With the introduction of HPV child vaccination programmes, there will have to be a shift from cytology to HPV testing as the main technology involved in primary cervical screening, say the contributors to an on-coming special issue of Preventive Medicine (Tota & Ratnam I) (T&R). Why?  Well, first, because of the inevitable decline in the positive predictive value of the test (i.e. proportion of positive results that are true positives) that comes with declining prevalence of HPV sequelae.  This is an important consideration given the reality of the potential ‘harm’ resulting from false positive diagnoses.  But it is also necessary to take into account the impact on diagnosis (which, of course, in the case of cytology, takes place through the judgment of fallible human cytotechnologists) of the ever-dwindling proportion of abnormalities – an effect well described by T&R as a reduced ‘signal-to-noise ratio’.  This, our authors argue, will inevitably lead to ‘fatigue’.

Yet the transition to HPV primary screening is very much to be welcomed, it seems.  Tota & Ratnam I comprehensively review recent trials – in Canada, US and Europe – which all demonstrate that primary HPV screening (in combination with various ‘triage’ regimes for positive cases) offers more security, even at more distant testing intervals, than a cytology-based regime.  Also one that is less prone to human error, more cost-effective, as well as capable (unlike cytology) of being adapted to ‘self-testing’ regimes that could allow wider access (especially in limited resource settings).

Another paper in this on-coming special issue reviews trials (Canadian HPV FOCAL, and Montreal-based VASCAR) testing different ‘triage’ regimes (Tota & Ratnam II).  These involve cytology, with or without HPV genotyping.  Genotyping allows the discrimination of different levels of risk according to HPV genotype, giving health services the option of a differentiated approach to more or less ‘high risk’ strains (i.e. retesting after a year, referral to cytology, or to colposcopy). Whether or not genotyping is included in the regime, the combination of primary HPV screening in combination with triage seems to offer a much more reliable test than cytology – at the possible cost of some relatively minor increase in needless colposcopy referral.

Yet cervical screening policy must, in practice, be informed by more than epidemiological evidence – as the editor of this special issue (Schiffman) reminds us.  It will also depend on available resources and the willingness of a particular system to assume a degree of risk.  The US is particularly good example.  As Kinney & Huh show, in another study in this issue same special issue, the very marginal increment in safety demonstrated by five-yearly co-testing over stand-alone HPV is one that US appears not to be willing to relinquish, even at considerable cost both economic and in terms of ‘harms from screening’.

At the other extreme, of course, are the medium and limited-resource settings in which, for various reasons the aspiration to offer affordable protection through traditional forms of screening (e.g.  visual inspection with acetic acid (VIA)), may currently be delivering ‘sub-optimal’ results (see, for example, Sibanda & Cowan (STIs)).  (For an evaluation of HPV screening as against VIA, see Mitchell & Ogilvie (STIs).)  The special issue includes papers that consider the possibility of diverse screening algorithms in limited resource settings (Maza & Gage; Kuhn & Denny).  Where there are problems of access, the self-collection of samples, which becomes a possibility with HPV primary screening may offer a more feasible alternative to clinician based approaches.  Vallely & Caldor (STIs) makes the case for screening based on self-sampling using CepheidXpert.  Nelson & Arnold (STIs)  review 24 studies of HPV self-sampling across five continents.