PDF《ARTICLES》

958 ARTICLES Journal of the National Cancer Institute, Vol. 85, No. 12, June 16,

1993 Subjects and Methods Study Subjects We selected

500 case subjects with CIN and

500 control subjects from a cohort of

21 146 women included in an ongoing natural history study of HPV infection at the Kaiser Permanente prepaid health plan in Portland, Ore. This cohort, established between April 1, 1989, and November 2, 1990, was chosen to be representative of all women presenting for routine cervical cylologic (Pap smear) screening at Kaiser Permanente, which serves about one fourth of the adult female population of Portland. The

500 case subjects were selected from among the

673 women found at enrollment to have cytologic evidence of CIN or condylomatous atypia [a diagnosis now combined with CIN

1 in the Bethesda cytologic classification system (//)]. (We excluded from consideration as case subjects the

2819 women [13%] of the cohort who had a medical history of CIN but were currently cytologically negative, often because of ablative treatment.) The

500 control subjects were randomly selected from the

17654 women with normal cytologic diagnoses and no known history of CIN. We selected the

1000 subjects on the basis of the original enrollment cytologic diagnoses, regardless of subsequent cytologic review or histopathologic confirmation. The case group included all

181 women in the cohort with enrollment cytologic diagnoses of CIN

1 or greater (23 with CIN

3 or cytologic changes suspicious for cancer,

27 with CIN 2, and

131 with CIN 1) and

319 women chosen randomly from the

492 with condylomatous atypia. Among the control subjects, the enrollment cytologic diagnoses of

413 were negative, and

87 (17.4%) were classified as having benign reactive atypia ;

there was no statistically significant difference between these two control subgroups with regard to HPV test results or demographic characteristics. Telephone Interview and Cytologic Sampling All subjects were contacted by telephone. If oral informed consent was obtained, trained interviewers administered a 20-minute telephone question- naire that measured the established epidemiologic risk factors for CIN except diet. For HPV testing, a 10-mL cervicovaginal lavage was obtained from the study participants as part of the cervical cytologic screening procedure, with approval from the Kaiser Permanente Institutional Review Board. The lavage consisted of a directed rinse of the cervical os (using sterile physiologic saline administered via a 10-mL syringe with a 3'

/i-inch flexible intracatheter extender), recollected after pooling in the posterior vaginal fornix (12). The lavage was obtained after the routine one-slide cervical smear, consisting of a cervical scrape performed with an Ayre spatula and a cytobrush sample of the endocervix. HPV test data and at least some questionnaire data were obtained from

438 (88%) of the case subjects and

424 (85%) of the control subjects. Four in the case group and

10 in the control group were ineligible for study because of a previous hysterectomy or past CIN determined after their selection. Of those eligible,

17 case subjects and

30 control subjects had cervical specimens that were inadequate for HPV testing, mainly because there were no apparent cervical cells, as assessed by the appearance of the samples and the lack of human DNA (B-globin) amplification. An addi- tional

41 case subjects and

36 control subjects refused the interview or were lost to follow-up. In the results presented below, all available data are included;