysis has two major strengths. First, the linear mixed model allowed us to address the research question directly regarding whether or not there is an association between the size of anal warts and cervical Papanicolaou test results over time while other statistical methods cannot. Second, the size of the largest anal wart, as the continuous outcome variable, was treated appropriately. If we had used other statistical methods such as logistic regression or Cox proportional hazards regression, we could not model the change in size of anal warts over time. A limitation in our study is the use of the size of the largest anal wart as an outcome variable. Since the largest wart measured at one visit might not be the same in subsequent visits, it is difficult to follow-up the same anal wart over time. We, however, thought that the overall disease burden posed by anal warts may well depend on the largest one present at any given visit.

In summary, we found that women with abnormal cervical Papanicolaou test results (ASCUS/LSIL) had larger anal warts at baseline and slower growth rate of the largest anal warts than those with normal results. The association between the size of anal warts and cervical Papanicolaou test results over time in HIV-infected women, however, becomes insignificant in the adjusted model. Our findings, however, provide more information for further consideration of screening programs using anal Papanicolaou testing in HIV-infected women. Currently, data on the natural history of anal HPV infection and its clinical endpoints (both warts and precancerous lesions) in HIV-infected women are limited. Additional studies among this particular population are required to address this question and support the implementation of an anal Papanicolaou screening program.

Acknowledgments

Data in this manuscript were collected by the Women's Interagency HIV Study (WIHS) Collaborative Study Group with centers (Principal Investigators) at New York City/Bronx Consortium (Kathryn Anastos); Brooklyn, NY (Howard Minkoff); Washington, DC, Metropolitan Consortium (Mary Young); The Connie Wofsy Study Consortium of Northern California (Ruth Greenblatt); Los Angeles County/Southern California Consortium (Alexandra Levine); Chicago Consortium (Mardge Cohen); Data Coordinating Center (Stephen Gange). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.