A new HPV-DNA test for cervical-cancer screening in developing regions: a cross-sectional study of clinical accuracy in rural China.

Publication Type:

Journal Article


Lancet Oncol, Volume 9, Issue 10, p.929-36 (2008)


Adult, Alphapapillomavirus, Cervical Intraepithelial Neoplasia, China, Cross-Sectional Studies, DNA, Viral, Female, Humans, Mass Screening, Middle Aged, Papillomavirus Infections, Predictive Value of Tests, ROC Curve, Rural Health Services, Sensitivity and Specificity, Tumor Markers, Biological, Uterine Cervical Neoplasms


<p><b>BACKGROUND: </b>A new test (careHPV; QIAGEN, Gaithersburg, MD, USA) has been developed to detect 14 high-risk types of carcinogenic human papillomavirus (HPV) in about 2.5 h, to screen women in developing regions for cervical intraepithelial neoplasia (CIN). We did a cross-sectional study to assess the clinical accuracy of careHPV as a rapid screening test in two county hospitals in rural China.</p><p><b>METHODS: </b>From May 10 to June 15, 2007, the careHPV test was done locally by use of self-obtained vaginal and provider-obtained cervical specimens from a screening population-based set of 2530 women aged 30 to 54 years in Shanxi province, China. All women were assessed by visual inspection with acetic acid (VIA), Digene High-Risk HPV HC2 DNA Test (HC2), liquid-based cytology, and colposcopy with directed biopsy and endocervical curettage as necessary. In 2388 women with complete data, 441 women with negative colposcopy, but unsatisfactory or abnormal cytology or who were positive on HC2 or the new careHPV test, were recalled for a second colposcopy, four-quadrant cervical biopsies, and endocervical curettage. An absence of independence between the tests was not adjusted for and the Bonferroni correction was used for multiple comparisons.</p><p><b>FINDINGS: </b>Complete data were available for 2388 (94.4%) women. 70 women had CIN2+ (moderate or severe CIN or cancer), of whom 23 had CIN3+. By use of CIN2+ as the reference standard and area-under-the-curve analysis with a two-sided alpha error level of 0.0083, the sensitivities and specificities of the careHPV test for a cut-off ratio cut-point of 0.5 relative light units, were 90.0% (95% CI 83.0-97.0) and 84.2% (82.7-85.7), respectively, on cervical specimens, and 81.4% (72.3-90.5) and 82.4% (80.8-83.9), respectively, on vaginal specimens (areas under the curve not significantly different, p=0.0596), compared with 41.4% (29.9-53.0) and 94.5% (93.6-95.4) for VIA (areas under the curve significantly different, p=0.0001 and p=0.0031, for cervical and vaginal-specimen comparisons for the careHPV test, respectively). The sensitivity and specificity of HC2 for cervical specimens were 97.1% (93.2-100) and 85.6% (84.2-87.1), respectively (areas under the curve not significantly different from the careHPV test on cervical specimens, p=0.0163).</p><p><b>INTERPRETATION: </b>The careHPV test is promising as a primary screening method for cervical-cancer prevention in low-resource regions.</p>