Comprehensive, up-to-date information on HIV/AIDS treatment, prevention, and policy from the University of California San Francisco

Anogenital neoplasia, which includes both cervical and anal cancer and their likely precursor lesions, cervical and anal squamous intraepithelial lesions (SIL), is an increasing problem among HIV-infected individuals. Human papillomavirus (HPV) infection is one of the most important risk factors associated with anogenital neoplasia. In addition, both HIV and HPV are sexually transmitted diseases with similar risk factors for acquisition. The prevalence of HPV infection and anogenital neoplasia is increased in immunocompromised patients in general. The pathogenesis of anogenital neoplasia in HIV-infected patients may not only be related to the generalized impairment of cell-mediated immunity caused by HIV but also may be due to a direct interaction between the two viruses. Pathogenesis, clinical manifestations, and recommendations for management and treatment of HIV-infected persons with anogenital neoplasia will be presented in this chapter.

HPV "High Risk" Oncogenic Types

Human papillomavirus is a double-stranded DNA tumor virus. Over 70 different types of HPV that infect only mucocutaneous tissues have been identified on the basis of unique DNA sequences. Each of these types has a particular tropism for a specific type of tissue and at least 30 types have been identified with a high predilection for the anogenital tract.(1) Specific types, such as 6 and 11, have been classified as "low risk" and are associated with benign disease including condyloma acuminata or common venereal warts. "High risk" types 16 and 18 are associated with high grade cervical or anal SIL and cervical and anal cancer. HPV DNA has been identified in more than 90 percent of cervical squamous cell carcinomas.(2) Type 16 is most frequently detected and together with type 18 accounts for 64 percent of cases. Types 31, 33, 35, 45, 51, 52, and 56 are associated with an "intermediate risk" of malignancy. This "intermediate risk," however, may actually reflect the lower prevalence of these types.

Several methods can be used to detect HPV infection. HPV DNA can be obtained from cervicovaginal lavage fluid or Dacron® swabs that are inserted either into the cervix or anus. The most sensitive assay method is the polymerase chain reaction (PCR), in which the signal is amplified and hence detects qualitatively low and high level infection, but the result is not quantitative. Alternatively, HPV DNA can be analyzed by dot blot, Southern blot, or Hybrid Capture (HC) (Digene Diagnostics, Inc., Silver Spring, MD), which detect high level infection only. HC is a nonamplification, semiquantitative technique. HC is analogous to branched chain DNA testing used to determine HIV viral load. The sample DNA is not amplified but instead binds to the corresponding complementary strands, which are attached to a plastic dish. Another probe is added that provides a chemiluminescent signal that can easily be detected. Specific probes can be used with HC to distinguish "high" and "low risk" types of HPV. Although HC is approved by the Federal Drug Administration (FDA) for clinical use, the usefulness of HC and PCR to detect and type HPV in the clinical management of patients is still being determined.

Much of the knowledge of HPV infection has come from the study of young women with cervical disease. Typically, women acquire HPV shortly after becoming sexually active. In a group of sexually active college women, the cumulative incidence of new HPV infections was 43 percent over 36 months.(3) Viral infection usually occurs in the transformation zone, the junction of the squamous epithelium of the exocervix with the columnar epithelium of the endocervix. The transformation zone is where most of the HPV-related lesions and cervical cancer develop. HPV infects the basal, replicating keratinocytes.

HPV-related lesions of the cervix and anus are detected using Papanicolaou smears and direct observation. The entire anogenital region with the exception of the internal anal canal is routinely assessed in women with evidence of HPV-related abnormalities by colposcopy and biopsy.

In persons at risk for HPV, the anus can be evaluated in the outpatient setting using similar techniques. During each study visit in the UCSF Anal Neoplasia Study, subjects first have a Pap smear (Table 1) samples are taken for HPV DNA, and then anoscopy is performed with acetic acid and magnification using a colposcope. Any lesions seen are biopsied using forceps from an endoscope or other small forceps that can be inserted through the anoscope. The procedure is generally well tolerated and not associated with significant complications. Anesthesia is not required unless the lesion is located close to the anal verge. Biopsies are not performed in subjects who are leukopenic, thrombocytopenic, chronically taking aspirin, or have any purulent discharge.

Currently, most cervical and anal cytological abnormalities are classified according to the Bethesda system.(4) Histopathologic abnormalities due to HPV infection at the benign end of the spectrum include condyloma and HPV-related cellular changes, such as koilocytosis, cells with an irregular, enlarged nucleus with a clear "halo." Condyloma, koilocytosis, mild dysplasia, cervical intraepithelial neoplasia (CIN I), and anal intraepithelial neoplasia (AIN I) are all known as low grade SIL (LSIL). In contrast, high-grade SIL (HSIL) encompass moderate and severe dysplasia, CIN or AIN II, CIN or AIN III, and carcinoma-in-situ. The clinical significance between LSIL and HSIL is that almost all of the invasive cancers arise from HSIL and not LSIL. The risk of invasive cancer developing increases as the severity of dysplasia increases.

Advances in the molecular biology of HPV have elucidated a likely mechanism for its oncogenicity. Two viral oncoproteins, E6 and E7, have been identified that are thought to promote tumorigenesis by binding to host tumor suppressor gene products, p53 and Rb, respectively. These tumor suppressor gene products are responsible for the normal regulation of DNA replication, repair, and cellular growth. Increased expression of E6 and E7 has been demonstrated in HSIL resulting in abrogation of the normal controls of cellular proliferation. In anogenital cancers, HPV is often integrated into the chromosome, which results in dysregulation and increased transcription of E6 and E7. This leads to increased chromosomal instability and proliferation of cells, contributing to a setting where other potential carcinogens, such as those in tobacco smoke, can lead to development of invasive cancer. These viral genes can also transform cells and generate tumors in animal models including transgenic mice. It is important to note that in spite of the high prevalence of infection with "high risk" types of HPV which lead to increased cellular proliferation, the majority of infections do not progress to invasive cancer because other molecular changes must occur or because of other host factors that determine susceptibility.

Prior to the discovery and understanding of the molecular biology of HPV, it was speculated that cervical cancer was a sexually transmitted disease. There is now strong evidence both biologically and epidemiologically for a causal relationship between HPV and cervical cancer.(5,6) Numerous case-control and cohort studies have found a relative risk (RR) for the association between HPV infection and cervical cancer from 20 to 70, a relationship that is stronger than the association between smoking and lung cancer.

Patients with diseases characterized by defects in cell-mediated immunity, such as organ transplantation, iatrogenic immunosuppression, or congenital immunodeficiencies, have an increased prevalence of anogenital neoplasia. A 100-fold increase in cancers of the vulva and anus were reported in renal transplant patients.(7) Although it is not clear if there is an actual increase in cervical or anal cancer among HIV-positive individuals, multiple studies have clearly documented an increase in prevalence of HPV infection and CSIL in HIV-positive women. Similarly, the prevalence rate of HPV infection and ASIL is also increased in both HIV-positive men and women. It is not entirely clear why there has not been a significant increase in the incidence of invasive cervical cancer. It may be due to several factors: the time required for progression of CSIL to cancer is longer than the life span of HIV-infected individuals, under-reporting or poor access to care of women in this population, or the effectiveness of cervical cytology as a screening tool in facilitating treatment of CSIL before it progresses to cancer.

Because it is likely that CSIL and ASIL are precursor lesions to cervical and anal cancer, these cancers are both theoretically preventable. This certainly adds a public health perspective to the problem, and implies a need for both developing adequate screening programs and defining treatment. Additionally, because dysplastic lesions are more friable and bleed easily, they potentially present an additional risk factor for the transmission of HIV.

The increase in anogenital HPV infection and SIL in HIV-infected patients may be explained by several interactions between HIV and HPV. The observation of increased anogenital neoplasia in patients who are iatrogenically immunocompromised underscores the role of the immune response to HPV infection. Women with cervical lesions have decreased cellular proliferative responses to HPV type 16 E6 and E7 proteins compared to women without lesions, further evidence of the importance of cell-mediated immunity.(8) Therefore, loss of specific cell-mediated immune responses to HPV due to HIV-related immunosuppression may account in part for the increased prevalence of HPV infection and SIL in HIV-infected patients.

At the cellular level, a direct interaction between HIV and HPV may occur. Although HPV infects the epithelium, HIV may be found in Langerhans' cells, stromal cells, and infiltrating T cells. HIV-1 tat has been shown in vitro to transactivate E6 and E7 leading to increased expression of these oncoproteins.(9) Abnormal secretion of cytokines by HIV-infected lymphocytes may also facilitate increased neoplasia.

Early reports of cervical cancer in HIV-infected women, which was often their initial presentation of HIV disease, revealed aggressive clinical behavior with poor response to traditional therapy and an unusual pattern of metastases. In 1993, the CDC revised the AIDS surveillance case definition and added cervical dysplasia (moderate or severe, also known as high grade SIL) to Category B (symptomatic not A or C conditions). Invasive cervical cancer was added as an AIDS-indicator condition or category C. Since that time, further epidemiologic studies do not support the observation of an increased incidence of invasive cervical cancer. Unlike non-Hodgkin's lymphoma and Kaposi's sarcoma, the degree of increase in cervical cancer in HIV-seropositive women compared to HIV-seronegative women is unclear. A four-fold increase in incidence in cervical cancer among HIV-seropositive women was seen 5 years prior to an AIDS diagnosis that did not increase after a diagnosis of AIDS. This finding suggests that some of the increased incidence may in addition be due to factors other than HIV infection and immunosuppression such as lifestyles associated with increased sexual activity and exposure to HPV.(10)

In the United States and worldwide, women make up an increasingly large proportion of people diagnosed with AIDS, from 8 percent of all cases in the US between 1981 and 1987 to 17.5 percent of cases from 1993 to 1995. In some areas, such as New York City, the number of women who are seropositive reaches the magnitude of infection as in homosexual men in San Francisco; 53 percent of a selected group of women who attended a methadone maintenance clinic were seropositive. In 1994, the HIV seropositivity rate in New York City for childbearing women was 1.0 percent and for runaway teens in shelters was 2.6 percent.(11)

Multiple studies have consistently documented a higher prevalence of cervical HPV infection, cervical squamous epithelial lesions (CSIL), and an increase in latent infections (HPV positive without CSIL on cytology or histology) in HIV-infected women as opposed to women not infected with HIV. A large prospective cross-sectional study from New York of 344 HIV-seropositive and 325 seronegative women who were recruited from an AIDS clinic, a methadone maintenance program, and a sexually transmitted disease clinic confirmed these results.(12) The prevalence of HPV infection of any type was significantly greater in women who were seropositive compared to seronegative, 60 percent vs. 36 percent. As well, multiple types of HPV were present in 51 percent of the seropositive women compared to 26 percent of the seronegative. Latent HPV infections were also significantly increased in the HIV-infected women.

The prevalence of abnormal Pap smears, reported in a number of studies, is consistently higher in HIV-seropositive women and ranged from 5 to 63 percent compared to a range of 2 to 24 percent in seronegative women.(11) In a study reported from San Francisco General Hospital (SFGH) of 114 women who used IV drugs, 48 percent were seropositive for HIV and 52 percent seronegative.(13) CSIL was detected in 9 of 44 (20 percent) HIV-seropositive women compared to 3 of 52 (4 percent) HIV-seronegative women. All 3 of the women with cervical high grade SIL (HSIL) were HIV-infected . In the New York cohort, 80 of 398 (20 percent) HIV-seropositive women had CSIL, 52 (13 percent) with LSIL and 28 (7 percent) with HSIL; only 15 of 357 (4 percent) of the HIV-seronegative women had CSIL, including 2 patients with HSIL.(14)

The importance of a cell-mediated immune response to HPV is clear. Persistence of HPV infection is thought to be requisite for the development of SIL. In a recent update of the New York cohort, the natural history of cervical HPV infection in HIV-infected women was described.(15) Persistent HPV infection with "high risk" types was found in 20 percent of the seropositive women and 3 percent of the seronegative women. The likelihood of persistent infection was related to the level of immunosuppression, such that women with CD4 counts less than 200/mm3 were more than twice as likely to have persistence as women with CD4 counts greater than 500/mm3.

The anal and cervical epithelium have similar embryologic origins and histology ranging from columnar epithelium to a transitional zone, an area with increased metaplastic activity, to a more differentiated squamous epithelium. The histopathology of squamous cell carcinomas from the anus and cervix are also similar . Pap smears appear almost identical and are graded using the same criteria. The anal and cervical epithelia are infected by the same types of HPV producing similar manifestations from condyloma to squamous intraepithelial lesions to squamous cancer. Although anal cancer and its precursor lesion, anal squamous intraepithelial lesion (ASIL), have not been as well studied as cervical cancer and CSIL, there is similar evidence linking HPV to anal cancer. Using PCR as opposed to the inherently less sensitive in situ hybridization, HPV DNA was detected in 85 percent (11 of 13) of anal cancer specimens.(16) The predominant type was HPV 16 and this type was exclusively found in high grade ASIL. In a case-control study of anal cancer in Denmark and Sweden, 388 specimens were analyzed for the presence of HPV. "High-risk" types of HPV were detected in 89 percent of the tumors from women and 65 percent of the tumors from men as determined by PCR.(17)

Anal cancer is a rare malignancy. Incidence in the United States is about 0.8 per 100,000 persons, with almost twice as many women affected as men. Rates have increased over the last 10 years both in this country and in Denmark, primarily in women, unmarried men, and persons from large urban areas. The relative risk in homosexual men is estimated to be 80-fold higher and this increase was reported prior to the AIDS epidemic.(10) Because epidemiologic studies do not track homosexuality, the surrogate marker, never married men , has been used. The risk ratio for anal cancer was 5.68 in never married men 20 to 59 years old prior to the AIDS era (defined as 1973 to 1976 in the Surveillance, Epidemiology, and End Results (SEER) data and 1940-1976 in the Connecticut tumor registry).(18) There was not an increased risk for either colon cancer or prostate cancer. A study of 15,656 homosexual men enrolled in a hepatitis B cohort study in New York and San Francisco compared cancer incidence identified from the state registries to the expected number among adult males derived from the SEER data from 1984 to 1988. The standardized incidence ratio for anal cancer was significantly increased to 24.2. Interestingly, anal cancer did not correlate with HIV-antibody status nor occur proximate to an AIDS diagnosis.(19) Other studies, however, have reported increased rates of anal cancer close to an AIDS diagnosis.(20) Although anal cancer is an uncommon cancer, the incidence in men prior to the AIDS epidemic with a history of receptive anal intercourse is estimated to be 35 in 100,000. This rate is similar to the incidence of cervical cancer in women prior to the introduction of Pap smear screening and several-fold higher than current rates of cervical cancer in either HIV-negative or HIV-positive women (approximately 8/100,000 in the United States). It is not clear whether the incidence of anal cancer is higher in HIV-positive than HIV-negative homosexual men. At the 1998 AIDS Malignancy Conference, however, Biggar reported that the risk for anal cancer in HIV-seropositive men doubled from 15 to 30-fold as an AIDS diagnosis was approached.(21)

Genital HPV infection is multicentric and may occur outside the cervix in the vulva, vagina, or anus. Women with cervical cancer are at increased risk for the development of subsequent anal cancer with a relative risk of 4.6. Conversely, women who have been diagnosed with a primary anal cancer have a relative risk of 1.3 for subsequent invasive cervical cancer and 3.4 for cervical carcinoma in situ.(22)

Anal cytology was evaluated in 114 women in the SFGH study and abnormalities were found in 11 of 55 (20 percent) HIV-seropositive women compared to 4 of 59 (7 percent) seronegative women (OR 3.4, 95 percent CI 0.9 to 15.5). The single case of anal HSIL was in an HIV-seropositive patient. Anal HPV infection detected by dot blot was found in 17 of 54 (32 percent) HIV-seropositive women versus 8 of 59 (14 percent) of the seronegative women and almost twice as frequently as cervical infection (OR 2.5, 95 percent CI 0.9 to 7).(23) In the New York cohort, there was also a significantly higher percentage of HIV-seropositive women with anal HPV, 29 percent versus 2 percent of the seronegative women. As for anal cytologic abnormalities, 26 percent (27 of 102) of HIV-seropositive women compared to 6 percent (6 of 96) of the HIV-seronegative women had disease but no high-grade smears were seen.(24) The most important risk factor for both anal HPV infection and anal cytologic abnormalities was HIV seropositivity and lower CD4 counts.

There is a higher prevalence of ASIL and HPV infection among HIV-seropositive men compared to seronegative homosexual or bisexual men, both of which correlate with the severity of immunosuppression.

The natural history of anal HPV infection in homosexual or bisexual men is being characterized by an ongoing prospective study conducted by Palefsky et al. in San Francisco, which has enrolled 346 HIV-seropositive and 262 HIV-seronegative men. At study baseline, 93 percent of the seropositive men and 61 percent of the seronegative men had anal HPV infection detected by PCR, with HPV 16 the most common type in both groups. Infection with multiple types of HPV was more frequent in the seropositive men, 73 percent compared to 23 percent of the seronegative men. Higher levels of infection as determined by HC and multiple HPV types were found in the seropositive men, which inversely correlated with the CD4 count. HIV-negative men who engaged in receptive anal intercourse or recreational use of rectal drugs were more likely than men who did not to have anal HPV detected. Because the majority of seropositive men were infected with HPV, risk factors for detection could not be determined.(25) LSIL and HSIL were present in 36 percent (124/346) of seropositive men and 7 percent (19/262) of seronegative men. Of the 18 cases of HSIL, only one was found in a seronegative man. The relative risk (RR) of ASIL inversely correlated with CD4 count and was 3.8 (95 percent CI = 2.1 to 6.7) for a CD4 > 500/mm3, 5.6 (95 percent CI = 3.4 to 9.0) for a CD4 between 200 and 500, and 7.3 (95 percent CI = 4.6 to 12) for a CD4 < 200/mm3. HSIL was more likely to be found in men with lower CD4s. Among seronegative men, the most important risk factor for ASIL was HPV infection of any type.(26)

Cervical cancer is thought to develop from HSIL as opposed to LSIL, and it is the treatment of high-grade lesions that is likely responsible for the decrease in morbidity and mortality from cervical cancer. Based on this analogy, determining which patients are at risk for anal HSIL is particularly important. Critchlow et al.(27) performed a prospective study of 158 HIV-seropositive and 147 HIV-seronegative homosexual men who initially had no evidence of disease by anoscopy or cytology with an average follow-up of 21.1 + 11.9 months. High-grade SIL developed in 24 (15.2 percent) HIV-seropositive and 8 (5.4 percent) HIV-seronegative men. Infection with HPV 16 or 18 was significantly associated with risk for development of HSIL, with a RR of 16.6 for HIV-seronegative and 3.0 for HIV-seropositive men. Moreover, persistent HPV positivity was a risk factor, particularly at high levels. Increased immunosuppression with a CD4 < 500 /mm3 was an independent predictor and significantly associated with a greater risk of HSIL (RR 7.5; 95 percent CI= 2.2 to 25.2).

New cases of HSIL developed in 38 percent (106/277) of the seropositive men and 15 percent (34/221) of the seronegative men over a 4-year period in the San Francisco cohort.(28) Seropositive men who had LSIL or atypical cells at baseline were more likely to develop HSIL compared to men with normal cytology and no visible lesions (57 percent vs. 38 percent) (p = .001), which was also true for seronegative men (33 percent vs. 14 percent) (p = .001). Risk factors for progression to HSIL were similar among seropositive and seronegative men, including infection with multiple HPV types, persistent anal infection, and high level infection determined by HC with "high risk" types of HPV. In HIV-seropositive men risk increased with lower CD4 counts. Seronegative men who had both "low risk" and "high risk" types detected at any time had a significant risk of developing HSIL (RR 9.1, 95 percent CI = 3.3 to 352), which was similar to seropositive men with CD4 > 500/mm3 (RR 11, 95 percent CI = 4.0 to 438). The highest risk for progression was among men with a CD4 < 200/mm3 with infection with both groups (RR 21, 95 percent CI = 8.1 to 832). A small number of HIV-seronegative men developed HSIL without ever detecting HPV by PCR, 3 of 55 or 5 percent. In these cases, either HPV was not involved or the HPV types that were present were undetectable with current technology. Defining who is at risk to develop HSIL provides the basis to develop a screening protocol.

The purpose of screening is to prevent the development of invasive cancers by identifying and treating individuals at high risk. The benefit of screening and treatment protocols for cervical neoplasia in HIV-infected women is well established. If anal HSIL is a precursor to anal cancer, then identification and treatment of persons with anal HSIL may prevent the development of anal cancer. Large natural history studies of HIV-seropositive men and women are in progress, but the benefits and economic impact of a screening program to prevent anal cancer have yet to be determined. The increased prevalence and risk factors for progression of HPV infection and associated squamous intraepithelial lesions in HIV-seropositive patients have been described. It is also important to determine the clinical efficacy of routine tests such as cytology in accurately identifying people with a high risk of ASIL who may benefit from screening.

Widespread screening of women with cervical Pap smears has led to a decline in morbidity and mortality from cervical cancer. In HIV-seronegative women, abnormal cytology, typically a diagnosis of CSIL, is evaluated by colposcopy. Direct visualization and biopsy of lesions is necessary because the severity of disease is not reliably indicated by cytology. When compared to histologic results as the standard of diagnosis, cervical cytology has a sensitivity of only about 80 percent. Concerns have been raised about the reliability of Pap smears in HIV-infected women. One early study reported a sensitivity of 8 percent in HIV-infected women and others have reported discordance between cytology and histology. Subsequently, several large studies that included colposcopy as part of the routine examination of all women have found no difference in sensitivity between HIV-seropositive and seronegative women. Sensitivity was decreased, however, if cytology classified as atypical squamous cells of undetermined significance (ASCUS) was not considered positive.(13)

When Pap smears alone are used to monitor HIV-infected women, it is necessary to refer women for colposcopy who have any degree of abnormality, including ASCUS. In one study, 38 percent of seropositive women with ASCUS had CSIL detected by biopsy compared to 14 percent of uninfected women. Reasons for this discrepancy in HIV-infected women may include inflammation secondary to other sexually transmitted diseases or altered shedding of cervical endothelial cells.(29) When seropositive women are first evaluated, it is recommended that the Pap smear be repeated within about 6 months to decrease the possibility of a false-negative initial smear, given the high incidence of HSIL in HIV-positive women.

The sensitivity and specificity of anal cytology was determined in the San Francisco cohort by comparing cytology to the results of biopsy of any visible lesions from almost 3,000 examinations.(30) The sensitivity of anal cytology was 80 percent in seropositive men and 51 percent in seronegative men for the detection of ASCUS, LSIL, and HSIL, which is similar to the sensitivity of cervical Pap smears. The grade determined on cytology did not correlate well with the grade determined by histopathology of biopsied lesions. Most high-grade lesions demonstrated by cytology, however, had HSIL on biopsy (70 percent). The problem is that HSIL would have been missed if cytology alone was used to follow patients, because 20 percent of lesions shown to be LSIL by cytology were actually HSIL on biopsy. Anal cytology should be combined with anoscopy with magnification and acetic acid to accurately determine the grade of lesions and maximize the detection of HSIL.

Anal Pap smears are relatively easy to perform as described in (Table 1). For clinicians with limited experience collecting anal cytology, however, we recommend the ThinPrep method of preparing slides. After the anal canal is swabbed the Dacron swab is immediately immersed into a preservative fluid instead of smearing the cells directly onto a glass slide. From the suspension of cells a special instrument is used to create a monolayer of cells on the slide. This ThinPrep method results in decreased drying artifact and better quality specimens.(31,32)

After the Pap smear has been performed an anoscope is placed using K-Y jelly. A wooden Q-tip wrapped with a single layer gauze pad soaked with 3 percent acetic acid (vinegar diluted 2:1 with tap water) is inserted into the anal canal for 1 minute. The anoscope is withdrawn and re-inserted after removing the swab. A colposcope, consisting of a light source and binocular lenses to provide magnification of 10. to 40., is brought to the opening to visualize the anal mucosa. Alternatively, if a colposcope is not available, any light source and magnification may be used, such as loupes worn by some vascular surgeons or, in the operating room, an operating microscope. The anoscope is withdrawn slowly to visualize the entire anal canal. The transformation zone should be clearly delineated. Normal epithelium is smooth and pink. Application of 3 percent acetic acid produces a change known as acetowhitening, particularly in SIL, so that the lesions become visible and can be distinguished from nomad epithelium.(33) Lugol's solution (strong iodine), which is also frequently used as an adjunctive dye in colposcopy of the cervix, may also be used with the acetic acid. In the cervix, when this dye stains tissues a deep, mahogany brown, lesions are more likely to be low-grade or normal. When the dye instead does not stain tissues darkly but is more mustard yellow, then lesions are more likely to be high-grade. It has not been determined if this is true for the anal canal.

After the internal anal canal is examined, then acetic acid is applied to the external perianal area to visualize external lesions.

Guidelines have been published for screening HIV-seropositive women for HPV-related cervical lesions and are listed below.(34) The value of screening for the detection and treatment of HPV-related anal lesions has not been determined. The time required for progression of anal HSIL to invasive anal cancer is not known. The morbidity of treatment of anal lesions is greater and patient acceptance less than treatment of similar lesions found in the cervix. Paradoxically, patients with HIV infection who are now living longer as a consequence of improved therapy may be at greater risk to develop anal cancer if therapy does not alter the course of HPV infection. It is our recommendation at UCSF that patients with HSIL determined by biopsy, particularly those with severe dysplasia or AIN III, who appear to have a reasonable life expectancy, be referred to an interested anal surgeon for treatment. The following represents a suggested approach for screening based on the risk of HPV infection and HSIL.

Cervical Pap Smears in HIV-Seropositive Women

When HIV infection is first diagnosed, women should have a thorough gynecologic examination, including a Pap smear. If the initial Pap smear is normal, then a second smear should be repeated at approximately 6 months. If both smears are normal, then women should have a Pap smear annually. Some clinicians feel that women with a CD4 < 500/mm3 should be monitored more frequently. If either smear shows severe inflammation or reactive changes, then any infections, if present, should be treated, and a Pap smear repeated in 3 months. If any smear shows ASCUS or SIL, then the patient should be referred for colposcopy with appropriate biopsy of abnormal areas.

Anal Pap Smears in HIV-Seropositive Women and Homosexual or Bisexual Men

When HIV is first diagnosed, an anal Pap smear could be part of the initial evaluation. If the initial Pap smear is normal, then a second smear could be repeated at approximately 6 months. If both smears are normal, then patients could have a Pap smear annually. Some clinicians feel that patients with a CD4 < 500/mm3 should be monitored more frequently. If any smear shows ASCUS or SIL, then the patient should be referred for anoscopy with appropriate biopsy of abnormal areas.

Anal Pap Smears in HIV-Seronegative Women and Homosexual or Bisexual Men

Men who have had anal warts or have receptive anal intercourse may benefit from an anal Pap smear. Women who have had anal warts, a history of either cervical or vulvar HSIL or cancer, or have ever had receptive anal intercourse may benefit from an anal Pap smear. If the initial Pap smear is normal, then a second smear could be repeated at approximately 6 months. If both smears are normal, then patients could have a Pap smear every 3 to 5 years. If any smear shows ASCUS or SIL, then the patient should be referred for anoscopy with appropriate biopsy of abnormal areas.

The majority of HIV-infected patients have evidence of infection with HPV and an increased prevalence and incidence of high-grade SIL of the cervix and anus. Paradoxically, the improvement in survival for HIV-infected patients may increase the incidence of invasive anogenital cancers. HSIL is the likely precursor to both invasive cervical and anal cancer. Detection and treatment of these lesions may prevent the development of invasive cancers. The results of large, ongoing, longitudinal natural history studies of anogenital neoplasia will provide important information to better prevent HPV-related malignancies in all patients.