icon star paper   HIV Articles  
Back grey_arrow_rt.gif
High Resolution Anoscopy Findings for Men Who Have Sex with Men: Inaccuracy of Anal Cytology as a Predictor of Histologic High-Grade Anal Intraepithelial Neoplasia and the Impact of HIV Serostatus
  Clinical Infectious Diseases May 2004
Lori A. Panther,1 Katiri Wagner,1 JoAnn Proper,1 Dana K. Fugelso,1 Pamela A. Chatis,1,2 Will Weeden,1 Imad A. Nasser,1 John P. Doweiko,1 and Bruce J. Dezube1
1Beth Israel Deaconess Medical Center, Harvard Medical School, and 2PerkinElmer Life Sciences, Boston, Massachusetts
We compared the pathological diagnoses obtained by anal Papanicolaou (Pap) smear with those obtained by anal biopsy or by surgical excision for 153 men who have sex with men (MSM). Analysis of these paired specimens showed that anal Pap smears were an inaccurate predictor of high-grade anal dysplasia, regardless of human immunodeficiency virus (HIV) serostatus. The presence of any abnormal anal cytological finding indicates a potential for high-grade dysplasia on histological examination of MSM.
Received 3 November 2003; accepted 17 December 2003; electronically published 21 April 2004.
In men, the occurrence of anal squamous cell carcinoma (SCC) is strongly associated with a history of receptive anal intercourse and anogenital human papillomavirus (HPV) infection. Concurrent HIV infection has increased the incidence of anal SCC in HIV-positive men who have sex with men (MSM) to almost 9 times the rate of cervical cancer in women and to 37 times that of HIV-negative MSM. In the HIV-infected population, there is no clear evidence that restoration of immunity with HAART allows high-grade anal intraepithelial neoplasia (AIN) to regress or resolve.
Because HAART has allowed HIV-infected patients to live for decades in relative immunosuppression (compared with the general population), the incidence of AIN and SCC in HIV-positive persons who are coinfected with HPV is expected to increase. Current estimates predict that 5% of HIV-positive men with high-grade AIN may have progression to SCC. Cytological screening for AIN and SCC using anal Papanicolaou (Pap) smear testing in HIV-infected men is cost-effective, although the predictive value of a screening anal Pap smear to detect histologically proven high-grade dysplasia may be low.
Women with HIV infection appear to have an increased incidence and prevalence of anogenital condylomata, poorer response to standard therapies, and a well-documented increase in risk for high-grade cervical dysplasia and cervical carcinoma. Although a similar epidemiological pattern of anogenital condylomata is suspected in HIV-positive men, there are few data describing the occurrence of high-grade AIN or SCC in patients with anal condylomata in this group. Because anal SCC appears to be increasing in the HIV-positive population, screening for HPV-related anogenital disease is coming under closer scrutiny.
We recorded the medical histories, high-resolution anoscopy (HRA) findings and pathology reports in a cohort of MSM referred to an anal dysplasia clinic to determine the accuracy of anal Pap smear as a predictor of paired anal histology results, as well as whether HIV serostatus had an impact on these findings.
Methods. Clinical and laboratory characteristics were reviewed for MSM who underwent anal Pap smears and anal histological examinations at or within 3 months of their initial HRA at the Beth Israel Deaconess Infectious Diseases Dysplasia Clinic (Boston, MA) during the period of April 1999 through April 2003. Patients were referred to this clinic by their primary physicians in the infectious diseases, hematology-oncology, and general medicine clinics affiliated with the medical center. Information was collected regarding HIV serostatus, medical history, HRA findings, and paired cytology and histology results, according to the highest grade reported. For all subjects, anal Pap smears were taken during their initial visit for HRA. Pap smear cytological diagnoses were paired with the histological diagnosis of anal biopsies obtained at the initial visit or with surgically excised anal tissue obtained within 3 months of the initial visit. This project was approved by the Beth Israel Deaconess Medical Center's Institutional Review Board, and the investigators followed the policies and procedures adopted by the institution to protect the privacy of its patients. Significant differences were assessed using Fisher's exact test and Spearman rank correlation coefficient, when appropriate.
Results. Of 153 MSM, 100 (65%) were HIV positive and 53 (35%) were HIV negative. The basis for referral was an abnormal screening anal Pap smear finding (24 [24%] of 100 HIV-positive and 22 [42%] of 53 HIV-negative subjects), high-grade AIN on pathological examination of surgically removed anogenital condylomata (19 HIV-positive subjects [19%] and 6 HIV-negative subjects [11%]), or a history of internal and/or external anal condylomata for 40 HIV-positive subjects [40%] and 14 HIV-negative subjects [26%]). The remainder of the cohort was referred for anal bleeding, palpable internal and/or external anal mass, or screening. Anogenital symptoms (i.e., mass, bleeding, itching, dyspareunia, or changes in bowel patterns) were reported more often by HIV-positive subjects (69 [77%] of 90) than HIV-negative subjects (4 [8%] of 51) (P < .0001, by Fisher's exact test).
The paired results of cytological and histological tests did not correlate when analyzed by detailed grades (rs = -0.37, by Spearman rank correlation coefficient) or when histological grades were categorized as "low grade" and "high grade," according to the treatment implication of each grade (rs = -0.53, by Spearman rank correlation coefficient). More than one-third of all subjects with low-grade dysplasia noted on a Pap smear had high-grade AIN noted during histological examination, and this proportion did not differ between HIV-positive and HIV-negative groups (data not shown).
Detection of high-grade squamous intraepithelial lesions (HSILs) on an anal Pap smear had a sensitivity of 47% (95% CI, 35%59%) and specificity of 90% (95% CI, 81%96%) for detection of a high-grade histological finding (AIN level 2, AIN level 3, or SCC) in the paired specimen. Likewise, detection of low-grade squamous intraepithelial lesions (LSILs) on an anal Pap smear had a sensitivity of 68% (95% CI, 56%78%) and a specificity of 48% (95% CI, 36%59%) for detection of AIN level 1 by the paired examination. A Pap smear finding of atypical squamous cells of uncertain significance showed essentially an equal distribution of pathological grades on paired histological examinations. The positive and negative predictive values were 81% (95% CI, 66%92%) and 65% (95% CI, 55%74%), respectively, for HSIL on Pap smears predicting high-grade histological findings (P < .01). The positive and negative predictive values for LSIL were 59% (95% CI, 48%69%) and 57% (95% CI, 44%70%), respectively, for prediction of AIN1 with paired histological examinations (P = .07). Overall, 36% of MSM with LSIL and 69% of MSM with HSIL noted on an anal Pap smear had high-grade AIN noted with paired histological examinations.
Two subjects had newly diagnosed invasive SCC of the anal canal; both subjects were HIV positive, had a history of anal condylomata, and had anal condylomata noted during HRA. Both subjects had HSIL on anal Pap smear, which represents a 6% incidence of histologically confirmed anal SCC in persons with HSIL on cytological examination in this cohort. These 2 subjects presented with significant symptoms that led to referral (one patient had experienced anal bleeding for weeks, and the other had a rapidly enlarging anal mass).
Discussion. There are few sizeable cohorts at risk for anal carcinoma that are being studied longitudinally; the largest such cohort is at the University of California at San Francisco. Cross-sectional and longitudinal studies of the San Francisco cohort have shown that HIV-positive MSM have the highest incidence of abnormal Pap smear findings, with a large proportion of this group progressing from LSIL to HSIL on Pap smears over a 4-year follow-up period, and the most immunosuppressed subjects have the highest risk of progression to HSIL. The cross-sectional data presented here independently confirm a substantial incidence of histologically proven high-grade anal dysplasia in MSM who present with minimally abnormal anal Pap smear findings. Unlike the San Francisco cohort, there was no difference in incidence between the groups when HIV serostatus was considered; perhaps the demographic characteristics of the HIV-negative MSM in the present cohort put them at higher risk for acquisition of oncogenic HPV types; thus, more of them presented with high-grade pathological findings than in the San Francisco cohort. Alternatively, there may be a geographic variation in the distribution of HPV types among MSM that has yet to be described, accounting for the difference in findings.
Given the increased incidence of invasive SCC of the anal canal among MSM, eradication of high-grade AIN is under evaluation as a method for prevention of anal cancer, analogous to the management of high-grade cervical dysplasia in women. The data presented here regarding the correlation between anal cytological findings and histological findings are strikingly similar to cervical cytological and histological findings. As such, our data indicate that MSM with screening anal Pap smears that are not normal should be considered for HRA and submission of tissue for pathological examination if indicated. The impact of ablating histologically confirmed high-grade AIN on the incidence of anal SCC is currently under investigation.
This study independently confirms that a substantial proportion of low-grade dysplasia on anal Pap smears occurs with high-grade AIN in MSM. Although randomized trials are lacking, ablation of high-grade AIN has been suggested as an intervention for cancer prevention. This study confirms that abnormal anal cytological findings of any grade should suggest the possibility of high-grade histological findings. These findings could have an impact on the use of HRA as a screening tool for high-grade AIN.
  icon paper stack View Older Articles   Back to Top   www.natap.org