Because persistent infection with high-risk human papillomavirus (HPV) types generally is accepted as the etiology of cervical carcinoma,1–6 recent research has focused on the role of HPV DNA testing in cervical cancer prevention programs. Studies in women suggest that testing for high-risk HPV types is a sensitive screen for severe cervical dysplasia and that when used as an adjunct to cervical cytology screening, HPV testing may spare women with abnormal cytology the inconvenience, anxiety, and cost of repeated cytological testing and/or colposcopy.5,7–12 Recently published consensus guidelines for the management of abnormal cytology and cervical intraepithelial neoplasia (CIN) state that HPV testing is an option in the following settings: as an adjunct to cervical cytology screening for women with atypical squamous cells of undetermined significance (ASC-US); in the follow-up of women with atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesions or low-grade squamous intraepithelial lesions (LSIL) after a negative colposcopy; in the follow-up of women with biopsy-confirmed CIN 1; and in the surveillance of women with biopsy-confirmed CIN 2 and 3 after treatment.13,14
A potential benefit of HPV testing in cervical cancer screening and follow-up protocols is that a speculum examination may not be needed if clinician- or patient-collected swabs for HPV DNA can be used instead of a cervical cytology specimen. Young women who find the speculum examination uncomfortable or embarrassing may be more likely to adhere to recommended screening for cervical cancer if it involves self-collected vaginal specimens rather than clinician-collected cervical specimens; in addition, HPV self testing may allow for testing in nonclinical settings.
However, it is unknown whether adolescent-collected HPV DNA test results are accurate. The results of studies in adult populations pertaining to the accuracy of self testing for detection of HPV DNA differ depending on the population studied, the analytic methods used, and the outcomes measured.15–19 In general, studies in adults have demonstrated that self testing detects rates of HPV DNA similar to those of clinician testing and that concordance between the results is fairly high. However, it is unknown whether adolescent and young adult women can self-collect specimens accurately and whether accuracy of collection is influenced by factors such as HPV type or age. Therefore, the objectives of this study were to estimate the concordance between adolescent and young adult self-collected and clinician-collected HPV DNA test results and to explore the effects of HPV type and adolescent age on concordance.
MATERIALS AND METHODS
The study population consisted of sexually active female patients aged 14 to 21 years seen at an urban, hospital-based teen health center between July 2002 and January 2003. After approval of the study protocol by the hospital’s institutional review board, informed consent was obtained from parents/guardians for patients who were aged less than 18 years and directly from patients aged 18 years and older. Exclusion criteria included a history of cervical dysplasia documented by colposcopically directed biopsy, current pregnancy, or a mental or physical health issue that would prevent the participant from being able to assent to or participate in the study. Of the 179 adolescents and young adults who expressed interest in the study and were eligible to participate, 9 patients declined participation after consenting, 11 could not be reached by telephone to schedule study visits or had moved, 36 scheduled but did not keep study visits, and 2 had positive pregnancy tests before the first scheduled visit. Thus, 121 (68%) completed both study visits. The first 20 participants were excluded from the analyses to allow pilot testing of the PGMY09/11 consensus primer system used for HPV testing. The final study sample, therefore, consisted of 101 participants.
At the baseline visit, the research coordinator first reviewed the study protocol with participants and described the pelvic examination and the method for self collection. Participants completed a brief survey instrument that was designed for this study. It consisted of 5 items assessing sociodemographic characteristics, 7 items assessing gynecologic and medical history, and an 11-item HPV knowledge scale developed by the investigators and used in a previous study.20 Participants then received a standardized educational protocol, which was given by the research coordinator and lasted approximately 20 minutes. The protocol included detailed information regarding the transmission of HPV infection, HPV types, and sequelae of infection.
After providing education, the research coordinator again reviewed methods for self testing. Participants were instructed to insert a swab (BBL Culture Swab; Becton Dickinson and Company, Sparks, MD) into the vagina as far as possible, then gently to move the swab up and down for 5 seconds and side to side for 5 seconds. One of the 2 clinicians (L.M.K. or J.A.K.) then inserted a speculum and obtained a specimen for HPV DNA by rotating an identical Dacron swab in the endocervix for 5 seconds and then swabbing the ectocervix and vaginal fornices surrounding the ectocervix for 5 seconds. The clinician then used a plastic spatula and cytobrush to obtain a specimen for cervical cytology using standard procedures recommended by the manufacturer (Cytyc Corporation, Boxborough, MA). Participants were scheduled for a return visit to obtain test results 2 weeks after testing. The research coordinator gave subjects their test results and explained the significance of any normal or abnormal results and recommended follow-up procedures using a results protocol that was specific to the participant’s test results. Participants with abnormal cervical cytology followed the usual clinic protocol for follow-up. Because there was no standard protocol in the clinic for positive HPV DNA results at that time, subjects with positive results were encouraged to continue cervical cytology screening or follow-up visits as recommended.
Self-obtained and clinician-obtained specimens were assessed for HPV by polymerase chain reaction (PCR) amplification with the use of the PGMY09 and PGMY11 consensus primer system and a reverse-line blot detection strip that individually identifies HPV types (Roche Molecular Systems, Alameda, CA).21,22 The HPV types detected include: 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 57, 58, 59, 66, 68, 73, 82, 83, and 84. Specimens were collected with a BBL Culture Swab. The DNA isolation procedure was based on magnetic bead technology. The samples were lysed by incubation with a special buffer, Glass Magnetic Particles (Roche Molecular Systems) were added, and total nucleic acids contained in the sample were bound to their surface. Each sample was amplified with 5′ biotinylated PGMY09 and PGMY11 (100 pM of each primer pool), as well as 2.5 pM each of the β-globin primers GH20 and PC04, in the presence of 1X PCR Buffer II; 4 mM MgCl2; 200 μmol each of deoxyadenosine triphosphate, deoxycytidine triphosphate, and deoxyguanosine triphosphate; 600 μmol of deoxyuridine triphosphate; and 7.5U of AmpliTaq Gold (Roche Molecular Systems, Alameda, CA) DNA polymerase. Amplifications were performed in a Perkin-Elmer 2400 (Perkin Elmer Life and Analytical Sciences Inc, Boston, MA) thermal cycler using AmpliTaq Gold activation. The protocol used for detection of products from both assays was performed as described previously,21 in which the PCR products were denatured in 0.4 N NaOH and were hybridized to an immobilized HPV probe array, with positive hybridization detected by streptavidin-horseradish peroxidase-mediated color precipitation at the probe site.
Cervical cytology was performed with a liquid-based cytology system (ThinPrep, Cytyc Corporation). The sample was collected by using a spatula and cytobrush device, which was then placed in a ThinPrep vial containing transport medium. The sample vial was capped, labeled, and sent to the laboratory. A monolayer slide was prepared per manufacturer’s directions. Experienced cytologists read the slides and pathologists reviewed any abnormal findings. Cytologists and pathologists were not aware of HPV testing results. The Bethesda II System was used for classification of cytology results.
Data were analyzed by using SAS version 8.2. High-risk HPV types were defined as types 16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, and 82, and low-risk types were defined as types 6, 11, 40, 42, 54, 55, 57, 83, and 84.23 Age was dichotomized into 14–17 years and 18–21 years for analysis. Race was categorized as white, black, and all other, and health insurance was categorized as Medicaid, private, and all other based on the distribution of responses. Number of sexual partners was analyzed as a continuous variable and also categorized into 5 categories because of the skewed nature of the results. Age was dichotomized into 14–17 years and 18–21 years because of the distribution of responses (17.2 was the mean age of participants), and because this dichotomy corresponds to the difference between middle and late adolescence, stages characterized by distinct cognitive and developmental skills that could relate to ability to collect the specimen correctly.24 Descriptive statistical methods were used to assess study sample characteristics, including sociodemographic characteristics, risk behaviors, and gynecologic history.
The accuracy of self versus clinician testing was assessed as follows. First, the results of self-obtained and clinician-obtained specimens were categorized and compared. A κ statistic was calculated to measure agreement (corrected for chance) between self and clinician test results, with results categorized as negative for HPV types, low-risk HPV types only, and 1 or more high-risk HPV type. A McNemar’s test was performed to assess whether the results of self and clinician testing were significantly different when results were categorized as negative/low-risk HPV versus high-risk HPV. Second, we examined level of agreement between self and clinician testing considering all HPV types. We assessed whether the results of HPV self testing and clinician testing for individual participants agreed completely, agreed partially, or did not agree at all. Chi-squared testing was used to determine whether the level of agreement was associated with HPV type (low-risk HPV type[s] only versus 1 or more high-risk HPV type) or age.
To examine whether test characteristics differed by collection technique, we constructed 2 × 2 tables to assess the sensitivity, specificity, and negative and positive predictive values of high-risk HPV DNA in predicting abnormal cervical cytology. We compared these test characteristics for self-testing versus clinician testing, examining 2 thresholds for abnormal results: ASC-US and LSIL.
Study sample characteristics are shown in Table 1. The mean age of study participants was 17 years, mean age of first sexual intercourse 14 years, number of lifetime sexual partners 5.9, and age of the most recent sexual partner 20 years. Those aged 18–21 years, compared with those aged 14–17 years, had initiated sexual intercourse later, had a higher number of lifetime sexual partners, had older sexual partners, and were more likely to have had a history of abnormal cytology (P < .05, data not shown).
Twenty-five different HPV types were identified by self or clinician testing: 17 high-risk and 8 low-risk types. Data were unavailable for 2 self-collected samples and 1 clinician-collected sample because the samples did not amplify. Overall, 50.5% of the participants’ results were positive for at least 1 HPV type identified by either self or clinician testing. Of those who were HPV positive, 53% had 1 HPV type, 25% had 2 types, and 22% had 3 or more types identified. Specific HPV types identified by self-testing and clinician-testing are shown in Table 2. Testing performed by a clinician was more likely than self-testing to identify low-risk types (29.6% versus 26.4%), and self-testing was more likely than testing performed by a clinician to identify high-risk types (73.6% versus 70.4%). Self-testing was slightly more likely than testing performed by a clinician to identify only 1 HPV type (Table 3), but the difference was not statistically significant (27.3% versus 23.0%, P = .65).
To examine agreement between HPV results obtained by self versus clinician testing, we first categorized participants into 3 groups: negative for HPV types, positive for low-risk HPV types only, and positive for at least 1 high-risk HPV type. The data in Table 4 demonstrate that results were concordant in 84 (85%) of 99 participants. Testing performed by a clinician identified 5 high-risk HPV types that were missed by self testing, and self testing identified 8 high-risk types that were missed by tests performed by a clinician. Compared with testing performed by a clinician, self testing detected any HPV type in a higher percentage of participants (45% versus 42%) and detected high-risk HPV types in a higher percentage of participants (38% versus 35%) but detected low-risk HPV in the same number of participants (7% versus 7%; P value for these comparisons = .65). A κ statistic, which measures agreement between the results of self testing and clinician testing, was 0.72; this indicates moderately high agreement.25 Agreement was also estimated after categorizing results into HPV negative/low-risk types versus 1 or more high-risk type. A McNemar’s test, performed to examine whether the results of self and clinician testing differed when results were classified in this way, was not statistically significant (P = .41). Finally, we examined level of agreement between self and clinician testing considering all HPV types identified (Table 5). There was complete agreement between self and clinician testing (that is, all types were identical) in 62% of participants who had only low-risk HPV compared with 51% who had at least 1 high-risk HPV type (P = .59). Differences in level of agreement by age group similarly were not statistically significant.
Overall, 78 (77.2%) of participants had negative cytology: 15 (14.8%) ASC-US, 7 (6.9%) LSIL, and 1 (.99%) high-grade squamous intraepithelial lesions (HSIL). Cytology results for those positive versus negative for high-risk HPV are shown in Table 6. Sensitivity, specificity, positive predictive value, and negative predictive value of high-risk HPV for the detection of abnormal cytology were similar for self testing and clinician testing. For example, the sensitivity for detection of both ASC-US and LSIL was identical for self-collected versus clinician-collected specimens (0.70 versus 0.70 for ASC-US and 0.63 versus 0.63 for LSIL). The specificity for detection of ASC-US and LSIL was slightly lower for self-collected versus clinician-collected specimens (0.62 versus 0.66 for ASC-US, and 0.56 versus 0.60 for LSIL). Sensitivity was higher and specificity lower when the results of self and clinician tests were combined.
In this sample of adolescent girls and young women attending a hospital-based teen health center, agreement between the results of self and clinician testing for HPV DNA varied depending on how results were categorized. The results of self-testing and clinician testing were in complete agreement in only half of those positive for HPV when all HPV types were considered. Explanations for this finding may include the relatively high percentage of participants with multiple HPV types and the possibility that self-collected and clinician-collected specimens sample different areas of the genital tract: self testing may be more likely to detect vulvar HPV types and clinician testing ectocervical and endocervical HPV types. In contrast, agreement between the results of self testing and clinician testing was high when test results were categorized into negative/low-risk HPV types versus high-risk HPV types. Human papillomavirus type and age were not associated with level of agreement, and the sensitivity and specificity of self- versus clinician-collected HPV DNA specimens for the detection of abnormal cytology were almost identical.
Efforts to compare our results regarding self testing in adolescents with those previously reported in adult populations are complicated by the wide variation in study samples, procedures, analytic methods, and outcome measures chosen. Most studies of women have demonstrated that self-testing detects similar or lower rates of HPV DNA than clinician testing.16,18,19,26 However, we demonstrated that self testing detected a higher percentage of women with high-risk HPV types than clinician testing; these results are consistent with 1 previous study of adults, which reported that self testing detected HPV DNA in 47% of women and clinician testing in 38% of women.15 Furthermore, the level of agreement between the results of self testing and clinician testing, as measured by a κ statistic, was relatively high in our study (.72) compared with adult studies (range 0.45 to 0.74).16,18,19
Using molecular techniques to identify women who are positive for high-risk HPV DNA and who consequently are at higher risk than other women for development of cervical dysplasia and carcinoma is important for clinicians. If future studies confirm that self-collected swabs are accurate in the detection of high-risk HPV DNA in adolescents and young adults, there may be important implications for testing young women at risk for HPV infection and its sequelae. In clinical settings, HPV testing may be used as an adjunct to cervical cytology screening and in the follow-up of selected women with abnormal cytology or biopsy-confirmed CIN.13,14 Self testing thus has the potential to eliminate the need for a speculum examination in some adolescents and young adults. Recently, the need for routine speculum examinations in adolescents has been questioned in light of the fact that other common sexually transmitted infections (including gonorrhea, Chlamydia, and trichomonas) may be diagnosed by self- or clinician-directed vaginal swabs.27–29 Disadvantages of the speculum examination include the time, equipment, and clinician expertise required to perform an examination and the costs associated with the examination. In addition, adolescents who experience pain and embarrassment during a speculum examination may avoid seeking gynecologic health care if they think it will involve an examination.30–32 Even adolescents who do not find the speculum examination uncomfortable may have different preferences regarding testing methods and differing comfort levels in terms of medical procedures. Self testing therefore has the potential to expand testing options and increase adherence to screening and follow-up protocols.
Implications for research protocols are less clear. Human papillomavirus testing is used with increasing frequency in epidemiologic studies of HPV infection and in HPV vaccine protocols. If self-testing is offered to adolescent and young adult women, some may be more likely to enroll in or adhere to these protocols. The results of our study suggest that in protocols that define HPV positivity as 1 or more high-risk HPV type, self testing may be adequate. However, if the study design requires that multiple and specific HPV types are taken into account, the level of agreement between self-collected and clinician-collected specimens may not be adequate.
There are a number of limitations to this study. First, the study sample was relatively small and consisted primarily of African-American, low-income adolescent girls and young women and, thus, the findings may not be generalizable to other adolescent and young adult populations. For example, the prevalence of HPV in this study sample was high, and therefore the specificity of HPV testing for detection of cervical dysplasia may be lower in this population compared with other adolescent and young adult populations. In addition, the prevalence of HPV 16 was lower than has been reported in previous studies. This finding may be explained in part by the fact that many of the participants in the study come from the same urban location, that is, transmission of HPV may be occurring within a relatively small group of young women. Thus, the relative proportions of specific HPV types identified may differ from that of other populations, and certain types that have not been found to be prevalent in other populations, such as HPV 84, may emerge as relatively more prevalent in this sample. As is typical in adolescent and young adult samples, there were relatively few with HSIL, limiting our ability to assess test characteristics of high-risk HPV for the detection of HSIL. Furthermore, we did not randomize the order in which samples were taken. However, detection of high-risk HPV was not dependent on the order in which sample was taken in a previous study.18 We did not perform colposcopy on all participants in this study; future studies in adolescents should include colposcopy as an outcome measure to better define the sensitivity and specificity of HPV testing in detecting CIN 2 and 3. Finally, we compared the results of self testing to clinician testing using 1 swab for each test but did not develop a criterion standard for HPV testing. In the future, it will be critical to determine a criterion standard for HPV detection in both female adolescents and adults.
In conclusion, our data suggest that adolescent and young adult women attending an urban teen health center were able to self-collect specimens for HPV DNA at least as well as women were able to and that self testing may be sufficiently sensitive for the detection of high-risk HPV in clinical settings.
1. Hildesheim A, Schiffman MH, Gravitt PE, Glass AG, Greer CE, Zhang T, et al. Persistence of type-specific human papillomavirus infection among cytologically normal women. J Infect Dis 1994;169:235–40.
2. Remmink AJ, Walboomers JM, Helmerhorst TJ, Voorhorst FJ, Rozendaal L, Risse EK, et al. The presence of persistent high-risk HPV genotypes in dysplastic cervical lesions is associated with progressive disease: natural history up to 36 months. Int J Cancer 1995;61:306–11.
3. Moscicki AB, Shiboski S, Broering J, Powell K, Clayton L, Jay N, et al. The natural history of human papillomavirus infection as measured by repeated DNA testing in adolescent and young women. J Pediatr 1998;132:277–84.
4. Clavel C, Masure M, Levert M, Putaud I, Mangeonjean C, Lorenzato M, et al. Human papillomavirus detection by the hybrid capture II assay: a reliable test to select women with normal cervical smears at risk for developing cervical lesions. Diagn Mol Pathol 2000;9:145–50.
5. Nobbenhuis MA, Walboomers JM, Helmerhorst TJ, Rozendaal L, Remmink AJ, Risse EK, et al. Relation of human papillomavirus status to cervical lesions and consequences for cervical-cancer screening: a prospective study. Lancet 1999;354:20–5.
6. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–9.
7. Manos MM, Kinney WK, Hurley LB, Sherman ME, Sheih-Ngai J, Kurman RJ, et al. Identifying women with cervical neoplasia: using human papillomavirus DNA testing for equivocal Papanicolaou results. JAMA 1999;281:1605–10.
8. Cox JT. Evaluating the role of HPV testing for women with equivocal Papanicolaou test findings. JAMA 1999;281:1645–7.
9. The ASCUS-LSIL Triage Study (ALTS) Group. Results of a randomized trial on the management of cytology interpretations of atypical squamous cells of undetermined significance. Am J Obstet Gynecol 2003;188:1383–92.
10. Association of Reproductive Health Care Professionals Human Papillomavirus (HPV) and Cervical Cancer. AHRP Clinical Proceedings. Washington (DC): Association of Reproductive Health Professionals; 2001. p. 22.
11. Kim JJ, Wright TC, Goldie SJ. Cost-effectiveness of alternative triage strategies for atypical squamous cells of undetermined significance. JAMA 2002;287:2382–90.
12. The ASCUS-LSIL Triage Study (ALTS) Group. A randomized trial on the management of low-grade squamous intraepithelial lesion cytology interpretations. Am J Obstet Gynecol 2003;188:1393–400.
13. Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ. 2001 Consensus Guidelines for the management of women with cervical cytological abnormalities [review]. JAMA 2002;287:2120–9.
14. Wright TC Jr, Cox JT, Massad LS, Carlson J, Twiggs LB, Wilkinson EJ, for the American Society for Colposcopy and Cervical Pathology. 2001 consensus guidelines for the management of women with cervical intraepithelial neoplasia. Am J Obstet Gynecol 2003;189:295–304.
15. Hillemanns P, Kimmig R, Huttemann U, Dannecker C, Thaler CJ. Screening for cervical neoplasia by self-assessment for human papillomavirus DNA. Lancet 1999;354:1970.
16. Wright TC, Denny L, Kuhn L, Pollack A, Lorincz A. HPV DNA testing of self-collected vaginal samples compared with cytologic screening to detect cervical cancer. JAMA 2000;283:81–6.
17. Sellors JW, Lorincz AT, Mahony JB, Mielzynska I, Lytwyn A, Roth P, et al. Comparison of self-collected vaginal, vulvar and urine samples with physician-collected cervical samples for human papillomavirus testing to detect high-grade squamous intraepithelial lesions. CMAJ 2000;163:513–8.
18. Harper DM, Noll WW, Belloni DR, Cole BF. Randomized clinical trial of PCR-determined human papillomavirus detection methods: self-sampling versus clinician-directed–biologic concordance and women’s preferences. Am J Obstet Gynecol 2002;186:365–73.
19. Lorenzato FR, Singer A, Ho L, Santos LC, Batista Rde L, Lubambo TM, et al. Human papillomavirus detection for cervical cancer prevention with polymerase chain reaction in self-collected samples. Am J Obstet Gynecol 2002;186:962–8.
20. Kahn JA, Rosenthal SL, Hamann T, Bernstein DI. Attitudes about human papillomavirus vaccine in young women. Int J STD AIDS. 2003;14:300–6.
21. Gravitt PE, Peyton CL, Apple RJ, Wheeler CM. Genotyping of 27 human papillomavirus types by using L1 consensus PCR products by a single-hybridization, reverse line blot detection method. J Clin Microbiol 1998;36:3020–7.
22. Gravitt PE, Peyton CL, Alessi TQ, Wheeler CM, Coutlee F, Hildesheim A, et al. Improved amplification of genital human papillomaviruses. J Clin Microbiol 2000;38:357–61.
23. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et al, for the International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–27.
24. Keating DP, Clark LV. Development of physical and social reasoning during adolescence. Dev Psychol 1980;16:23–30.
25. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York (NY): John Wiley & Sons, Inc; 1981.
26. Moscicki AB. Comparison between methods for human papillomavirus DNA testing: a model for self-testing in young women. J Infect Dis 1993;167:723–5.
27. Blake DR, Duggan A, Quinn T, Zenilman J, Joffe A. Evaluation of vaginal infections in adolescent women: can it be done without a speculum? Pediatrics 1998;102:939–44.
28. Shafer MB. Annual pelvic examination in the sexually active adolescent female: what are we doing and why are we doing it? J Adolesc Health 1998;23:68–73.
29. Kahn JA. With urine-based screening, do sexually active adolescent girls still need annual pelvic examinations? Yes: pelvic examination continues to play a key role. West J Med 2000;173:292.
30. Zabin LS, Clark SD. Why they delay: a study of teenage family planning clinic patients. Fam Plann Perspect 1981;13:205–17, 211–7.
31. Kahn JA, Chiou V, Allen JD, Goodman E, Perlman SE, Emans SJ. Beliefs about Papanicolaou smears and compliance with Papanicolaou smear follow-up in adolescents. Arch Pediatr Adolesc Med 1999;153:1046–54.
© 2004 The American College of Obstetricians and Gynecologists
32. Millstein SG, Adler NE, Irwin CE. Sources of anxiety about pelvic examinations among adolescent females. J Adolesc Health Care 1984;5:105–11.