HPV-negative CIN3 and cervical cancer in Switzerland: any evidence of impact on screening policies?

New practice guidelines for cervical cancer screening in Switzerland recommend testing for the presence of DNA from high-risk types of human papillomavirus (hrHPV) as the primary screening test for all women 30 years and older. However

or not evaluable (n = 19) owing to inhibitors of the polymerase chain reaction (PCR) or lack of material.Confirmatory testing of 20 available samples (18 initially negative samples and two initially nonevaluable samples) was performed by the WHO Global Reference Laboratory, which found 16 retested samples to be hrHPV positive.The final results found: 97.0% (745/768) hrHPV positive, 2.3% (18/ 768) nonevaluable and 0.7% (95% confidence interval [CI] 0.2-1.5%)(5/768) hrHPV negative.Only two of these five negative samples could be subjected to confirmatory testing.The hrHPV-positivity rate amongst all evaluable samples was therefore 99.3% (95% CI 98.5-99.8%,745/750).The number of hrHPV-negative precancers and cancers in the CIN3+ study was low but shows that a small percentage of samples will have negative results with hrHPV testing.The findings from retrospective studies and archival material cannot be directly extrapolated to the performance of hrHPV testing in cervical samples in screening populations because these studies include samples from both women who have presented with clinical symptoms and asymptomatic women undergoing screening.Predictive values of hrHPV testing depend on the prevalence of the disease, which is not comparable between these two populations of women.In this article, we summarise the rationale for hrHPV testing as the primary test for screening, explain potential reasons for false-negative hrHPV test results in the presence of CIN3+ lesions, describe true HPV-negative precancer and cancer, and comment on the relevance of the debate about co-testing versus hrHPV-only screening approaches in Switzerland.

Primary hrHPV screening for cervical cancer
Randomised controlled trials and cohort studies have consistently demonstrated the high sensitivity of clinically validated hrHPV tests for the detection of clinical lesions.The high negative predictive value allows longer screening intervals than cytology alone [8].Furthermore, randomised controlled trials [9][10][11][12][13][14][15] showed that hrHPV testing detects more CIN3 lesions than cytology at the first screening round and leads to an overall reduction in the incidence of these lesions and of invasive cervical cancer in the following years among the screened population.The protective effect of HPV testing for CIN3 and invasive cancer was not observed in the first 2.5 years after screening but increased significantly thereafter.The results of most randomised controlled trials also showed that primary HPV screening applies only to women from 30 years.In younger women, HPV testing has a lower specificity for underlying precancer and cancer.Based on these data and on cost-effectiveness modelling analyses [16], several European countries (Italy, Denmark, The Netherlands, UK, Estonia, Germany, Portugal, Spain, Sweden, Turkey), as well as the USA, Australia, New Zealand, and 10 other countries have now adopted or plan to implement primary hrHPV testing as a stand-alone screening modality without cytology.
Possible reasons for a negative hrHPV test result in a CIN3+ lesion  [20] have limited analytic sensitivity.To minimise these deficiencies, clinicians are advised to use sufficiently sensitive and clinically validated hrHPV tests for primary screening [21] in qualified laboratories accredited by authorised accreditation bodies and in compliance with international standards.However, even very good HPV tests may miss CIN3 lesions associated with hrHPV types in 1-3% of cases [22].3. True hrHPV-negative CIN3 and cervical cancer.Although hrHPV is a necessary causal factor in the pathway to cervical cancers [23], hrHPV is not always detected in tumour specimens from women diagnosed with CIN3 or invasive cervical cancer.Nearly all CIN3 lesions are associated with hrHPV types, with a very small proportion associated only with low risk HPV types [24].
True primary cervical cancers not linked to HPV account for a small portion of hrHPV-negative testing [25,26].They include endocervical adenocarcinomas of the gastric type, minimal deviation adenocarcinomas, as well as clear cell and mesonephric adenocarcinomas [27,28].A small group of hrHPV-negative endocervical carcinomas have features of endometrial adenocarcinomas.A recent study systematically evaluated 777 cervical cancer tissues from several US-based cancer registries and found carcinogenic HPV in 91% of the cases.Nearly 60% of the HPV-negative cases could not be distinguished histologically from endometrial primary adenocarcinomas, suggesting that these tumours may not be of cervical origin [29].Histological findings of cervical adenocarcinoma with hrHPV-negative testing should be verified using immunostaining for p16 and additional biomarkers to exclude an origin from other sites.There are limited data regarding the performance of the Pap test for hrHPV-negative adenocarcinomas.Because false negative Pap results are common with endocervical adenocarcinomas [30], it is unclear how accurate cytology is for the detection of these rare tumours.

Co-testing versus hrHPV-only primary screening
Prospective European randomised screening trials have shown that adding cytology to hrHPV primary testing (cotesting) offers minimal increased protection against the subsequent development of cervical disease, at the expense of a considerable loss in specificity (colposcopy is performed if positive by either test), compared with hrHPVonly primary screening [12,14,[31][32][33].Similar results have been found in a large prospectively conducted US FDA registration trial of primary hrHPV screening [15].Another US study, which included approximately one million women aged 30-65 years, demonstrated that co-testing confers only a very slight marginal gain (0.003%) in reassurance (lower cancer risk) of safety against cancer over 5 years.Most of the reassurance provided by co-testing was derived from the hrHPV test component.The study also demonstrated that the 5-year reassurance after a negative co-test is approximately the same as the 4-year reassurance after a negative HPV test alone [14].
From a population management perspective, screening modalities must be projected over the screening lifetime to estimate what strategy is best for women in terms of benefits, harms and costs.Primary screening with two tests, especially in places like Switzerland where the prevalence of disease in screened populations is very low, may lead to over referral and overtreatment.Another consequence is that two test results obtained simultaneously are difficult to interpret resulting in significant algorithmic complexity for the management of screened women [34].Furthermore, combined cytological and hrHPV screening is not cost-effective, since there is an increased cost without significantly increasing sensitivity [35].

Conclusion
Testing for hrHPV DNA, like other cancer screening tests, cannot detect all cases of prevalent or incipient cervical cancer.Most missed cases are a small subset of adenocarcinomas that are not linked to HPV.The evidence base demonstrated that a validated hrHPV test is sufficiently accurate for clinical use and could reduce the complexities of interpretation and management of co-test results as well as resource expenditure inherent in screening with two tests.

Disclosure statement
No financial support and no other potential conflict of interest relevant to this article was reported.