The human papillomavirus (HPV) vaccine has dramatically reduced precancerous cervical lesions – and the most carcinogenic types, HPV16 and HPV18, have almost disappeared at the point where diagnoses are made. Precancerous changes still occur, however, driven by other high-risk types. This raises an unavoidable question: how should screening be organised when the biological risk behind the diagnosis has changed?
The risk of cervical cancer has diminished – not because cervical cancer has disappeared but because its driving force has shifted. The HPV vaccine has removed the HPV types that previously accounted for most cervical cancer cases.
In a new Danish cohort study, HPV16 and HPV18 have almost disappeared from both cervical cell samples and the biopsies in which the most advanced precancerous stages are detected. In addition, the risk of what doctors call cervical intraepithelial neoplasia grade 2 or worse (CIN2+) – tissue changes with an increased risk of progressing to cancer – is markedly lower among vaccinated women.
“The most striking thing is how rarely we now see HPV16 and HPV18 in both cell samples and in the more serious precancerous stages,” says Elsebeth Lynge, Professor Emerita at the University of Copenhagen and initiator of the Trial23 study.
That is precisely why success is reshaping the diagnostic landscape: once the most aggressive HPV types are removed, fewer screening tests return positive results, and a positive sample may not indicate the same risk as it once did – and this challenges the foundations of current screening.
Screening was designed for a disease biology that is now fading
For more than half a century, screening has been the main reason why cervical cancer in Denmark has become both rarer and less deadly. Cervical cell samples have enabled doctors to detect precancerous changes long before cervical cancer becomes life-threatening.
The HPV vaccine, however, has fundamentally altered the biological reality on which screening programmes were built.
“Screening was developed at a time when up to one in five young women were infected with HPV16 or HPV18,” says Elsebeth Lynge. “These were the types that carried the greatest risk of precancerous lesions and cancer.”
When Denmark began offering HPV vaccination as part of the childhood vaccination programme in 2008, the aim was to prevent infection with precisely these HPV types. Vaccination therefore had to be given before sexual debut and before exposure to HPV.
Vaccination shifts which HPV types actually drive cancer risk
Today, the first fully vaccinated cohorts have reached screening age, making it possible for the first time to examine how vaccination affects the diagnoses that are actually made in clinical practice.
“Our studies show that the prevalence of HPV16 and HPV18 has fallen sharply among vaccinated women. But in clinical practice, it is not enough simply to know which HPV types are circulating. What matters is which types actually drive cell changes to progress — and therefore represent the real cancer risk.”
The consequence is that the composition of the precancerous lesions that still occur is changing. They will be linked to other high-risk HPV types — not necessarily because these types have become more common, but because they may previously have been obscured by HPV16 and HPV18.
“It is only in the biopsies – where diagnoses are actually made – that you can see what vaccination has changed biologically,” says Elsebeth Lynge.
A natural experiment inside the national screening programme
When the meaning of a CIN2+ diagnosis changes, simply counting positive HPV tests is no longer sufficient. Women must be followed through the same screening system used in clinical practice – all the way into the tissue samples from which diagnoses are actually made and risk determined.
“The key was to use a fully population-based design,” says Elsebeth Lynge. “We wanted to see what happens in the national screening programme as it functions in real life.”
“We started with women born in 1993 – the first cohort to be offered HPV vaccination – and compared them with women born in 1983, who had undergone the same screening but had not been vaccinated. “Here we used only register data,” she explains.
The study then examined women born in 1994 – the next birth cohort offered HPV vaccination. Here, register data were supplemented with HPV testing of cervical cell samples and, later, HPV genotyping of biopsies.
“The study includes all women born in 1994 who were living in Denmark and who, from the age of 23 years, had at least one cervical cell sample analysed at five of the country’s seven pathology departments.”
In total, 15,668 women were included. Almost 95% had received at least one dose of the quadrivalent HPV vaccine at the age of 14 years.
“When it comes to HPV genotyping of biopsies, we do not have data from a birth cohort that was never offered vaccination at all. This means that we have only been able to compare vaccinated and unvaccinated women within the 1994 cohort.”
From screening to tissue diagnosis – where risk is finally decided
In the studies of women born in 1993 and 1994 respectively, participants were followed throughout the national screening programme using Danish health registries, including the population register, pathology registries and vaccination registries.
For women born in 1994, these register data were further supplemented through collaboration with pathology departments on HPV testing of cervical cell samples and subsequent HPV genotyping of biopsies for which the diagnosis was CIN2 or worse.
“It is a strength that we work with histologically confirmed diagnoses,” says Elsebeth Lynge. “We are not just looking at infection but at changes in the tissue that can actually progress to cancer.”
HPV genotyping included both high- and low-risk types. Each diagnosis was classified hierarchically – according to the HPV type most clearly associated with cancer risk. If HPV16 or HPV18 was present, the lesion was therefore attributed to these types, even if other HPV types were detected at the same time.
“The hierarchy reflects our biological understanding of which types drive risk,” explains Elsebeth Lynge. “It is a pragmatic way of handling the many mixed infections we see in clinical practice.”
Among women born in 1994, the risk of developing moderate to severe dysplasia associated with different HPV genotypes was compared between vaccinated and unvaccinated women.
The risk of developing CIN2+ was calculated as incidence rates and hazard ratios using statistical survival models. Among women born in 1994, both the incidence of CIN2+ and the distribution of HPV genotypes were compared between vaccinated and unvaccinated women.
“We have deliberately kept the method simple,” says Elsebeth Lynge. “This is not about calculating risk for each individual woman but about seeing how vaccination shifts the disease pattern in the population.”
The risk is falling – and the diagnosis may no longer mean the same
The results from Trial23 show not only that HPV vaccination reduces the risk of serious cell changes but also that the disease pattern behind the diagnoses has shifted.
“At the overall level, we see a very clear protective effect,” says Elsebeth Lynge.
Among the 15,668 women in the cohort, 584 cases of CIN2+ were diagnosed during the follow-up period. Among vaccinated women, this corresponded to 5.3 cases per 1,000 person-years versus 12.1 among unvaccinated women. The incidence of new CIN2+ cases was thus reduced by 56% among vaccinated women compared with those who were unvaccinated.
The largest difference emerges when looking at which HPV types were actually behind the diagnoses – and therefore at what a CIN2+ diagnosis now tells us about cancer risk.
“This is where the vaccine shows its strength,” says Elsebeth Lynge. “HPV16 and HPV18, which historically accounted for a large share of serious precancerous lesions, have been reduced to very low levels among vaccinated women.”
HPV16 and HPV18 have almost vanished from confirmed diagnoses
The risk of CIN2+ caused by HPV16 or HPV18 was reduced by around 95% among vaccinated women compared with those who were unvaccinated. In absolute terms, only about 4% of serious precancerous lesions among vaccinated women were linked to these types, whereas HPV16 and HPV18 continued to dominate among unvaccinated women.
“This shows that the protection is not just theoretical,” says Elsebeth Lynge. “It can be seen directly in the biopsies where the disease is actually present.”
Among vaccinated women, other high-risk HPV types make up a relatively larger share – particularly HPV31, HPV33, HPV45, HPV52 and HPV58. Together, these types accounted for around half of the remaining CIN2+ cases in vaccinated women.
“The other HPV types may have been there all along,” says Elsebeth Lynge. “But when the most aggressive types are removed, the others become more visible. It is a classic shift in the disease pattern.”
When the most dangerous types disappear, others come into view
For these other high-risk HPV types, a tendency towards lower risk was observed among vaccinated women, but the differences were smaller and not statistically significant. The same was true for HPV types not covered by any of the current vaccines. In addition, there were no signs that low-risk HPV types had taken over as the driving force behind serious precancerous lesions.
“Vaccination does not just change how many precancerous lesions we see,” says Elsebeth Lynge. “It may also change the biological risk they represent.”
Taken together, the results tell a story of major success – but also of a prevention landscape that has fundamentally changed.
“I have worked with screening throughout my professional life,” she continues. “And I have never before seen such a marked biological shift in the disease pattern.”
A success that reshapes the logic of screening
The results from Trial23 not only point backwards to what the HPV vaccine has already achieved. They also point forwards towards a new question: what should cervical cancer prevention look like at a time when the biological basis of the disease has changed?
“We are in a situation in which screening programmes are designed for a different disease biology than the one we see today,” says Elsebeth Lynge. “That in itself is a sign of success – but also a challenge.”
Vaccination against cervical cancer was originally designed with a focus on HPV16 and HPV18, because these types were both widespread and carried a high risk of progressing to cancer.
Now that these types have almost disappeared among vaccinated women, the balance between the benefits and potential harms of frequent screening has shifted – between what is detected and the risk of overtreatment.
The HPV types that now account for just over half of CIN2+ cases among vaccinated women – HPV31, HPV33, HPV45, HPV52 and HPV58 – are also covered by the 9-valent HPV vaccine.
“This vaccine has been used in Denmark since 2018, and those cohorts will reach screening age at 23 years in 2029,” says Elsebeth Lynge. “So we have built a screening programme for a disease biology that today looks different.”
“This does not mean that the programme is wrong,” she adds. “But it does mean that the conditions have changed.”
A less aggressive disease biology calls for a different kind of screening
If the precancerous lesions that still occur are fewer and largely caused by HPV types with slower disease progression and a greater tendency to resolve on their own, this has direct implications for how screening should be organised.
“It gives us reason to rethink both the starting age and the screening intervals,” says Elsebeth Lynge. “The aim is to avoid overdiagnosis and unnecessary interventions – without compromising safety.”
A central perspective in the study is therefore that screening can be more closely aligned with actual biological risk rather than treating all serious findings as biologically equivalent. Vaccination status is not merely an administrative detail but a biological signal of the risk a woman actually carries.
“Vaccination status tells us something fundamental about a person’s risk profile,” says Elsebeth Lynge.
Screening will not stop – but it cannot continue unchanged
The older birth cohorts who were never offered HPV vaccination still make up the majority of women currently included in the screening programme. In addition, the first HPV-vaccinated cohorts – those born from 1993 onwards – have now entered screening with a markedly lower risk, as Trial23 clearly shows.
“In these women, we still see new cases of CIN2+ – around 5 per 1,000,” says Elsebeth Lynge. “That is a level that opens up consideration of a later starting age and longer screening intervals.”
Furthermore, trends in the younger generations who have received the 9-valent vaccine should be followed closely, as these women are expected to develop even fewer precancerous lesions than what we see in vaccinated women today. Our results also indicate that younger women who have not been vaccinated against HPV continue to face a substantial risk.
“The message is not that we can stop screening,” says Elsebeth Lynge. “But that we need to screen more intelligently. And in my view, younger women who have not been vaccinated should now be offered a special approach.”
Denmark as a natural laboratory for prevention
Few countries have had the opportunity to systematically follow an almost fully childhood-vaccinated birth cohort through vaccination, screening and histological diagnosis. That is precisely the opportunity Denmark has had.
“What is special about Trial23 is that we are looking at real women, followed all the way to the tissue diagnosis,” says Elsebeth Lynge. “It is only at that point that you can see what has actually happened.”
Over time, the interaction between vaccination and screening may once again reshape the disease pattern. The HPV types that now account for a relatively larger share of precancerous lesions are covered by the 9-valent vaccine, which has been part of the Danish childhood vaccination programme since 2018.
“Prevention is not static,” says Elsebeth Lynge. “It moves in step with both biology and technology.”
Strategies therefore need to be adjusted continuously as disease risk changes. And that is why a screening programme that was fit for purpose yesterday may be too blunt tomorrow.
