Epidemiologic Determinants of Seroreactivity to Human Papillomavirus (HPV) Type 16 Virus-Like Particles in Cervical HPV-16 DNA—Positive and —Negative Women

Journal of Infectious Diseases, Nov 1996

The epidemiologic determinants of seroreactivity to human papillomavirus (HPV) type 16 L1/L2 virus-like particles (VLPs) were assessed separately in HPV-16 DNA–positive and –negative women participating in a nested case-control study of incident cervical neoplasia. Seventy-four women with cervical HPV-16 DNA and 656 cytologically normal HPV-16 DNA–negative subjects were interviewed and tested at two time points for viral DNA and once (at the later time) for VLP seroreactivity. Among subjects who were currently HPV-16 DNA–negative, seroreactivity odds ratios increased from 2.9 for 2–5 male sex partners (vs. 0 or 1) to 5.4 for 6–9 partners and 14.0 for ⩾10. Thus, prior cervical infection may be a major determinant of seroreactivity in HPV-16 DNA–negative women. This trend was not observed in HPV-16 DNA-positive subjects. Seroreactivity was independently associated with oral contraceptive use, particularly in HPV-16 DNA–negative subjects with use for ⩾10 years. Consequently, a possible role for virus–steroid hormone interactions in seroconversion is suggested.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://jid.oxfordjournals.org/content/174/5/937.full.pdf

Epidemiologic Determinants of Seroreactivity to Human Papillomavirus (HPV) Type 16 Virus-Like Particles in Cervical HPV-16 DNA—Positive and —Negative Women

L. Wideroff 0 1 2 M. H. Schiffman 0 1 2 R. Hoover 0 1 2 R. E. Tarone 0 1 2 B. Nonnenmacher 0 1 2 3 N. Hubbert 0 1 2 R. Kirnbauer 0 1 2 3 C. E. Greer 0 1 2 " A. T. Lorincz 0 1 2 M. M. Manos 0 1 2 3 A. G. Glass 0 1 2 D. R. Scott 0 1 2 M. E. Sherman 0 1 2 J. Buckland 0 1 2 D. Lowy 0 1 2 J. Schiller 0 1 2 0 The Journal of Infectious Diseases 1996; 174:937 -43 1996 by The University of Chicago. All rights reserved. 0022-1899/96/7405-0006$01.00 1 Received 20 February 1996; revised 12 June 1996. Presented in part: 14th International Papillomavirus meeting , 23-28 July 1995, Quebec , Canada. Informed consent was obtained from participants, and the study was con ducted in accordance with human subjects research guidelines of the National Institutes of Health. The HPV-16 serologic assay has been patented by J. Schiller, R. Kirnbauer, and D. Lowy. A. Lorincz is scientific director of Digene Diagnostics. Executive Blvd. , Bethesda, MD 20892-7374 2 Epidemiology and Biostatistics Program and Laboratory of Cellular Oncology, National Cancer Institute , Bethesda , and Digene Diagnostics and Information Management Services, Silver Spring, Maryland; Cetus Corporation, Emeryville, California; Kaiser Permanente Center for Health Research , Portland, Oregon; Department of Pathology, George Washington University , Washington, DC 3 Current affiliations: Servicio de Oncologia, Hospital de Clinicas de Porto Alegre , Porto Alegre , Brazil ( B.N.); Division of Immunology , Allergy, and Infectious Diseases , Department of Dermatology, University of Vienna Medi cal School , Vienna , Austria ( R.K.); Chiron Corporation , Emeryville, California (C.E.G. ); Division of Research, Kaiser Foundation Research Institute , Broad way, Oakland, California (M.M.M.) The epidemiologic determinants of seroreactivity to human papillomavirus (HPV) type 16 LlIL2 virus-like particles (VLPs) were assessed separately in HPV-16 DNA-positive and -negative women participating in a nested case-control study of incident cervical neoplasia. Seventy-four women with cervical HPV-16 DNA and 656 cytologically normal HPV-16 DNA-negative subjects were interviewed and tested at two time points for viral DNA and once (at the later time) for VLP seroreactivity. Among subjects who were currently HPV-16 DNA-negative, seroreactivity odds ratios increased from 2.9 for 2-5 male sex partners (vs. 0 or 1) to 5,4 for 6-9 partners and 14.0 for ~10. Thus, prior cervical infection may be a major determinant ofseroreactivity in HPV-16 DNAnegative women. This trend was not observed in HPV-16 DNA-positive subjects. Seroreactivity was independently associated with oral contraceptive use, particularly in HPV-16 DNA-negative subjects with use for s-10 years. Consequently, a possible role for virus - steroid hormone interactions in seroconversion is suggested. - Oncogenic human papillomavirus (HPV) infection has been clearly established as the main epidemiologic risk factor for cervical neoplasia and as a major risk factor for several other anogenital cancers as well [1, 2]. In epidemiologic studies of HPV-associated disease, the standard for measurement of HPV infection is the detection of viral DNA in epithelial cells, using polymerase chain reaction (PCR) [3], hybridization methods [4], or both. The recent development of serologic tests for antibodies to synthetic or recombinant HPV antigens provides additional biomarkers for studying patterns of viral infection, immune response, and disease in human populations [5-7]. In one approach to serologic testing, an ELISA has been used to detect antibody to empty HPV-16 capsids, or viruslike particles (VLPs), that self-assemble in insect cells after expression of the L1 and L2 structural genes via recombinant baculoviruses [8]. This assay detects virion antibodies in about one-half to two-thirds of women with HPV-16-associated squamous intraepithelial lesions of the cervix and up to 15% of cytologically normal women without detectable HPV-16 DNA [9, 10]. Moreover, in women who were followed prospec tively and tested by PCR at two points in time for the presence of HPV-16 DNA in the cervix, antibody was detected in 22% of those who were DNA-positive only once but in 83% of those who were repeatedly DNA-positive. A significantly higher seroprevalence in women with current HPV-associated cervical lesions, or with repeatedly detectable HPV-16 DNA, would suggest that virus load and persistence of viral DNA in the cervical epithelium are the primary factors for the development of an anti-virion humoral immune response [5]. However, secondary influences on the development of se roreactivity by epidemiologic factors, such as sexual behavior, oral contraceptive (OC) use, pregnancy, and lifestyle, have not been critically examined. The objective of this study was, therefore, to determine which epidemiologic factors were asso ciated with an antibody response among 656 cytologically nor mal HPV-16 DNA-negative control women presumed to be representative of a healthy population and also, separately, among 74 women with or without squamous intraepithelial lesions (SIL) who tested positive for HPV-16 DNA. Methods The subjects were participants in a case-control study of incident cervical neoplasia, nested within a prospective cohort of 17,654 women who underwent routine Pap screening at Portland Kaiser Permanente Center for Health Research in 1989-1990. The study population, design, and protocols for DNA and serologic testing and cytopathology review are described elsewhere in detail [IOJ. Briefly, incident cases with a confirmed diagnosis of high-grade SIL, low-grade SIL, or atypical squamous cells of undetermined significance (ASCUS) were each matched on age and other factors to ~ 3 cohort members who were consistently cytologically normal up to the time of case diagnosis. Cervicovaginal lavage samples, collected at two points in time (at the cohort study enrollment and the diagnostic follow-up visit), were tested for a broad spectrum of genital HPV DNA types, by either a PeR-based method [11,12] or hybrid capture (HC) [13]. Serum drawn at the follow-up visit was tested by the HPV-16 ELISA, and a questionnaire was administered that covered sexual, reproductive, and lifestyle factors of etiologic relevance to cervical neoplasia. In 1994, a subset of subjects enrolled in the main inci dent case-control study who had a confirmed pathology diagnosis, an available serum sample for antibody testing, and DNA measure ments from both enrollment and follow-up was selected for the HPV serology study. This subset consisted of 840 women (60% of all final incident cases and controls), including 688 cytologically normal controls and 152 incident cases, who had been followed for an average of 727 days between enrollment and detection of the cases' cervical cytologic abnormalities. For the present study, subjects were classified as HPV -16 DNA positive if viral DNA was detected in at least 1 of the 2 lavage samples by PCR or HC and as HPV-16-seropositive if the ELISA optical density reading was above a previously determined cut point of 0.75. The main statistical analyses focused on subgroups assumed to represent HPV-16-uninfected and -infected popula tions. As the likely uninfected subgroup, the 656 of 688 cytologi cally normal controls who were HPV-16 DNA - negative at enroll ment and follow-up (but possibly positive for other HPV DNA types) were selected. As the likely infected subgroup, the 74 HPV 16 DNA -positive subjects were selected independent of their cyto pathologic status, since previous work showed that DNA detection rather than cytopathologic status was the primary determinant of seroreactivity. Thus, the HPV-16 DNA-positive group included cytologically normal controls and cases with incident high- or low grade SIL or ASCUS lesions. Conditional maximum likelihood estimators of the prevalence odds ratios (ORs) and exact 95% confidence intervals (CIs) were calculated for the HPV-16 DNA-positive and -negative sub groups to examine associations of seroreactivity with the hypothe sized epidemiologic determinants [14]. The ORs were assumed to approximate the ratio of percentage of subjects who were seroposi tive in subjects with and without a hypothesized determinant under conditions of low seroprevalence; these ORs would differ some what under conditions of high seroprevalence. The sexual behavior covariates considered in the analysis were total lifetime number of male sex partners, number of partners since enrollment in the study, age at first sexual intercourse, and a self-report of sexually transmitted disease during study follow up (i.e., trichornonas. Chlamydia infection, scabies, genital herpes, syphilis, gonorrhea. venereal warts. other). The reproductive vari ables that were examined included OC use (ever, and years of use), number of pregnancies, and number of vaginal births. Lifestyle and demographic factors included smoking status, number of alcoholic drinks per week, age at interview, race, years of education, and total family income. Seroprevalence in the 656 HPV-16 DNA-negative, cytolog ically normal controls increased markedly with increasing life time number of male sex partners, from 3.7% in subjects with o or 1 partner to 11.1% in those with 2 - 5, 19.8% in those with 6-9, and 38.5% in those with ~ 10 partners. With 0 or 1 lifetime partner as the referent group, ORs of seroreactivity were 2.9 for subjects with 2-5 partners, 5.4 for subjects with 6-9, and 14.0 for those with ~ 10 (table 1), after adjustment for an additional determinant, ever having used OCs on a regu lar basis. A notably weaker trend was observed for an increas ing number of recent male sex partners (e.g., since enrollment in study), which was not statistically significant after adjust ment for lifetime number of partners (table 1). In contrast, adjustment for recent partners did not alter the effect of lifetime partners on seroreactivity. Age < 18 years at first sexual intercourse and occurrence of a sexually transmitted disease other than HPV infection were not associated with HPV-16 seroreactivity after adjustment for lifetime number of male sex partners and ever OC use. No associations were observed for age at first intercourse even when the < 18 years category was further subdivided into ~ 16 or ~ 15 years (data not shown). A different pattern emerged in the 74 HPV-16 DNA-posi tive subjects, a group that included 32 cytologically normal controls and 42 cases who were cytologically normal at enroll ment but developed an abnormality detected at follow-up (13 with ASCUS, 21 with low-grade SIL, and 8 with high-grade SIL). When subjects with 2-5 lifetime sex partners were used as the referent group (no HPV-16 DNA -positive subjects had 0 or 1 sex partner), no significant relationship was noted between seropositivity and lifetime number of sex partners (table 1). As with the DNA-negative subjects, significant associations with recent number of sex partners and other sexual factors were not observed in the adjusted models. HPV testing method (PCR vs. HC) could theoretically affect the associations observed, since HC is designed to detect only the higher levels of HPV DNA predictive of SIL, and subjects classified as HPV-16 DNA-negative could be PCR-positive. In this situation, a high number of sex partners could be a proxy for HPV-16 DNA positivity among HC-negative sub jects. Thus, the relationship of sex partners and seroreactivity was further explored in the subset of subjects who were tested with the peR method. The PCR-tested subset included 290 controls who were HPV-16 DNA-negative by PCR throughout the study, as well as 18 subjects who were HPV-16 DNA -positive once at enroll ment or follow-up and another 18 who were repeatedly HPV OR (95% (CI) HPV-16 DNA-negative (n = 656) HPV-16 DNA-positive (n = 74) NOTE. CI, confidence interval; STD, sexually transmitted disease. * Adjusted for ever use of oral contraceptives. t Adjusted for lifetime no. of sex partners and ever use of oral contraceptives. : Insufficient data to derive valid risk estimates. 16 DNA-positive at enrollment and follow-up (figure 1). Most important, a strong trend of rising seroprevalence with increas ing number of sex partners was observed in the PCR DNA negative controls. As with the combined PCR and HC data, no trend was observed in the once- or twice- DNA-positive subgroups. Among those DNA-positive, seroprevalence was more than twice as high in the repeatedly DNA-positive sub group relative to the once- DNA-positive group, for each level of sex partner numbers. Ever use ofOCs was positively associated with seroreactivity after adjustment for lifetime number of male sex partners in the HPV DNA-negative controls. An adjusted seroreactivity OR of 3.1 was observed among subjects who reported ever regularly using OCs for at least 1 year relative to those who had not, which rose to 5.8 for OC use of ~ 10 years (table 2). Further adjustment for barrier contraceptive use of at least 1 year's duration did not nullify the association of seroreactivity with ever OC use (OR, 2.8; 95% CI, 1.2-7.7). In the HPV-16 DNA-positive subgroup, the OR was 2.3 for ever use, although the 95% CI included 1.0, perhaps due to small numbers, and there was no statistically significant trend for increasing years of use. After adjustment for sex partners and OC use, age was not significantly associated with seroreactivity, although ORs ap peared to decrease somewhat with increasing age in HPV-16 DNA-positive subjects (table 2). In the HPV-16 DNA-nega tive subjects, 9.1% <20 years old were seropositive, after which seroprevalence increased to 14.9% in those 20-29 years old and remained relatively steady in those 30-39 (15.1%) and ~40 (13.8%) (test for trend, P = .2). Given the absence of a strong age effect and sample size limitations in the subgroup analyses, age was removed from the other models of epidemio logic determinants, resulting in no appreciable change in the risk estimates of seroreactivity. Additional reproductive, socioeconomic, and lifestyle vari ables were not associated with seroreactivity in either the HPV 16 DNA-negative or -positive subgroup after adjustment for the main epidemiologic determinants of lifetime sex partners and ever OC use (data not shown). Also, seroprevalence did not vary by the time interval from last Pap smear to the diagnos tic date, a potential confounding variable in all studies of risk factors for screening-detected cervical neoplasia. Furthermore, as shown above for lifetime sex partners, the associations of other epidemiologic determinants with seroreactivity did not % Seropositive 2-5 6-9 Lifetime Number of Male Sex Partners Figure 1. Seroprevalence by polymerase chain reaction HPV-16 DNA status within sex partner levels. Seropositive frequencies (%) for 0 or 1,2-5,6-9, and ~ 10 sex partner categories were as follows: HPV-16 DNA-negative (-1-; n = 290, including 5 with unknown partner number): 2/87 (2.3%), 11/105 (10.5%), 10/41 (24.4%),22/52 (42.3%); HPV-16 DNA-positive once (-1+ or +1-; n = 18, includ ing 1 unknown): 0/0, 3/10 (30.0%), 1/4 (25.0%), 0/3 (0%); HPV-16 DNA-positive twice (+1+; n = 18, including 1 unknown): 0/0,5/6 (83.3%), 1/1 (100.0%), 8/10 (80.0%). notably vary by HPV DNA testing method, since results similar to those described above were obtained when PCR- and HC tested subgroups were analyzed separately. Given the sharp trend of increasing seroreactivity ORs with greater numbers of lifetime sex partners in the HPV-16 DNA negative controls, a concern arose as to whether these associa tions were HPV -16-specific or perhaps partly attributable to antigenic cross-reactivity from other genital HPV types. To assess this possibility, the HPV-16 DNA-negative controls were further stratified into subjects with at least one other type of HPV DNA (n = 104) and those without any detectable HPV DNA (n = 552), and adjusted ORs were obtained for comparison purposes (table 3). The 6-9 and ~ 10 lifetime sex partner categories were combined to permit calculation of valid adjusted risk estimates in the subgroup positive for other HPV DNA. In the latter subgroup, seroreactivity was detected in 12.5% of those with 0 or 1 lifetime sex partner, 15.4% of those with 2-5, and 46.4% with ~6. Corresponding percentages in the HPV DNA-negative subgroup were 3.3%, 10.3%, and 22.0%. In both subgroups, similar significant increases in sero reactivity ORs were observed with increasing numbers of sex partners. This study assessed the epidemiologic determinants of sero reactivity to HPV-16 VLPs, a biomarker of the humoral im mune response to HPV-16 infection, in 74 women with detectable cervical HPV-16 DNA and 656 cytologically normal HPV 16 DNA-negative controls. Lifetime number of male sex part ners emerged as the major determinant of HPV-16 seroreactiv ity in the HPV-16 DNA-negative controls, with a sharp trend observed for increasing number of partners and the strongest association found among subjects reporting ~ 10 partners (table 1). This pattern is consistent with associations observed by other investigators of sex partners and seroreactivity to HPV 16 and other types (Dillner J, personal communication). In earlier epidemiologic studies done before the establishment of reliable HPV DNA detection methods, multiple sex partners was identified as a major risk factor for cervical cancer, serving as a surrogate for oncogenic HPV infection [15]. The strong association of multiple sex partners with seroreactivity, ob served in this study among women without current cytopatho logic evidence of cervical HPV infection or HPV-16 DNA, implies that seroreactivity may be an indicator of prior as well as current infection. A significant association of seroreactivity with multiple sex partners was not observed in HPV-16 DNA-positive women. Even among subjects tested by PCR methods, the sharp trend of increasing seroprevalence with increasing number of male sex partners, which was clearly evident in the HPV-16 DNA negative controls, was absent in those who were HPV-16 DNA-positive once or twice during follow-up (figure 1). This lack of an association probably reflects the diminished value of multiple sex partners as an HPV surrogate in a subgroup in which everyone shows molecular evidence of viral infection, although the study had statistical power too low to detect a weak association because of small numbers of HPV-16 DNA positive subjects. Interestingly, the number of recent sex partners (i.e., since enrollment in the prospective cohort component of this study) was not associated with seroreactivity after adjustment for life time number of sex partners. One explanation for this is that exposure to a high number of sex partners, with subsequent seroconversion, occurred at an early age, although early age at first sexual intercourse was not a significant determinant of seroreactivity. Alternatively, the lack of an association may reflect the relatively short follow-up period, which averaged 727 days, during which time the number of sex partners and age (a correlate of number of sex partners) did not vary substan tially. In contrast, the greater variability in the number of life time sex partners, which includes both pre- and postenrollment partners, may provide increased statistical power to detect a strong association with seroreactivity. Within each level of lifetime sex partner numbers in the PCR-tested subgroup, seropreva1ence was consistently more than twice as high in those repeatedly HPV-16 DNA-positive relative to those who were DNA-positive just once. This sup ports the hypothesis that prolonged exposure to viral antigens may be a primary determinant of a detectable VLP antibody response in women infected with cervical HPV-16 DNA [10]. HPV-16 DNA-negative (n = 656) HPV-16 DNA-positive (n = 74) Seropositive (no.) Seronegative (no.) Seropositive (no.) Seronegative (no.) 5 17 6 6 2 3 5 27 1 o NOTE. CI, confidence interval. * Adjusted for lifetime no. of sex partners. t Adjusted for lifetime no. of sex partners and ever use of oral contraceptives. ~ Insufficient data to derive valid risk estimates. The average duration of exposure required for seroconversion cannot be determined in this study, in which time of first HPV infection is unknown. Instead, prospective studies of subjects who are uninfected at enrollment and tested repeatedly for HPV DNA and VLP antibodies during the follow-up period are needed to further clarify the time relationship of DNA and serologic markers [16, 17]. Whereas lifetime number of sex partners appears to explain increased seroprevalence in HPV-16 DNA-negative subjects through its role as a surrogate of past virus exposure, a different mechanism may account for the elevated seroprevalence in ever users of OCs. In HPV-16 DNA-negative subjects, ORs approximated 3 in subjects using OCs for 1-9 years and 6 in those using OCs for ~ 10 years, even after adjustment for lifetime number of male sex partners. This association was not explained by an absence of protective barrier contraceptive methods, such as condoms and diaphragms, in OC users. In addition, the association with OC use was also observed within the HPV-16 DNA-positive subgroup, although confidence in tervals included 1.0 and there was no evidence of a dose re sponse. While the crude OR among HPV-16 DNA-positive subjects with ~ 10 years of OC use was 6.0, which is similar to that among HPV-16 DNA-negative subjects, the OR after stratification by number of lifetime sex partners was 1.3. Be HPV DNA-negative (n = 552) Other HPV DNA-positive (n = 104) Seropositive (no.) Seronegative (no.) Seropositive (no.) Seronegative (no.) OR (95% CI) NOTE. CI, confidence interval. * Adjusted for ever use of oral contraceptives. cause of the small numbers in this subgroup, the adjusted OR was unstable, as evidenced by the width of the 95% Cl, and should be interpreted with caution. Thus, ever use of OCs does not appear to function as another surrogate for virus exposure, which implies that the relationship of seroreactivity with OCs may involve a biologic link between steroid hormones and antibody response. OC use, particularly of long-term duration, has been associ ated with cervical carcinoma after adjustment for sexual factors in many previous epidemiologic studies [18, 19], and increased detectability of HPV DNA has been suggested in OC users [20]. Several laboratory studies have reported the presence of multiple steroid hormone response elements in the viral genome [21 - 24], as well as estrogen-induced increases in HPV-16 transforming proteins [25], which suggests a possible influence of these hormones on viral gene expression. While steroid hormone-viral gene transcription interactions may not occur for all genes in the HPV genome [26], the results of the present study do suggest a link with LI/L2 VLP seroreactivity, either through an OC-related increase in viral capsid gene expression, which in tum leads to increased antibody production, or through another unknown mechanism. Although some nonsignificant variability in seroreactivity by age was apparent, age was not an independent determinant of HPV -16 seroreactivity after adjustment for number of sex partners and OC use, in either HPV -16 DNA -positive or negative subjects. The slightly lower seroprevalence in HPV 16 DNA-negative controls <20 years old suggests less prior exposure to this HPV type relative to older controls. Additional associations with VLP seroreactivity were postu lated for other cervical neoplasia risk factors that are correlates of HPV infection, such as low socioeconomic status, ethnicity, and other sexually transmitted disease; pregnancy, during which there may be increased detection of HPV infection [27]; and other factors that possibly have an impact on the host immune response to infection by promoting immune system contact with capsid antigens (e.g., cervical trauma from vaginal childbirth or ulceration caused by other infectious agents) or by suppressing host immune response to infection (e.g., cigarette smoking or alcohol intake) [28]. However, none of these factors showed an association with seroreactivity within either DNA subgroup after adjustment for lifetime sex partners and OC use. Early age at first sexual intercourse, which is a risk factor for invasive cervical disease hypothesized to represent either latency or an age window of increased susceptibility of the cervical epithelium to neoplasia [15], was also not significantly associated with VLP seroreactivity in the multivariate models using various age cut points. ELISA cross-reactivity of the HPV-16 test with capsid anti gens of other genital HPV types could potentially confound the observed associations between HPV -16 seroreactivity and various epidemiologic determinants. In fact, seroprevalence was consistently higher in those who were positive for other HPV DNA than among those HPV DNA -negative within each level of number of sex partners, which initially raised the possi bility of cross-reactivity. To further examine this, the HPV-16 DNA-negative controls were categorized into subgroups that were DNA-positive or -negative for other HPV types, and ad justed ORs of HPV-16 seroreactivity and lifetime number of sex partners were calculated. Under conditions of cross-reactiv ity, it was assumed that in the other HPV DNA-positive sub group, as in the HPV-16 DNA-positive subgroup, lifetime number of sex partners would lose its surrogate value for HPV DNA positivity, and therefore, associations would be observed mainly in the HPV DNA -negative controls. Adjusted seroreac tivity ORs for lifetime sex partners, however, were not mark edly different among other HPV DNA-positive controls, com pared with those for the HPV DNA-negative controls, which suggests that the assay is primarily type-specific. Although antibody cross-reactivity may have occurred, the higher HPV 16 seroprevalence in the other DNA-positive subgroup could be due to more frequent prior HPV-16 infection (e.g., DNA positivity to other HPV types may be a proxy for prior HPV 16 infection). In summary, in cytologically normal HPV-16 DNA-nega tive women, the number of sex partners was the primary deter minant of HPV-16 seroreactivity, suggesting that the latter is a marker of prior HPV infection. In HPV-16 DNA-positive women, repeated detection of viral DNA, which may be an indicator of relatively long-term duration of infection, was the primary determinant of HPV-16 seroreactivity. In addition, ever use of OCs, particularly for ~ 10 years, was independently associated with seroreactivity in HPV-16 DNA-negative women. A larger sample size than what was available in the present study would be required to best evaluate this association in HPV-16 DNA-positive women. We thank the field staff (Brenda Rush, Patty Lawler, Leilani Wilson, and Chris Eddy) of the Kaiser Permanente Center for Health Research in Portland, Oregon, for their invaluable assis tance.


This is a preview of a remote PDF: https://jid.oxfordjournals.org/content/174/5/937.full.pdf

L. Wideroff, M. H. Schiffman, R. Hoover, R. E. Tarone, B. Nonnenmacher, N. Hubbert, R. Kirnbauer, C. E. Greer, A. T. Lorinez, M. M. Manos, A. G. Glass, D. R. Scott, M. E. Sherman, J. Buckland, D. Lowy, J. Schiller. Epidemiologic Determinants of Seroreactivity to Human Papillomavirus (HPV) Type 16 Virus-Like Particles in Cervical HPV-16 DNA—Positive and —Negative Women, Journal of Infectious Diseases, 1996, 937-943, DOI: 10.1093/infdis/174.5.937