Secondary Logo

Journal Logo

Self-Collected Versus Clinician-Collected Anal Cytology Specimens to Diagnose Anal Intraepithelial Neoplasia in HIV-Positive Men

Cranston, Ross D. MBChB*; Darragh, Teresa M. MD; Holly, Elizabeth A. PhD; Jay, Naomi RN, NP§; Berry, J. Michael MD§; Costa, Maria Da MS§; Efird, Jimmy T. PhD; Palefsky, Joel M. MD

JAIDS Journal of Acquired Immune Deficiency Syndromes: August 1st, 2004 - Volume 36 - Issue 4 - p 915-920
Clinical Science

Background: Anal intraepithelial neoplasia (AIN) is common in men who have sex with men (MSM) and may be diagnosed by anal cytology screening.

Methods: One hundred two MSM with clinician-collected anal cytology and histopathology specimens were assessed from a cohort study of AIN at the University of California at San Francisco. The men were given a cytology self-collection kit with written instructions for use and requested to collect the sample 1 month after the clinic visit.

Results: Ninety-one percent of self-collected and 99% of clinician-collected anal cytology samples were adequate for interpretation. The sensitivity of abnormal anal cytology to detect AIN by histology was 68% in self-collected samples and 70% in clinician-collected samples, and the sensitivity to detect AIN 2 or AIN 3 was 71% and 74%, respectively. Cytologic results did not differ by grade between self-collected and clinician-collected samples. Among MSM diagnosed with AIN 2 or 3 by biopsy, 33% of self-collected and 39% of clinician-collected cytology samples were high-grade. The sensitivity of both self-collected and clinician-collected samples to detect AIN 2 or 3 was higher among HIV-positive MSM than among HIV-negative MSM.

Conclusions: MSM with biopsy-proven AIN can self-collect anal cytology samples with sensitivity comparable with that of experienced clinicians. This may facilitate screening for AIN.

From the *Department of Medicine; †Department of Pathology; ‡Department of Epidemiology and Biostatistics; and §Department of Stomatology, University of California at San Francisco, San Francisco, CA.

Received for publication December 11, 2003; accepted March 29, 2004.

Supported by grant NCI R01CA54053. These studies were carried out in the General Clinical Research Center, University of California at San Francisco, with funds provided by the Division of Research Resources 5 M01-RR-00079, US Public Health Service.

Reprints: Joel M. Palefsky, Box 0126, Department of Medicine, University of California at San Francisco, San Francisco, CA 94143.

The incidence of anal cancer in the United States is increasing and is 0.8–0.9 case per 100,000 persons among the general population. 1,2 Before the HIV epidemic, the incidence of anal cancer was estimated to be as high as 35 cases per 100,000 persons among men who have sex with men (MSM), 3 with recent data indicating that the incidence among HIV-positive MSM is approximately twice that among HIV-negative MSM. 4,5 Thus, the incidence of anal cancer among both HIV-negative and HIV-positive MSM is higher than the current incidence of cervical cancer among women and is similar to the incidence of cervical cancer before the introduction of cervical cytology screening. 6 Although the magnitude of the increase in anal cancer compared with the general population is greater among HIV-positive MSM than among HIV-positive women, the latter are also at increased risk of anal cancer. 7

Anal cancer and cervical cancer are similar histologically, and like cervical intraepithelial neoplasia and cervical cancer, anal intraepithelial neoplasia (AIN) and anal cancer are associated with oncogenic human papillomaviruses. 8–12 It is believed that AIN 2 or 3 precedes the development of anal cancer in the same way that cervical intraepithelial neoplasia 2 or 3 precedes cervical cancer (J.M. Palefsky, unpublished observations). Similarly, Bowen disease (perianal AIN 3) may progress to invasive anal cancer. 13,14 Analogous to the treatment of cervical intraepithelial neoplasia 2 or 3 to prevent the development of cervical cancer, treatment of AIN 2 or 3 may prevent progression to anal cancer.

Anal and cervical cytology findings are both described using the Bethesda classification system 15 and are reported, in increasing severity, as atypical squamous cells of undetermined significance, atypical squamous cells suggestive of high-grade change, low-grade squamous intraepithelial lesions, and high-grade squamous intraepithelial lesions (HSIL). 16

Anal cytology has been studied as a screening test for AIN in MSM. 17–19 In the largest study, using anal histopathology as a gold standard, sensitivity of 69% for HIV-positive individuals and sensitivity of 47% for HIV-negative individuals were found using high-resolution anoscopy (HRA) and biopsy to detect AIN. 19 HRA is colposcopy of the anus and perianal region using 3% acetic acid. The sensitivity is comparable with that of cervical cytology screening for the detection of cervical intraepithelial neoplasia in the general female population. 19 The method by which anal cytology samples are processed has also been investigated. In addition to conventional cell fixation and staining by the Papanicolaou method, cytology samples may be prepared using the ThinPrep technique before staining with the Papanicolaou method. 20 Both sampling methods have been validated for the anal canal and found to be comparable in the assessment of anal cytology. 20

Although somewhat theoretical and not universally accepted, using a state-transition Markov model, we projected that anal cytology screening to identify individuals with HSIL with subsequent treatment of the lesion should be cost-effective to prevent anal cancer when performed annually for HIV-positive MSM and every 2–3 years for HIV-negative MSM. 21,22

Several barriers currently exist to the implementation of an anal cytology screening program for at-risk individuals. One obstacle is the lack of studies to document that treatment of AIN reduces the incidence of anal cancer. Performance of such studies is hampered by the large numbers of subjects needed, difficulties in assembling an appropriate control group, and the long follow-up needed to assess the impact on cancer incidence. Other obstacles include a paucity of clinicians trained in collecting anal cytology samples, reluctance of some clinicians to address the issue of AIN with their patients, and patients’ feelings of embarrassment and fear of potential discomfort associated with the collection of anal specimens. Collection of anal cytology samples by subjects themselves has the potential to obviate some of these concerns.

The purpose of this study was to assess the performance of results from anal cytology samples collected by study participants at home compared with samples collected by experienced research clinicians in a clinic of MSM with a high prevalence of AIN. Anal histopathology was used as the gold standard for this research comparison.

Back to Top | Article Outline



This study was performed with the approval of the University of California at San Francisco Committee on Human Research. Informed consent was obtained from all subjects. Participants were 505 HIV-positive MSM and 364 HIV-negative MSM enrolled the University of California at San Francisco Anal Neoplasia Study. Of these men, 469 HIV-positive subjects and 256 HIV-negative subjects underwent biopsy at least once during the study. Study participants were examined every 3–6 months with anal cytology and HRA with biopsy of any suspicious lesions. Demographic information was collected and blood samples were obtained for HIV antibody testing.

One hundred twenty-five study participants from our ongoing cohort study of anal neoplasia were invited to participate. Criteria for study eligibility included having had an anal cytology sample taken using glass-slide smears and ethanol fixation and a simultaneous biopsy of any visible anal lesion that appeared dysplastic using HRA. The study participants had previously had anal cytology samples collected in the clinic by their practitioner. The men were consecutively chosen for participation during the study period, and neither cytology nor anal biopsy results were known at the time of enrollment. A clinician provided an explanation of the study, and informed consent was obtained from study subjects. Cytology specimen collection and HRA were performed as described previously with standard glass slides and ethanol fixation used to process the cytology specimens for clinician-collected samples. 19

Back to Top | Article Outline

Self-Collection of Samples

We previously showed that glass-slide cytology and liquid cytology have similar performance characteristics. 20 The study participants used liquid cytology methods to collect their specimens at home because it was a substantially easier process. The men were given written instructions on how to self-collect an anal cytology sample. The collection kit consisted of 1 bottle of Cytyc ThinPrep sample collection medium (Cytyc Corp., Boxborough, MA), 1 Dacron swab, 2 sealable plastic bags for disposal of sampling materials, and 1 pair of latex gloves. To allow for healing of the initial anal biopsy site, participants were requested to wait 1 month from the time of the original anal cytology and biopsy before collecting their cytology sample.

While wearing latex gloves, the men were instructed to remove the Dacron swab from its sterile package and moisten it with tap water before inserting it 2 inches into the anal canal. Applying gentle pressure to the walls of the anal canal, participants were instructed to remove the swab with a spiral motion over a 10-second period. The swab was to be immediately placed into an open Cytyc bottle containing Cytyc ThinPrep medium and agitated vigorously to disgorge the cells from the swab before it was removed from the bottle. The men were instructed to close the Cytyc bottle tightly, store it at room temperature, and return it to the clinic in a sealed plastic bag within 1 week of specimen collection.

Back to Top | Article Outline

Evaluation of Samples

The same pathologist (T.M.D.) analyzed all clinician-collected cytology and biopsy samples. Because of the different sampling method used, the pathologist knew whether the specimen was clinician collected or self-collected. However, all specimen findings were interpreted without reference to the paired specimen. Cytology findings were classified according to the Bethesda classification system as inadequate for interpretation (sample contained less than ~2000 to 3000 nucleated squamous cells), negative for squamous intraepithelial lesions, atypical squamous cells of undetermined significance, low-grade squamous intraepithelial lesions, or HSIL.

Back to Top | Article Outline

Statistical Analysis

Statistical analysis was performed using version 8 of the Statistical Analysis System (SAS, Inc., Cary, NC). The sensitivity of anal cytology to detect low-grade squamous intraepithelial lesions and HSIL was defined as the number of specimens with abnormal cytology divided by the number of individuals with AIN. The Fisher exact text was used to assess the level of statistical difference between sample specimens.

Back to Top | Article Outline


MSM who had had a prior clinician-collected anal cytology and HRA examination with anal biopsy were consecutively enrolled from the University of California at San Francisco Anal Neoplasia Study. Approximately 10% of the men invited to participate declined to do so. Of the men who agreed to participate, ~20% did not return the required materials in a timely fashion. Of the 106 remaining participants, 4 were excluded because they had incomplete clinician-collected cytology and clinic biopsy data. One hundred two subjects were therefore included in the analysis, and of these subjects, 82 were HIV positive and 20 were HIV negative. The mean age at the time of sampling was 45 years (range, 29–72 years). Most of the study participants were white non-Hispanic, (n = 89 [87%]); the rest were white Hispanic (n = 2 [2%]), black (n = 3 [3%]), Amerindian (n = 1 [1%]), Asian (n = 3 [3%]), and other (n = 4 [4%]).

Results of the clinician-collected cytology and the corresponding histology are shown in Table 1. Results of the self-collected cytology and the corresponding histology are shown in Table 2. Specimen adequacy was high in both clinician-collected (101/102, 99%) and self-collected (93/102, 91%) cytology specimens, with a significantly higher rate of adequacy seen in the clinician-collected samples (P = 0.02). The adequacy rate of self-collected samples was similar for HIV-positive (93%) and HIV-negative (85%) men (P = 0.37).





The sensitivity of any grade of anal cytology abnormality for detection of AIN 1, AIN 2, or AIN 3 was comparable between clinician-collected (70%) and self-collected (68%) samples. The sensitivity of any grade of anal cytology abnormality for detection of a high-grade lesion (ie, AIN 2 or 3) was also comparable between clinician-collected (74%) and self-collected (71%) samples. However, the sensitivity of anal cytology to detect AIN in self-collected samples was higher among HIV-positive men than among HIV-negative men. Three (20%) of 15 HIV-negative men with AIN had abnormal cytology compared with 56 (77%) of 73 HIV-positive men (P < 0.01). The sensitivity for detection of AIN 2 or 3 specifically in self-collected specimens was also higher among HIV-positive men than among HIV-negative men. Among the 57 HIV-positive men with biopsy-proven AIN 2 or 3, anal cytology was abnormal in 43 (75%), compared with 4 (40%) of 10 HIV-negative men (P = 0.05). The sensitivity of anal cytology to detect either AIN overall or AIN 2 or 3 specifically was also higher for HIV-positive men in the clinician-collected samples, and the magnitude of these differences in sensitivity was similar to those seen in self-collected samples (P = 0.15 for AIN overall; P = 1.0 for AIN 2 or 3). Overall cytologic results did not differ by grade between clinician-collected and self-collected samples. Among men diagnosed with AIN 2 or 3 by biopsy, 39% with AIN 2 and 33% with AIN 3 had HSIL by cytology.

Back to Top | Article Outline


Recent guidelines from the US Public Health Service for the treatment of opportunistic infections in HIV-positive individuals indicate that although anal cytology screening has not yet been formally recommended, anal cytology screening should be considered for HIV-positive men and women. 23 Anal cytology should be used as a screening tool to identify individuals who would benefit from HRA to detect and treat AIN. 19 The current recommendation is to refer patients with any grade of anal cytologic abnormality for HRA. 23,24 Moreover, our group previously showed that detection of HSIL by anal cytology has a high positive predictive value for detection of biopsy-proven high-grade AIN. 19 Thus, inability to confirm high-grade AIN by HRA-directed biopsy in a patient with HSIL by cytology should prompt a repeated examination.

In this study, we compared overall rates of abnormality between samples collected by clinicians and study participants. The results of this study strongly suggest that this population of MSM with previous experience of having had anal swabs used to collect cytology specimens and anal biopsy specimens of lesions clinically suspicious for anal dysplasia obtained by a clinician are capable of self-collecting samples with sensitivity comparable with that of experienced clinicians. This was accomplished with only written instructions on how to self-collect an anal cytology specimen in this experienced population.

The primary goal of anal cancer prevention is to detect and treat high-grade AIN before it progresses to cancer. To this end, we also compared the sensitivity of cytology specimens collected by clinicians and subjects to detect AIN 2 or 3 by biopsy and found that the sensitivity for the 2 groups was comparable. This ability to self-collect anal cytology specimens may allow high-risk populations to be screened outside of a medical setting and allow for subsequent evaluation with HRA and treatment of lesions in individuals with abnormal cytology findings. It could also provide a useful tool for epidemiologic studies of anal cytology and possibly anal human papilloma-viruses in large population-based cohorts.

The rates of clinician-collected and self-collected anal cytology specimen adequacy were similar to those seen in a previous study that validated anal cytology as a screening test for AIN. 19 In the present study, specimens collected by clinicians had a slightly lower inadequacy rate than those collected by study subjects. However, this should not affect the role of self-screening because patients with inadequate specimens could be informed of the need to repeat the test. Notably, the sensitivity of anal cytology to detect AIN overall and AIN 2 or 3 specifically was significantly higher among HIV-positive men than among HIV-negative men, a finding consistent with our earlier study. 19 We speculate that this reflects the larger size of lesions typically found in HIV-positive men, leading to a higher likelihood of detection. Another explanation is that HIV-negative men were not as vigorous in obtaining specimens as the HIV-positive men. This is less likely because rates of specimen sufficiency were similar between HIV-positive and HIV-negative men. Further, the clinician-collected specimens were less sensitive for HIV-negative men than for HIV-positive men, similar to our findings for the self-collected specimens. Because participants were enrolled in the study only if they had undergone anal biopsy, the prevalence of AIN in the study was very high, and there were too few participants with normal anoscopic examination findings to assess differences in other measures of screening validity (eg, specificity and positive and negative predictive values) between sample specimens.

We do not yet have any data on the performance of self-collected anal cytology specimens for a population with a lower prevalence of disease. If a population has a lower prevalence of disease than the population in the present study because it has a higher proportion of HIV-negative men, then this might lead to lower sensitivity of the test for this population. In addition, it is likely that the positive predictive value for such a population would be lower than for our study population and the negative predictive value would be higher.

As with previous studies of anal and cervical cytology, a substantial proportion of normal cytology specimens was falsely negative. In this study, ~30% of subjects with biopsy-proven AIN had normal cytology findings, suggesting that, similar to cervical screening, sequential anal cytology screening may be necessary to detect AIN in this population. In this context, the role of testing for high-risk human papillomavi-ruses as an adjunct to anal cytology has not been characterized. Previous studies have shown that detection of multiple human papillomavirus types is a risk factor for progression of low-grade squamous intraepithelial lesions to HSIL, although the risk of progression associated with individual high-risk human papillomavirus types has not been fully addressed. 25

The findings for self-collected anal cytology specimens were encouraging in this study, although the results should be interpreted with caution. The study participants had experienced the collection of multiple anal cytology samples by clinicians in the past and were aware of how an adequately collected sample felt while being taken. This experience may have enhanced their ability to take an adequate specimen. Self-collected samples were taken 1 month after clinic anal cytology and anal biopsy. It is possible, although unlikely given the slow progression of dysplastic lesions in the cervical and anal canal, that the degree of dysplasia may have changed during this time. It is possible, although unlikely, that taking an anal biopsy specimen could completely have removed the dysplastic lesion or changed the natural history of the lesion. Clinician-collected cytology specimens were obtained using a glass slide and were preserved in ethanol, whereas self-collected specimens were processed using liquid cytology methods. The liquid method may have facilitated improved specimen adequacy by reducing processing problems associated with fecal contamination and low cell count. However, when previously assessed using a split-sample technique, both methods were shown to be comparable, although this may not have been the case in this study. 20

One of the strengths of this study was that the same pathologist analyzed all of the cytology and biopsy specimens, facilitating comparison of the results. The pathologist was also aware if the cytology specimen was clinician or subject collected because of the different sampling methods used. Therefore, a potential bias may have been introduced into the interpretation of cytology results. However, we do not believe that this affected the results in a substantive way because the pathologist interpreted the samples at least 1 month apart and would not have known the findings for the other samples collected from the same participant.

In summary, these data indicate that self-collected anal cytology specimens, when adequate for interpretation, had sensitivity similar to that of specimens collected by experienced clinicians. The sensitivity of anal cytology sampling to detect AIN 2 and 3 is good among HIV-positive men, but further effort is needed to improve sampling among HIV-negative men. If successful, this technique could greatly facilitate institution of anal cytology screening to prevent anal cancer among at-risk individuals and would be a useful tool to define the prevalence of abnormal anal cytology findings among large populations. Further research is needed to determine whether this technique works as well when performed by individuals with no prior experience with anal cytology and among a population with a lower prevalence of AIN. In addition, the sensitivity, specificity, and positive and negative predictive values should be studied for populations with a lower prevalence of AIN and with a larger number of HIV-negative men.

Back to Top | Article Outline


1. Palefsky JM. Anal human papillomavirus infection and anal cancer in HIV-positive individuals: an emerging problem. AIDS. 1994;8:283–295.
2. Frisch M, Glimelius B, van den Brule AJ, et al. Sexually transmitted infection as a cause of anal cancer. N Engl J Med. 1997;337:1350–1358.
3. Daling JR, Weiss NS, Hislop TG, et al. Sexual practices, sexually transmitted diseases, and the incidence of anal cancer. N Engl J Med. 1987; 317:973–977.
4. Goedert JJ, Cote TR, Virgo P, et al. Spectrum of AIDS-associated malignant disorders. Lancet. 1998;351:1833–1839.
5. Melbye M, Cote TR, Kessler L, et al. High incidence of anal cancer among AIDS patients. The AIDS/Cancer Working Group. Lancet. 1994;343: 636–639.
6. Qualters JR, Lee NC, Smith RA, et al. Breast and cervical cancer surveillance, United States, 1973-1987. Mor Mortal Wkly Rep CDC Surveill Summ. 1992;41:1–7.
7. Frisch M, Biggar RJ, Goedert JJ. Human papillomavirus-associated cancers in patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome. J Natl Cancer Inst. 2000;92:1500–1510.
8. Zaki SR, Judd R, Coffield LM, et al. Human papillomavirus infection and anal carcinoma. Retrospective analysis by in situ hybridization and the polymerase chain reaction. Am J Pathol. 1992;140:1345–1355.
9. Fenger C, Frisch M, Jass JJ, et al. Anal cancer subtype reproducibility study. Virchows Arch. 2000;436:229–233.
10. Palefsky JM. Anal squamous intraepithelial lesions: relation to HIV and human papillomavirus infection. J Acquir Immune Defic Syndr. 1999;21: S42–S48.
11. Palefsky JM, Holly EA, Hogeboom CJ, et al. Virologic, immunologic, and clinical parameters in the incidence and progression of anal squamous intraepithelial lesions in HIV-positive and HIV-negative homosexual men. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;17:314–319.
12. Melbye M, Frisch M. The role of human papillomaviruses in anogenital cancers. Semin Cancer Biol. 1998;8:307–313.
13. Bowen JT. Centennial paper. May 1912 (J Cutan Dis Syph 1912;30:241–255). Precancerous dermatoses: a study of two cases of chronic atypical epithelial proliferation. By John T. Bowen, M.D., Boston. Arch Dermatol. 1983;119:243–260.
14. Fazio VW, Tjandra JJ. The management of perianal diseases. Adv Surg. 1996;29:59–78.
15. Wright TC Jr, Cox JT, Massad LS, et al. ASCCP-Sponsored Consensus Conference. 2001 consensus guidelines for the management of women with cervical cytological abnormalities. JAMA. 2002;287:2120–2129.
16. Solomon D, Davey D, Kurman R, et al. The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA. 2002;287: 2114–2119.
17. Haye KR, Maiti H, Stanbridge CM. Cytological screening to detect subclinical anal human papillomavirus (HPV) infection in homosexual men attending a genitourinary medicine clinic. Genitourin Med. 1988;64:378–382.
18. de Ruiter A, Carter P, Katz DR, et al. A comparison between cytology and histology to detect anal intraepithelial neoplasia. Genitourin Med. 1994; 70:22–25.
19. Palefsky JM, Holly EA, Hogeboom CJ, et al. Anal cytology as a screening tool for anal squamous intraepithelial lesions. J Acquir Immune Defic Syndr Hum Retrovirol. 1997;14:415–422.
20. Darragh TM, Jay N, Tupkelewicz BA, et al. Comparison of conventional cytologic smears and ThinPrep preparations from the anal canal. Acta Cytol. 1997;41:1167–1170.
21. Goldie SJ, Kuntz KM, Weinstein MC, et al. The clinical effectiveness and cost-effectiveness of screening for anal squamous intraepithelial lesions in homosexual and bisexual HIV-positive men. JAMA. 1999;281:1822–1829.
22. Goldie SJ, Kuntz KM, Weinstein MC, et al. Cost-effectiveness of screening for anal squamous intraepithelial lesions and anal cancer in human immunodeficiency virus-negative homosexual and bisexual men. Am J Med. 2000;108:634–641.
23. United States Public Health Service (USPHS). Infectious Disease Society of America (IDSA) Guidelines for the Treatment of Opportunistic Infections in Adults and Adolescents Infected with the Human Immunodeficiency Virus. 2004; in press.
24. Chin-Hong P, Palefsky JM. The natural history and clinical management of anal HPV disease in HIV-positive men and women. Clin Infect Dis. 2002;35:1127–1134.
25. Palefsky JM, Holly EA, Ralston ML, et al. High incidence of anal high-grade squamous intra-epithelial lesions among HIV-positive and HIV-negative homosexual and bisexual men. AIDS. 1998;12:495–503.
© 2004 Lippincott Williams & Wilkins, Inc.