Human papillomavirus (HPV) testing for the two most dangerous strains (HPV16 and HPV18) detects more high-grade pre-cancerous lesions (which can lead to cervical cancer) than current cervical cancer screening using cytology alone. The findings published Online First in The Lancet Oncology, suggest that HPV testing should become the primary screening tool to rule out disease, with cytology reserved for triage of women who test positive for HPV, to decide which women need immediate colposcopy.
It is well established that HPV DNA testing is more effective than cytology-based primary cervical cancer screening, but the best strategy for managing women infected with HPV remains unclear. It is estimated that 27% of women in the USA are infected with HPV. HPV 16 and 18 are the most common high-risk types of HPV, responsible for about 70% of invasive cervical cancer. HPV testing with HPV16 and HPV18 detection has been suggested as a triage technique in HPV-positive women to increase the accuracy of identifying women with CIN3 or worse lesions (cervical cancer and its most serious precursor lesions) who need immediate colposcopy.*
The ATHENA trial was designed to assess the performance of HPV testing (including the detection of high-risk HPV strains 16 and 18) compared with liquid-based cytology, and to establish more effective management strategies for HPV-positive women.
The study recruited over 47 000 women aged 25 years or older who attended routine cervical screening in the USA between May, 2008, and August, 2009. Two samples were taken from each woman for conventional cytology and HPV testing. All women with atypical squamous cells of underdetermined significance (ASC-US) cytology or worse, all those who had normal cytology but were HPV positive, and a subset of women negative for both were referred to colposcopy.
In women who had colposcopy, many more existing high-grade precancers were detected in women given HPV testing (92%) compared with those given cytology (53.3%).
Importantly, combining HPV testing with cytology provided little benefit compared with HPV testing alone—increasing sensitivity (the proportion of true positives correctly identified) by only 4.7%, but the number of screen positives by more than a third.
The authors also identified potentially useful combinations of tests that could be used for triage of women with HPV infection. They explain:"The use of HPV16 or HPV18 detection as an additional or alternative triage strategy to reproducible cytological abnormalities (LSIL or worse, or HSIL or worse)…resulted in increased, more reliable (interlaboratory) performance for identification of women with CIN3 or worse compared with the use of ASC-US or worse cytology alone."
They also point out that because the cobas HPV test identifies HPV16 and HPV18 and other cancer-causing HPV genotypes in one test, testing for HPV16 and HPV18 to triage women infected with HPV could be very efficient and reduce manpower requirements in laboratories compared with cytology.
They conclude: "Rational use of HPV testing (and genotyping for HPV16, or HPV18, or both) with or without liquid-based cytology can provide potentially cost-effective and safe cervical cancer screening."
In a Comment, Guglielmo Ronco from the Centre of Cancer Prevention, Turin, Italy and colleagues point out: "Co-testing of HPV and cytology will probably be replaced by standalone HPV testing as the primary screening test in high-income countries, because addition of cytology seems to provide little gain, according to Castle and colleagues' findings…and the results of longitudinal studies."
They add that the findings: "Also provide useful information about triage strategies for parts of the world where high-quality cytology has been difficult to implement and combinations of HPV tests might eventually offer a more sustainable option."
Dr Philip E Castle, American Society for Clinical Pathology Institute, Washington, DC, USA. T) +1 202 347 4450 Ext 2902 E) Philip.castle@ascp.org
Professor Mark Stoler, University of Virginia Health System, Virginia, USA. T) +1 434 982 0284 or +1 434 284 2555 (mobile) E) mhs2e@virginia.edu
Dr Guglielmo Ronco, Centre of Cancer Prevention, Turin, Italy. T) +39 3357 553701 E) guglielmo.ronco@cpo.it
Notes to Editors: *Currently, US guidelines recommend that women with HPV who have positive cytology are immediately referred for colposcopy, but in women with HPV who have normal cytology they recommend rescreening in 1 year. However, some of these women might develop invasive cancer during the screening interval and up to half of these women are lost to follow-up. Therefore, a more efficient screening option is required that could identify all women with lesions needing immediate referral.
Journal
The Lancet Oncology