Human papillomavirus (HPV) is the central etiologic agent in the development of most cervical neoplasms, including invasive cervical cancer. HIV-positive women are at substantially elevated risk for HPV infection as well as cervical tumorigenesis, and this risk increases with diminished CD4+ T-cell and/or higher HIV RNA levels. Surprisingly few studies, however, have assessed the natural history of HPV infection in HIV-positive women. Gaps in our knowledge include the persistence of HPV DNA in HIV-positive women and its relation with risk of cervical neoplasms. Reactivation of latent HPV infections has been suggested to be important in immune compromised individuals, but clear evidence that this occurs is lacking. Recent widespread use of highly active anti-retroviral therapy (HAART) has dramatically affected HIV natural history. However, its impact on HPV natural history is unknown. It will be of clinical importance to determine whether greater longevity with improved but still diminished immunity with HAART ultimately increases or decreases risk of cervical disease in HIV-positive women. The purpose of this application is to study the long term effects of HIV on the natural history of HPV infection and the development of cervical neoplasms, using specimens obtained from the Women=s Interagency HIV Study (WIHS). The WIHS cohort is a large, geographically and ethnically diverse population of HIV-positive (n=2056), and risk-matched HIV-negative women (n=569). Since enrollment (October, 1994 - November 1995) WIHS subjects have been followed at 6 month intervals, and follow-up is funded through 2002. WIHS clinical data and specimens are available on a collaborative basis for use in investigator initiated research. Our research group has analyzed HPV and cytologic results at enrollment, but HPV DNA testing for most planned visits (years 2 - 7 of follow-up) has not been arranged. Under this application, we will test all untested cervical specimens, extending to 5 years the follow-up period since the widespread use of HAART began in 1997. Specimens positive for HPV 16, 18 and/or 31 will be further tested to identify the type-specific variant. Serum antibodies to papillomavirus epitopes will be measured to relate humoral immunity with HPV natural history and the development of cervical disease. These tests will be conducted to address several specific aims:

$ The association of HIV-serostatus, CD4+ T-cell count and HIV plasma RNA level with risk of 3 outcomes, namely, incident type-specific HPV DNA detection, persistence/duration of HPV DNA, and incident cervical neoplasms.

$ The effects of HAART on these same 3 outcomes, comparing (i) subjects receiving and not receiving HAART, and (ii) women before and during their use of HAART.

$ The relation of HPV type, type-specific variants, and coinfection with multiple HPV types on risk of 2 outcomes, namely, persistence/duration of HPV infections and incidence of cervical neoplasms in HIV-positive and -negative women.

$ Detection of the same HPV type-specific variant at two time points that are separated by sequential negative specimens, focusing on women who reported being sexually inactive for that entire period.

3) To study humoral immune responses to HPV and their relation with the natural history of HPV infection and development of cervical disease in HIV-positive and -negative women, focusing on:

$ Disrupted bovine papillomavirus VLP (group-specific) antibodies and their inverse (protective) association with cervical neoplasms.

$ E610 peptide antibodies and their positive association with cervical neoplasms.

$ Antibodies to HPV virus-like particles (VLP) and their association with type-specific HPV infection.

 

1. Human papillomavirus and cervical neoplasia - Human papillomavirus (HPV) is a sexually transmitted DNA virus, widely accepted to be the central etiologic agent in most cervical tumorigenesis.^1-3 Greater than 90% of cervical cancers contain HPV DNA, as do the vast majority of precancerous cervical neoplasms when tested using sensitive DNA hybridization methods.^4-6 HPV DNA detection predicts development of precancerous lesions,^{7, 8} and nested case-control studies have found that antibodies to HPV predict development of invasive cervical cancer years later.^{9, 10} The biologic plausibility of HPV oncogenicity in humans has also been well documented. HPV early proteins, E6 and E7, can immortalize human keratinocytes in vitro, in part, by binding the human tumor suppressor proteins p53 and Rb, and interfering with their functions.²

2. Risk factors for cervicovaginal HPV infection and age-specific prevalence - HPV is a sexually transmitted virus.¹¹ Most studies of virgins have failed to detect cervicovaginal HPV DNA,¹² and HPV DNA prevalence increases with life time number of sexual partners.^{7, 11, 13-15} Other factors may also be associated with risk of HPV infection, including the sexual history of heterosexual sex partners, age at first intercourse, oral contraceptive use, single marital status, and smoking.^{7, 11, 13-15} Cervicovaginal infection with HPV is common among young, sexually active women (e.g., 30 - 50% prevalence), much more common than detection of cervical neoplasms. After its initial peak in the late teens or early twenties, the age-specific prevalence of HPV begins to decrease. Recent data suggest that a second, smaller peak, may also occur among elderly women (e.g., above age 65).¹⁶

3. HPV types and type-specific variants - Over 70 individual HPV types have been defined, and more than 20 of these are known to commonly infect anogenital epithelium.^5-7 These types vary considerably in their prevalence as wells as the strength of their association with cervical cancer. HPV 16, 18, 31 and 45 are frequently referred to as AHigh Risk@ HPV types because of their high prevalence in cancer specimens. HPV 33, 35, 39, 51, 52, 56, 58, 59 and 73 are frequently termed AIntermediate Risk@ HPV types.^{5, 6} HPV types considered ALow Risk@ for the development of high grade neoplasms include HPV 6, 11, 53 and others. Among the oncogenic HPV types, particular genetic variants may be associated with especially high risk for the development of cervical cancers.^17-20 In particular, increased risk of advanced cervical neoplasms was found associated with non-European variants of HPV 16 and 18.²¹

4. Natural History of HPV Infections and Cervical Neoplasms - The decreased prevalence of HPV infection with age suggests that the vast majority of HPV infections either resolve or become latent and undetectable, and this has been confirmed in several recent cohort studies.^{7, 22, 23} The duration of HPV DNA detection, however, varies by HPV type and, consistent with expectations, the high risk HPV types are more likely to persist and to be associated with subsequent development of cervical neoplasms.⁷ Other factors associated with the likelihood of HPV DNA persistence include older age and infection with multiple HPV types.^{7, 24} HPV persistence is important, as duration of infection is associated with risk of cervical neoplasia.^{7, 22} Carcinoma in situ is a cancer precursor lesion.^25,26 Most precancerous cervical neoplasms, however, like most HPV infections, are more likely to regress than to persist or progress.²⁶ A recent historical cohort study found that regression to normal occurred in 44% of women with mild and 33% of women with moderate dysplasia within the first two years.²⁷ Higher grade neoplasms were more likely to progress, however. The risk of progression from moderate dysplasia was 16% within 2 years and 25% within 5 years.

5. Immune Response and Immune Suppression to HPV-Associated Lesions - Clearance of HPV infections and neoplastic cells is thought to involve primarily cellular immunity. Diminished T-cell responses to HPV epitopes are associated with presence of cervical neoplasms.^{28, 29} Antibodies increase in the presence of current infection, but may protect against later HPV type-specific exposures. Iatrogenically immune suppressed individuals are at high risk of HPV infection and cervical neoplasms. In a recent retrospective cohort study of 8,215 renal allograft recipients, the risk of cervical carcinoma was three-fold higher than in population-based controls.³⁰ In related studies, transplant patients were shown to have higher prevalence of cervical HPV DNA and precancerous neoplasms relative to population based controls,³¹ as well as higher prevalence of anal HPV infection and neoplasms.³² Moreover, HIV-induced immune suppression is associated with development HPV-induced cervical neoplasms.

6. Human Immunodeficiency Virus and Cervical Cancer - The incidence of acquired immunodeficiency syndrome (AIDS) has increased more in women than men in the past several years. In 1995 women constituted 19% of new AIDS cases, with rates as high 50.1 per 100,000 among African American women. The epidemic now effects women in all regions of the country, and heterosexual contact has become the major route of HIV transmission.³³ Women with AIDS have high rates of cervical cancer,³⁴ and in1993 the Centers for Disease Control and Prevention added invasive cervical neoplasms to the case definition for AIDS.³⁵ Nevertheless, it remains uncertain how much of this might be due to the shared sexual transmission of HPV and HIV or other non-biological factors.³⁴

6.1 Cross-sectional Studies of HPV Infection and Cervical Neoplasms in HIV-positive Women - HIV-infected women are at elevated risk for HPV infection as well as precancerous cervical neoplasms, and this risk increases with diminished CD4+ T-cell and/or higher HIV RNA levels.^36-45 At enrollment in WIHS, HIV-positive women with a CD4+ cell count of less than 200/mm³were at highest risk of HPV infection, as compared with HIV-negative controls, regardless of HIV RNA levels (OR=10.13; 95% CI: 7.32 - 14.04), followed by women with a CD4+ cell count of greater than 200/mm³ and an HIV RNA load greater than 20,000 copies/mL (OR=5.78; 95% CI: 4.27-8.08) and women with a CD4+ count greater than 200/mm³and an HIV RNA load less than 20,000 copies (OR=3.12; 95% CI: 2.36-4.12) 200/mm³.⁴⁶ The baseline prevalence of cervical neoplasms was also increased among the HIV-positive subjects in WIHS.⁴⁷ Cervical neoplasms were detected by cytology in 17% of HIV-positive women as compared with 4% of HIV-negative women. These lesions were highly associated with detection of HPV DNA, and in multivariate analyses both CD4+ cell count and HIV RNA levels were significant risk factors (See Preliminary Studies).

6.2 Previous Prospective Studies of HPV Infection in HIV-positive Women - Surprisingly few prospective studies of HPV infection in HIV-positive women have been reported. Although carefully conducted, the three studies we are aware of involved relatively small numbers of subjects and had short follow-up times. In addition, none evaluated the impact of highly active antiretroviral therapy (HAART) on HPV natural history. In the most recent study, preliminary data were reported from one of six WIHS study sites, and the findings highlight the need for large prospective investigations. Specifically, Minkoff et al.,⁴⁸ studied 268 HIV-positive and 265 HIV-negative women for a median of 10.4 and 13.5 months, respectively before these women, were later Arolled over@ into the WIHS cohort. Prevalence of HPV was 34% in HIV-positive and 14% in HIV-negative women, and both prevalence and persistence of HPV were strongly associated with low CD4+ T-cell levels. Unexpectedly, however, the rate of incidently detected oncogenic HPVs was highest in women with high (>500/mm³) versus low (<200/mm³) CD4+ T-cell levels. The authors suggested, this inverse association could be explained by loss of HPV specific immunity and reactivation of latent HPV infections early in the course of HIV infection. That is, possibly by the time a woman=s CD4 count drops to <200 mm³, if an old infection was going to be manifest it would already have been. Alternatively, however, sample size in this initial study, with only 29 new HPV infections among HIV positive subjects, may have been too small to properly characterize these relationships. In this connection, it was not possible to examine these data by individual HPV type. HIV RNA levels were also not studied, nor the relation of HPV DNA detection with development of cervical neoplasms.

7.1 Antibodies to HPV Virus Like Particles (VLPs) - Enzyme linked immunosorbent assays (ELISAs) based on HPV 16 virus-like particles (VLPs) have been shown to detect antibodies strongly associated with anogenital HPV 16 infection,^{59, 60} and laboratory evidence supports the premise that these antibodies are type-specific.⁶¹ Interlaboratory agreement is good to moderate (kappa = 0.4 - 0.6).⁶² Seropositivity is greatest in women with persistent HPV 16 infections ⁶⁰ and in those who progress to high grade cervical neoplasia ⁶³, both important risk factors in the development of cervical cancer. Moreover, nested case-control studies have found that HPV VLP antibodies predict the development years later of invasive cervical cancer, as well as other HPV-associated tumors.^{9, 10, 64-66} In women with current invasive cervical cancer the prevalence of HPV 16 VLP antibodies is about 50%, as compared with <5% for individuals with benign, non-gynecologic conditions.⁶⁷ Age-specific seroprevalence increases with older age but may plateau, and it is greater in women than in men.⁶⁸ The number of different HPV VLP types in common use is increasing. HPV 18, 33, 6 and 11 have all been studied.^{69, 70} Little is known regarding their natural history. However, a recent investigation by Carter et al.,⁷⁰ found that the kinetics of these antibodies was similar to that with HPV 16 VLPs.^{60, 70} Seroconversions usually occur within 8 months from the time of incident type-specific HPV, and these antibodies generally persist for an extended period, although some seroreversions occur.⁷⁰

7.2 Antibodies to HPV Peptides and to Disrupted Bovine Papillomavirus (BPV) VLPs - Antibodies to different HPV epitopes have different characteristics, providing independent information. The advantage to linear peptides is that they allow determination of the components to the humoral immune response. Antibodies to a number of HPV-derived synthetic linear peptides are associated with cervical cancer.^71-74 However, only a subset of these antibodies are also associated with precancerous cervical neoplasms.^{75, 76} The antibody most strongly associated with invasive and precancerous lesions was E610 (derived from the carboxyterminus of the E6 open reading frame of HPV 16).⁷⁶ Interestingly, E610 antibodies and antibodies to HPV 16 VLPs were not correlated, despite both being strongly associated with cervical neoplasia. However, when E610 and HPV 16 VLP antibodies were considered together, few cases and controls were misclassified.⁷⁶

In contrast to the above, a significant inverse association between antibodies to disrupted bovine papillomavirus-1 (BPV-1) VLPs and current cervical neoplasms was observed in two cross-sectional investigations.^{72, 76} In Strickler et al.,⁷⁶ controls (83%) were almost twice as likely as cases (47%) with pre-cancerous cervical lesions to be positive for IgG to disrupted BPV-1 antibodies. However, no prospective studies have been conducted to assess whether this inverse association truly reflects a protective immune response.

Surprisingly few prospective studies of HPV infection in HIV-positive women have been reported. Although carefully conducted, these earlier studies had small sample sizes, short follow-up time, and limited numbers of incident HPV infections. The currently planned study, with its large sample size and extensive follow-up, is intended to be the definitive investigation of the long term effects of HIV on HPV natural history. The population we will study, the Women=s Interagency HIV Study (WIHS) cohort, is a geographically and ethnically diverse population of HIV-positive and risk-matched HIV-negative women, that accurately reflects the demographic, social and biologic characteristics of women infected with HIV in the United States. This will be the first study to measure duration of incident instead of prevalent HPV infections, which will provide much greater accuracy, and the first study to directly investigate the relation of HPV duration with the development of cervical neoplasms. In addition, the planned study will examine the effects of HIV viral load and anti-retroviral treatment on the natural history of HPV, which have not been done previously. We summarize below several important unresolved issues we will address:

a) Persistent HPV infections in HIV-positive women and its relation with cervical neoplasms. Whether the longer duration of HPV infection in HIV-positive women is directly associated with risk of incident cervical neoplasms is not known. It is possible that the higher prevalence of cervical neoplasms in HIV-positive women is due to other factors (e.g., greater persistence of lesions), and that persistent HPV infections in HIV-positive women do not have the same significance as in HIV-negative women.

b) HPV types may vary in their association with HIV and CD4+ cell count. A group of phylogenetically-related oncogenic HPV types, including HPV 18, 45, 51, 53, 54, 56, 59 were strongly associated with CD4+ T-cell levels. However, this initial observation has not been prospectively investigated.

c) The effects of coinfection with multiple HPV types - Multiplicity of HPV infection is common in HIV-positive men and women. It is associated with incident anal neoplasms and their progression in men, but no prospective studies of multiple cervical HPV infections have been reported in HIV-positive women.

d) The effects of HPV type-specific variants in HIV-positive women. Non-European variants of HPV 16 and 18 are associated with detection of high grade cervical neoplasms, but few prospective studies have been conducted and none, to our knowledge, among HIV-positive women.

e) Reactivation of HPV - Reactivation of latent HPV infection has been suggested to be important in immune compromised women, but clear evidence of this has not been reported.

f) The effects of HAART on HPV natural history and risk of cervical neoplasms - Whether greater longevity with improved but still diminished immunity with HAART will ultimately increase or decrease risk of cervical disease in HIV-positive women is currently unknown.

g) HPV seroepidemiology in HIV infected women. HPV VLP antibodies are associated with current detection of type-specific HPV DNA, but little is known regarding their natural history. Whether or not these antibodies persist for years is relevant to current vaccine development efforts. Antibodies to disrupted BPV VLPs, a group-specific (rather than type-specific) response, had a strongly inverse association with cervical neoplasms in our cross-sectional studies. It is a priority to determine whether this antibody is truly associated (directly or as a biomarker) with protection against development of new cervical lesions. Antibody to E610 is strongly associated with detection of cervical neoplasms and may be useful in combination with VLP antibodies for screening populations (e.g., lacking access to cytology).

1. The Women=s Interagency HIV Study (WIHS)

The WIHS cohort is a large, geographically and ethnically diverse population of HIV-positive, and risk-matched HIV-negative women. Between October,1994 and November, 1995 HIV-positive and -negative women were enrolled from similar clinical and outreach sources ⁷⁷ at each of six clinical consortia (Reprint # 1). The two groups of subjects were frequency matched on demographics and several known risk factors for HIV (see Methods for details). Selected recent data comparing the HIV-positive and HIV-negative subjects are shown in Table 1 (not on web). Overall, HIV-seropositive and HIV seronegative cohorts are comparable in age, race/ethnicity, education, income, median number of sexual partners in the last 6 months, and substance abuse. In addition, the resulting HIV seropositive cohort was found to be similar in terms of race/ethnicity, exposure status, and age to national AIDS cases among U.S. women reported in 1995. Thus, it is truly a representative sample of HIV/AIDS cases among women in this country.

At baseline, the majority of WIHS subjects report being sexually active. However, approximately 855 (35%) of HIV-positive subjects and 24% of HIV-negative subjects reported not having been sexually active with a male in the previous 6 months. Some of this difference may reflect efforts by HIV-positive women to reduce risk of virus transmission, or poorer health among HIV-infected individuals. Approximately 10% of both groups had used injected drugs in the past 6 months. Following the baseline visit, follow-up visits continue every 6 months for 16 total visits (see Research Design and Methods for details).

CD4+ T-cell count and HIV-RNA levels are measured at each visit. At enrollment the median CD4+ cell count among HIV-positive women was 330 cell/mm³, and 30% had <200 cell/mm³. Approximately half of patients had HIV RNA levels above 20,000 copies/mL, and 25% had levels above 100,000 copies/mL.

The use of anti-retroviral therapy in WIHS subjects has been preliminarily examined by Ahdieh et al. ⁷⁸. Therapy was classified as monotherapy (i.e., a single nucleoside revere transcriptase inhibitor, including idovudine, stavudine, zalcitabine, didanosine, and lamivudine, or a single protease inhibitor), combination therapy (i.e., two or more nucleoside reverse transcriptase inhibitors, or a protease inhibitor plus zidovudine and stavudine), and highly active antiretroviral therapy (HAART) (i.e., two or more nucleoside reverse transcriptase inhibitors with either a protease inhibitor, such as indinivir, saquinavir, ritonavir, or nelfinavir, or a non-nucleoside reverse transcriptase inhibitor, such as nevirapine or delavirdine. or two or more protease inhibitors). At baseline, 65% of patients were receiving no antiretroviral therapy, 28% were receiving monotherapy, 7% were receiving combination therapy and <1% were receiving HAART. Widespread use of HAART in the United States began in 1997, and this was reflected in WIHS. During the 6 months surrounding January 1998, 148 women initiated HAART. Figure 1. shows the increasing probability of HAART use over time in WIHS given a fixed CD4+ cell count of 500 cells/mL and HIV RNA level of 5,000 copies/ml (estimated by Ahdieh and colleagues, using a multivariate logistic regression model).

2. HPV DNA Testing in WIHS - Cross-sectional data from HPV DNA testing of enrollment CVL samples were described in a recent report in the Journal of the National Cancer Institute ⁴⁶ (Reprint #2). In brief, HPV testing was performed using PCR with L1 consensus primers MY09/MY11 (described in Research Design and Methods). Amplification of β-globin DNA was performed as a positive control for the presence of amplifiable DNA in specimens. Testing was performed in two different laboratories (Chicago, Los Angeles, and San Francisco sites by J. Palefsky and Bronx/Manhattan, Brooklyn, and Washington D.C. by R. Burk). In 92% of HIV-positive and 92% of HIV-negative patients with available specimens, β-globin was successfully amplified, and HPV DNA data were analyzed from these 1778 HIV-positive and 500 HIV-negative patient specimens.

Duplicate testing of 129 randomly chosen samples was conducted to measure interlaboratory agreement. There was agreement in 110 (85%) of the samples overall with respect to the presence or absence of HPV, and when samples (n=25) classified as equivocally positive by either of the laboratories were excluded, there was 94% agreement. Among the 41 samples tested by both laboratories in which both reported specific HPV types, there was agreement for at least one HPV type in 39 (95%) samples. Among cases in which both laboratories reported one or two types, Palefsky detected 78% of the specific types detected by Burk, and Burk detected 80% of the types detected by Palefsky. Drs. Burk and Palefsky will continue their roles in this project.

The relation of HPV DNA detection with HIV infection, CD4+ cell count and HIV RNA levels was analyzed. HPV DNA was detected in CVL specimens from 149 (30%) of 500 HIV-negative women, as compared with 1127 (63%) of HIV-positive women. Detection of HPV DNA was associated with both lower CD4+ level (p<0.001) and higher HIV viral load (p<0.0001). At CD4 levels above 200 cells/mm³, a higher prevalence of HPV DNA was detected among women with an HIV viral load of >20,000 copies/mL. At CD4+ levels below 200 cells/mm³, HIV viral load had no effect on HPV DNA prevalence (Figure 2). In addition, infection with multiple HPV types was common and was strongly associated with HIV infection and CD4+ cell level.

Selected oncogenic HPV types were more significantly associated with CD4+ cell levels than others. HPV 18, 45 (two highly related types), as well as HPV 51, 53, 54, 56, and 59 which are all phylogenetically related to HPV 18 and 45, were as a group strongly associated with CD4+ cell levels. In contrast, HPV 16, the HPV type most commonly associated with cervical cancer, was not strongly associated with CD4+ cell count, and neither were the HPV 16-related types (HPV 31 and 35). Although, HPV 33 was associated with CD4+ cell levels, it is not as closely related to HPV 16 as HPV 31 or 35. These data raise the possibility that one or more shared epitopes may render phylogenetically related HPV types to be more or less sensitive to the loss of immune control, as reflected by declining CD4+ cell levels. If correct, this could be an important observation that will add substantially to our understanding of the interaction of immune function and HPV natural history. Alternatively, these data may reflect the multiple comparisons that were made with so many HPV types. Serial testing and prospective analysis in this population will be necessary to assess these relationships further.

3. Cervical Neoplasia in WIHS - The relation of HPV DNA detection with presence and grade of cervical neoplasms (based on cytology) was evaluated at enrollment in 1951 subjects with complete data at the time of analysis (Figure 3). Among both HIV-positive and HIV-negative women cervical neoplasms were highly associated with detection of HPV DNA. Although the prevalence of HPV DNA among women with apparent neoplasms was greater in HIV-positive women, the number of HIV-negative women with cervical lesions was small (n=14), making these estimates imprecise. In multivariate analyses, the presence of cervical neoplasms among HIV positive women was associated with CD4+ T-cell count (p<0.001), younger age (p=0.01), HPV infection (p<0.001), history of abnormal cytology (p<0.001), and history of genital warts (p=0.02). HIV RNA levels were associated with cervical neoplasms in univariate but not multivariate analyses. High risk HPV types were more strongly associated with detection of cervical neoplasms, and their prevalence increased with grade of lesion. Multiple HPV types were also significantly associated with increased risk of cervical neoplasia (P<0.001), and were significantly associated with more advanced grade of cytologic abnormality (p<0.001).

4.1 HPV 16 VLP antibodies in relation with cervical HPV infections among college women

An ELISA was employed to measure antibodies to HPV-16 VLPs in sera from sexually active college women (Reprint #3). The study subjects were members of a cohort of university women 18-35 years old, with a median age of 22. Every six months, demographic and behavioral data were collected and each woman provided a cervical specimen for Pap smear and for HPV diagnosis by PCR. Serum specimens for antibody assays were obtained at enrollment and every 12 months thereafter. Between October, 1992 and December, 1993, 414 women were enrolled in the study. At least one serum specimen was available from 376 (84%) of the 416 women.

The 376 women were classified according to their HPV PCR test results from cervical specimens collected on or before the date of the first serum specimen as follows: group 1 (n=247) negative for HPV (No HPV); group 2 (n=101), positive for an HPV type other than 16 (Other HPV); group 3 (n=28), HPV-16 positive (HPV 16). The differences in OD were statistically significant in comparisons of group 3 with the other two groups (p<0.001). The association of seropositivity with HPV DNA positivity was also analyzed as dichotomous comparisons of antibody prevalence. The seroreactivity of children between 15 and 60 months of age was used to set the cutoff point. The highest antibody prevalence was observed with sera from the HPV-16 group (46%). The antibody prevalence in women with no HPV DNA in the genital tract (19%) was significantly lower than that in the HPV-16 group (OR, 3.7; 95% CI, 1.5 to 8.9). The seroprevalence among women who harbored other HPVs in the genital tract (30%) was lower than that of women with HPV-16 infection (OR, 0.5; 95% CI 0.2 to 1.25) but greater than that in women who were HPV DNA negative (OR, 1.8; 95% CI 1.02 to 3.16).

These studies show that serum HPV-16 antibody as measured by VLP ELISA correlates with cervical HPV-16 infection; however, not all women with an HPV-16 infection had detectable antibody to HPV-16 VLPs. The reasons for this may relate to a low antibody response to infection or detection of infection by PCR prior to seroconversion. A significant number of women without a current HPV-16 infection had antibodies to HPV-16 VLPs. This maybe due to previous HPV-16 infection with persistence of an antibody response after clearance of virus from the genital tract.

4.2 Inter-laboratory agreement of HPV-16 VLP ELISAs To assess the generalizability of HPV 16 VLP results, we recently investigated the level of agreement in ELISA results from three laboratories. Each laboratory tested replicate serum specimens obtained from a heterogeneous group of patients in two separate studies: vulvar cancer case-control subjects (n=256), and consecutive blood donors (n=107) (Reprint #4). Each laboratory (R. Viscidi, Johns Hopkins; J. Schiller, NCI; and D. Waters, SAIC laboratory at Frederick ,MD) used ELISAs. Inter-laboratory correlations ranged between 0.61 and 0.80 for specimens from the case-control study and healthy blood donors, indicating that ELISA OD values in the three laboratories were linearly related regardless of the population. Kappa coefficients (k), based on each laboratory=s categorical interpretation of their results (as positive or negative), showed good agreement (k>0.6) in the case-control subjects and moderate agreement (k>0.4) in blood donors, a population that had few strongly positive sera. When OD values near the seropositive cutoff were treated as indeterminate, there was little discordance between the laboratories in either population. The data suggest that each laboratory measured the same humoral immune response, and support the use of the HPV -16 VLP ELISA as proposed for this study.

The current proposal is to investigate the long term natural history of HPV infection and the development of cervical neoplasms in HIV-positive and -negative women, using specimens obtained from the Women=s Interagency HIV Study (WIHS), a multi-center prospective cohort investigation. We will conduct HPV DNA testing in cervical lavage (CVL) samples serially collected at 6 month intervals from all subjects during visits 5 through 16. CVL samples from visits 1 through 4 (12 years of follow-up) have already been tested by our research group. The HPV polymerase chain reaction (PCR) assays we will utilize distinguish 39 individual HPV types. These cumulative data (visits 1-16 from all subjects) will be used to study the incident detection and persistence of type-specific HPV DNA, and incidence of cervical disease throughout the course of HIV/AIDS. Subjects positive for highly oncogenic types HPV 16, 18 and/or 31 will be further tested to identify the HPV type-specific variants present (using the first positive CVL), to study the relation of variants with duration of HPV infection and incidence of cervical disease. HPV type-specific variants will also be used to distinguish suspected reactivation from new infections. To study humoral immune responses to HPV antigens and their relation with HPV natural history, enzyme linked immunosorbent assays (ELISAs) will be used to measure antibodies to HPV virus-like particles (VLP), disrupted bovine papillomavirus VLPs, and peptide E610. Testing will involve first and last visit serum samples from all subjects, and additional follow-up samples from specific subjects (e.g., to assess antibody kinetics).

The WIHS is the largest multi-center prospective cohort studies in the United States, designed to understand the natural history and pathogenesis of HIV/AIDS and its complications in women. The NIH has determined in previous reviews that the primary importance of the WIHS cohort lies in the study of women-specific issues as they relate to HIV infection, and that these studies should specifically emphasize gynecologic manifestations and outcomes. However, resources for the relevant scientific investigations were not part of the original core grant.

The WIHS cohort is a large, geographically and ethnically diverse population of HIV-positive, and risk-matched HIV-negative women. The WIHS study population has been shown to accurately reflect demographic and social characteristics of women with HIV in the United States. As described below, the subjects are followed every 6 months with physical and gynecological examinations, questionnaires, and collection of laboratory specimens, including cervical lavage (CVL) specimens (which will be used for HPV DNA testing), cervical cytology, and serum. These specimens and related data are the materials for the proposed study, in which supplemental laboratory assays will be performed and additional research questions not included in the original WIHS protocol will be investigated (see enclosed letter from WIHS).

HIV-positive and -negative women were enrolled in the WIHS between October 1, 1994, and November 15 1995 from similar clinical and outreach sources at each of six WIHS study sites (Table 3 - missing from web)⁷⁷. These included HIV primary care clinics, hospital-based programs, research programs, community outreach sites, women=s support groups, drug rehabilitation programs, HIV testing sites, and referrals from previously enrolled participants. To be eligible to participate women had to be at least 13 years of age and give informed consent. They also had to agree to give blood and to be tested for HIV. To enhance participation and then maintain enrollment, there are several incentives to participation in WIHS. Each subject receives $40 per visit, and gifts such as a calendar date book are given every year at the holidays, key chains at the completion of visit 7, and a WIHS Dance for Life pin. In addition, WIHS subjects are provided access to dental care at five of the six sites, as well as access to other health services.

To select the WIHS cohort of 2058 HIV-positive and 568 HV-negative women, the recruited subjects were frequency matched on demographics and known HIV risk factors, including age, race/ethnicity, level of education, injection drug use since 1978, and total number of sex partners since 1980. Details regarding the characteristics of WIHS subjects are summarized above (see Preliminary Studies / Progress Report). In brief, the data support the comparability of HIV-positive and -negative subjects in WIHS (Reprint #1).

There was an initial 21% loss to follow-up after the baseline examination. However, the cohorts stabilized after that point, and at two years (the most recent visit with complete data) 92% of those who had presented for the first follow-up (visit 2) were in attendance (Table 4 - missing from web). Loss to follow-up has been approximately 8% greater for HIV-negative than HIV-positive women, and was associated with recent use of crack, cocaine, heroin or injection drugs (unpublished data).

To guarantee best use of the HPV DNA and serologic data generated by the proposed study, a working group has been formed to focus on studies of HPV and cervical neoplasms in the WIHS. The working group will consist of the co-investigators, as well as a representative from each WIHS site and the WIHS data coordinating center, WDMAC. Twice each year the group will meet to set the HPV research agenda. Travel expenses for working group meetings are paid by the WIHS core grant, and no additional funding is sought for this purpose. Dr. Strickler and the co-investigators in this proposal have had long standing collaborative relationships, which have resulted in numerous co-authored publications regarding HPV infection, immunity and cervical neoplasms ^{6, 46, 67, 79-82}.

The baseline visit involved a 1-1.5 hour structured interview addressing sociodemographic factors, medical and health history, obstetrical/gynecological and contraceptive history, as well as alcohol/drug use and sexual history. Medical record abstraction was conducted to confirm information regarding all hospitalizations and AIDS-defining and other HIV-related conditions. Subjects underwent a physical and gynecologic examination, involving collection of laboratory specimens. At each follow-up visit, most activities are repeated except that the questionnaire is shortened to 2 hour and some specimens and clinical data are not included. The full gynecologic examination, however, is completed at each visit. Detailed information regarding current medications is obtained, and patients are asked to bring all medications with them for review. Specimen collection includes blood, cervical lavage (CVL), pap smear, urine, and several samples for detecting sexually transmitted diseases (STDs). Table 5 summarizes all clinical and laboratory testing, included at baseline and each follow-up visit.

All pap smears are interpreted at Kyto Diagnostics (New York, NY) using the Bethesda System criteria for cytologic diagnosis. All smears are screened by two independent cytotechnologists, with 10% of all negative smears and all abnormal smears read by the cytopathlogist. These specimens are interpreted by a single research cytopathologist, and this individual is blinded to HIV status, as well as other patient characteristics, but not to the cytotechnologists= interpretations.

By protocol, all subjects with abnormal cytology, including abnormal squamous changes of undetermined significance (ASCUS), abnormal glandular changes of undetermined significance (AGUS), low and high grade squamous intraepithelial lesions (SIL), and cancer are referred for colposcopy. Details of this protocol are given in Appendix A. In brief, colposcopy is performed as soon as possible following initial detection of an abnormal pap smear. During colposcopy acetowhite lesions observed under magnification are biopsied and fixed in formaldehyde for histopathologic evaluation. If no significant lesion is found at colposcopy the subject returns to every 6 month pap smear follow-up. If a low grade lesion is detected then coloposcopy is repeated every 6 months until two consecutive normal colposcopy and pap smear results are obtained. Treatment is at the discretion of the physician. If a high grade lesions is detected at any time the patient receives definitive treatment and colposcopy every 6 months until the conclusion of the investigation, regardless of the subsequent pap smear or colposcopy findings. A recent analysis showed that approximately 66% of subjects complied with colposcopy within 6 months of an abnormal pap smear (Dr. Helen E. Cejtin, manuscript in preparation).Conventional histopathology is performed by the local institutions. Similarly, the treatment plan is determined individually by the clinician in consultation with the patient, and no uniform protocol has been imposed. However, all clinical, diagnostic and treatment information related to the patient=s care are collected using standardized WIHS reporting forms.

We will use the highest grade of neoplasia determined by cytology or histology as the final cervical diagnosis. Each final diagnosis will also be assigned a level of certainty, to account for variability between the cytologic and histologic diagnoses. When the two methods agree exactly the final diagnosis is assigned certainty=1. If both suggest the presence of abnormality (including ASCUS and AGUS) but disagree by one grade of neoplasia, or if only cytology is available, then certainty=2. Certainty=3 when greater differences are present. We have had substantial success with this approach in other observational studies ^{6, 83}, taking into account the level of certainty for each diagnosis in our analyses.

Standardized data collection forms are used at all WIHS sites, and standardization of data management was established through development of a data management manual of operations, uniform coding conventions, and monthly conference calls to resolve data issues. Data are entered at the clinical and laboratory sites using data entry software developed for WIHS. All data are transmitted to the WIHS Data Management and Analysis Center (WDMAC), at Johns Hopkins, which acts as the WIHS data repository.

D.2.11 Centralized Training and Quality Assurance Monitoring

All WIHS staff were centrally trained in standardized data collection and processing procedures for each data collection component. A detailed field manual for operations was developed, and all protocol refinements are systematically incorporated.

a) Untested CVL specimens (visits 5-16) will be assayed using L1 consensus primer PCR with MY09/MY11 primers. Testing every 6 months is necessary because the natural history of HPV infection is often less than one year, and less frequent testing would result in substantial misclassification. Testing for the full follow-up period of WIHS is warranted because little is known regarding the long term effects of HIV on HPV natural history. In addition, the recent widespread use of highly active anti-retroviral therapy (HAART), since 1997, has dramatically effected HIV natural history, further supporting the need to test all samples for HPV DNA. On the average, it is expected approximately 1800 subjects will have 12 specimens available for HPV testing. The total number of samples for PCR is 1,800 x 12 = 21,600.

a) Subjects positive for high risk types HPV 16, 18 and/or 31 will be further tested to identify the HPV type-specific variants present, using direct sequencing of virus regulatory regions to test the first of the serial CVL specimen from each subject that was positive for the virus. Recent studies have suggested that non-European variants of HPV 16 and 18 are associated with increased risk of cervical disease. However, little prospective data are available, especially in HIV-positive women. Whether HIV potentiates the tumorigenic effects of these high risk variants is not known. Therefore, we will study the three most important cancer associated HPV types (HPV 16, 18, 31) to determine whether detection of a high risk HPV variant is associated with HPV persistence and risk of later neoplasms.

D.4.1 Specific Aim 1: To study the effects of HIV on the natural history of HPV infection

Incident Detection of HPV infections in HIV-positive and -negative women

We use the term incident detection, rather than incidence, because the relative contribution of incident HPV infections versus reactivation of latent HPV infections is not known. Incident detection of an HPV infection is defined as (1) a positive PCR result, even if untyped, in a subject who was negative for any HPV DNA in all earlier CVL samples, and (2) a positive PCR result for a specific HPV type in a subject who was negative for that HPV type in all earlier CVL samples. Subjects who were negative at enrollment for HPV DNA or the relevant HPV type(s) will be included in the analysis. An event will be defined as: the new detection of any HPV DNA (i.e., among subjects who were previously negative for all HPV types); the new detection of any of a category of HPV types, such as high risk HPV-types or a group of phylogenetically-related HPV types; or, the new detection of an individual HPV type. Mid-interval will be used to estimate time of the event. Subjects will be stratified by baseline HIV, CD4+ cell and HIV RNA levels. In each stratum, Kaplan-Meier analysis will be conducted to estimate the cumulative proportion of subjects with new HPV infections over time. Kaplan-Meier curves of the different groups will be compared by the log rank test. Time-dependent Cox proportional hazards models will be used to identify risk factors for incident HPV infection. Time-dependent risk factors that may change over time include HIV serostatus, CD4+ T-cell count, HIV plasma RNA level, the use of HAART (also see below), sexual behavior in the last 6 months, and smoking. Time-independent factors include such factors as race/ethnicity.    <Sample Size Tables Missing From Web>

To determine sample size and statistical power for these comparisons, we have estimated the likely rate of new HPV infections in subjects. Minkoff et al. ⁴⁸, found the rate of incident infections with any of 14 different oncogenic HPV types to be 31 per 100 person-years among HIV-positive subjects (median follow-up = 14 months) and 12 per 100 person-years among HIV-negative controls (median 10 months follow-up), using short term data from one of the six WIHS sites (Kaplan Meier; p=0.01). Rates in Sun et al.⁴⁹, were 11 and 9 per 100 visits (every 6 months) for HIV-positive and -negative women, respectively. The reasons for the differences in these two studies is not known, but if either set of incidence rates were fixed we would expect nearly all individuals to eventually become HPV DNA positive during the 7 2 years of follow-up in this investigation. The rate of new HPV DNA detection, however, is likely to diminish over time. It is, therefore, reasonable to assume that cumulative incidence of any new HPV DNA detection will be 60% - 90% in HIV-positive subjects. The rate of new infections for individual HPV types will obviously be much lower.

Table 6 (missing from web) shows the number of HIV-positive (n=1778) and -negative (n=500) subjects at baseline who were negative for: (1) any HPV DNA type; (2) negative for high risk HPV types; and (3) negative for HPV 16. HPV 16 is used here as an example of analyses that will be conducted of common individual HPV types. Statistical power was calculated using the log rank test, making varied assumptions regarding the rate of new HPV DNA detection. Note that we assumed very conservative differences between HIV-positive and -negative women. These data indicate that we will have substantial power to detect the effect of HIV on new detection of HPV infections for a wide range of possible outcomes, including analysis of individual HPV types (or even individual HPV type-specific variants). Thus, we will have adequate data for most multivariate analyses, to also assess the effects of demographic and behavioral cofactors on incident detection of HPV DNA.

Sample Size for the Effect of CD4+ Cell levels - Among HIV-positive women analyzed for HPV DNA at enrollment (n=1778), approximately 27% had high CD4+ cell count (>500 cells/mm³), 43% intermediate levels (200-500 cells/mm³), and 30% had low levels (<200 cells/mm³). It is difficult to accurately predict the proportion with each CD4+ cell count throughout the course of follow-up, especially given the competing effects of HIV progression and HAART. However, for purposes of sample size/statistical power calculations we assume, that over the course of the investigation there will be a substantial downward shift in CD4+ cell levels in the cohort: on average, the proportion with high CD4+ will decrease 30%, the proportion with intermediate CD4+ levels will be unchanged, and the proportion with low CD4+ cell levels will increase 30%. The expected sample size is shown in Table7 (missing from web), assuming 70% follow-up.

In Minkoff et al.⁴⁸, incidence was highest among HIV-positive women with high CD4+ counts (45 per 100 person years) versus low (21 per 100 person years) (discussed in B.6.2, Background). We estimated the minimal detectable difference between subjects with high versus low CD4+ T-cell count in terms of the cumulative incidence of new HPV DNA detection. Based on a cumulative incidence of 60% - 80% among subjects with low CD4+ counts (from Table 6, above - missing from web), we would have 80% power to detect an OR ~ 1.6-1.8 for the relation of any HPV DNA with high CD4+ count.

To assess sample size and statistical power for these comparisons, we have estimated the likely rates of persistent high risk HPV infections in subjects. Sun et al.⁴⁹ found persistence of HPV infection in approximately 20% of high and 6% of intermediate risk HPV types among HIV-infected women, compared with 3% and 1% among HIV-negative women (OR>7). These differences were highly significant in that study, with only 220 HIV-positive and 231 HIV-negative women, and approximately 2 years follow-up. We based our calculations on the very conservative assumption that 20% (n= 356) of 1778 HIV positive (n=356) and 10% (n=50) of HIV-negative subjects analyzed for HPV DNA at baseline will have new detection of high risk HPV types during the course of investigation (see above, New Detection of HPV Infections).

Table 8. (missing from web) shows that there is sufficient power to detect differences in persistence among HIV-positive and -negative subjects; much smaller than expected based on Sun et al. Thus, we will have adequate data for most multivariate analyses (e.g., to assess demographic/behavioral cofactors).

Incident Detection of Cervical Neoplasms Among HIV-positive and -negative Women

Subjects who were negative at enrollment for cervical abnormality (including ASCUS and AGUS) will be included in the analysis. An event will be defined as first time detection of cervical neoplasms, and mid-interval will be used to estimate time to the event. Statistical analyses to determine risk factors for incident cervical neoplasms will be similar to those described in the previous two sections; using time to incident cervical neoplasm as the outcome.

To determine sample size and statistical power for these comparisons, we have estimated the likely rates of incident neoplasms in subjects. In Six et al.⁵¹, the incidence of newly detected cervical neoplasms was 21% in 271 HIV-positive subjects as compared with 5% in 171 HIV-negative subjects, based on a cohort followed for just two 6-monthly visits. The prevalence of cervical neoplasms at enrollment (19% in HIV-positive subjects) was similar to that in the WIHS (17%) based on a recent report by Massad et. al ⁴⁷, suggesting some comparability. If we assume similar rates of incidence in WIHS, we will have substantial statistical power to measure the relation of newly detected cervical neoplasms with HIV infection (Table 9 - missing from web). Thus, we will have adequate data for most multivariate analyses to also assess demographic and behavioral cofactors.

Note: These are conservative values given the long follow-up in this investigation, as compared with Six et al. Assuming higher incidence and greater differences, our power statistical power will be even greater.

For analyses by type-specific variants involving the high risk types (i.e., HPV 16, 18, 31), data for this analysis will be available from enrollment as well as from incidently detected infections with these HPV types. Among HIV-positive subjects, we estimate that there will be 645 infections with high risk HPV types (i.e., 289 at enrollment and an estimated 356 incidently detected infections, as above). For simplicity, we will initially compare differences in the cumulative incidence of cervical neoplasms between European and non-European genotype variants. Assuming that the cumulative incidence is 30% (slightly elevated) among the 2/3rds of subjects expected to have non-European variants, the minimal detectable effect (80% power) would be OR=1.6.

Use of HAART will be treated as a cofactor in the above multivariate, time-dependent Cox proportional hazards models, since HAART may independently effect the natural history of HPV infection and development of cervical neoplasms. Therefore, the above analyses will be used to compare women receiving and not receiving HAART; i.e., to assess the effects of HAART on incident HPV DNA detection, persistence/clearance of HPV infection, and the incidence of cervical neoplasms. However, that will not be our primary approach in evaluating the effects of HAART. There are substantial biases in who receives different forms of therapy (Aconfounding by indication@). Therefore, the most relevant comparison will be an analysis limited to the women who initiate HAART, comparing outcomes before and during HAART usage. Specifically, we will compare the risks of incident and persistent HPV DNA detection among women currently using HAART, with the risks in this same group of women for the period (similar duration) just prior to HAART usage. Persistence will be defined as positive results at two or more consecutive visits, following incident HPV DNA detection. Initial analyses will utilize the one sample proportion test to compare the cumulative incidence and cumulative frequency of persistent HPV DNA detection before and during HAART usage. Generalized estimating equations (GEE) will be used to conduct multivariate analyses, controlling for within individual intraclass/serial correlations, as well as other factors such as high versus intermediate or low risk HPV type, similar to the analyses we reported in earlier studies (Reprint #6). Incident cervical neoplasms will not be examined in these analyses, since a subject can only have an incident neoplasm once (whereas you can have repeated incident infections with many different HPV types). Therefore, the relation of HAART with incident cervical neoplasms will only be studied using HAART as a cofactor in the time-dependent Cox proportional hazards models, as described above. We will define HAART as two or more nucleoside reverse transcriptase inhibitors with either a protease inhibitor, such as indinivir, saquinavir, ritonavir, or nelfinavir, or a non-nucleoside reverse transcriptase inhibitor, such as nevirapine or delavirdine.

To estimate sample size and power based on the one sample proportion test, we determined the likely number of subjects who will initiate HAART during the investigation and the expected effect of HAART on HPV incidence, and HPV persistence/clearance. At baseline, 65% of patients were receiving no antiretroviral therapy, 28% were receiving monotherapy, 7% were receiving combination therapy and <1% were receiving HAART. Widespread use of HAART in the United States began in 1997, and this was reflected in WIHS. During the 6 months surrounding January 1998, 148 women initiated HAART. The number ever initiating HAART is now approximately 800 subjects.

a) Incident HPV DNA detection B For simplicity we have ignored the serial correlations in this analysis. We assume that 600 women will eventually be treated with HAART for at least 12 years (a conservative number given >1,000 women with CD4+ cell counts <500 cells/mm³), and that the before HAART incidence of HPV is 31 per 100 person-years (see above). Thus, approximately 50% of subjects (1.5 * 0.31 = 0.47) are expected to have at least one incident detection of a new HPV DNA type in the 12 years prior to therapy. Table10 (missing from web) shows that the power to detect a difference from this baseline cumulative incidence (based on the one sample proportions test) is adequate for even relatively small effects.

b) Persistence versus Clearance of HPV Infections - Persistence of HPV DNA detection will be examined in the subset of women who developed new HPV infections both before and after initiation of HAART. From the 600 initial subjects we estimate that 50% will develop at least one new infection before HAART (see previous), and if we assume half that proportion (25%) develop a new HPV infection after HAART we would have approximately (600 * 25% = 150) 150 subjects for analysis of HPV persistence. We further assume that 15% of infections (all HPV types) before HAART will persist (see HPV DNA Persistence, above). Table 10 shows that the power to detect differences from this baseline while on HAART is adequate to detect even relatively small effects. In addition, if we included HPV infections prevalent at the start of the two 12 year observation periods, the number of subjects analyzed would be substantially increased.

We will seek detection of the same HPV type-specific variant at two time points that are separated by sequential negative specimens, limiting investigation to women who reported being sexually inactive for that entire period. We limit study to women who were sexually inactive to minimize the possibility that these observations could reflect reinfection due to repeated exposure from the same sexual partner or other infected individuals in the community. The questionnaire specifically asks, AHave you been sexually active with a male [or a female] in the past 6 months.@ Only HPV variants that are 100% homogeneous in the amplified region will be considered the same. The steps that will be taken are shown in Table 11 (missing from web).

To estimate the likely sample size we determined the number of women with appropriate characteristics (as described above). The HIV-positive women in WIHS have a median age of 36 years (a mature age-group), consistent with the characteristics of HIV/AIDS cases among women nationally. At enrollment, 35% (n=720) of 2,058 HIV-positive and 24% (n=136) of 568 HIV-negative subjects reported they had not been ASexually active with a male in the previous 6 months@. Some HIV-positive women may be sexually inactive to reduce risk of virus transmission, or due to poor health. If even 10% of WIHS subjects (less than one third of those who had abstained for 6 months at the time of enrollment) abstained from sexual activity for 2 years during any of the 72 year study period, that will provide a sample of approximately 190 women (based on 70% follow-up of the WIHS cohort). Most of these women (from above, 720 / (720+136) = 84%) will be HIV-positive. Moreover, we expect that most HPV DNA in HIV-positive women will become undetectable for at least a part of their follow-up. Sun et al.⁴⁹ found persistence of HPV infection (positive results in sequential serial samples) in approximately 20% of high and 6% in intermediate risk HPV types among HIV-infected women, compared with 3% and 1% among HIV-negative women. Thus, we would expect the following number of sexually inactive HIV-positive women to at one point resolve HPV infections:

Assuming HPV infection, latency and reactivation occurs in just 10-20% of these HIV-positive cases we can expect to observe 10-25 clear cases of reactivation. Statistically, if reactivation occurs on average in just 10% of cases, the probability of zero observations of clear cases of reactivation would be (1.0 - 0.1)¹²⁷ = p<0.0001. The probability of $10 cases is 79%. Thus, we have ample sample size to detect reactivation in this study. If sufficient number of cases are observed we will also assess the effects of CD4+ cell count and HIV RNA levels on the risk of reactivation. We note that it is doubtful that other studies with smaller sample sizes or less extensive follow-up will have sufficient data in women at high risk of reactivation (e.g., HIV-positive) to conduct such an analysis.

D.4.3 Specific Aim 3: To study humoral immune responses to HPV and their relation with the natural history of HPV infection and development of cervical disease in HIV-positive and -negative women

a) Disrupted BPV-1 VLP Antibodies (1) To study the inverse (protective) relation of antibodies to disrupted BPV-1 VLPs with detection of current as well as subsequent cervical neoplasms, we will test baseline serum samples from all subjects, and compare seropositivity in women who did and did not develop neoplasms. (2) To describe the frequency of changes in serostatus, we will also test all last visit serum samples. The natural history of these antibodies is currently unknown. However, if the suggested model is correct, that these antibodies are group-specific (common to all HPV) rather than type-specific, there should be few seroconversions or seroreversions, as exposure to HPV is nearly universal. (3) To describe the timing (kinetics) of any changes in serostatus, among the subset of subjects whose serostatus does change we will test specimens from all their additional visits. Based on testing first and last specimens there will be approximately 5000 tests conducted; 2600 first + 1800 last + 600 total samples from a small number of seroconverters (i.e., 600 = 45 seroconverters * 14 follow-up visits, not including first and last). We assessed sample size and power for measuring the inverse association of disrupted BPV VLP antibodies with development of cervical neoplasms, both cross-sectionally and prospectively.

i) Cross-sectional analysis of the inverse relation with cervical neoplasia B We will compare seropositivity in all subjects with and without cervical neoplasms at baseline. In our previous cross-sectional investigation, IgG to disrupted BPV particles was detected in 83% of controls as compared with 47% of cases (OR = 5.8 or actually 1/5.8 = 0.17 as this is an inverse association). At baseline in WIHS there were 312 subjects with cervical neoplasms and 1520 with normal cytology. Assuming 80% seroprevalence in subjects with normal cytology (as above) we will have 80% power to detect OR=1.5 (or 1/1.5 = 0.66). Among HIV-positive subjects alone the minimal detectable effect would be OR=1.6 (or 1/1.6 =0.63). Multivariate logistic regression will be used to consider the effects of cofactors.

ii) Prospective analysis of the inverse relation with cervical neoplasia B Subjects who were negative at enrollment for cervical abnormality (including ASCUS and AGUS) will be included in the analysis. An event will be defined as first time detection of cervical neoplasms. Mid-interval will be used to estimate both times for entry and event. Time-dependent Cox proportional hazards models will be used with time to incident cervical neoplasm as the outcome. Baseline BPV serostatus will be used as an independent variable. The other time-dependent and independent risk factors that will be examined are similar to those described in earlier sections (see 4.2). Analysis will be stratified by HIV serostatus. We expect 70% overall seroprevalence in the total cohort and that the cumulative incidence of neoplasms will be 30% in HIV-positive and 10% in HIV-negative subjects. Table 12 (missing from web). shows that there is substantial power to detect a protective effect of antibody to disrupted BPV-1 particles among HIV-positive, and adequate power among HIV-negative subjects.

i) Cross-sectional at enrollment - We will compare serostatus in subjects at enrollment with and without cervical neoplasms. In our previous cross-sectional investigation, IgG to E610 was detected in approximately 84% of patients with cervical neoplasms as compared with 28% of controls (OR~17.0). Comparing seroprevalence between subjects with and without cervical neoplasms at enrollment we will have 80% power to detect an OR=1.4 (based on 312 cases of neoplasms at enrollment, and assuming a seroprevalence of 30% among controls). If we study HIV-positive subjects alone (n=295 with neoplasms at enrollment), and assume even much higher background prevalence among HIV-positive controls (i.e., seroprevalence = 50%) we would still have 80% power to detect an OR=1.5 (or 1/1.5=0.66).

iii) Seroconversion with new neoplasms B We will describe the temporal relationship of incident development of cervical neoplasms and seroconversion for E610 antibody. No power calculations are indicated for these descriptive analyses.

c) HPV Virus-like particle Antibodies - Testing of enrollment serum samples for antibodies to 12 different HPV VLPs (HPV 16, 18, 31, 33, 35, 45, 52, 53, 58, 6, 11, and AE7) is already ongoing under an R01 held by our group (Dr. Viscidi #AI42058), and no funds for that effort are requested herein. Dr. Viscidi=s grant addresses the question: Once HPV capsid (i.e., VLP) antibodies form in response to type-specific HPV infection, do they protect against later infection with the same HPV type? This questions is important since HPV VLPs are the basis of several HPV vaccines that are in development. However, Dr. Viscidi=s grant does not provide for HPV DNA testing, and without the current application there would only be 12 years of HPV DNA data follow-up. In addition, there were no provisions in the earlier grant for repeated measures of HPV VLP antibodies. This is important because little is known regarding the natural history of most HPV VLP antibodies and their relation with type-specific HPV infection, especially in HIV-positive women. Information regarding the frequency of seroreversions is particularly relevant to HPV vaccine development efforts, as at least some seroreversions have been documented ⁷⁰, and the frequency with which women lose VLP antibodies during long follow-up is not known.

The PCR detection and typing of HPV DNA is described in detail in the enclosed manuscripts (Reprint #7). Briefly, cervical samples are processed in a BioSafety Cabinet in a laboratory physically separated from where the PCR amplification is performed as previously described ⁸⁴. Ten _l of this material is then amplified using the polymerase chain reaction (PCR) with the MY09/MY11 L1 consensus primers including HMB01 which amplifies a 450 bp HPV DNA fragment and a control primer set (PC04/GH20) which simultaneously amplifies a 268 bp cellular beta-globin DNA fragment and serves as an internal control for amplification. Ten _l of the PCR reaction mix is analyzed by gel electrophoresis in 3% NuSieve/0.5% SeaKem agarose (FMC BioProducts, Rockland, ME) and transferred to nylon filters. The filters are hybridized overnight with radiolabeled generic probes for HPV and an oligonucleotide for beta-globin as described ^{85, 86}. The filters are washed in 2 X SSC (1X SSC = 0.15 mol/L sodium chloride and 0.015 mol/L sodium citrate), 0.1% SDS at 55oC and exposed to X-ray film.

Samples hybridizing to the beta-globin probe but negative for the generic probe are considered HPV negative. PCR products which are positive with the HPV generic probe are analyzed for HPV DNA type. Seven _l aliquots of the initial PCR reaction are denatured in 0.4 M NaOH, 25 mM EDTA and applied to 10 replicate filters using a 96 well dot blot apparatus (Bio Rad, Hercules, CA). Filters are individually hybridized using biotinylated type-specific oligonucleotide probes for multiple HPV types including HPV 6, 11, 13, 16, 18, 26, 31, 32, 33, 34, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 57, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 72, 73 (PAP238A), 83 (PAP291), AE2, AE5 -8, W13B, PAP155 as described ^{24, 87, 88}. Samples which are positive by the generic probe mix but negative by all type-specific probes are considered to represent "uncharacterized" HPV types. Hybridization signals are recorded using a 1 -5+ scale for signal intensity.

In each of the two HPV DNA laboratories, every 40^th sample will be a replicate control specimen, and 5% of all test specimens will be retested in a masked fashion. On an ongoing basis, Dr. Strickler will be responsible for selecting the test specimens for repeat testing and for monitoring the performance of the laboratories. The masked duplicate specimens will be prepared by the WIHS repository according to Dr. Strickler=s instructions, and labeled in such a way that they are indistinguishable from all other specimens, using unique ID numbers that are consistent with the standard WIHS format.

To monitor inter-laboratory agreement, each year we will repeat the earlier comparison between the laboratories of Drs. Palefsky and Burk, showing good to excellent concordance. Specifically, 150 randomly chosen samples will be tested in both laboratories. The results will be assessed with respect to the presence or absence of HPV, and among positive samples for the HPV types present. When more than one type is present we will determine the number and specific types detected in both laboratories. In all comparisons, we expect better than 80% concordance, and better than 90% concordance after excluding results that were identified as equivocally positive. If more substantial discordance is observed at any point, HPV testing will be halted until the problems is resolve.

89-92. In addition, since papillomaviruses have evolved over a long period of time and recombinant HPVs are not known, changes in one part of the genome are found in cis with other genetic changes ^{91, 93}. To determine HPV genome variation, we will amplify the URR region and subject this region to automated sequencing. The sequence from each clinical sample will be provided as a DNA sequence file. The sequence will be reviewed for quality by analysis and alignment with other sequences of the same type. This later method is important since it is rare to have unknown sequence variation, more commonly specific DNA sequence variants are found in phylogenetically related HPVs ^{90 92}. Variant classification will be determined by phylogenetic tree analyses. The trees will contain previous characterized Aprototype@ sequence variants representative of each major branch from previous characterization of HPV 16, 18, and 31 variants⁹².

We have baculovirus constructs for the expression of the L1 and L2 proteins of HPV-16, -18, -11, -6, -31 and -45. Constructs for HPV 33, 35, 52, 53, 58, and AE7 will be obtained from our collaborator Dr. Robert Burk. We will purify VLPs by the method of Kirnbauer ⁹⁴. Sf9 cells will be grown at 27⁰C in 1-liter spinner flasks in 500 ml of Grace=s medium supplemented with 10% fetal calf serum. One liter of cells will be harvested and infected with recombinant baculoviruses at a multiplicity of infection of ~10 in 100 ml of Grace=s medium for 1 h. An aliquot of 5 ml of infected cells will be plated into each of 20 tissue culture plates (245 X 245 mm) containing a volume of 90 ml of Grace=s medium. After 72 h, the cells will be harvested by scraping and centrifuged at 1,100 x g for 5 min. They will be washed once in cold phosphate-buffered saline (PBS) and once in PBS containing a cocktail of protease inhibitors. The final pellet will either be stored at -70⁰C or processed immediately. For purification, the cell pellet will be resuspended in 24 ml of PBS-protease inhibitors and sonicated on ice for 75 s, and the total cell lysate will be loaded onto six 40% sucrose-PBS cushions and centrifuged in a SW-28 rotor at 25,000 rpm (110,000 x g) for 2.5 h. Each of the resulting pellets will be resuspended in 2ml of 27% CsCl-PBS, sonicated for 75 s, pooled into 2 ultracentrifuge tubes, and centrifuged for 20 h at 28,000 rpm (141,000 x g) in a SW-28 rotor. The visible band (at a density of ~1.29 g/ml) will be harvested, centrifuged again by using the identical conditions, dialyzed against PBS, and stored at -70⁰C. For purification of HPV-18 VLPs the sucrose and cesium chloride solutions will be prepared in PBS/0.5M NaCl. The pellet from the 40% sucrose cushion will be centrifuged through a 40%/70% sucrose step gradient and the material banding at the interface will be collected and resuspended in 27% cesium chloride. The purity and quality of each particle preparation is then determined by several means, including analysis by an SDS-page gel, electron microscopy, the ability of particles to hemagglutinate mouse erythrocytes,⁹⁵ reactivity with polyclonal antiserum and monoclonal antibodies.

Wells of a flat bottom microtiter plate will be coated with 100ul of a particle preparation diluted in PBS. Control wells will be coated with the same concentration of the particle preparation diluted in 0.06 M carbonate buffer (pH 10.6). Particles disrupted by treatment with an alkaline buffer will serve as controls for nonspecific binding. The plates will be covered and incubated overnight at 4⁰C and then washed five times. These and subsequent washes will be done with PBS-0.05% Tween 20 using an automated ELISA plate washer. Human sera (2.5ul) will be diluted into 100ul of PBS containing 0.5% dry milk (serum dilution of 1:40) and added to the wells. Each sample will be assayed in duplicate on the same plate in wells coated with intact VLPs and denatured VLPs. After incubation for 1 h at 37⁰C, the plate will be washed five times and 100ul of horseradish peroxidase (HRP) conjugated Protein G diluted in PBS-0.05% Tween 20 will be added to the wells. The plate will be incubated at room temperature for 30 min and then washed five times. The peroxidase substrate ABTS, prepared in accordance with the manufacturer=s instructions, will be added to the wells and the plates will be incubated at room temperature for 30 min. Optical densities (OD) will be read at 405nm in a microtiter plate reader. An OD value will be calculated by subtracting the mean OD of wells coated with denatured VLP from the mean OD of wells coated with intact VLP. For a categorical interpretation of ELISA results as positive and negative, a cut-off point will be determined from OD values of control populations whose exposure to HPV is expected to be low (e.g., sexually naive college women and young children).

A positive and negative control will be included on every plate. The positive control will be prepared by pooling human serum samples previously found to be reactive. The negative control will be prepared by pooling human sera previously found to give a reactivity below the cutoff point defined above. The positive control serum will be used to assess the amount of variability between plates. The interplate variability will be estimated by the coefficient of variation (CV) for all the control OD values. The intraplate variability will be assessed by calculating the CV for the duplicate OD values for all the sera tested. For individual serum samples, the CV will be calculated using OD values of duplicate wells. If the CV exceeds 20%, the sample will be retested. In addition, as with HPV DNA testing (see above), 5% of all test specimens will be retested in masked fashion, and Dr. Strickler we be responsible for monitoring laboratory reproducibility.

There are several design issues that need to be considered throughout our statistical analyses. Although study subjects are followed at 6-month intervals it is possible that some incident HPV infections and cervical neoplasms will be missed due to their short natural history. Subjects who have short duration of HPV/neoplasia are more likely than those with long duration to resolve it before the next visit. Therefore, detection of HPV/neoplasia at any given visit is likely to be biased towards observance of infections/lesions with long duration. As this will effect all groups, any significant differences found in the incident detection of HPV/neoplasia between risk factor groups will be conservative estimates. The true/unbiased differences or relative risks are likely to be greater. Similarly, the duration of persistent HPV infections can never be exactly measured. Instead, we aim to assess relative duration; that is, compare duration among groups by HIV serostatus, CD4+ cell, etc. These issues apply to a lesser degree to serology, because we expect antibodies to persist for periods measured in months/years, rather than days, weeks and months.

Loss to follow-up and increasing numbers of deaths in HIV-positive women over time are potential additional issues. Although follow-up has been stable for several years since the first follow-up visit, it could begin to decrease. Several incentives, including financial awards, are used to minimize this possibility. However, we will need to monitor differences in the cohort over time. Loss of women with advanced HIV-infections due to death could reduce differences measured between HIV-positive and -negative women.

An approximate time table is shown. Since the laboratories have all the necessary reagents for HPV DNA testing and for most serologic assays, work will begin almost immediately. Therefore, quality assurance work and data monitoring will also begin immediately. Several cross-sectional research analyses will begin shortly thereafter. These initial studies will include the relation of HPV type-specific variants with neoplasms at baseline, and several cross-sectional studies of HPV antibodies in HIV-positive and -negative women.

1) This proposed study will utilize clinical samples and questionnaire data that had been/are being collected in an NIH funded multicenter cohort investigation designed to understand the natural history and pathogenesis of HIV/AIDS and its complications in women, the Women=s Interagency HIV Study (WIHS). In WIHS, 2056 HIV-positive and 569 HIV-negative women have been followed every 6 months since enrollment (October, 1994 - November, 1995), and observation will continue through 2002.

2) At each visit, questionnaire data, blood, and cervicovagnial cells for HPV DNA assays are collected. This proposed study will analyze these existing questionnaire data, cervical cell samples, and plasma samples.

3) The WIHS study has been approved by NIH IRB, as well as by human subjects review committees at all participating local institutions. All subjects signed an informed consent, which stated that the blood and cervical cell samples would be stored and used in future investigations, before they were recruited into the study.

4) This proposed study will use data and specimens already being collected as part of the WIHS. No risk is involved.

5) All data files and laboratory results are kept confidential by the WIHS.

6) No justification for risk is necessary, since no risk is involved.

 

 

G. Literature Cited

1) M. H. Schiffman, H. M. Bauer, R. N. Hoover, A. G. Glass, D. M. Cadell, B. B. Rush, D. R. Scott, M. E. Sherman, R. J. Kurman, S. Wacholder, C. K. Stanton and M. M. Manos. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993; 85:958-964.

2) IARC Group. Human Papillomavirus. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol. 64. Lyon: IARC, 1995.

3) P. H. Levine, H. D. Strickler, A. Manns and L. Diamondstone. Epidemiology of viruses associated with human cancers. In: D. J. McCance, ed. Human Tumor Viruses. Washingon, D.C.: ASM Press, 1998:1-49.

4) F. X. Bosch, M. M. Manos, N. Munoz, M. Sherman, A. Jansen, J. Peto, M. H. Schiffman, V. Moreno, R. Kurman, K. V. Shah and I. S. Group. Prevalence of HPV DNA in Cervical Cancer in 22 Countries. J Natl Cancer Inst 1995; 87:796-802.

5) A. T. Lorincz, R. Reid, A. B. Jenson, M. D. Greenberg, W. Lancaster and R. J. Kurman. Human papillomavirus infection of the cervix: Relative risk associations of 15 common anogenital types. ObstetGynecol 1992; 79:328-337.

6) C. Rattray, H. D. Strickler, C. Escoffery, B. Cranston, C. Brown, A. Manns, M. H. Schiffman, J. M. Palefsky, B. Hanchard and W. A. Blattner. Type-specific prevalence of human papillomavirus DNA among Jamaican colposcopy patients. J Infect Dis 1996; 173:718-721.

7) G. Y. Ho, R. Bierman, L. Beardsley, C. J. Chang and R. D. Burk. Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998; 338:423-8.

8) L. A. Koutsky, K. K. Holmes, C. W. Critchlow, C. E. Stevens, P. A. J. Paavonen, A. M. Beckmann, T. A. DeRouen, D. A. Galloway, D. Vernon and N. B. Kiviat. A cohort study of the risk of cervical intraepithelial neoplasia grade 2 or 3 in relation to papillomavirus infection. N Engl J Med 1992; 327:1272-1278.

9) K. V. Shah, R. P. Viscidi, A. J. Alberg, K. J. Helzlsouer and G. W. Comstock. Antibodies to human papillomavirus 16 and subsequent in situ or invasive cancer of the cervix. Cancer Epidemiol Biomarkers Prev 1997; 6:233-237.

10) J. Dillner, M. Lehtinen, T. Bjorge, T. Luostarinen, L. Youngman, E. Jellum, P. Koskela, R. E. Gislefoss, G. Hallmans, J. Paavonen, M. Sapp, J. T. Schiller, T. Hakulinen, S. Thoresen and M. Hakama. Prospective seroepidemiologic study of human papillomavirus infection as a risk factor for invasive cervical cancer. J Natl Cancer Inst 1997; 89:1293-9.

11) A. Hildesheim, P. Gravitt, M. H. Schiffman, R. J. Kurman, W. Barnes, S. Jones, J. G. Tchabo, L. A. Brinton, C. Copeland and J. Epp. Determinants of genital human papillomavirus infection in low- income women in Washington, D.C. Sex Transm Dis 1993; 20:279-285.

12) A. Andersson-Ellstrom, B. M. Hagmar, B. Johansson, M. Kalantari, B. Warleby and L. Forssman. Human papillomavirus deoxyribonucleic acid in cervix only detected in girls after coitus. Int J STD AIDS 1996; 7:333-6.

13) H. M. Bauer, A. Hildesheim, M. H. Schiffman, A. G. Glass, B. B. Rush, D. R. Scott, D. M. Cadell, R. J. Kurman and M. M. Manos. Determinants of genital human papillomavirus infection in low- risk women in Portland, Oregon. Sex Transm Dis 1993; 20:274-278.

14) C. M. Wheeler, C. A. Parmenter, W. C. Hunt, T. M. Becker, C. E. Greer, A. Hildesheim and M. M. Manos. Determinants of genital human papillomavirus infection among cytologically normal women attending the University of New Mexico student health center. Sex Transm Dis 1993; 20:286-289.

15) R. D. Burk, G. Y. Ho, L. Beardsley, M. Lempa, M. Peters and R. Bierman. Sexual behavior and partner characteristics are the predominant risk factors for genital human papillomavirus infection in young women. J Infect Dis 1996; 174:679-89.

16) E. Franco. Epidemiology and Natural History of HPV-Induced Disease, 17th International Papillomavirus Conference, Charleston, South Carolina, 1999.

17) L. F. Xi, L. A. Koutsky, D. A. Galloway, J. Kuypers, J. P. Hughes, C. M. Wheeler, K. K. Holmes and N. B. Kiviat. Genomic variation of human papillomavirus type 16 and risk for high grade cervical intraepithelial neoplasia. J Natl Cancer Inst 1997; 89:796-802.

18) M. Vand Duin, P. Snijders, E. Klaassen, T. Helmerhorst, R. Verheijen, C. Meijer and J. Walboomers. HPV 16 variant analysis in a prospective follow-up study of women with abnormal cervical cytology, 17th International Papillomavirus Conference, Charleston, South Carolina, 1999.

19) I. Zehbe, H. Delius, J. Voglino, E. Wilander and M. Tommasino. Human papillomavirus 16 E6 polymorphisms: geographically related risk of cervical cancer development, 17th International Papillomavirus Conference, 1999.

20) I. Nindl, K. Rindfleisch, T. Binder, B. Horn, A. Schneider and M. Durst. Sequnce variations of the oncogenes HPV 16 E6 and E7 and HLA-typing in patients with normal histology, CIN and cervical cancer, 17th International Papillomavirus Conference, Charleston, South Carolina, 1999.

21) L. Sichero, P. Rahal, L. Villa, A. Ferency and E. Franco. Molecular Variants of HPV-16 and -18 associated with high-grade cervical lesions, 17th International Papillomavirus Conference, Charleston, South Carolina, 1999.

22) A. B. Moscicki, S. Shiboski, J. Broering, K. Powell, L. Clayton, N. Jay, T. M. Darragh, R. Brescia, S. Kanowitz, S. B. Miller, J. Stone, E. Hanson and J. Palefsky. The natural history of human papillomavirus infection as measured by repeated DNA testing in adolescent and young women. J Pediatr 1998; 132:277-84.

23) M. Evander, K. Edlund, A. Gustafsson, M. Jonsson, R. Karlsson, E. Rylander and G. Wadell. Human papillomavirus infection is transient in young women: a population-based cohort study. J Infect Dis 1995; 171:1026-30.

24) A. Hildesheim, M. H. Schiffman, P. E. Gravitt, A. G. Glass, C. E. Greer, T. Zhang, D. R. Scott, B. B. Rush, P. Lawler and M. E. Sherman. Persistence of type-specific human papillomavirus infection among cytologically normal women. J Infect Dis 1994; 169:235-240.

25) L. G. Koss, F. W. Stewart, F. W. Foote, M. J. Jordan, G. M. Bader and E. Day. Some histological aspects of behavior of epidermoid carcinoma in situ and related lesions of the uterine cervix. Cancer 1963; 16:1160-1211.

26) W. Chanen. The CIN saga--the biological and clinical significance of cervical intraepithelial neoplasia. Aust N Z J Obstet Gynaecol 1990; 30:18-23.

27) P. Holowaty, A. B. Miller, T. Rohan and T. To. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 1999; 91:252-8.

28) A. S. Kadish, G. Y. Ho, R. D. Burk, Y. Wang, S. L. Romney, R. Ledwidge and R. H. Angeletti. Lymphoproliferative responses to human papillomavirus (HPV) type 16 proteins E6 and E7: outcome of HPV infection and associated neoplasia. J Natl Cancer Inst 1997; 89:1285-93.

29) T. Tsukui, A. Hildesheim, M. H. Schiffman, J. Lucci, 3rd, D. Contois, P. Lawler, B. B. Rush, A. T. Lorincz, A. Corrigan, R. D. Burk, W. Qu, M. A. Marshall, D. Mann, M. Carrington, M. Clerici, G. M. Shearer, D. P. Carbone, D. R. Scott, R. A. Houghten and J. A. Berzofsky. Interleukin 2 production in vitro by peripheral lymphocytes in response to human papillomavirus-derived peptides: correlation with cervical pathology. Cancer Res 1996; 56:3967-3974.

30) C. K. Fairley, A. G. Sheil, J. J. McNeil, A. M. Ugoni, A. P. Disney, G. G. Giles and N. Amiss. The risk of ano-genital malignancies in dialysis and transplant patients. Clin Nephrol 1994; 41:101-5.

31) C. K. Fairley, S. Chen, S. N. Tabrizi, J. McNeil, G. Becker, R. Walker, R. C. Atkins, N. Thomson, P. Allan, C. Woodburn and et al. Prevalence of HPV DNA in cervical specimens in women with renal transplants: a comparison with dialysis-dependent patients and patients with renal impairment. Nephrol Dial Transplant 1994; 9:416-20.

32) O. A. Ogunbiyi, J. H. Scholefield, A. T. Raftery, J. H. Smith, S. Duffy, F. Sharp and K. Rogers. Prevalence of anal human papillomavirus infection and intraepithelial neoplasia in renal allograft recipients. Br J Surg 1994; 81:365-7.

33) P. M. Wortley and P. L. Fleming. AIDS in women in the United States. Recent trends. JAMA 1997; 278:911-6.

34) J. J. Goedert, T. R. Cote, P. Virgo, S. M. Scoppa, D. W. Kingma, M. H. Gail, E. S. Jaffe and R. J. Biggar. Spectrum of AIDS-associated malignant disorders. Lancet 1998; 351:1833-1839.

35) CDC. From the Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. JAMA 1993; 269:729-30.

36) F. H. Sillman, R. G. Fruchter, Y. S. Chen, L. Camilien, A. Sedlis and E. McTigue. Vaginal intraepithelial neoplasia: risk factors for persistence, recurrence, and invasion and its management. Am J Obstet Gynecol 1997; 176:93-9.

37) A. R. Feingold, S. H. Vermund, R. D. Burk, K. F. Kelley, L. K. Schrager, K. Schreiber, G. Munk, G. H. Friedland and R. S. Klein. Cervical cytologic abnormalities and papillomavirus in women infected with human immunodeficiency virus. J Acquir Immune Defic Syndr 1990; 3:896-903.

38) M. Maiman, R. G. Fruchter, E. Serur, J. C. Remy, G. Feuer and J. Boyce. Human immunodeficiency virus infection and cervical neoplasia. Gynecol Oncol 1990; 38:377-82.

39) L. K. Schrager, G. H. Friedland, D. Maude, K. Schreiber, A. Adachi, D. J. Pizzuti, L. G. Koss and R. S. Klein. Cervical and vaginal squamous cell abnormalities in women infected with human immunodeficiency virus. J Acquir Immune Defic Syndr 1989; 2:570-5.

40) A. Spinillo, P. Tenti, R. Zappatore, G. Barbarini, A. Maccabruni, L. Carratta and S. Guaschino. Prevalence, diagnosis and treatment of lower genital neoplasia in women with human immunodeficiency virus infection. Eur J Obstet Gynecol Reprod Biol 1992; 43:235-41.

41) M. Laga, J. P. Icenogle, R. Marsella, A. T. Manoka, N. Nzila, R. W. Ryder, S. H. Vermund, W. L. Heyward, A. Nelson and W. C. Reeves. Genital papillomavirus infection and cervical dysplasia--opportunistic complications of HIV infection. Int J Cancer 1992; 50:45-8.

42) T. C. Wright, Jr., T. V. Ellerbrock, M. A. Chiasson, N. Van Devanter and X. W. Sun. Cervical intraepithelial neoplasia in women infected with human immunodeficiency virus: prevalence, risk factors, and validity of Papanicolaou smears. New York Cervical Disease Study. Obstet Gynecol 1994; 84:591-7.

43) R. S. Klein, G. Y. Ho, S. H. Vermund, I. Fleming and R. D. Burk. Risk factors for squamous intraepithelial lesions on Pap smear in women at risk for human immunodeficiency virus infection. J Infect Dis 1994; 170:1404-9.

44) X. W. Sun, T. V. Ellerbrock, O. Lungu, M. A. Chiasson, T. J. Bush and T. C. Wright, Jr. Human papillomavirus infection in human immunodeficiency virus- seropositive women. Obstet Gynecol 1995; 85:680-6.

45) G. Y. Ho, R. D. Burk, I. Fleming and R. S. Klein. Risk of genital human papillomavirus infection in women with human immunodeficiency virus-induced immunosuppression. Int J Cancer 1994; 56:788-92.

46) J. M. Palefsky, H. Minkoff, L. A. Kalish, A. Levine, H. S. Sacks, P. Garcia, M. Young, S. Melnick, P. Miotti and R. Burk. Cervicovaginal human papillomavirus infection in human immunodeficiency virus-1 (HIV)-positive and high-risk HIV-negative women [see comments]. J Natl Cancer Inst 1999; 91:226-36.

47) L. Massad, K. Riester, K. Anastos, R. Fruchter, J. Palefsky, R. Burk, D. Burns, R. Greenblatt, L. Muderspach and P. Miotti. Prevalence and predictors of squamous cell abnormalities in papanicolaou smears from women infected with human immunodeficiency virus-1. In Press.

48) H. Minkoff, J. Feldman, J. DeHovitz, S. Landesman and R. Burk. A longitudinal study of human papillomavirus carriage in human immunodeficiency virus-infected and human immunodeficiency virus- uninfected women. Am J Obstet Gynecol 1998; 178:982-6.

49) X. W. Sun, L. Kuhn, T. V. Ellerbrock, M. A. Chiasson, T. J. Bush and T. C. Wright, Jr. Human papillomavirus infection in women infected with the human immunodeficiency virus [see comments]. N Engl J Med 1997; 337:1343-9.

50) S. D. Vernon, W. C. Reeves, K. A. Clancy, M. Laga, M. St. Louis, H. E. Gary, Jr., R. W. Ryder, A. T. Manoka and J. P. Icenogle. A longitudinal study of human papillomavirus DNA detection in human immunodeficiency virus type 1-seropositive and -seronegative women. J Infect Dis 1994; 169:1108-12.

51) C. Six, I. Heard, C. Bergeron, G. Orth, J. D. Poveda, P. Zagury, P. Cesbron, C. Crenn-Hebert, R. Pradinaud, M. Sobesky, C. Marty, M. L. Babut, J. E. Malkin, A. Odier, S. Fridmann, J. P. Aubert, J. B. Brunet and I. de Vincenzi. Comparative prevalence, incidence and short-term prognosis of cervical squamous intraepithelial lesions amongst HIV-positive and HIV-negative women. AIDS 1998; 12:1047-56.

52) I. Heard, V. Schmitz, D. Costagliola, G. Orth and M. D. Kazatchkine. Early regression of cervical lesions in HIV-seropositive women receiving highly active antiretroviral therapy. AIDS 1998; 12:1459-64.

53) I. Heard, C. Bergeron, D. Jeannel, R. Henrion and M. D. Kazatchkine. Papanicolaou smears in human immunodeficiency virus-seropositive women during follow-up. Obstet Gynecol 1995; 86:749-53.

54) K. U. Petry, D. Scheffel, U. Bode, T. Gabrysiak, H. Kochel, E. Kupsch, M. Glaubitz, S. Niesert, H. Kuhnle and I. Schedel. Cellular immunodeficiency enhances the progression of human papillomavirus-associated cervical lesions. Int J Cancer 1994; 57:836-40.

55) A. Adachi, I. Fleming, R. D. Burk, G. Y. Ho and R. S. Klein. Women with human immunodeficiency virus infection and abnormal Papanicolaou smears: a prospective study of colposcopy and clinical outcome. Obstet Gynecol 1993; 81:372-7.

56) P. Belafsky, R. A. Clark, P. Kissinger and J. Torres. Natural history of low-grade squamous intraepithelial lesions in women infected with human immunodeficiency virus [letter]. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 11:511-2.

57) M. Maiman, R. G. Fruchter, E. Serur, P. A. Levine, C. D. Arrastia and A. Sedlis. Recurrent cervical intraepithelial neoplasia in human immunodeficiency virus-seropositive women. Obstet Gynecol 1993; 82:170-4.

58) R. G. Fruchter, M. Maiman, A. Sedlis, L. Bartley, L. Camilien and C. D. Arrastia. Multiple recurrences of cervical intraepithelial neoplasia in women with the human immunodeficiency virus. Obstet Gynecol 1996; 87:338-44.

59) R. Kirnbauer, N. L. Hubbert, C. M. Wheeler, T. M. Becker, D. R. Lowy and J. T. Schiller. A virus-like particle enzyme-linked immunosorbent assay detects serum antibodies in a majority of women infected with human papillomavirus type 16. JNatlCancer Inst 1994; 86:494-499.

60) J. J. Carter, L. A. Koutsky, G. C. Wipf, N. D. Christensen, S. K. Lee, J. Kuypers, N. Kiviat and D. A. Galloway. The natural history of human papillomavirus type 16 capsid antibodies among a chort of university women. J Infect Dis 1996; 174:927-936.

61) Z. Wang, N. Christensen, J. T. Schiller and J. Dillner. A monoclonal antibody against intact human papillomavirus type 16 capsids blocks the serological reactivity of most human sera. J Gen Virol 1997; 78:2209-15.

62) H. D. Strickler, A. Hildesheim, R. P. Viscidi, K. V. Shah, B. Goebel, J. Drummond, D. Waters, Y. Sun, N. L. Hubbert, S. Wacholder, L. A. Brinton, C. L. Han, P. C. Nasca, R. McClimens, K. Turk, V. Devairakkam, S. Leitman, C. Martin and J. T. Schiller. Interlaboratory agreement among results of human papillomavirus type 16 enzyme-linked immunosorbent assays. J Clin Microbiol 1997; 35:1751-1756.

63) T. D. de Gruijl, H. J. Bontkes, J. M. M. Walboomers, J. T. Schiller, M. J. Stukart, B. S. Groot, M. M. R. Chabaud, A. J. Remmink, R. H. M. Verheijen, T. J. M. Helmerhorst, C. J. L. M. Meijer and R. J. Scheper. Immnoglobulin G responses against human papillomavirus type 16 virus-like particles in a prospective nonintervention cohort study of women with cervical intraepithelial neoplasia. J Natl Cancer Inst 1997; 89:630-638.

64) M. Lehtinen, J. Dillner, P. Knekt, T. Luostarinen, A. Aromaa, R. Kirnbauer, P. Koskela, J. Paavonen, R. Peto, J. T. Schiller and M. Hakama. Serologically diagnosed infection with human papillomavirus type 16 and risk for subsequent development of cervical carcinoma: nested case-control study. BMJ 1996; 312:537-539.

65) T. Bjorge, J. Dillner, T. Anttila, A. Engeland, T. Hakulinen, E. Jellum, M. Lehtinen, T. Luostarinen, J. Paavonen, E. Pukkala, M. Sapp, J. Schiller, L. Youngman and S. Thoresen. Prospective seroepidemiological study of role of human papillomavirus in non-cervical anogenital cancers. BMJ 1997; 315:646-9.

66) H. D. Strickler and M. H. Schiffman. Is human papillomavirus an infectious cause of non-cervical anogenital tract cancers? [editorial; comment]. BMJ 1997; 315:620-621.

67) H. D. Strickler, M. H. Schiffman, K. V. Shah, C. S. Rabkin, J. T. Schiller, S. Wacholder, B. Clayman and R. P. Viscidi. A survey of human papillomavirus 16 antibodies in patients with epithelial cancers. Eur J Cancer Prev 1998; 7:305-313.

68) H. D. Strickler, G. D. Kirk, J. P. Figueroa, E. Ward, A. R. Braithwaite, C. Escoffery, J. Drummond, B. Goebel, D. Waters, R. McClimens and A. Manns. HPV 16 antibody prevalence in Jamaica and the United States reflects differences in cervical cancer rates. Int J Cancer 1999; 80:339-344.

69) G. van Doornum, M. Prins, A. Andersson-Ellstrom and J. Dillner. Immunoglobulin A, G, and M responses to L1 and L2 capsids of human papillomavirus types 6, 11, 16, 18, and 33 L1 after newly acquired infection. Sex Transm Infect 1998; 74:354-60.

70) J. Carter, L. Koutsky, S. Lee, J. Kruypers, M. Ravenello, N. Kiviat and D. Galloway. Natural history of human antibody responses following infection with the genital HPV types 6, 11, 16 or 18, 17th International Papillomavirus Conference, Charleston, SC, 1999.

71) J. Dillner. Antibody responses to defined HPV epitopes in cervical neoplasia. Papillomavirus Report 1994; 5:35-41.

72) J. Dillner, F. Wiklund, P. Lenner, C. Eklund, V. Frederiksson-Shanazarian, J. T. Schiller, M. Hibma, G. Hallmans and U. Stendahl. Antibodies against linear and conformational epitopes of human papillomavirus type 16 that independently associate with incident cervical cancer. IntJCancer 1995; 60:377-382.

73) J. Dillner, P. Lenner, M. Lehtinen, C. Eklund, P. Heino, F. Wiklund, G. Hallmans and U. Stendahl. A population-based seroepidemiological study of cervical cancer. Cancer Res 1994; 54:134-141.

74) P. Heino, C. Eklund, V. Fredriksson-Shanazarian, S. Goldman, J. T. Schiller and J. Dillner. Association of serum immunoglobulin G antibodies against human papillomavirus type 16 capsids with anal epidermoid carcinoma. J Natl Cancer Inst 1995; 87:437-440.

75) H. D. Strickler, J. Dillner, M. H. Schiffman, C. Eklund, A. G. Glass, C. Greer, D. R. Scott, M. E. Sherman, R. J. Kurman and M. Manos. A seroepidemiologic study of HPV infection and incident cervical squamous intraepithelial lesions. Viral Immunol 1994; 7:169-177.

76) H. D. Strickler, M. H. Schiffman, C. Eklund, A. G. Glass, D. R. Scott, M. E. Sherman, S. Wacholder, R. J. Kurman, M. M. Manos, J. T. Schiller and J. Dillner. Evidence for at least two distinct groups of humoral immune reactions to papillomavirus antigens in women with squamous intraepithelial lesions. Cancer Epidemiol Biomarkers Prev 1997; 6:183-188.

77) S. E. Barkan, S. L. Melnick, S. Preston-Martin, K. Weber, L. A. Kalish, P. Miotti, M. Young, R. Greenblatt, H. Sacks and J. Feldman. The Women's Interagency HIV Study. WIHS Collaborative Study Group. Epidemiology 1998; 9:117-25.

78) L. Ahdieh, S. Gange and A. Munoz. Selection by indication of potent antiretroviral therapy usage in a large cohort of HIV-infected women. Submitted.

79) H. D. Strickler, C. Rattray, C. Escoffery, A. Manns, M. H. Schiffman, C. Brown, B. Cranston, B. Hanchard, J. M. Palefsky and W. A. Blattner. Human T-cell lymphotropic virus type I and severe cervical neoplasia of the cervix in Jamaica. Int J Cancer 1995; 61:23-26.

80) H. D. Strickler, J. Goedert, R. Burk, R. Viscidi, J. Schiller, G. Pizza, F. Bertoni, A. Jackson, R. Metcalf, W. Qu and K. Shah. A multifaceted study of human papillomavirus and prostate cancer. Cancer 1998; 82:1118-25.

81) H. D. Strickler, A. Hildesheim, R. P. Viscidi, K. V. Shah, B. Goebel, J. Drummond, D. Waters, Y. Sun, N. L. Hubbert, S. Wacholder, L. A. Brinton, C. L. Han, P. C. Nasca, R. McClimens, K. Turk, V. Devairakkam, S. Leitman, C. Martin and J. T. Schiller. Inter-laboratory agreement in HPV 16 enzyme linked immunosorbent assays. J Clin Microbiol 1997; 35:1751-1756.

82) H. D. Strickler, R. Burk, K. Shah, R. Viscidi, A. Jackson, G. Pizza, F. Bertoni, J. T. Schiller, A. Manns, R. Metcalf, W. Qu and J. J. Goedert. A multifaceted study of human papillomavirus and prostate carcinoma. Cancer 1998; 82:1118-1125.

83) H. D. Strickler, C. Rattray, C. Escoffery, A. Manns, M. H. Schiffman, C. Brown, B. Cranston, B. Hanchard, J. M. Palefsky and W. A. Blattner. Human T-cell lymphotropic virus type I and severe neoplasia of the cervix in Jamaica. Int J Cancer 1995; 61:23-26.

84) R. D. Burk, G. Y. F. Ho, L. Beardsley, M. Lempa, M. Peters and R. Bierman. Sexual behavior and partner characteristics are the predominant risk factors for genital human papillomavirus infection in young women. J Infect Dis 1996; 174:679-689.

85) G. Y. F. Ho, R. D. Burk, S. Klein, A. S. Kadish, C. J. Chang, P. Palan, J. Basu, R. Tachezy, R. Lewis and S. Romney. Persistent genital human papillomavirus infection as a risk factor for persistent cervical dysplasia. J Natl Cancer Inst 1995; 87:1365-1371.

86) R. Tachezy, M. A. Van Ranst, Y. Cruz and R. D. Burk. Analysis of short novel human papillomavirus sequences. Biochem Biophys Res Commun 1994; 204:820-827.

87) J.N. Kaye, W.G. Starkey, B. Kell, C. Biswas, K.S. Raju, J. Cason. Human papillomavirus 16 in infants: use of DNA sequence analysis to determine the source of infection. J Gen Virol 1996; 77:1139-43.

88) G. Jiang, W. Qu, H. Ruan and R. D. Burk. Elimination of false-positive signals in enhanced chemiluminescence (ECL) detection of amplified HPV DNA from clinical samples. Biotechniques 1995; 19:566-8.

89) T. Yamada, M. M. Manos, J. Peto, C. E. Greer, N. Munoz, F. X. Bosch and C. M. Wheeler. Human papillomavirus type 16 sequence variation in cervical cancers: a worldwide perspective. J Virol 1997; 71:2463-72.

90) L. Ho, S. Y. Chan, V. Chow, T. Chong, S. K. Tay, L. L. Villa and H. U. Bernard. Sequence variants of human papillomavirus type 16 in clinical samples permit verification and extension of epidemiological studies and construction of a phylogenetic tree. J Clin Microbiol 1991; 29:1765-72.

91) S.-Y. Chan, L. Ho, C.-K. Ong, V. Chow, B. Drescher, M. Durst, J. ter Meulen, L. Villa, J. Luande, H. Mgaya and H.-U. Bernard. Molecular variants of human papillomavirus type 16 from four continents suggest ancient pandemic spread of the virus and its coevolution with humankind. J Virol 1992; 66:2057-2066.

92) A. C. Stewart, A. M. Eriksson, M. M. Manos, N. Munoz, F. X. Bosch, J. Peto and C. M. Wheeler. Intratype variation in 12 human papillomavirus types: a worldwide perspective. J Virol 1996; 70:3127-3136.

93) C. K. Ong, S. Y. Chan, M. S. Campo, K. Fujinaga, P. Mavromara-Nazos, V. Labropoulou, H. Pfister, S. K. Tay, J. ter Meulen, L. L. Villa and H.-U. Bernard. Evolution of human papillomavirus type 18: an ancient phylogenetic root in Africa and intratype diversity reflect coevolution with human ethnic groups. J Virol 1993; 67:6424-6431.

94) R. Kirnbauer, J. Taub, H. Greenstone, R. Roden, M. Durst, L. Gissmann, D. R. Lowy and J. T. Schiller. Efficient self-assembly of human papillomavirus type 16 L1 and L1-L2 into virus-like particles. J Virol 1993; 67:6929-6936.

95) R. B. Roden, N. L. Hubbert, R. Kirnbauer, F. Breitburd, D. R. Lowy and J. T. Schiller. Papillomavirus L1 capsids agglutinate mouse erythrocytes through a proteinaceous receptor. J Virol 1995; 69:5147-5151.

 

There are two subcontracts for laboratory testing. The first subcontract is with Dr. Raphael Viscidi, Johns Hopkins University, to conduct HPV serology. Dr. Viscidi was previously awarded an NIH grant to test all enrollment serum samples in the WIHS. The current proposal will extend these efforts, and maintain the continuity of HPV serologic results in the WIHS. The second subcontract is with Dr. Joel Palefsky, University of California San Francisco, to conduct HPV DNA testing. Half the HPV DNA testing will be conducted by Dr. Palefsky under this subcontract. The other half will be tested by Dr. Burk, who is at the same institution as the Primary Investigator (Albert Einstein College of Medicine). Drs. Palefsky and Burk conducted the HPV DNA testing in specimens collected during enrollment and the first three follow-up visits (Chicago, Los Angeles, and San Francisco sites by J. Palefsky and Bronx/Manhattan, Brooklyn, and Washington D.C. by R. Burk). The same division of HPV DNA testing is important, because of the substantial amount of HPV tests that will be involved in this project, and for maintaining the continuity of HPV DNA data in the WIHS.

Dr. Linda Ahdieh is a PhD epidemiologist/biostatistician working for WDMAC at Johns Hopkins, the data center for the WIHS. She is uniquely familiar with the WIHS data and will be essential in providing ready access to the relevant WIHS data bases and, more importantly, help in interpreting these data sets. Dr. Ahdieh is especially familiar with the analysis of serial HPV data, and will participate in efforts to analyze the laboratory results. However, only travel support for meetings with Dr. Strickler are requested for Dr. Ahdieh.