Secondary Logo

Share this article on:

Anal Cytology as a Screening Tool for Anal Squamous Intraepithelial Lesions

Palefsky, Joel M.*†; Holly, Elizabeth A.; Hogeboom, Charissa J.; Berry, J. Michael; Jay, Naomi*†; Darragh, Teresa M.§

Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology: April 15th, 1997 - Volume 14 - Issue 5 - p 415-422
Clinical Science

Anal squamous intraepithelial lesions (ASIL) are common in homosexual and bisexual men, and high-grade ASIL (HSIL) in particular may represent an anal cancer precursor. Cervical cytology is a useful screening tool for detection of cervical HSIL to prevent cervical cancer. To assess anal cytology as a screening tool for anal disease, we compared anal cytology with anoscopy and histopathology of anal biopsies. A total of 2958 anal examinations were performed on 407 HIV-positive and 251 HIV-negative homosexual or bisexual men participating in a prospective study of ASIL. The examination consisted of a swab for anal cytology and anoscopy with 3% acetic acid and biopsy of visible lesions. Defining abnormal cytology as including atypical squamous cells of undetermined significance and ASIL, the sensitivity of anal cytology for detection of biopsy-proven ASIL was 69% (95% confidence interval: 60 to 78) in HIV-positive and 47% (95% confidence interval: 26 to 68) in HIV-negative men at their first visit and was 81% and 50%, respectively, for all subsequent visits combined. The absence of columnar cells did not affect the sensitivity, specificity, or predictive value of anal cytology. Anal cytology may be a useful screening tool to detect ASIL, particularly in HIV-positive men. The grade of disease on anal cytology did not always correspond to the histologic grade, and anal cytology should be used in conjunction with histopathologic confirmation.

Departments of *Laboratory Medicine, †Stomatology,‡Epidemiology and Biostatistics, and §Pathology, University of California, San Francisco, San Francisco, California, U.S.A.

Address correspondence and reprint requests to Joel Palefsky, M.D., F.R.C.P.(C), Box 0100, Department of Laboratory Medicine, Univérsity of California, San Francisco, San Francisco, CA 94143, U.S.A.

Manuscript received September 18, 1996; accepted January 12, 1997.

Before the onset of the HIV epidemic, the incidence of anal squamous cell cancer (SCC) was estimated to be approximately 35 cases per 100,000 men with a history of receptive anal intercourse(1), rendering it nearly as common as cervical cancer in women before the introduction of cervical cytology (Papanicolaou) screening. HIV infection has been associated with several malignancies, including Kaposi's sarcoma and non-Hodgkin's and Hodgkin's lymphoma(2-5). Anal SCC is the fourth most common reportable malignancy in men with HIV infection(4), and data suggest that the risk of developing anal SCC may be increased in association with HIV-related immunosuppression(6). However, the degree to which HIV-related immunosuppression has contributed to this increase is not clear(4), and the true incidence of anal SCC among homosexual and bisexual HIV-positive men is unknown.

Anal SCC shares many features with SCC of the cervix. Both cancers are associated with human papillomavirus(HPV) infection(7-10). Cervical SCC is preceded by a precursor lesion known as cervical squamous intraepithelial lesion (CSIL), which like cervical SCC, is strongly associated with HPV infection. Anal squamous intraepithelial lesions (ASIL), which includes the categories of low-grade SIL (LSIL) and high-grade SIL (HSIL), are also associated with HPV infection and are similar histologically to CSIL. There have been no studies of the natural history of HSIL. However, based on the known natural history of HSIL of the cervix and its relation to cervical SCC, many investigators consider HSIL of the anus to be a precursor to anal SCC.

Given the similarities between ASIL and CSIL, an approach similar to that used for the detection of CSIL was adopted by several investigators for the detection of ASIL. Early studies using anal cytology confirmed high rates of ASIL in homosexual and bisexual men(11-16). Studies have confirmed that the rates of ASIL were higher among HIV-positive men than HIV-negative men and that the rates increased inversely with CD4 cell count, suggesting an important role for HIV-related immunosuppression(16,17).

The incidence of cervical SCC has decreased considerably in the past several decades, with much of the decline attributable to cervical cytology screening and the detection of precursor lesions. The detection of CSIL relies on the use of cervical cytology as a screening tool. Typically, a cytologic result showing any grade of CSIL and, in some cases, detection of atypical squamous cells of undetermined significance (ASCUS) triggers colposcopy. The latter procedure permits the source of the abnormal cells to be localized under magnification. Because the grade of CSIL on cytology is not considered to be a reliable indicator of how far the disease has advanced and because identification of the site of the lesion is important to guide therapy, biopsies of the abnormal area typically are obtained, and the histopathologic result constitutes the gold standard for diagnosis. In addition to its poor reliability as an indicator of degree of disease, cytology has limited sensitivity. In most studies using colposcopy or histopathology (or both) as the gold standard, the sensitivity of cervical cytology is only approximately 80%(18,19) and can vary widely (20). To minimize the negative impact of the limited sensitivity of cervical cytology, women at risk undergo repeated screenings at regular intervals, reducing the likelihood of a repeated false-negative result and increasing the likelihood that CSIL can be detected before it progresses to invasive SCC.

Unlike most other malignancies that develop in HIV-positive men and women, anal SCC may be preventable. Based on the cervical model, anal cytology could potentially be a valuable screening tool to prevent the development of anal SCC in individuals at risk. Before the introduction of such a screening program, data on the performance of the proposed screening tests are necessary. The goal of this study was to assess the sensitivity, specificity, and positive and negative predictive values of anal cytology in a cohort study of HIV-positive and HIV-negative homosexual and bisexual men, using anoscopy and anoscopically directed anal biopsy as the standard.

Back to Top | Article Outline

METHODS

Anoscopic Examinations and Biopsies

Informed consent was obtained from each subject using a protocol approved by the Committee on Human Research of the University of California, San Francisco. Subjects in this study included 658 HIV-positive and HIV-negative homosexual or bisexual men participating in a natural history study of ASIL. Each examination began with insertion of a moistened Dacron anal swab for collection of an anal cytology specimen. Swabs were inserted as far into the anal canal as possible and were removed with a twirling motion and gentle pressure on the walls of the anal canal to maximize cellular yield. The cellular material was smeared onto a glass slide and immediately immersed in a bottle containing 100% ethanol. After insertion of two more swabs for HPV testing, a disposable plastic anoscope was inserted. A Dacron swab stick wrapped in gauze that had been soaked in 3% acetic acid was then inserted through the anoscope, and the anoscope was removed to permit direct contact of the gauze with the anal mucosa. The gauze-wrapped stick was left in place for 1 minute and gently removed. The anoscope was reinserted, and a colposcope was brought to the opening of the anoscope to visualize the anal mucosa at magnifications ranging from×10 to ×40.

Consenting subjects underwent anal biopsy if one or more lesions were visualized using the colposcope. If more than one lesion was seen, biopsies were obtained from as many areas with different colposcopic appearances as possible. Biopsy specimens were fixed in 10% formalin and were paraffin-embedded for routine histopathologic examination. During some examinations, one or more lesions were seen, but no biopsy was obtained. Reasons for this included subject refusal and medical contraindications such as recent intake of acetylsalicylic acid, neutropenia (<1000 cells/mm3), thrombocytopenia (<75,000 platelets/mm3), or concurrent bacterial or viral infection in or around the anus.

Subjects with LSIL or HSIL diagnosed by cytology or histology were followed every 3 months with repeat cytology, anoscopy, and biopsy of visible lesions. HIV-positive subjects with normal cytology findings and no signs of disease at anoscopy were reexamined at 6-month intervals. HIV-negative subjects without disease were reexamined every 12 months.

Back to Top | Article Outline

Cytologic and Histopathologic Assessment

All cytologies and biopies were interpreted by one pathologist and were independently read without knowledge of the results of the other tests. Anal cytology results were classified using standard Bethesda criteria for cervical cytology(21). The normal category encompassed benign cellular changes, including inflammation, reactive changes, and squamous metaplasia. A smear was defined as insufficient when fewer than 200 well-preserved nucleated cells were present on the slide(22). Condyloma and anal intraepithelial neoplasia (AIN) 1 were combined into the LSIL category, and AIN 2 and 3 were combined into the HSIL category. For analysis, the cytology results were categorized as normal, ASCUS, LSIL, HSIL, or insufficient. The presence or absence of rectal columnar cells was also noted for each smear. Anal histopathology was graded according to standard criteria for cervical histopathology(23), and the categories were identical to those used for cytology. If multiple biopsies were obtained, the result with the highest grade of disease was used for analysis.

Because the purpose of this study was to assess the adequacy of anal cytology, only internal disease was considered; external lesions or their biopsy results were not included. Results were excluded from analysis when a lesion was seen at anoscopy but no biopsy was obtained. Cytologic findings were excluded when insufficient for diagnosis. Histologic results were excluded when insufficient for diagnosis or when they did not meet any of the previously described criteria and were classified as atypical.

Back to Top | Article Outline

Statistical Analysis

Four parameters—sensitivity, specificity, positive predictive value(PPV), and negative predictive value (NPV)—were estimated as measures of the performance of cytology, using the highest histologic diagnosis for a given visit as the estimate of the subject's true disease state.

We considered two ways to define cytologic abnormality for analysis. The first abnormality group included ASCUS and ASIL (i.e., LSIL or HSIL). The second abnormality group consisted of ASIL only and excluded ASCUS. Histologic results were classified as normal when subjects had no lesions seen on colposcopy or had a biopsy result that was interpreted as normal. An abnormal histologic finding was defined as LSIL or HSIL.

The sensitivity of cytology was defined as the proportion of cytology test results that were positive among those with histologically confirmed anal disease. The specificity of cytology was defined as the proportion of colposcopically and histologically disease-free samples that produced a negative cytology result. PPV was defined as the proportion of positive cytology results that corresponded with histologically proven disease. NPV was defined as the proportion of negative cytology results that corresponded with a disease-free state, as evaluated by colposcopy and histology.

All parameters were estimated separately for HIV-positive and HIV-negative subjects. To avoid bias and statistical dependency arising from variable numbers of unequally spaced visits per subject, estimates were obtained from a cross-sectional examination consisting of the subject's first visit after colposcopic criteria were instituted. The performance of anal cytology was also evaluated in a separate cross-sectional examination from a visit 2 years later, if available. Confidence intervals were calculated using the binomial distribution.

Point estimates using data from all visits after the first visit, as defined previously, were obtained after weighting each observation inversely by the number of visits for a given subject, thereby reducing potential bias arising from the variable number of visits per subject. No confidence intervals were estimated for these values because of the difficulty in estimating within-subject correlations for ordinal data when time intervals are variable and variance is likely to be nonuniform.

Back to Top | Article Outline

RESULTS

Study Population

Of the 658 subjects studied, 407 (62%) were HIV-positive and 251 (38%) were HIV-negative. Each was homosexual or bisexual. The mean age of the HIV-positive men was 41 years, with a range of 21 years to 66 years, and that of the HIV-negative men was 44 years, with a range of 25 to 73 years. The racial composition of the HIV-positive men was 86% Caucasian, 6% Hispanic, 2% African American, 1% "other," and information on ethnicity was missing for 6% of the subjects. The racial composition of the HIV-negative men was 88% Caucasian, 4% Hispanic, 1% African American, 3% other, and information on ethnicity was missing for 4% of the subjects. Fifty-five percent of the HIV-positive men and 62% of the HIV-negative men reported having had a college degree.

Three clinicians conducted a total of 2958 examinations on the 658 subjects, for a median of three examinations on the 658 subjects, for a median of three examinations each (range 1 to 14 examinations). One hundred twenty-two (77 HIV-positive and 45 HIV-negative) subjects had one visit only, and 207 (166 HIV-positive and 41 HIV-negative) subjects had biopsies on more than one occasion. The mean time between examinations was 6 months, and the median was 5 months (range 1 to 31 months). Of the 2300 examinations performed after the first visit, 456 examinations were excluded from analysis, leaving a total of 1844 examinations after the first visit available for analysis. Reasons for exclusion from analysis included cytology specimens insufficient for diagnosis, tissue biopsies insufficient for diagnosis, biopsies diagnosed as atypical, examinations in which lesions were observed but no biopsies were obtained, and intervals between visits of less than 45 days.

Back to Top | Article Outline

Sensitivity and Specificity of Anal Cytology

The results of cytology, anoscopy, and histology at the first visit that included colposcopic criteria are presented separately for the HIV-positive and HIV-negative subjectsin Tables 1 and 2, respectively. A higher prevalence of anal disease was found among the HIV-positive men than the HIV-negative men by cytology and anoscopy. Among the HIV-positive subjects at the first visit, the prevalence of ASIL by cytology was 27%. Excluding subjects who had visible lesions but no biopsies, ASIL was diagnosed by histology in 27% of subjects. Among the HIV-negative subjects, the prevalence of ASIL diagnosed by cytology was 4%, and ASIL was found in 8% by histologic methods.

The sensitivity, specificity, PPV, and NPV of anal cytology at the first visit are presented in Tables 3 and 4 for the HIV-positive and HIV-negative subjects, respectively. Inclusion of ASCUS in the abnormal category raised the sensitivity and lowered the specificity of anal cytology, because the definition of a positive test result became more inclusive. Among the HIV-positive men, the sensitivity of anal cytology for the detection of ASIL when ASCUS was included in the abnormal category was 69%, compared with 46% when ASCUS was excluded. The specificity was 59% when ASCUS was included, compared with 81% for that of ASIL only. Among the HIV-negative men, the sensitivity for detection of ASIL was lower than among HIV-positive men. When ASCUS was included as abnormal, the sensitivity of anal cytology among HIV-negative men was 47%, and when ASCUS was excluded, the sensitivity was 26%. However, the specificity of anal cytology was higher among HIV-negative men than HIV-positive men with ASCUS (92%) and without ASCUS (98%).

TABLE 1

TABLE 1

The results of cytology and their corresponding histopathologic findings for 1844 visits (1284 of HIV-positive men and 560 of HIV-negative men) after the first visit were also examined. The sensitivity of anal cytology among HIV-positive men was 81% when ASCUS was included as abnormal and 56% when only ASIL was used in the definition of abnormal. The specificity of anal cytology among the HIV-positive men was 63% and 81%, respectively. When data from all subsequent visits were considered, the sensitivity of anal cytology among HIV-positive men increased compared with that of the first visit, with no change in the specificity. Among the HIV-negative men, the sensitivity of anal cytology was 50% when ASCUS was included as abnormal and 27% when only ASIL was used in the definition of abnormal. The specificity of anal cytology among the HIV-negative men was 92% and 98%, respectively. These estimates were not different from those obtained at the first visit.

TABLE 2

TABLE 2

Back to Top | Article Outline

Positive and Negative Predictive Value of Anal Cytology

The PPV of abnormal anal cytology for detection of ASIL was similar among the HIV-positive (38%) and HIV-negative men (35%), and it was higher in both groups of men when ASCUS was included in the normal category (46% and 56%, respectively). The NPV was slightly higher among the HIV-negative men than the HIV-positive men when ASCUS was categorized as normal(94% versus 80%) or abnormal (95% versus 84%).

During the follow-up of this cohort, anal disease incidence and progression was seen among HIV-positive and HIV-negative men. To assess the performance of anal cytology in a population with a higher disease prevalence, we assessed sensitivity, specificity, PPV, and NPV of anal cytology for all subjects who remained in the study 2 years after the first visit. Three hundred seventy-six subjects were seen at the 2-year visit, of whom 210(56%) were HIV positive and 166 (44%) were HIV negative. At the 2-year visit, the prevalence of histologically proven ASIL was higher among HIV-positive men(50%) than in the first visit data, but the prevalence at 2 years among HIV-negative men (12%) showed only a small absolute increase compared with first visit data. The PPV of anal cytology among HIV-positive men when ASCUS was included as abnormal was 70%. When ASCUS was included in the normal category, the PPV was 78% among HIV-positive men. The NPV among HIV-positive men was 79% when ASCUS was included in the abnormal category and 70% when ASCUS was included in the normal category. Among HIV-negative men, the PPV of anal cytology was 43% when ASCUS was included as abnormal and 50% when ASCUS was excluded. The NPV was 92% when ASCUS was included in the abnormal category and 90% when ASCUS was included in the normal category. The PPV increased substantially among the HIV-positive men in association with a higher disease prevalence when compared with that of the first visit. However, among HIV-negative men, there was little change in the PPV of anal cytology at 2 years, consistent with a less marked absolute change in disease prevalence from the first visit.

TABLE 3

TABLE 3

TABLE 4

TABLE 4

Back to Top | Article Outline

Effect of Presence or Absence of Rectal Columnar Cells

A sample is considered inadequate when interpreting cervical cytology unless endocervical columnar cells are present. Similarly, when interpreting anal cytology, the presence of rectal columnar cells indicates that the anal swab collected cells up to and beyond the anorectal transformation zone. The results for how the presence or absence of rectal columnar cells influenced the performance of anal cytology are shown in Tables 3 and 4. When smears with no columnar cells were excluded from analysis, there was little change in the sensitivity, specificity, PPV, and NPV of anal cytology among the HIV-positive men and HIV-negative men.

Back to Top | Article Outline

Correlation Between Grade of Disease on Anal Cytology and Histology

We examined the correlation between grade of disease determined by anal cytology and by histopathology. A substantial number of cases of HSIL would have been missed had clinicians relied only on the results of cytology. Of 147 biopsy-proven cases of HSIL at all visits of HIV-positive men for whom cytology was abnormal(with ASCUS in the abnormal category), only 39 (27%) showed HSIL on cytology. Of 22 biopsy-proven cases of HSIL at all visits of HIV-negative men for whom cytology was abnormal (with ASCUS in the abnormal category), only 4 (18%) showed HSIL on cytology. For HIV-positive and HIV-negative men, of the 61 visits when disease was proven by biopsy and HSIL was diagnosed by cytology, 43 (70%) had biopsy-proven HSIL. Of the 397 visits when disease was proven by biopsy and LSIL was diagnosed by cytology, 81 (20%) had biopsy-proven HSIL.

Back to Top | Article Outline

DISCUSSION

It is likely that ASIL (HSIL in particular) represents the precursor lesion to invasive anal SCC, analogous to the relation between CSIL and cervical SCC. High rates of ASIL in several previous studies suggest that a substantial number of HIV-positive men may be at risk for the development of invasive anal SCC. Moreover, this risk may increase as HIV-positive individuals live longer because of improvement in medical therapy and because progression to invasive SCC may require several years. Although the prevalence of ASIL is lower among HIV-negative homosexual men, this group is also at risk given their normal life span.

Screening for precancerous anal disease in a manner analogous to that of cervical disease may prevent development of invasive SCC among those at highest risk. A thorough understanding of the performance and limitations of anal cytology, including PPV and NPV, would be of great importance before its implementation as part of a screening program for anal disease. Early studies of the sensitivity of cervical cytology in HIV-positive women have suggested that the method in this population may be less sensitive than for HIV-negative women(24). However, subsequent studies have suggested that the sensitivity of cervical cytology did not vary with HIV status(25,26). We therefore analyzed our anal cytology data separately for HIV-positive and HIV-negative homosexual men to determine whether measures such as sensitivity, specificity, NPV, and PPV differed by HIV status, because both groups may eventually participate in a screening program.

In the cervix, the term atypia has been superseded in the Bethesda classification system by the term ASCUS, and a large proportion of women with ASCUS determined on cytology do not have HPV infection or HPV-associated disease. The significance of ASCUS on anal cytology is also undetermined. However, in our study, a substantial proportion of men with ASCUS had detectable lesions on anoscopy. These data are consistent with an earlier study in which a high proportion (79%) of men with ASCUS had anal HPV infection in proportions similar to those for ASIL(13). ASCUS may reflect very early stages of anal disease and may be present at a time when some anal lesions are too small to be seen by anoscopy. Inclusion of ASCUS into the abnormal category therefore seems justified, especially given that an important requirement for a screening test is to have high sensitivity.

Several investigators have reported data on the sensitivity of anal cytology from a smaller number of subjects, with the sensitivity of cytology ranging from 42% to 98%(11,12,14), but only one previous study used anoscopy with routine histologic confirmation of anal disease as the criterion for assessing anal cytology. In that study of 56 men with biopsy-proven ASIL, anal cytology had a sensitivity of 88%, but it had a very low specificity(16%), low PPV (37%), and low NPV (70%)(15). Restriction of the definition of a positive cytology to include ASIL alone reduced the sensitivity to 37% but increased the specificity to 73%. In another report, cytologic abnormalities were detected in 98% of patients with visible warts of the anal canal and 90% of patients with intraanal acetowhite lesions(27). However, the specificity of anal cytology in that study was very low, with 81% of subjects with a colposcopically normal examination having abnormal anal cytology results.

The data from our study confirm that the sensitivity of anal cytology for the detection of biopsy-proven anal disease in HIV-positive men is equivalent to that of cervical cytology for the detection of cervical disease. However, the sensitivity of anal cytology among HIV-negative men was lower. Although the reason for this difference is unclear, it may reflect the presence of larger, more widespread lesions in the HIV-positive men, facilitating cytologic detection. This may also explain the higher sensitivity seen among HIV-positive men when all data after the first visit were combined, because many of the subjects experienced increasing immunosuppression, and it is likely that lesion size increased over time. In contrast, we saw relatively less change in anal disease among the HIV-negative men during the course of the study (data not shown), and there was little change in the sensitivity of anal cytology for this group.

Detection of a cytologic abnormality in the absence of visible disease probably reflects a false-negative examination result using anoscopy. Possible explanations include small lesion size; hemorrhoids; disease deep in the glands, anal folds, or mucus; and stool obscuring a lesion. These findings are consistent with an earlier report of histologically confirmed HSIL in a small group of men with normal anoscopic results(28).

In our study, disease progression was seen among the HIV-positive and HIV-negative men, although more markedly among the former group (data not shown), probably in conjunction with increased HIV-related immunosuppression over time. We therefore had the opportunity to measure the PPV and NPV of anal cytology in a subset of the same samples with a higher prevalence of anal disease 2 years after the first visit. The PPVs of ASIL on anal cytology among the HIV-positive men and HIV-negative men were relatively low at the first visit. However, analysis of our data at a 2-year follow-up visit showed a PPV of 70% for HIV-positive men, consistent with the higher prevalence of ASIL at this time. Our data indicate that anal cytology has a high PPV for populations with a high prevalence of anal disease.

Among HIV-negative men, the NPV of anal cytology was high, but among HIV-positive men, anal disease would be missed at approximately 20% of visits at which the anal cytology result was negative. These data illustrate the need to perform anal cytology on more than one occasion to maximize the chance of identifying individuals with anal disease.

It may be clinically important to differentiate HSIL from LSIL, because detection of the former would mandate closer follow-up and treatment, but the latter diagnosis may merit follow-up only with other clinical circumstances such as discomfort or itching. To determine the reliability of the cytologic grade of disease in predicting the histologic grade of disease, we compared these two tests and found a low correlation. Although HSIL on cytology was usually accompanied by HSIL on histologic specimens, 20% of examinations with LSIL found on cytology in subjects with biopsy-proven disease showed HSIL on histologic review. These data indicate that determination of the grade of the lesion must be made on histology, not cytology, and that patients with abnormal anal cytology findings, especially LSIL, require close follow-up.

In summary, among HIV-positive men, anal cytology has a sensitivity equivalent to that of cervical cytology for the detection of CSIL when ASCUS is included in the abnormal category. Among HIV-negative men, the sensitivity was lower than among HIV-positive men. As expected, the PPV of ASIL on anal cytology was high in an HIV-positive sample with high disease prevalence. However, among HIV-positive and HIV-negative men with normal anal cytology results, serial collection of cytology samples at intervals that remain to be determined is needed to exclude the possibility of anal disease. The performance of anal cytology was not affected by the presence or absence of rectal columnar cells. The accuracy of the grading of disease severity on cytology was not sufficient to establish the true grade of anal disease, indicating the need for anoscopy and biopsy of anal disease whenever abnormal cytology is detected.

Acknowledgments: This work was supported by grants NCI R01CA54053 and by funds provided by the Cigarette and Tobacco Surtax Fund of the State of California through the Tobacco Related Diseases Research Program of the University of California (grant RT484). These studies were carried out in the General Clinical Research Center, University of California, San Francisco with funds provided by the Division of Research Resources 5 M01-RR-00079, U.S. Public Health Service. We gratefully acknowledge the participation and dedication of the subjects of the Anal Neoplasia Study and the expert assistance of Rachel Remler, Robert Havens, and Michelle Tanner. Dedicated to the memory of David Jaramillo.

Back to Top | Article Outline

REFERENCES

1. Daling JR, Weiss NS, Hislop TG, et al. Sexual practices, sexually transmitted diseases, and the incidence of anal cancer. N Engl J Med 1987;317:973-7.
2. Harnly ME, Swan SH, Holly EA, Kelter A, Padian N. Temporal trends in the incidence of non-Hodgkin's lymphoma and selected malignancies in a population with a high incidence of acquired immunodeficiency syndrome (AIDS). Am J Epidemiol 1988;128:261-7.
3. Horn PL, DeLorenze GN, Brown SR, Holly EA, West DW. Re: "Temporal trends in the incidence of non-Hodgkin's lymphoma and selected malignancies in a population with a high incidence of acquired immunodeficiency syndrome (AIDS) [letter]." Am J Epidemiol 1989;130:1069-71.
4. Rabkin CS, Yellin F. Cancer incidence in a population with a high prevalence of infection with human immunodeficiency virus type 1. J Natl Cancer Inst 1994;86:1711-6.
5. Holly EA, Lele C. Non-Hodgkin's lymphoma in HIV positive and HIV negative homosexual men in the San Francisco Bay area: allergies, prior medication use and sexual practices. J Acquir Immune Defic Syndr Hum Retrovirol 1997 (in press).
6. Melbye M, Cote TR, Kessler L, Gail M, Biggar RJ. High incidence of anal cancer among AIDS patients: The AIDS/Cancer Working Group. Lancet 1994;343:636-9.
7. Holly EA, Whittemore AS, Aston DA, Ahn DK, Nickoloff BJ, Kristiansen JJ. Anal cancer incidence: genital warts, anal fissure or fistula, hemorrhoids, and smoking. J Natl Cancer Inst 1989;81:1726-31.
8. Beckmann AM, Daling JR, Sherman KJ, et al. Human papillomavirus infection and anal cancer. Int J Cancer 1989;43:1042-9.
9. Palefsky JM, Holly EA, Gonzales J, Berline J, Ahn DK, Greenspan JS. Detection of human papillomavirus DNA in anal intraepithelial neoplasia and anal cancer. Cancer Res 1991;51:1014-9.
10. Zaki SR, Judd R, Coffield LM, Greer P, Rolston F, Evatt BL. Human papillomavirus infection and anal carcinoma: retrospective analysis by in situ hybridization and the polymerase chain reaction. Am J Pathol 1992;140:1345-55.
11. Frazer IH, Medley G, Crapper RM, Brown TC, Mackay IR. Association between anorectal dysplasia, human papillomavirus and human immunodeficiency virus infection in homosexual men. Lancet 1986:657-60.
12. Haye KR, Maiti H, Stanbridge CM. Cytological screening to detect subclinical anal human papillomavirus (HPV) infection in homosexual men attending genitourinary medicine clinic. Genitourin Med 1988;64:378-82.
13. Palefsky JM, Gonzales J, Greenblatt RM, Ahn DK, Hollander H. Anal intraepithelial neoplasia and anal papillomavirus infection among homosexual males with group IV HIV disease. JAMA 1990;263:2911-6.
14. Law CL, Qassim M, Thompson CH, et al. Factors associated with clinical and sub-clinical anal human papillomavirus infection in homosexual men. Genitourin Med 1991;67:92-8.
15. de Ruiter A, Carter P, Katz DR, et al. A comparison between cytology and histology to detect anal intraepithelial neoplasia. Genitourin Med 1994;70:22-5.
16. Palefsky JM, Shiboski S, Moss A. Risk factors for anal human papillomavirus infection and anal cytologic abnormalities in HIV-positive and HIV-negative homosexual men. J Acquir Immune Defic Syndr 1994;7:599-606.
17. Critchlow CW, Surawicz CM, Holmes KK, et al. Prospective study of high-grade anal squamous intraepithelial neoplasia in a cohort of homosexual men: influence of HIV infection, immunosuppression and human papillomavirus infection. AIDS 1995;9:1255-62.
18. van der Graaf Y, Vooijs GP. False negative rate in cervical cytology. J Clin Pathol 1987;40:438-42.
19. Centers for Disease Control. Improving the quality of clinician Pap smear technique and management, client Pap smear education and the evaluation of Pap smear laboratory testing. Bethesda: U.S. Department of Health and Human Services Office of Family Planning 1989:A-3—A-4.
20. Joseph MG, Cragg F, Wright VC, Kontozoglou TE, Downing P, Marks FR. Cyto-histologic correlates in a colposcopic clinic: a 1-year prospective study. Diagn Cytopathol 1991;7:477-81.
21. Kurman RJ, Solomon D. The Bethesda system for reporting cervical/vaginal cytologic diagnoses: definitions, criteria and explanatory notes for terminology and specimen adequacy. New York: Springer-Verlag, 1994.
22. Kiviat N, Critchlow C, Holmes K, et al. Association of anal dysplasia and human papillomavirus with immunosuppression and HIV infection among homosexual men. AIDS 1993;7:43-9.
23. Ferenczy A, Winkler B. Cervical intraepithelial neoplasia and condyloma. In: Kurman R, ed. Blaustein's pathology of the female genital tract. 3rd ed. New York: Springer-Verlag, 1987:177-217.
24. Maiman M, Tarricone N, Vieira J, Suarez J, Serur E, Boyce JG. Colposcopic evaluation of human immunodeficiency virus-seropositive women. Obstet Gynecol 1991;78:84-8.
25. Adachi A, Fleming I, Burk RD, Ho GY, Klein RS. Women with human immunodeficiency virus infection and abnormal Papanicolaou smears: a prospective study of colposcopy and clinical outcome. Obstet Gynecol 1993;81:372-7.
26. Boardman LA, Peipert JF, Cooper AS, Cu-Uvin S, Flanigan T, Raphael SI. Cytologic-histologic discrepancy in human immunodeficiency virus-positive women referred to a colposcopy clinic. Obstet Gynecol 1994;84:1016-20.
27. Sonnex C, Scholefield JH, Kocjan G, et al. Anal human papillomavirus infection: a comparative study of cytology, colposcopy and DNA hybridisation as methods of detection. Genitourin Med 1991;67:21-5.
28. Surawicz CM, Critchlow C, Sayer J, et al. High grade anal dysplasia in visually normal mucosa in homosexual men: seven cases. Am J Gastroenterol 1995;90:1776-8.
Keywords:

Anal cytology; Dysplasia; Intraepithelial neoplasia; Anal squamous intraepithelial lesion; Anal cancer; Colposcopy

© Lippincott-Raven Publishers.