Dr Euphemia McGoogan outlines the recent developments in the detection of cervical cancer
The exact cause of cervical cancer is unknown but an important factor is infection with some types of a virus called the human papillomavirus (HPV). Infection with this virus is extremely common in the population, particularly among younger women. It causes no symptoms and almost all women completely clear the infection without ever knowing they have had it and suffer no consequences. However a few women do not appear to be able to get rid of the virus which eventually inserts fragments of its DNA into the DNA strands of the human cells thus altering the genetic code in these cells, some of which may become malignant (cancer). It takes at least 10 to 15 years for a cancer to develop after infection with the virus. During this period the surface of the cervix looks entirely normal but pre-cancer changes can be identified if cells are scraped from the surface of the cervix and studied under a microscope. This is the cervical smear test. The pre-cancer changes are called cervical intraepithelial neoplasia or CIN. CIN is classified into three grades. Most cases of CIN1 (low grade) get better on their own. About 50 per cent of CIN2 and CIN3 cases are likely to progress to an invasive cancer if left untreated. We cannot predict which cases of CIN2/3 will get better on their own so all cases are treated by removing the surface covering of the cervix. This is done at a Colposcopy Clinic. Colposcopy is simply a process of looking at the cervix using magnifying binoculars. When certain chemicals are applied to the surface of the cervix, the CIN becomes visible. If all the affected tissue can be removed, cancer will not develop.
Many studies from Scandinavia and the Grampian region in Scotland in the 1960s showed that regular screening of a high proportion of the population at risk results in a dramatic reduction in the incidence and mortality from cervical cancer.
treating as many women as possible who develop CIN2/3. This is done by offering all women aged 20 to 65 years of age a cervical smear test at regular intervals. The aim of cervical screening is to sort out those women attending for a smear test who probably have CIN2/3 from those who probably do not.
A basic principle of screening where the vast majority of people participating are entirely normal, is that the chance of benefit must outweigh the chance of harm - both physical harm and psychological harm. Another "truth" is that no screening programme can be 100 per cent effective. All screening programmes have false positive and false negative results. The more one tries to decrease the false negative rate, the more one increases the false positive rate. False positive results cause great anxiety among normal women and may result in unnecessary treatment (i.e. harm). The challenge for those managing screening programmes is to balance false positives and false negatives.
Screening has been implemented throughout the UK since the 1950s but in a rather haphazard way. By the mid 1980s about four million cervical smears were being taken each year but, despite all this activity, the incidence of cervical cancer remained high. On investigation it was shown that less than a third of eligible women were participating in screening, these were mainly under 30 years of age and many were having smears more often than three yearly. Two thirds of women with invasive cancer had never been screened. Indeed over 90 per cent of women over 40 years of age with invasive cancer had never been screened.
The NHS Cervical Screening Programme was re-launched in 1987 with organised call and recall of all women every three to five years. Incentive payments to General Practitioners were introduced in 1990 to encourage the screening of as many women as possible. By 1992 the population coverage had risen to almost 80 per cent and the incidence of cervical cancer fell dramatically. Over the next five years the excellent population coverage was maintained and strict quality standards for each component of the programme were introduced (smear taking, laboratory, colposcopy, programme management) and a further fall in the invasive cancer rate was noted. Indeed by 1997 it had fallen to about half the rate in 1987 and mortality was decreasing at seven per cent per annum. And all this was achieved without a significant increase in the number of smears each year!
Cervical screening during this period was the subject of recurrent ad-verse publicity as quality assurance processes identified a few areas where standards had fallen below an acceptable level. This led to low morale among those staff who were performing well. There was a lack of understanding among the general public and some health professionals that cervical screening can never be 100 per cent effective and every case of invasive cancer was seen as an error or somebody's fault. Indeed some people took the view that someone had caused the woman to develop cervical cancer!
Most attention focused on laboratory staff. Little attention was paid to the smear test itself since there was little understanding of its inherent deficiencies. Analysis of published articles in peer-reviewed literature has repeatedly shown that a single smear test can only identify half the women who have CIN in their cervix (sensitivity). The reason that the cervical screening programme is so successful is that the pre- cancer is present for more than 10 years and women return for at least one if not two smears during that period thus overcoming this low test sensitivity.
The smear test has remained essentially unchanged since the 1930s since it was first described by a Greek gynaecologist, George Papanicolaou, who was doing research into cervical cancer in America. In the USA his name has been given to the test - the "Pap smear". Cells are scraped from the surface of the cervix using a wooden spatula and spread onto a glass slide. The cells are then "fixed" with a preservative fluid or spray to prevent them degenerating in transit to the laboratory. In the laboratory the cells are stained with a series of dyes to make them visible down the microscope. If the preservation is poor, the cells do not stain crisply and they are difficult to assess. All the coloured cells in the smear are then examined under the microscope by a cytology screener who methodically searches the whole slide for the presence of any abnormal cells that might indicate that a pre-cancer is present. There are more than 500,000 cells in an average smear, many of which may not be clearly visible because other cells, pus or blood obscure them. Screening a cervical smear is a highly skilled and laborious task and the efforts of our dedicated laboratory staff go largely unrecognised.
No matter how expert the smear taker, not all the cells removed from the cervix can be placed on the glass slide and a large proportion of the cells are discarded with the spatula. It has been shown that up to 90 per cent of the cells removed from the cervix may end up in the trash can with the spatula. These may include most or all of the abnormal cells taken from the cervix!
False negative results may not be the biggest problem in organised cervical screening programmes. In response to the adverse publicity and in order to avoid false negative smears, many laboratories have increased the false positive rate by increasing the number of borderline and inadequate smear test results. The inadequate rate is now almost 10 per cent, which means that one woman in 10 is recalled for a repeat test since the smear is considered unsatisfactory for assessment. The "emotional harm" to the woman apart, there are serious economic costs associated with increased referral of normal women for repeat smear or further investigation. This adds an ever-increasing burden to primary care teams, recall systems and laboratories. Further improvement in the quality of the cervical smear must begin in the doctor's clinic with improved techniques of specimen collection to allow better quality samples, better slide preparations and possibly also ancillary tests.
It has been suggested that a new technique called "liquid based cytology" (LBC) would provide such an improvement. This technique involves collecting the sample with a plastic broom that is immediately rinsed into a container or vial of preservative fluid. Thus all the cells removed from the cervix are sent to the laboratory as a cell suspension.
In the laboratory excess blood or pus can be removed and a randomised aliquot of cells placed as a thin layer in a small circle on a glass slide using robotic devices. This randomised deposit contains a proportional representation of all the cells removed from the cervix. These preparations result in fewer unsatisfactory results since the cells are well preserved and clearly visualised. Abnormal cells are not hidden in thick areas of the slide even when very few in number or very small in size. These LBC slides are quicker and easier to screen than conventional smears and can increase the laboratory throughput by up to 50 per cent. However there is a steep learning curve for assessment of LBC slides even for very experienced laboratory staff and additional training is required.
Additionally there are sufficient cells left in the vial for further tests on the same sample such as looking for evidence of the types of human papillomavirus associated with cervical cancer, microbiological infections such as Chlamydia or gonorrhoea as well as a whole range of molecular biological or cytogenetic tests that may prove of value in the future.
What is happening in the UK?
The National Institute for Clinical Excellence (NICE) commissioned a literature review and modelling in 1999 to consider the implementation of LBC in the NHS Cervical Screening Programme. This showed that there were no randomised control trials of LBC using invasive cancer or mortality as the outcome and there were no published studies that could provide direct evidence regarding cost-effectiveness of LBC for cervical screening in the UK. Modelling was used to estimate cost implications but this was considered too inaccurate. However the authors concluded that LBC would reduce the number of inadequate samples, reduce number of false negative results and decrease the time required for examination of specimens by laboratory staff. Therefore they recommended a full cost effectiveness study of LBC on a trial of its introduction in a low prevalence population for routine screening to provide more definite information than was possible by modelling studies. In June 2000, NICE announced its decision that three regional pilots should be set up in England by March 2001 to confirm the cost effectiveness of LBC and that these pilots should, in addition, evaluate of use of reflex human papillomavirus testing in routine screening. That meant that all samples showing a borderline or mildly abnormal result would be tested for evidence of the viral types associated with cancer and this information used to manage these women more appropriately. The LBC arm of the pilots was to report by May 2002 and the viral arm by the following year to allow for some follow up studies.
Scotland does not come under the jurisdiction of NICE and set up its own pilots. Four very different laboratories were part-converted to LBC. The pilots ran for six months, included 30,000 routine cervical samples and reported in November 2001. The results showed that the technique was preferred by smear takers and laboratory staff alike. More importantly the inadequate smear result rate dropped to less than two per cent, the high grade rate doubled and the borderline rate was not increased. This was considered an improved quality of screening test for women.
Since some of the robotic devices can handle very large numbers of samples each year, it is recognised that the implementation of LBC offers a unique opportunity for modernisation of laboratory services through rationalisation of the preparation and staining of LBC samples to a smaller number of centres. This would fit well with the imminent introduction into the market of computers that can automatically scan LBC slides and subsequently identify to the microscopist potentially abnormal cells. The use of such technology could increase laboratory throughput by another 50 per cent making centralised preparation and scanning laboratories cost effective.
Malcolm Chisholm MSP, the Scottish Minister for Health, announced his decision in April 2002 that LBC would be implemented throughout Scotland as the method for routine cervical screening by April 2004 and identified £2.75 million to assist with the implementation.
The use of reflex human papillomavirus testing for the management of women whose cervical sample shows a low grade abnormality is much more controversial and Scotland is awaiting the results of the English pilots before considering whether to include this as routine in the Scottish Cervical Screening Programme.
NICE will receive the results of the LBC arm of the English Pilots in September 2002 and their decision whether to follow Scotland in rolling out LBC as a better quality routine screening test is awaited with interest.
