Objective. To determine the value of human papillomavirus (HPV) testing for HPV-associated cervical disease (HPV-AD) and overall disease (atypical glandular cell [AGC]-associated cervical disease) in women with AGCs.
Materials and Methods. A literature search was conducted from January 1993 through September 2007 using various AGC-related terms with the exploded Medical Subject Heading (MeSH) term "HPV." Findings from 7 studies were used to calculate disease rates according to HPV status.
Results. The rate of AGC-associated cervical disease for 661 cases of AGC with concurrent HPV testing was 23.3%. The rate of HPV-AD was higher in HPV-positive, versus negative, cases (53% vs 3%, respectively). Human papillomavirus-positive, versus negative, status predicted a higher likelihood of a cervical intraepithelial neoplasia 2/3 lesion (odds ratio = 39.6, 95% CI = 17.9-87.4, p <.001). The rate of HPV-nonassociated cancers was significantly higher in patients who were negative, versus positive, for HPV (4% vs 0.4%; p =.016). Human papillomavirus testing had an overall 90% sensitivity, 79% specificity, 53% positive predictive value, and 97% negative predictive value for cases of HPV-AD. Atypical glandular cell with concurrent atypical squamous cell (ASC) or squamous intraepithelial lesion (SIL) (ASC/SIL) had higher rates of disease than AGC alone. The positive predictive value of HPV testing for AGC with concurrent ASC/SIL was higher than that for AGC alone.
Conclusions. All women with AGC should undergo a comprehensive initial examination regardless of HPV status. The presence of HPV identifies a group of women at higher risk for cervical disease who should be followed closely. Women positive for human papillomavirus with AGC and concurrent ASC/SIL are at even higher risk. If, after a comprehensive initial examination, women with AGC not-otherwise-specified and positive HPV have no identifiable disease, a cervical conization may be considered.
Human papillomavirus testing of women with atypical glandular cells on Pap testing identifies those at high risk for cervical disease.
1Baystate Medical Center, Springfield, MA, 2Hartford Hospital, Hartford, CT, and 3University of Connecticut, Farmington, CT
Reprint requests to: Peter F. Schnatz, DO, FACOG, Hartford Hospital, Conklin Bldg 203, 80 Seymour St, Hartford, CT 06102. E-mail: email@example.com
This research was unfunded.
There are no conflicts of interest.
These data were presented orally at the 2008 American Society for Colposcopy and Cervical Pathology Biennial Meeting on March 17, 2008, in Lake Buena Vista, FL.