Recruitment and clinical methods were detailed previously . Briefly, women undergoing routine screening (n = 252)(“screened”) or had not been screened in the last three years (n = 191) (“under-screened”) were recruited as part of a larger study of self-collection and HPV DNA testing in 4 counties (Tallahatchie, Leflore, Sunflower, and Washington) located in northwest Mississippi and part of the Mississippi Delta. Women aged 26 to 65 years of age, non-pregnant, with a cervix, and willing to provide written, informed consent were enrolled into the study. Institutional review boards from the NCI, UAB, State of Mississippi Health Department, and Westat approved this study.
For all participants, cervical specimens were collected into PreservCyt (Hologic, Bedford, MA, USA) for routine cervical cytology and the residual was retained for HPV testing. Women were given a kit for self-collection at home, including a self-collection device [20, 21] (http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=6,475,165.PN.&OS=PN/6,475,165&RS=PN/6,475,165) and a vial of Scope™ (Proctor and Gamble, Cincinnati, OH, USA) mouthwash for specimen transport . The self-collection device physically and mechanically resembles a tampon and includes an outer sheath to shield the collection from vaginal contamination and irrelevant vaginal HPV infections. Women were instructed to insert the device into their anterior vagina, expose the Dacron collection tip to obtain a sample of cervical cells, retract the tip into the sheath, remove the device, and eject the tip into the transport medium. We used mouthwash as our transport medium to avoid giving kits containing toxic liquid-based cytology medium to our participants to take home .
Specimens were tested for 37 HPV genotypes (6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51–56, 58, 59, 61, 62, 64, 66–73, 81–84, 82v, and 89) using Linear Array (LA; Roche Molecular Systems, Pleasanton, CA, USA) [23, 24]. HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68 were classified as the certain or probable carcinogenic HPV genotypes . HPV genotypes were also classified according to broad branches in the phylogenetic tree for HPV genotypes [26, 27]: 1) HPV genotypes 6, 11, 40, and 42, which are classified in the alpha 1, 8, and 10 group (alpha 1/8/10 genotypes). HPV 6 and 11 cause 90% of condyloma accuminata (genital warts); 2) HPV genotypes 16, 18, 26, 31, 33, 35, 39, 45, 51–56, 58, 59, 64, 66, 67, 68, 69, 70, 73, 82, and 82v, which are classified in the alpha 5, 6, 7, 9, and 11 group (alpha 5/6/7/9/11 genotypes). All the carcinogenic and borderline carcinogenic HPV genotypes are found in these species; and 3) HPV genotypes 61, 62, 71, 72, 81, 83, 84, and 89, which are classified in the alpha 3, 4, and 15 group (alpha 3/4/15 genotypes). There is some evidence that these HPV genotypes in these species have a predilection for vaginal tissue [28–31].
Logistic regression was used to calculate odds ratios (OR), 95% confidence intervals, and p values to test for differences between study groups (screened vs. under-screened) and linear trends with age groups (26–39, 30–39, 40–49, and 50–65 years) for testing HPV positive for categories of HPV. For categories of HPV prevalences, both crude and adjusted for study group and age group, were calculated. Binomial exact 95% confidence intervals were used where noted. An exact version of McNemar chi-square test or a symmetry chi-square test was used to test for differences in prevalence for individual or specific group of HPV genotypes from the same specimen or for same individual or specific group of HPV genotypes from different specimens.
A post-hoc power calculation based on a McNemar chi-square test indicated that 1) a sample size of 406 pairs for the whole analysis population achieves 80% power (alpha = 0.05) to detect a difference in prevalence of 4.5% between specimens collections if the discordance was 10%, of 6.4% if the discordance was 20%, and of 7.8% if the discordance was 30%, and 2) a sample size of 312 pairs for women 30 years and older achieves 80% power (alpha = 0.05) to detect a difference in prevalence of 5.2% between specimens collections if the discordance was 10%, of 7.3% if the discordance was 20%, and 8.9% if the discordance was 30%.
A p value of < 0.05 was considered statistically significant. STATA Version 11.1 was used for these analyses (StataCorp, College Station, Texas, USA).