Rh. Kaufman et al., RELEVANCE OF HUMAN PAPILLOMAVIRUS SCREENING IN MANAGEMENT OF CERVICALINTRAEPITHELIAL NEOPLASIA, American journal of obstetrics and gynecology, 176(1), 1997, pp. 87-92
OBJECTIVE: To evaluate the utility of human papillomavirus detection i
n identifying women with abnormal Papanicolaou smears who can be safel
y followed up with cytologic study only, we conducted a study to deter
mine the sensitivity, specificity, and negative and positive predictiv
e values of a Food and Drug Administration-approved human papillomavir
us test kit for detection of cervical intraepithelial neoplasia in col
poscopically directed biopsy specimens. STUDY DESIGN: We enrolled wome
n with abnormal Papanicolaou smears referred to a colposcopy clinic se
rving indigent patients. All 1128 women had a referral Papanicolaou sm
ear, a clinic Papanicolaou smear, and a sample for human papillomaviru
s deoxyribonucleic acid test; 1075 underwent colposcopically directed
biopsies and endocervical curettage. We used the HPV Profile kit for h
uman papillomavirus testing. RESULTS: Of 486 women with low-grade squa
mous intraepithelial lesions on Papanicolaou smear, 35.4% had high-ris
k human papillomavirus deoxyribonucleic acid detected, and of 592 with
high-grade lesions, 44.4% had high-risk human papillomavirus detected
. Among 527 women with biopsy specimens showing cervical intraepitheli
al neoplasia and in 267 with cervical intraepithelial neoplasia grades
2 or 3, 38.7% and 56.2% had high-risk human papillomavirus deoxyribon
ucleic acid detected. However, the sensitivity of human papillomavirus
deoxyribonucleic acid detection to identify biopsy-confirmed cervical
intraepithelial neoplasia grades 2 or 3 was 55.7%, and the positive p
redictive value of the test was only 34.9%. CONCLUSION: Human papillom
avirus appears to be causally associated with cervical cancer but huma
n papillomavirus screening does not appear to be of value to identify
women with abnormal Papanicolaou smears who can be safely followed up
with cytologic study alone.