Prevalence and genotype distribution of human papillomavirus in cytology specimens containing atypical glandular cells: a case-control study.

Incidence of glandular cell cancers has risen. While atypical glandular cell (AGC) grade cytology results represent only a small percentage of all Pap test results reported annually in the US, a significant percentage represents a corresponding high-grade lesion on follow-up biopsy. The 2006 ASCCP consensus guidelines for AGC-grade cytology results include colposcopy, endocervical sampling, and high-risk (HR) HPV testing for patient management.Determine HPV prevalence and genotype distribution in AGC-grade cytology specimens (n=53) compared to cytology specimens negative for intraepithelial lesion or malignancy (n=338).DNA extracted from residual, de-identified liquid-based cytology specimens, using QIAamp MinElute Media Kit was analyzed by PCR using Roche Linear Array HPV Genotyping and Detection Test Kits. Multivariate logistic regression compared HPV prevalence and genotype distribution between cases and controls to generate age-adjusted odds ratios (ORadj) and 95% confidence intervals (CI).HR-HPV DNA was found in 34.0% of cases and 7.4% of controls (ORadj=9.11; 95% CI: 4.08-20.33, p-value<0.001). Limiting analysis to HPV-16 and/or -18 resulted in finding HPV DNA in 20.8% of cases and 1.2% of controls (ORadj=40.10; 95% CI: 10.73-149.88, p-value<0.001). In contrast, prevalence of low-risk HPV DNA was similar between groups: 13.2% of cases and 17.2% of controls (ORadj=0.91; 95% CI: 0.35-2.31, p-value=0.834).AGC-grade cases contained a significantly higher rate of HR-HPV compared to controls, supporting earlier recommendations for HPV testing of AGC-grade cytology specimens. Our findings also suggest that follow-up genotyping of HR-HPV containing AGC cases for HPV-16 and/or -18 specifically would be useful in patient management.