Ep44: Cervical Screening–Less Is More

Ep44: Cervical Screening–Less Is More
12 February 2019

Fellows of the RACP can claim CPD credits via MyCPD for listening to this episode and reading the resources below.

Cervical cancer is the fourth most common cancer in women worldwide, but it’s almost entirely preventable. Incidence in Australia and New Zealand has fallen by half since national Pap testing programs were implemented almost thirty years ago, and it now it sits between 6 and 7 cases per 100,000 women. But this rate has been at plateau for over a decade, and Pap cytology now plays second fiddle to HPV testing. In December 2017 Australia seconded the Netherlands to adopt this as the primary tool in cervical screening, and New Zealand plans to follow suit in 2021.

The most noticeable shift is that the interval between screens will now be five years rather than two. And women will enter the program at age 25 instead of 18. Modeling shows that this could half the incidence of cervical cancer further and at much lower cost than the previous program. Some, however, are concerned about the costs and risks involved in the triage pathway. This episode of Pomegranate Health will answer some questions that women and health professionals might have about the HPV-based National Cervical Screening Program.

Guests: Associate Professor Julia Brotherton AFPHM (VCS Foundation, University of Melbourne), Professor Ian Hammond AM FRANZCOG (Chair, Renewal of the National Cervical Screening Program Implementation Committee, University of Western Australia)


Audio Appendix, including outtakes from interviews with Ian Hammond and Julia Brotherton [RACP]
National Cervical Screening Update, Dr Ian Hammond [Cancer Council]

Correction added to transcript after publication: Since 2017, the nonavalent vaccine is offered free to both girls and boys between the ages of 9 and 26 [NZ Health]

Review Publications

Cervical screening in HPV-vaccinated populations [Canfell, Climacteric 2018]
Screening for Cervical Cancer With High-Risk Human Papillomavirus Testing Updated Evidence Report and Systematic Review for the US Preventive Services Task Force [JAMA 2018]
Cytology versus HPV testing for cervical cancer screening in the general population [Cochrane Library 2017] 
National Cervical Screening Program Renewal: Evidence Review [MSAC 2014]
Point-Counterpoint: Cervical Cancer Screening Should Be Done by Primary Human Papillomavirus Testing with Genotyping and Reflex Cytology for Women over the Age of 25 Years [JCM 2016]
HPV vaccine impact in Australian women: ready for an HPV-based screening program [Brotherton, MJA 2016]
HPV vaccine coverage is increasing in Australia [Brotherton, MJA 2017]
Guidelines for human papillomavirus DNA test requirements for primary cervical cancer screening in women 30 years and older [IJC 2008]

Modeling Analyses

Primary HPV testing versus cytology-based cervical screening in Australia vaccinated for HPV and unvaccinated: effectiveness and economic assessment for the National Cervical Screening Program [Lancet, 2017]
Projected future impact of HPV vaccination and primary HPV screening on cervical cancer rates from 2017–2035: Example from Australia [PLoS One 2018]
Screening for Cervical Cancer in Primary Care: A Decision Analysis for the US Preventive Services Task Force [JAMA 2018] 
Optimal Cervical Cancer Screening in Women Vaccinated Against Human Papillomavirus [JNCI 2017]
Effectiveness modelling and economic evaluation of primary HPV screening for cervical cancer prevention in New Zealand [PLoS One 2016]

Key Primary Literature

Effect of Screening With Primary Cervical HPV Testing vs Cytology Testing on High-grade Cervical Intraepithelial Neoplasia at 48 Months [HPV Focal, JAMA 2018]
Relative Performance of HPV and Cytology Components of Co-testing in Cervical Screening [Kaiser Permanente, JNCI 2018]
Cervical screening with primary HPV testing or cytology in a population of women in which those aged 33 years or younger had previously been offered HPV vaccination: Results of the Compass pilot randomised trial [Compass, PLoS One 2017]
Safety of extending screening intervals beyond five years in cervical screening programmes with testing for high risk human papillomavirus: 14 year follow-up of population based randomised cohort in the Netherlands [BMJ 2016]
Impact of HPV sample self-collection for under-screened women in the renewed Cervical Screening Program [MJA 2016]
Efficacy of HPV-based screening for prevention of invasive cervical cancer: follow-up of four European randomized controlled trials [Lancet 2014]
Long term predictive values of cytology and human papillomavirus testing in cervical cancer screening: joint European cohort study [BMJ 2008]

Other Literature

In a Draft Recommendation Statement for Cervical Cancer Screening, the U.S. Preventive Services Task Force had offered primary HPV testing (five yearly) or cytology (three yearly) but suggested the was cost-benefit analysis for co-testing was not favourable. Co-testing was reintroduced in the final statement published in August 2018, on the back of concerns from some in the cytology and epidemiology communities. A few of these publications are presented here for completeness:

An advocacy victory: final USPSTF cervical cancer screening recommendations revised to include co-testing option [JASC 2018]
HPV screening, invasive cervical cancer and screening policy in Australia; Response and Counter-response [JASC 2018]
The clinical and economic benefits of co-testing versus primary HPV testing for cervical cancer [J Womens Health 2016]
Molecular testing and cervical screening: will one test fit all? [JASC 2016]
Sensitivity, effect and overdiagnosis in screening for cancers with detectable pre‐invasive phase [IJC 2014]
The HPV test has similar sensitivity but more overdiagnosis that the Pap test — a randomised health services study on cervical cancer screening in Finland [IJC 2013]


Written and produced by Mic Cavazzini. Recording assistance in Perth from Meri Fatin and in Melbourne from Jon Tjhia of Paper Radio. Music courtesy of Free Music Archive; ‘Electro Cool’ by 4T Thieves, ‘Fryeri,’ ‘Headway,’ ‘Brand New World’ and ‘Mare’ by Kai Engel. Image courtesy of Wikimedia Commons. Executive producer Anne Fredrickson.

Editorial feedback for this episode was provided by members of the RACP’s Podcast Editorial Group Alan Ngo, Phillip Gaughwhin, Andrea Knox, Rhiannon Mellor, Michael Herd, Marion Leighton, Paul Jauncey and Adrienne Torda

Thanks also to Debra Letica, Melissa Cadzow, Hamza Vayani from the RACP’s Consumer Advisory Group and Megan Smith from Cancer Council NSW.


MIC CAVAZZINI: Cervical cancer is the fourth most common cancer in women worldwide, but it’s almost entirely preventable. Incidence in Australia and New Zealand has fallen by half since national Pap testing programs were implemented almost thirty years ago, and it now it sits between 6 and 7 cases per 100,000 women. But this rate has been at plateau for over a decade, and Pap cytology now plays second fiddle to HPV testing. In December 2017 Australia seconded the Netherlands to adopt this as the primary tool in cervical screening, and New Zealand plans to follow suit in 2021.

Some investigators predict HPV-based screening together with vaccination could halve the incidence of cervical cancer further. Others are worried about abandoning tried and true methods, and that new triage pathways will be costly. Today on Pomegranate Health, we’re going to answer some questions that women and health professionals might have about the new program. I’m Mic Cavazzini, from the Royal Australasian College of Physicians.

Even if you have nothing to do with gynaecology, this story is a really interesting example of how to weigh up clinical evidence and take brave decisions that will affect the health of millions. Which measures matter? And how do you change long-established clinical habits? The most noticeable shift is that the interval between screens will now be five years rather than two. And women will enter the program at age 25 instead of 18.

“How can that work?” I hear you say. Well, in the classic cytology pathway, cells collected with a Pap smear are examined for abnormalities that suggest some lesion is present in the cervical epithelium. But the sensitivity of the technique is low, which is why Pap tests are repeated every two years, so that any missed cases can be picked up at the following screen.

Almost every cervical lesion has its origins in a sexually-transmitted infection of human papilloma virus. HPV infections are as prevalent as the common cold and are usually cleared from the body within a few years. But sometimes viruses remain hidden from the person’s immune system and go on to insert themselves into the genome of fast-dividing squamous cells at the cervical transformation zone. This can disrupt two well-known oncogenes, which then allows cell division to go on uninhibited. The progression to invasive cancer takes one or two decades, which is why HPV screens can be conducted five years apart. Note, however, that women already familiar with the old program should have their first HPV test two years after their last Pap test.

There are forty or so variants of HPV that infect the anogenital areas, but most are completely innocuous. Of the fourteen strains classified as oncogenic, HPV types 16 18 account for over seventy per cent of cervical cancers. These variants are associated a 15 to 20-fold higher incidence of cervical cancer. That’s to say the risk conveyed by infection with oncogenic HPV is greater than risk for lung cancer brought on by smoking. But these highly oncogenic strains are rare, and later we’ll hear about the incredible impact of the HPV vaccination program.

For women undergoing HPV testing, the experience will be very similar to what they’ve been used to. A smear is taken from the cervix, but now the sampled fluid is run through a machine that can detect viral DNA. Only the oncogenic variants are tested for, then positive results are followed up by more familiar techniques to examine for abnormalities.

Before the new program was rolled out, there was concern among GPs, nurses and members of the public about the safety of a delayed starting age and the longer interval between screens. Public health physician Julia Brotherton explains the logic behind these changes. She is Medical Director of VCS Population Health at the VCS Foundation. She’s also an Associate Professor at the University of Melbourne.

JULIA BROTHERTON: Yeah, look, I think it’s a huge undertaking to turn around what has been an incredibly successful Pap-testing program that we’re all really familiar with. And this really is a monumental shift in thinking, if you like. When we started doing Pap-testing, we didn’t know that HPV was causing those changes in the cells that we were looking for. And now that we understand the natural history, and we understand that it is persistent infection with high risk types that causes those changes, we’re actually a lot more accurate in looking for the virus itself at those high levels that we know are associated with the disease.

So, the HPV-based screening that we now have in Australia, it is really not looking for any molecule of HPV. Because I think patients and clinicians both get confused about that. They go, “Well, HPV’s really common, how can we possibly use it as a screening test?” You’re absolutely right, it is really common. And that’s why the HPV tests we use are very carefully calibrated to be looking for levels of HPV of the most oncogenic types, that do have a proven predictive value for those high-grade lesions we’re trying to detect through screening.

If we have a look at the typical prevalence curves for HPV in a female population, in the years following sexual debut, almost everybody gets HPV. Cross-sectionally, the prevalence is around 70 per cent. And most women clear that, they clear any low-grade, or even high-grade lesions associated with that, and then their body generates sufficient immunity that it’s no longer an issue for them through that woman’s lifetime.

We have been treating a lot of women with low and high-grade lesions who are very young. That really was over-treatment. And we have really great data from several countries showing that, in fact, all this screening we’ve been doing in women under 25 hasn’t been doing anything. It doesn’t seem to have any protective effect against development of cervical cancer. Cervical cancer is incredibly rare in women under 25, and when it does happen it’s very aggressive— probably not screen-preventable at all.

So with that, and of course in the context of the mass HPV vaccine program, and the fact that those types 16 and 18 are now incredibly uncommon in our young women, it really is a safe and appropriate thing to do to start screening between 25 and I really do want to emphasise, though—that some women will have potentially had a persistent HPV infection for ten years by that time, and so I think it’s a great thing that every woman can do for her health.

MIC CAVAZZINI: And as you said, the peak in infection rates is early twenties, let’s say, whereas the highest rate of incidence of cervical cancer is in a woman’s forties. So, there’s still plenty of time to intervene before the pathology?

JULIA BROTHERTON: Absolutely. That’s correct. So, the natural history of the development of cervical cancer is I guess fortunately for us, reasonably forgiving. And that of course is why, doing a Pap test every two years has allowed us to prevent cancers, even if it’s not been that sensitive. If we miss it this time, it’s still likely that we’ll detect it the next time or the time after that. It’s usually decades before it has a chance to develop into cancer. But, of course, it doesn’t happen in all women. Probably of women with the highest grade of precancerous lesions, the CIN 3s, about 30 per cent of them would go on to develop cervical cancer within about 30 years.


MIC CAVAZZINI: To be approved for use in Australia, HPV assays must be able to detect the thirteen most oncogenic strains. If a test is positive for HPV 16 or 18, the women is referred for colposcopy. If any of the other strains are detected, cells collected on the same smear undergo cytological testing instead, to avoid unnecessary invasive examinations. Classically, samples would be prepared on a microscope slide and treated with a Pap stain. The technique is named after Georgios Papanikolaou, who described the link between cellular abnormalities and underlying lesions back in 1928. Though spare a thought for Romanian physician Aurel Babeş who independently published very similar findings in the same year.

In most modern labs, liquid-based cytology is now performed—this is a more automated procedure that reduces user variability in sample preparation. If abnormal cells are observed in this triage step, colposcopy is also performed. This involves examining the surface of the cervix with a magnifying device to look for signs of disease. In the past, lesions were assumed to be precancerous, and labelled cervical intraepithelial neoplasia grade 1, 2 or 3, depending on how much of the epithelium wall was affected. Nowadays it’s understood that almost all grade 1 abnormalities resolve by themselves. They’re simply the result of viral replication and cell lysis and are now called low-grade squamous intraepthelial lesions or LSIL. The term HSILs, for high-grade lesions, pools together the categories CIN 2 and 3, and does away with idea that stage progression is inevitable.

Back to the HPV assay. When we say that it’s a very sensitive clinical tool, we’re talking about its ability to detect high grade lesions. So the sensitivity of commercially available assays ranges from 90 to 98 per cent. To get this number, imagine a hundred vials, each containing a piece of biopsy tissue that’s been histologically confirmed to have an HSIL. A sensitivity of 95 per cent, say, means the assay will come up positive on 95 of these samples. The five missed cases are what you’d call false negatives. Now compare this to liquid-based cytology which averages 73 percent for sensitivity, and classic cytology which scores about 67 percent, according to a 2017 Cochrane study.

So, HPV testing doesn’t miss as many cases as cytology, but the flipside is that when presented with healthy specimens, it has a comparatively high rate of false positives. I’ll leave further explanation to gynaecological oncologist Professor Ian Hammond AM. After retiring from practice at King Edward Memorial Hospital in Perth, he chaired the steering committee for designing and implementing the renewal of the National Cervical Screening Program.

IAN HAMMOND: I think the important thing here is the differential. That you’ve got a test that is vastly superior with HPV compared with cytology. In fact, although you’ve quoted 70 per cent plus for cytology, that’s very laboratory dependent, and in fact some laboratories have rates of 45 to 50 per cent compared with well into the 90s for HPV.

MIC CAVAZZINI: Yeah, and so a super-sensitive threshold might read positive for every sample, and sure you’d detect every real lesion correctly, but you’d also flag up a lot of duds.

IAN HAMMOND: Oh, it would be crazy, yeah, it would be crazy.

MIC CAVAZZINI: So ‘specificity’ is the percentage of healthy women who are correctly identified as not having a lesion.

IAN HAMMOND: Correct. And HPV’s not great, by the way. HPV specificity is not as good as cytology. But that’s why we add cytology to the mix, once you have the positive HPV test. I mean, the sequence of things really is women will have an HPV test, and if they’re found to have oncogenic HPV, they all have cytology done. About 8 per cent of all of the tests will be positive for HPV, and the higher rate in younger women. But they will have the reflex liquid-based cytology. They are triaged according to the cytology report. So, if the triage comes back saying ‘negative cytology’, or ‘low-grade cytology’, then those women are asked to come back in 12 months to see their GP to have another HPV test to see if the HPV has gone away. And 80 per cent would have gone away within 12 months. I mean, 98 per cent of HPV goes away within five years.

And actually, when you look at the old program, we used to do exactly the same thing without having the HPV test. We had a low-grade abnormality on cytology, and then we repeated the test in 12 months. That was shown to be quite safe, a very safe thing, because that was monitored very closely for some years. So, the cytology is a predictive thing, and if they’ve got high grade abnormalities, or a glandular abnormalities, colposcopy is essential to make the diagnosis with the biopsy.

MIC CAVAZZINI: Because you need tissue, you can’t just …

IAN HAMMOND: —absolutely, absolutely, yeah.


MIC CAVAZZINI: Modern HPV assays are predictive not just of squamous cell cancer but also adenocarcinomas, which are typically missed by classic cytology screens. Now the numbers for sensitivity and specificity we’ve been talking about are derived from validated samples at a given point in time- what you’d call cross-sectional studies. But to measure the real-world predictive value of a screen you need longitudinal studies on large numbers of women.

A seminal example is the 2008 Joint European Cohort Study which followed over 24 thousand women recruited through screening programs in six different countries. For women who tested negative for HPV on entering the study, the incidence of CIN grade 3 six years later was only 28 percent of what it was for women who had been cytology negative. A negative HPV test is also more reassuring than a negative Pap test because the likelihood that it’s a false negative is much lower.

IAN HAMMOND: Now, I thought you might mention this, so I actually printed the slide which shows the graphical representation that you’re talking about. And if you look at the results, at two years, the group that was cytology negative have a rate of 40 per 10,000. But at five years with HPV they have a rate of 20. So, I say the HPV test is twice as good at five years as the cytology is at two years. And I reckon that’s more understandable actually to people. And I think this is actually the study that really provides tremendous support to the five-year interval. And that should be reassuring to most people.

MIC CAVAZZINI: But it is reported, however, that 10 to 15 per cent of invasive cancers are not HPV positive. Does this mean that the causal infection has since been resolved, or that perhaps the cancers were caused by mutagens other than viral insertion?

IAN HAMMOND: Well, I actually dispute that data. I know that’s reported by some studies in some areas. But in fact the original work—Walboomers’ study, way back now—said that 99.7 percent of all cervical cancers had HPV. Well, that might be a bit enthusiastic. But there are some cervical cancers that you don’t find HPV. It could be it was a hit-and-run episode, that’s possible.

There are some of course that occur in the cervix, they’re not necessarily squamous, they’re not glandular, they’re not adenosquamous, they’re unusual histological types, and they definitely don’t have HPV. And they weren’t found to be effectively protected by any program, they were difficult to detect even by cytology. But what was very interesting as you mention that study of course is they also looked at women who were both negative for HPV and negative for cytology, and they did have a slightly lower rate, but not one that was really clinically remarkable.

MIC CAVAZZINI: And this was a polemic or fear in the U.S. literature, that if you switched to an HPV testing program you’d skip over a bunch of prevalent cancers that aren’t HPV positive. But it was sort of put to bed in a large part by that massive Kaiser Permanente cohort, of over a million women. And their figure is that 5.9 per cent of invasive cancers and 3.5 per cent of pre-cancers had been HPV negative and cytology positive. But they conclude that given the rarity of cancers among screened women, the contribution of cytology to screening—co-testing—translated to earlier detection of at most five cases per million of women per year.

IAN HAMMOND: Correct. And certainly not cost effective if you were doing things at a government-based publicly funded program.


MIC CAVAZZINI:  So, it looks like a few unusual cancers could slip past an HPV screen, but that more high-grade are be missed by cytology screening. How do you weight these numbers up in a practical way? The proof of the pudding is in the randomised-controlled trial, where you measure incidence of disease following each intervention. Up to 2012 there were four major trials conducted in Italy, the Netherlands, the U.K. and Sweden. These consistently showed that primary testing for oncogenic HPV found high grade lesions at two to three times the detection rate of primary cytology. None of those RCTs was powered to show a reduction in incidence of cervical cancer, but a pooled analysis in The Lancet indicated that HPV screening provided 60 per cent greater protection from cancer after two and a half years.

This was corroborated by findings of the Canadian HPV Focal Trial published in JAMA last year. In that study, women were randomised to either an HPV or a cytology entry test, and then triaged and treated accordingly. Two years later they underwent exit screening with both tests, and women in the HPV-screening group were found to be better protected against high grade lesions. If we round the two cohorts up to 10,000 women, we see that 76 of them were diagnosed with HSIL at follow up, compared to 161 in the cytology group.

But to show a real-world impact on cervical cancer is tricky, first because baseline incidence is already low in developed countries and second because of the slow progression of disease. Instead, simulations are designed that crunch the numbers for test sensitivity along with other variables like screening intervals, attendance rates, age of entry and vaccine coverage. Researchers from Cancer Council NSW and the VCS Foundation performed a computational analysis for the Australian setting comparing different parameters and triage strategies. For example, one possibility under consideration had been primary liquid-based cytology every three years followed by reflex HPV testing. This is current practice in New Zealand.

The researchers looked at a hundred and thirty two different models and the one adopted in the renewal program was not only one of the most effective, but could be provided at three quarters the cost of the old biennial Pap test program, a saving of 50 million dollars per year. When the simulation was projected to 2035, incidence of cervical cancer was predicted to fall to just over 3 cases per 100,000 women from its current level at around 7. 

HPV vaccine coverage and screening attendance are two of the most important variables that affect the impact of a cervical screening program. In 2007, Australia was the first country in the world to implement nation-wide vaccination. At that time, the vaccine protected against HPV types 16 and 18 only, but from last year, kids receive a nonavalent vaccine that should protect against 90 per cent of cervical cancers. It also vaccinates against HPV 6 and 11 which are not oncogenic but do cause genital warts.

The Australian program targets girls and boys from the age of 12 or 13, while in New Zealand only girls are currently included [Correction added after publication- Since 2017, the nonavalent vaccine is offered free to both girls and boys between the ages of 9 and 26]. Vaccine coverage in NZ was reported at 55 percent for all eligible cohorts in 2015. Julia Brotherton describes where Australia has now gotten to in regards to vaccine coverage and screen participation.

JULIA BROTHERTON: We’re now sitting for both boys and girls up near 80 per cent. Now that’s actually very significant, and the reason that I say that is because HPV is very different to other viruses like measles et cetera that are airborne. Because it’s an STI, it’s actually a lot easier to control, and there’s a very nice meta-analysis of different transmission models, and they found that if you can sustain coverage over 80 per cent in males and females, you can eliminate transmission of type 16, which of course is the most oncogenic, within a 70 year time horizon.

So Australia’s really on track to one day not have HPV 16 in our population. Now of course we’re not a closed population, we’re a nation of immigrants after all, so it does depend very much on the scale up globally. I mean, we really want to live in a world where every young girl receives HPV vaccine.

MIC CAVAZZINI: There’s a study in The Lancet reporting that the crude prevalence of the big four HPV types in Australia dropped from 29 per cent to 7 per cent between the year 2005 and 2007 in three major metropolitan areas, and another Australian report found that by 2013 only 1.6 per cent of sampled swabs were positive. But the oldest recipients of that first wave of vaccination will now be 38 years old.


MIC CAVAZZINI: That’s too early to assess the real impact of a vaccination on incidence of cervical cancer. But tell us some of the other clinical measures of the program’s effectiveness that you’ve collected with the VCS.

JULIA BROTHERTON: Yeah. Look, I think the most significant finding really has been that impact on the high-grade lesions that we detect through screening. We always had the highest rates of those CIN 2 and 3 lesions in our 20-to-24-year-olds. But within a few years of the program they were dropping so dramatically that the peak rates were now in 25-to-29-year-olds, even by the end of 2009, and we published the world’s first data in The Lancet looking at that. And as time’s gone by, we can see that in the 30-to-34-year-olds now that those cohorts are aging through.

MIC CAVAZZINI: Let’s briefly chat about the Compass trial published in PLoS Medicine in 2017, an Australian study. They reported that CIN lesions of grade 2 or higher were identified the times more frequently in the group randomised to HPV screening compared to the conventional pap screening. But the novel finding they say, the novel finding of the study was that it doesn’t make a difference whether women are vaccinated or not. Does that safety margin change very much depending on the vaccine coverage?

JULIA BROTHERTON: The Compass trial is really the first international study to really assess HPV-based screening in a heavily vaccinated population—in a low HPV prevalence environment. It was predicted for a long time that cytology-based screening would start to be less accurate. And in fact we’ve just published a paper in the International Journal of Cancer in the last couple of weeks showing exactly that. Showing that in the Victorian registry data we’ve started to see a drop off in the positive predictive value of a cytology for predicting high grade disease in our young women.

And the data suggest that the tools that we have for HPV-based screening are more robust to those changes in prevalence than the cytology-based screening. You know, the wonderful thing about HPV testing, is that in fact, it applies equally to both vaccinated and unvaccinated women—it’s telling me her risk right now. So, whether she was only vaccinated after sexual debut, if she didn’t complete the course, if she wasn’t in the country at the time and missed out, whatever the reason, it gives me an indicator—a very good indicator—of her risk right now. And it really simplifies what we as practitioners need to do, rather than having for example different algorithms for whether you were vaccinated or not vaccinated.

MIC CAVAZZINI: So about 80 per cent of cervical cancers in Australia arise in women who have never been screened or don’t attend regularly. Before we get into the socio-cultural factors about that, I want to talk about the role of the registry. What is the participation rate in screening programs in Australia and New Zealand, and how does a registry help with that?

JULIA BROTHERTON: The two-year program we’ve been tracking around about 60 per cent, even a little bit lower, for that two-year attendance. That may lead you to think there’s 40 per cent of women who aren’t screening at all. In fact, that’s not the case. If you have a look at how many people come every three or every five years, you’re getting up towards 80 per cent of women. So, you know, women will go and do it, they just haven’t been doing it as often as recommended, they’ve been putting it off.

MIC CAVAZZINI: Oh, so that answers my next question. In the models they did estimate that a five-year program would have an 83 per cent participation rate.

JULIA BROTHERTON: Yeah. And I think that’s very evidence-based on what we might have been doing historically. We all know it’s quite challenging to convince women to come in and have a speculum inserted and a sample taken. It’s certainly not any woman’s favourite activity. And also, I have to say we’re a little bit victims of our own success. Cervical cancer is not common anymore. You no longer know someone on your street or in your community or in your family with cervical cancer. So, I think participation very much has been, and I think will be, an ongoing challenge for our programs.

And I think the other thing is to recognise that for some of our most vulnerable groups it’s really hard, and it’s just not on their radar. So, if they only needed to do it once every five years rather than every two, it’s much more likely that we can find a way to fit that into the woman’s life and make it accessible to her. So, it gives us a great opportunity to try and reach women who we’ve never been able to reach before.

MIC CAVAZZINI: Aboriginal and Torres Strait Islander women have twice the incidence of cervical cancer compared to non-Indigenous women, and their participation in screening is only at about 34 per cent. What are some other under-screened populations in Australia and New Zealand?

JULIA BROTHERTON: Look, I think we know that there are also some culturally and linguistically diverse communities where cervical screening is not a normal thing that’s offered in their country of origin, and a lot of that initially at least translates. We do know that for most migrant groups within about a decade they do start to take up the screening rates of the general Australian population. There are certainly some communities where there are layers of cultural issues that may not be at first obvious to us, factors around potential shame or issues around the collection of the test and not wanting a male practitioner.

And one of the most exciting things of course is that HPV testing offers the opportunity for self-collection, which is basically literally it’s a vaginal sample the woman takes with a cotton swab, and I think we’re very hopeful that that is going to really I guess empower women to feel that in fact they’re in control of their body, and that they can participate without needing to have a speculum exam, which for some women, whether it’s because of past sexual trauma, or bad experiences or whatever, is just something they’re not prepared to do.

Women who are not screening—there are so many reasons why they’ve never participated. And it’s really a conversation with a trusted provider that’s ever going to change that. And in our pilot study in Victoria looking at homeless women, drug-using women, some migrant and refugee women, who have known barriers to screening— we found in that context 85 per cent of them would do a self-collection. And that’s an absolute game-changer.

And I think we also know one of the main questions women have about doing the self-collection is women are worried about doing it right. And that’s what we found when we mailed out the packs, they were worried, “Have I done this correctly, and is it as good as a doctor’s test? “And the great thing now is that we can tell them that it is.

So there was a meta-analysis published last month in the BMJ showing that as long you use a PCR-based test, self-collection with a vaginal sample is just as good for detecting the presence of high grade lesions as a clinician collected sample. Particularly women who have never, ever screened, it could save your life. And then I feel confident that in the next five years we’re going to be looking more closely towards mainstreaming self-collection. We’re talking about a context where in 20 years’ time, we’re going to have young women who have received a nonavalent vaccine. Those women may very well only need two screens in their lifetime.

MIC CAVAZZINI: Just a final reflection on the program, again with Ian Hammond, we discussed all of the tweaks and variables that went into it. We discussed the screening rates and the initiation age. You can always make things “safer,” and I’m doing air quotes, by testing more often and at lower thresholds. I like how this is expressed by Mark Schiffman, a cancer epidemiologist at the National Cancer Institute in the US, who was behind the Kaiser Permanente cohort study. He says that, and I’m quoting here, “The choice of acceptable levels of safety compared with the disadvantages of frequent intensive screening is not a scientific question, rather it requires a societal and consensus or compromise.” And in different countries, well, is it a cultural difference, that in the Netherlands they’re talking potentially even a 10-year interval would be safe.

JULIA BROTHERTON: Correct. Yeah, I agree at that level, but I think it’s really important that we communicate to people that doing HPV testing more frequently than five years is just not supported by evidence, and it’s going to have a lower predictive value for disease. So, you’re not going to do yourself any favours, you’re going to over-refer, worry people, there’s a limit to that. And I agree, pulling back further absolutely is a question for society. But I think we just have to be careful not to say that more frequently is always better.

MIC CAVAZZINI: No, and I think those comments were rather on the very negligible benefit of contesting not more frequent testing, but in another JAMA editorial ,Professor Philip Castle at the Albert Einstein College of Medicine notes that primary HPV screening in a population well vaccinated against oncogenic HPV types provides a risk reduction of 98 to 99 per cent. But he writes, “Attempting to exceed those maximum risk reductions must be weighed not only against the costs of doing so, but also against the potential psychosocial impacts and medical harms associated with excessive screening and treatment of pre-cancerous lesions.”

JULIA BROTHERTON: Yeah, Phil is a very wise man, and he often says that perfection is the enemy of good, and I totally agree with him on that one.

MIC CAVAZZINI: Yeah, yeah, it was a great quote, perfection is the enemy of good, or in this case the very good.



MIC CAVAZZINI: The USA does not have a national screening program, and there has been much debate about the safety of giving up co-testing with cytology. But since August last year, the US Preventative Services Task Force has enthusiastically supported primary HPV testing in its recommendation statement. Researchers from the organisation conducted a systematic review of all the trials to date which found no compelling advantage to co-testing.

But one worrying finding of the analysis, published in JAMA, was that when HPV testing is used for primary screening, more women are referred for colposcopy—as a proportion of all women screened, the rates were 8 per cent for HPV screening versus 3 percent for cytology. This has generated some handwringing about over-testing and overtreatment.

No doubt clinicians want to avoid a repeat of the PSA story—where reported incidence of prostate cancer in men increased by over 40 per cent after the introduction of that new diagnostic marker. The profession was roundly scolded for a parallel increase in possibly unnecessary surgical interventions and complications. Ian Hammond addresses some of the concerns about the referral pathway in the renewed cervical screening program.

IAN HAMMOND: Well, yeah, I’ve been actively involved in this, because our committee has considered this information. In Australia it was predicted that we would have an initial rise in colposcopy over the first two, three years, and then it would fall actually to below the current levels, and that’s what we would expect. What we actually found, it was quite disturbing, was actually 150 per cent in some areas. And the reason for this wasn’t actually to do with the sensitivity of the test alone, at all. And this is your first question in our discussion today. It was to do with the fact that there’s a proportion of doctors who are not particularly accepting or still concerned, and they want a belt and braces approach, that they did both tests. And they weren’t really allowed to do co-testing.

And so, what they have done, there are specific indications for co-testing in our program. And one of them is symptomatic women, women who have symptoms that could be attributable to cervical cancer. Now by that we mean postmenopausal bleeding, persistent unexplained intermenstrual bleeding. I mean, many women get a bit of spotting in between their periods or occasionally after intercourse, but it’s not persistent. What was happening is the doctors were writing down ‘symptomatic co-test.’ And therefore, these women automatically went for colposcopy almost, irrespective of what happened in the laboratory testing.

MIC CAVAZZINI: Because the primary cytology is no longer covered by Medicare?


MIC CAVAZZINI: So, you need to have one of those indications in order to order it.

IAN HAMMOND: The other thing is of course—and women under 25 are not supposed to be tested. Well, the reality, 1,000 women every week were being tested. So, I think there was some cynicism about it all, and a bit of misuse of the referral system—personally, that’s my personal view, it’s not the government view.

MIC CAVAZZINI: Well, without being cynical, it’s an interesting behavioural observation, isn’t it?

IAN HAMMOND: Yeah, it’s a behavioural thing, and listen, I understand it. But the reality is that’s halved now, that’s settled down quite significantly.

MIC CAVAZZINI: And that was found in the most recent Canadian HPV focal trial. Yes, colposcopy rates were higher for the first round of screening, but by four years it was the other way around, the cytology randomised group had a higher rate, and the numbers evened out. I’m not sure I entirely understand that, but is it also a difference between …

IAN HAMMOND: —no, no, I think I can explain it. I mean, the reason is when you first put a test out to a test-naive population, then you’re going to pick up all the women that are sitting there with HPV with nothing on cytology. And they will suddenly go into the mix. But as you move down the track, those women will have been triaged, and eventually you get a steady state. The people who had the cytology only, they get these minor abnormalities that are persistent, possible LSIL, the old CIN 1 or less than CIN 1, atypia. If you were a woman and you had a couple of years of low grade abnormalities, I reckon most women would turn around and say, “Listen, I just want to make sure there’s nothing there.” And the doctors felt the same. Even though the pickup rate, the actual detection rate, was incredibly low. So, I mean, that’s why the cytology arm will continue to show higher rates.

MIC CAVAZZINI: And women should take comfort from the fact that even the highest grade CIN 3 lesions, 65 per cent of these will not progress over 30 years.


MIC CAVAZZINI: Well, the question is what do we call the 65 per cent of CIN 3 that don’t progress? Do we call these false positives, or an example of over-detection, what do we do with them?

IAN HAMMOND: Well, I suppose the problem is, it’s like any abnormality. I mean, I think if you say to a woman sitting across the desk, “We have found an abnormality on the cervix which we call pre-cancerous, and we know that let’s just say 50 per cent will go on to cancer over the next few years, and 50 per cent won’t. But we have no idea which ones will or won’t.” I could tell you there’s not a woman on the planet who will not have it treated.

I must say in my own career, I had one woman only, she was a lovely person, and she believed that she could use herbs and other things to control this disease. And we observed her for three years, it did not go away, and then she allowed me to do a cone biopsy of the cervix and remove the abnormality, following which all cells were back to normal and she was very grateful.

MIC CAVAZZINI: And Brian Cox, who’s an Associate Professor at the University of Otago, he’s a sceptic in the sense that he says that classic Pap screening picks up proportionally more progressive lesions than HPV testing.

IAN HAMMOND: Well, they’re more obvious lesions, they have to be.

MIC CAVAZZINI: And so less false, fewer false positives. And therefore, is going to lead to less over-treatment, whether it’s biopsy or electrosurgical excision, that there’s a fourfold greater frequency of miscarriages in the second trimester after one of these procedures, more premature deliveries. So yeah, do you think that question can be answered mathematically, or is it more …

IAN HAMMOND: —well, I don’t know, I’m not a mathematician, but I can answer it from a practical sense anyway. I mean, I think the point is, I mean, he’s correct in the sense that yes, if you find a high-grade abnormality straight away on cytology, yes, you’ve found a high-grade abnormality. But cytology also finds a vast number of low-grade abnormalities that have no malignant potential whatsoever, not even CIN 1. You know, if a woman got repeated minor abnormalities, they would say, “Look, can you reassure me that I won’t get a problem?” And you can’t actually, because the cytology’s not sensitive enough. So, these women often had treatments, totally unnecessarily.

MIC CAVAZZINI: And if we had a biomarker that identified those lesions that would progress, we wouldn’t be having this conversation.

IAN HAMMOND: That would be fantastic, of course it would. And in fact, we used to treat all of these women who had CIN 2 and CIN 3. And there was a lot of discussion about whether CIN 2 is a reproduceable abnormality. But in the new order, women who are diagnosed as HSIL, will actually have P16, which is a marker, and if it’s positive for P16, then it’s much more likely to be a real pre-cancerous abnormality, they’re more at risk of progression to something if left untreated for a period of time.

What we did say about Cox’s paper, respectfully, their approach rested on the assumption that a negative HPV test provided the same protection over time as a negative cytology test. I mean, you yourself gave me the evidence from the European cohort study. The big thing about HPV testing, from my point of view anyway, is the negative predictive value of the test. That when it’s negative, the chance of you developing cancer of the cervix in the next five years is exceptionally low, and significantly better than any cytological studies anywhere in the world. And that’s been borne out everywhere in the world. And to me, that’s what I–I’m happy to go to bed at night thinking, “Well, this is safe for the women of Australia.” I have a young daughter, and she’s in her 20s, and I’m happy that she will be well-served by the program that we have going forward.


MIC CAVAZZINI: Thanks to Julia Brotherton and Ian Hammond for contributing to this episode of Pomegranate Health. There is a lot more fascinating material from these interviews attached in an audio appendix at our website racp.edu.au/podcast. The views expressed are those of the guests and may not represent those of the Royal Australasian College of Physicians. Julia Brotherton’s research at the VCS Foundation has been partly funded by Roche Molecular Systems along with NHMRC. Ian Hammond received travel allowances from the Commonwealth government in his role as chair of the implementation committee for the renewal program but has no other declared interests.

All the material discussed today is supported by references that can be found embedded throughout the transcript. Our website lists of the some of the main ones, starting with a review by Associate Professor Karen Canfell, one of the leads on the Compass Trial conducted between Cancer Council NSW and the VCS Foundation.

If you found today’s episode informative, subscribe to Pomegranate Health via any podcasting app, or the mailing list found at our web page. Please share with friends and colleagues, and feel free to send some feedback to podcast@racp.edu.au. I’m Mic Cavazzini, thanks for listening


Be the first to comment on this Podcast!

Thank you for posting your comments

23 May 2019
Close overlay