The clinical relevance of human papillomavirus testing: relationship between analytical and clinical sensitivity

Abstract

Given the fact that infection with high‐risk human papillomavirus (HPV) is causally involved in cervical cancer, addition of high‐risk HPV testing to a cervical smear may improve the efficacy of cervical cancer screening programmes, the triage of women with equivocal or borderline Pap smears, and the monitoring of women who have been treated for cervical intraepithelial neoplasia grade 3 (CIN 3). Compared to a cervical smear HPV tests revealed a superior sensitivity (ie clinical sensitivity) for lesions ≥ CIN 3, and a negative predictive value approaching 100%. However, a potential complication is the availability of several HPV testing methods, all displaying a different sensitivity and specificity to detect HPV‐positive women (ie analytical sensitivity and specificity). There is now compelling evidence that the clinical sensitivity and specificity of HPV tests are not simply synonymous to their analytical sensitivity and specificity, respectively. In fact, a distinction between so‐called clinically relevant and irrelevant high‐risk HPV infections should be made when considering HPV tests for primary screening, triage policies, or post‐treatment monitoring. Here, we discuss the potential importance of HPV load in the context of currently widely applied HPV detection methods, to distinguish clinically relevant from irrelevant HPV infections. From this it can be concluded that it is of utmost importance to define criteria, involving viral load threshold and the type of HPV detection method that should be fulfilled by an HPV test before implementation of such a test in clinical practice and population‐based cervical cancer screening programmes. Copyright © 2003 John Wiley & Sons, Ltd.