**3.1 Interrelationship of HIV and Human Papillomavirus**

Studies have shown that HIV-infected women have higher prevalence of HPV, higher incidence of HPV (Branca 2003; Ahdieh 2001), higher HPV viral load (Jamieson 2002), longer persistence of HPV (Ahdieh 2000; Sun 1997), higher likelihood of multiple HPV subtypes (Jamieson 2002; Firnhaber et al. 2009; Clifford et al. 2007), and greater prevalence of oncogenic subtypes (Minkoff 1998; Uberti-Foppa 1998; Acta Cytol 2009; 53: 10–17) than HIVuninfected women. HPV viral load is independently associated with HPV persistence (Ahdieh 2001). A recent meta-analysis found that the rate of cervical HPV infection in HIVinfected women with normal cervical cytology varied from more than 55% in South and Central America and Africa to over 30% in Asia, North America, and Europe (Clifford 2006). Furthermore, in HIV-positive women the prevalence and persistence of HPV infection increases with decreasing CD4 count and increasing HIV RNA levels (Palefsky 1999; Denny 2008) and some studies show that oncogenic HPV types may be more common with lower CD4 counts and/or higher viral loads. (Luque 1999; Minkoff 1998; Clifford et al. 2007). A recent cross-sectional study of 109 HIV+ women initiating ART in South Africa (median CD4 count 125/mm3) found a high-risk HPV (HR-HPV) prevalence of 78.9% (Moodley et al. 2009). In another South African cohort of over 123 women with HIV seroconversion HR-HPV infection doubled within 36 months of seroconversion (Wang et al. 2011). Higher HPV viral loads are also associated with lower CD4 counts (Heard 2000).

Cervical Cancer Screening and Prevention for HIV-Infected Women in the Developing World 235

In 1993, the Center for Disease Control and Prevention (CDC) expanded the case definition of AIDS to include invasive cervical cancer (ICC). As is the case with HIV- negative women, oncogenic HPV types play a central role in the relationship between HIV and cervical cancer. Recent African data found that without high-risk HPV present, the risk ratio for ICC between HIV-positive and HIV-negative women was approximately 1 (Hawes 2003). HPV types 16 and 18 were the most common HPV subtypes in a study of ICC in Kenyan HIV+ and HIV- women and were detected in 65% of ICCs in the HIV-infected patients. Almost half of the type 16 or 18

A study matching data from AIDS registries and cancer registries in 15 US regions found that persons with AIDS (information captured 4–60 months after AIDS diagnosis) had statistically significantly elevated risk of ICC compared to the general population, with standard incidence ratios (SIR) of 68.6, 95% CI = 59.7 to 78.4 ( Chaturvedi et al. 2009). During the period 1996—2004 (post-HAART introduction), ICC in women with low CD4 T-cell count was not significantly increased, possibly reflecting active screening but also not showing evidence of decline in incidence with HAART availability. There has been no evidence of increased incidence of ICC with the use of regular screening and appropriate evaluation and treatment of abnormal Paps (Massad et al. 2004; Massad et al. 2009). Casecontrol or cross-sectional studies in various African countries, including Cote d'Ivoire, Tanzania, South Africa, Kenya and Senegal have found that ICC was associated with HIV infection (Adjorlolo-Johnson et al. 2010; Kahesa et al. 2008; Stein et al. 2008; Hawes et al. 2003; Gichangi et al. 2003). However, studies evaluating the strength of association of HIV with cervical cancer among African women have shown conflicting results, possibly reflecting the competing risk of dying from other HIV-related conditions or other illnesses (Adjorlolo-Johnson et al. 2010; Moodley 2006). A recent mathematical modeling simulation projected that, compared with no ART and no screening, the lifetime cumulative risk of dying from ICC approximately doubled with ART and no screening; however, screening

even when done once, had the potential to reduce ICC mortality (Atashili et al. 2011)

with HIV-negative women of similar stage (Klevens 1996; Maiman 1990).

**4. Cervical cancer screening in developed countries** 

When ICC does develop in the setting of HIV, it tends to occur at younger ages and with less immunosuppression as compared with HIV-positive women with other AIDS-indicator conditions. Women with HIV and cervical cancer also tend to be 10—15 years younger than HIV-negative women with cervical cancer (Lomalisa 2000; van Bogaert 2011). HIV-positive women with invasive cervical cancer may present at more advanced stages (especially with CD4 <200/mm3), may metastasize to unusual locations (e.g., psoas muscle, clitoris, meningeal involvement), have poorer responses to standard therapy, and have higher recurrences and death rates, as well as shorter intervals to recurrence or death, compared

In the U.S. there have been marked reductions in cervical cancer incidence and mortality over the past 60 years, largely the result of the development and widespread introduction of the Papanicolau (Pap) test in 1949 (Sawaya 1999; van der Graaf et al. 1986; Eddy 1990), based on cytologic examination of cells obtained from the cervix with a simple scraping using a wooden spatula or brush. It is estimated that 60% of the women who are diagnosed with ICC have never had cervical cytology testing or have not been screened within the 5 years

**3.3 Invasive Cervical Cancer (ICC) in HIV disease** 

associated cancers involved multiple HPV types (De Vuyst et al. 2007).

### **3.2 HIV and cervical dysplasia**

The prevalence and incidence of abnormal Pap smears are increased among HIV-infected women as compared to uninfected women, with up to 10-fold higher rates (Maiman 1998); abnormal cervical cytology is associated with the presence of HPV infection and the degree of immunosuppression. Both frequency and severity of abnormal Pap smears and histologically documented dysplasia increase with declining CD4 counts and have also been associated with higher HIV-RNA levels (Garzetti 1995; Shah 1996; Davis 2001; Ellerbrock 2000; AIDS Care 2007; 19: 1052–1057. Massad 2008; Massad 2001). Two-thirds of 109 HIV+ women initiating ART in South Africa with median CD4 125/mm3 had abnormal Pap smears (Moodley et al. 2009). Increased HPV viral load, seen in women with more advanced HIV, is also associated with increased frequency, severity, and incidence of cervical dysplasia (Heard 2000; Weissenborn 2003; Cohn 2001). HIV is also associated with more extensive/larger volume of cervical involvement, and are also more likely to involve other areas in the lower genital tract (e.g., vulva, vagina, anal regions) (Maiman 1990). Progression and regression of Pap smear abnormalities have also been associated with level of immunosuppression and plasma viremia, as reflected in CD4 count and HIV viral load (Massad 2001; Schuman 200).

The role of effective ART and immune reconstitution in reducing the incidence and progression and promoting the regression of HPV infection and cervical dysplasia remains unclear, but HPV-related lesions do not appear to respond to ART like other opportunistic illnesses. Studies examining this issue have mixed findings, which may be related to differences in study design, virologic and immunologic parameters, duration and type of ART use, length of follow-up or other factors. In one study use of ART was associated with increased likelihood of regression of cervical dysplasia after treatment for 12 months (Heard 2002). In the Women's Interagency HIV Study (WIHS), after adjustment for CD4 count and Pap status, use of ART was associated with increased regression and decreased risk of progression of cervical cytologic abnormalities (Minkoff 2001). The HIV Epidemiology Research Study (HERS), a U.S. observational, multisite cohort study, among women with preexisting abnormal cervical cytology, ART was associated with enhanced HPV clearance but not with regression of abnormal Pap results (Paramsothy et al. 2009). In a study of women initiating HAART there was a high prevalence of cervical HPV DNA at baseline, but this declined over 96 weeks of HAART (Fife et al. 2009). On the other hand, with 15 months of follow-up, persistence of high-risk HPV and progression of SIL were comparable among women without antiretroviral treatment, those treated with nucleoside analogues only, and those treated with ART (Lillo 2001). In a more recent analysis from Women's Interagency HIV Study (WIHS), the prevalence, incident detection, and clearance of HPV infection and/or SIL before versus after ART initiation were compared, using women as their own comparison group. The role of adherence, defined as use of HAART as prescribed > or = 95% of the time, and effective ART, defined as suppression of HIV replication, were also examined. ART initiation among adherent women and among women on effective ART was associated with a significant reduction in prevalence, incident detection of oncogenic HPV infection, and decreased prevalence and more rapid clearance of oncogenic HPV-positive SIL, although strength of these protective effects was only moderate. (Minkoff et al. 2010). Given these conflicting findings, HIV-positive women should continue to be followed closely for evidence of lower genital tract neoplasia, regardless of antiretroviral therapy or viral load.
