Frequency of Occult High-Grade Squamous Intraepithelial Neoplasia and Invasive Cancer Within Anal Condylomata in Men Who Have Sex With Men

Abstract

Human papillomavirus causes anal condylomata, high-grade intraepithelial neoplasia (HGAIN), and squamous cell cancer (SCCA). We found HGAIN or SCCA in 75/159 (47%) HIV-seropositive and 42/160 (26%) HIV-seronegative men who have sex with men (MSM) with anal condylomata meriting surgery (p<0.0001, chi-square). Anal condylomata in MSM often harbor HGAIN and SCCA.

The two main groups of human papillomavirus (HPV) types that infect the anogenital tract are low-risk HPV types, associated with development of anogenital condylomata, and high-risk HPV types, implicated in the evolution of anal squamous cell cancer (SCCA) and its putative precursor, high-grade anal intraepithelial neoplasia (HGAIN) [1]. Repeated exposure, infection, and shedding of genital HPV are common and overlapping events in sexually active populations, and simultaneous detection of multiple HPV types in the genital tract is a recognized phenomenon, especially in immunosuppressed populations [2]. HIV-related immunosuppression may increase the chances of both low-grade and high-grade HPV lesions coexisting in the same patient [3].

The rate of anal SCCA in HIV-seropositive MSM has increased in the decades following institution of highly active antiretroviral therapy, suggesting that longer life expectancies may permit established high-risk HPV infections to progress to SCCA in a small proportion of patients [4]. Reports of histopathological findings in anogenital condylomata in large MSM populations are limited. In one study, at least one-third of 74 MSM over half of 26 HIV-seropositive MSM had HGAIN within surgically excised warts [5]. We report the prevalence of occult HGAIN and SCCA in a large urban cohort of MSM with anal condylomata too extensive to be amenable to outpatient ablation.

METHODS

A retrospective cohort of all MSM referred to a single university-affiliated surgical practice for ablation of anal condylomata from May 2002 through April 2007 was evaluated. Patients were referred from their primary provider, or from the affiliated high resolution anoscopy (HRA) clinic for evaluation of anogenital warts deemed too widely distributed and/or too large to undergo office treatment. HRA was used in the operating room to identify all lesions, which were then surgically excised using a CO2 laser. All excised tissue was submitted for histopathological analysis. Histopathologic grade was recorded as anal intraepithelial neoplasia grade 1 (AIN1), AIN grades 2 or 3 (HGAIN), or SCCA.

The protocol for electronic medical record review underwent formal evaluation and waiver by the Beth Israel Deaconess Medical Center Committee on Clinical Investigations. Subjects with a history of anal HGAIN or SCCA were excluded. Subjects with documented HIV infection were recorded as HIV-seropositive, and those with either laboratory documentation or absence of HIV infection on the electronic medical record problem list were recorded as HIV-seronegative,. If available, presenting symptoms and anal cytology results within six months prior to surgery were documented. In HIV-positive subjects, CD4 cell count and HIV viral load results within six months of surgery were recorded.

Anal cytology specimens were collected from the transformation zone of the anal canal using moistened Dacron® swabs, then fixed in liquid ThinPrep® (Cytyc, Boxborough, MA) solution. Results were reported per the 2001 Bethesda classification system as normal, atypical squamous cells of undetermined significance (ASC-US), low-grade squamous intraepithelial lesion (LSIL), atypical squamous cells, cannot rule out high-grade lesion (ASC-H), and high-grade squamous intraepithelial lesion (HSIL) [6].

Data were stratified by HIV serostatus and histologic grade. Categorical variables were compared between groups using the Fisher’s Exact Test or Chi-Square Test where appropriate, and normally distributed continuous data were compared using the one-way ANOVA with multiple comparisons, or the two-tailed unpaired t test. For nonparametric comparisons, the Spearman rank correlation or Kruskal-Wallis tests were used. CD4 counts were log₁₀-transformed, with nondetectable viral load (VL) recorded as 50 copies/ml. Sensitivity and specificity of anal cytology for detection of HGAIN or SCCA were calculated (GraphPad InStat version 3.06 for Windows, GraphPad Software, San Diego California USA, www.graphpad.com).

RESULTS

Table 1 summarizes the clinical characteristics and histopathologic findings of the 319 cases included in the analysis. The cohort was evenly divided between HIV-seropositive (n = 159) and HIV-seronegative (n = 160) subjects. Most subjects were referred from their primary providers, and more HIV-seropositive subjects were referred from HRA clinic than HIV-seronegative subjects. Mean age of the entire cohort was 37 years (range, 18–73), and the HIV-seropositive group was significantly older (p<0.001, Unpaired t-test with Welch correction). Except for a greater proportion of HIV-seronegative subjects presenting with pain (p=0.02, Fisher’s Exact), presenting symptoms had similar frequencies regardless of HIV serostatus. There was no significant trend toward increasing frequency of symptoms with worsening histologic grade (data not shown).

TABLE 1.

Age, Presenting Characteristics, and Histopathological Findings in Men Who Have Sex with Men (MSM) with Anal Condylomata Meriting Surgical Excision

CHARACTERISTIC	TOTAL (n = 319)	HIV-SEROPOSITIVE (n = 159)	HIV-SERONEGATIVE (n = 160)	P VALUEa
Age in years:
Mean (SD)	37 (10)	41 (8)	34 (11)	<0.001
Median (IQR)	37 (30–44)	40 (36–45)	32 (24–40)	<0.001
Symptom, No. present/No. reporting (%):
Pain	75/286 (26)	29/145 (20)	46/141 (33)	0.02
Palpable Mass	234/286 (82)	115/141 (82)	119/145 (82)	NS
Bleeding	117/286 (41)	58/141 (41)	59/145 (41)	NS
Itching	58/286 (20)	31/141 (22)	27/145 (19)	NS
Referral Source, No. (%):				<0.0001
Primary Care Provider	209 (65)	78 (49)	131 (82)
HRA Clinic	110 (35)	81 (51)	29 (18)
Histological Grade, No. (%):				0.0002
AIN1, n (%):		84 (53%)	118 (74%)
HGAIN, n (%):		68 (43%)	41 (26%)
SCCA, n (%):		7 (4%)	1 (<1 %)

NOTE. Abbreviations: MSM, men who have sex with men; HIV, human immunodeficiency virus; SD, standard deviation; IQR, interquartile range; HRA, high-resolution anoscopy; AIN1, anal intraepithelial neoplasia grade 1; HGAIN, high-grade anal intraepithelial neoplasia (includes AIN2, anal intraepithelial neoplasia grade 2 and AIN3, anal intraepithelial neoplasia grade 3); SCCA, anal squamous cell carcinoma..

^ap value (2-sided) for unpaired t test with Welch correction (means), Mann-Whitney test (medians), Fisher’s exact test, or Chi-squared test for independence comparing HIV-seropositive and HIV-seronegative subjects.

In the entire cohort, 202 (63%) had AIN1, 109 (34%) had HGAIN, and 8 (3%) had anal SCCA on histopathological exam. HGAIN or SCCA was present in 47% of HIV-seropositive vs. 26% of HIV-seronegative subjects (p = 0.0002, Chi-square). Seven of the 8 subjects with anal SCCA were HIV-seropositive. Subjects with SCCA were older (mean 51 years, range 35–73) compared to those with either AIN1 (mean 36 years, range 18–69) or HGAIN (mean 38 years, range 19–71) (p = 0.002, Kruskal-Wallis).

Anal cytology findings were available for 123/319 (36%) of all subjects, and there was correlation between anal cytology and histopathology grades (p = 0.001, Chi-square). A finding of HSIL or ASC-H on anal cytology correlated with the presence of HGAIN or SCCA on histopathology (p = 0.002, Fisher’s Exact test). This association remained significant in the HIV-seropositive group (p < 0.0001, Fisher’s Exact test) but not the HIV-seronegative group (p = 0.353, Fisher’s Exact test). The sensitivity for HSIL on anal cytology for presence of HGAIN or SCCA on histopathology was 90% (95% CI, 81%–96%) over the entire cohort, 93% (95% CI, 83%–98%) for HIV-seropositive subjects and 89% (95%CI, 73%–98%) for HIV-seronegative subjects. However, the specificity for HSIL on anal cytology for HGAIN or SCCA was only 33% (95%CI, 21%–48%) over the entire cohort; 58% (95%CI, 37%–78%) for HIV-seropositive MSM, and 23% (95%CI, 5%–54%) for HIV-seronegative MSM.

Of the 159 HIV-seropositive subjects, CD4 counts and/or HIV viral loads within 6 months of surgery were available for 127 (80%). Overall, mean CD4 count was 431 ± 238 cells/mm³ (range, 4–1153) and 60/121 (50%) had detectable viral loads (range, 75–523,000 copies/ml). There was a trend toward correlation between CD4 count and the log₁₀(HIV viral load) that did not reach significance (p = 0.08, Pearson correlation). Though there was not a significant overall correlation between either CD4 count or HIV viral load and histopathologic grade. The mean CD4 count of the 7 HIV-positive subjects with SCCA was lower than the rest of the group (218 ± 124 cells/mm³ vs. 441 ± 237 cells/mm³, p = 0.02, Unpaired t-test). There was no correlation between HIV viral load and histopathologic grade.

COMMENT

Since anogenital condylomata are classically associated with low-risk HPV types [7], the presence of HGAIN within what clinically appears to be a low-grade condylomatous lesion is unsettling. This large observational cohort of MSM undergoing surgical excision of anogenital condylomata, demonstrated HGAIN or SCCA within 37% of subjects, with HIV-seropositive MSM almost twice as likely (47%) to harbor HGAIN or SCCA compared to HIV-seronegative MSM (26%). Smaller cohorts of MSM have reported HGAIN rates ranging from 16–71% in HIV-seropositive and 6–46% in HIV-seronegative populations [8–10].

This study shows similar sensitivity and specificity of HSIL on preoperative anal cytology for detection of HGAIN or SCCA as we have previously reported for the detection of HGAIN by HRA-assisted biopsy in MSM [11]. Thus, we have observed that the predictive value of HSIL on anal cytology for HGAIN is limited but consistent, regardless of whether the HPV-related disease is subclinical or severe enough to merit surgical attention.

Several factors limit our study. Since our cohort consisted of referrals for severe condylomatous disease, the pathological associations may not reflect what exists in MSM with smaller burdens of anogenital condylomata. In addition, only one-third of subjects had HRA as part of their evaluation. If routine use of HRA was more likely to result in surgical referrals having a higher likelihood of HGAIN, the prevalence of HGAIN in this population would be underestimated. However, it should be emphasized that operative excision and pathological analysis without routine preoperative HRA in the majority of the cohort still demonstrated a high prevalence of HGAIN. Associations taking into account racial demographics, smoking, or type of highly active antiretroviral therapy were not captured by this study. Though the age difference between the HIV-seropositive and HIV-seronegative groups may have favored a higher proportion of HGAIN in the HIV-seropositive group, the EXPLORE study did not show age to be a significant factor for the carriage of high-risk HPV types HIV-seronegative MSM [12]. Lastly, if a fundamental difference exists in the development of HGAIN depending on the location of the condylomatous lesion, detection bias could have influenced our results since the exact origin of the histopathological specimen (i.e., internal anal vs. perianal) was not available.

The present study demonstrates that, in a large urban population of MSM, condylomata requiring surgical excision frequently harbored occult HGAIN or SCCA. These data emphasize the importance of obtaining tissue for histopathological examination in MSM presenting for treatment of anogenital condylomata. Prevention of recurrences and careful clinical follow-up of anal condylomata harboring HGAIN may be a method of anal cancer prevention in MSM, particularly in those with HIV infection.