HART study findings suggest a bigger role for human papillomavirus
testing in cervical cancer screening
Oncogenic types of human papillomavirus (HPV) are the underlying
cause of almost all cervical cancer and its precursors (1). Human papillomaviruses
are sexually transmitted and data suggests that prevalence of high risk HPV
in women attending routine cervical screening in the UK in the 1990s was 6-7%
(2,3) . High risk (HR) oncogenic types of HPV can be detected through DNA
testing of epithelial cells obtained during routine cytological screening.
However HPV testing has not yet been introduced into routine cytological screening
programmes in Europe and the focus has been on cytological screening rather
than on the underlying sexually transmitted disease. The perceived wisdom,
drawn from ten years of trials on triage in cervical screening, has cytology
as the first step. More recently, in the USA, HR HPV testing has been advocated
as a triage step to improve the sensitivity of CIN2 detection. Drafts of new
European guidelines (European Cervical Cancer Screening Network website, http://www.cancer-network.de/cervical/index.html
recognize HPV as causal and include a section on HPV testing. They advocate
further research to better define the performance indicators of HPV DNA testing
(alone and with cytology) before launching any new screening approaches.
A recent paper presents findings from the recently published HART (HPV
in Addition to Routine Testing) study (4), and argues for initial screening
of high risk HPV before triaging positive cases with cytology. The HART
study is a multicentre randomised controlled trial of 11 085 women (aged
30-60 years) comparing, for women with minimal abnormalities (a borderline
smear or positive HR HPV test and negative cytology), strategies of immediate
colposcopy versus viral and cytological surveillance. The study confirmed
previous findings that HR HPV testing can raise sensitivity of CIN2 detection
by 20% compared with cervical cytology but also increases the false positive
rate by 5%. The researchers found that, for those with minimal abnormalities,
it was safe to repeat testing at 12 months rather than resort to colposcopy
immediately and suggested the possibility of lengthened screening intervals
for low risk women (negative by cytology and HPV) from 3 to 5 years. In
women positive for HPV at baseline, who later had virological and cytological
surveillance, a significant proportion (45% with negative cytology and 35%
with borderline cytology) had cleared their HPV infections by the retest
after 6-12 months. No CIN2 was found in these groups nor in women with an
initial negative HPV test with borderline or mild dyskaryosis on cytology.
The authors concluded that HPV positive women with normal or borderline
cytology (about 6% of women screened) could be managed with repeat testing
at 12 months, improving detection rates of CIN2 without increasing colposcopy
referral rates. They also provocatively advocate for initial screening with
HPV test followed by triaging of those HPV positive cases with cytology.
This study had number of limitations. The abnormality rate was 25% lower
than the national average possibly limiting the general applicability of
the results. Also compliance with follow-up was only approximately 70%.
A further important limitation of the trial is its concurrent design in
which all women received both tests, such that it did not compare HPV testing
with cytology as separate screening modalities. Further randomised controlled
trials are required to assess the duration of protection conferred by a
negative HPV result and the efficacy of a Pap-centred triage approach before
any conclusions can be drawn regarding HPV as a lone initial screening tool.
In Europe the triaging of women after baseline borderline smears is different
to the USA, with the guidelines suggesting a usual approach of repeat testing
at 6 months rather than immediate colposcopy (5). The impact of this study’s
findings in terms of suggesting deferring colposcopy in those with mild
dyskaryosis (and HPV negative) may therefore have different significance
for the European guidelines.
There have been very few randomised controlled trials of HPV testing in
population screening. These have generally focused on triage (the management
of patients with borderline or mildly dyskaryotic smears (ASCUS-atypical
squamous cells of undetermined significance)) rather than on initial screening.
In the USA, a recent trial ASCUS -LSIL (low grade squamous intra-epithelial
lesions) Triage Study (ALTS))(6,7) prompted changes in their guidelines
for cervical screening to highlight the usefulness of HPV testing in triaging
women with equivocal smears but not advocating it as initial screening test.
These changes were published as two consensus statements by the American
Medical Association (8,9) and were reviewed in a previous issue of Eurosurveillance
(10). New technologies in cervical screening including the role of HPV testing
are being studied in Europe under the auspices of the European Cervical
Cancer Screening Network and their progress can be reviewed on their website
If the HART study findings are borne out by further randomised controlled
trials, a radical shift towards STI detection and away from cytological
screening will be required. This will create a powerful new focus for STI
prevention for women and their partners and could have a major impact on
other STI prevention activities. The European Surveillance of STI (ESSTI)
network, a collaboration of 14 member states (and Norway) has been established
to address issues of sexually transmitted infection surveillance, prevention
and care issues across Europe (11-13). ESSTI has been highlighting the growing
concern that STIs are rising rapidly in Europe(14). This additional momentum
to tackle the problem would be welcomed by the ESSTI network.