First international proficiency study on human papillomavirus testing in cervical cancer screening

Abstract

BackgroundAlthough cervical screening using Human Papillomavirus (HPV) testing is globally recommended public health policy, there has been no international proficiency studies specifically targeting HPV testing for cervical screening. ObjectiveTo obtain the first global overview of the current proficiency of HPV testing services for cervical cancer screening. Study designA coded proficiency panel of 12 samples containing HPV types 16, 18, 31, 33, 45, 52, 58 or 35/39/51/56/59/68 in human DNA in varying amounts as well as control. Datasets detecting at least a) 10 International Units (IU) of HPV16 and 18, b) 1000 IU of HPV types 31, 33, 45, 52, 58 and c) having no false positives were considered proficient. ResultsIn total, 84 laboratories worldwide submitted 158 datasets (some laboratories used >1 HPV testing platform). Of those, 122 (77%) were 100% proficient. Only 14/158 datasets (9%) contained false positive results. Comparison of results with assays approved by the Food and Drug Administration (FDA) suggest that future proficiency requirements should also accommodate assays detecting only 100 IU of HPV16/18. A pool of low oncogenicity HPV types that contributed very little to sensitivity, but adversely affected specificity, was detectable by most datasets. ConclusionInternationally recognized proficiency studies of HPV screening, traceable to international standards, provided an overview of current testing performance. There was a high level of proficiency in terms of sensitivity and few false positives, but specificity was not optimal and further research on optimal specificity of HPV screening tests may be warranted.