VOL: 98, ISSUE: 24, PAGE NO: 32
Alan Shaw, BA, FIBMS, is chief biomedical scientist, cytology department, Morecambe Bay Hospitals NHS Trust
The basis of the cervical screening programme is the smear test or Pap test, which was named after George Papanicolaou. He was the first person to realise that cells exfoliated from the cervix and picked up in a vaginal smear could be seen under a microscope and used in the diagnosis of uterine cancer. His paper (Papanicolaou and Traut, 1943) encouraged the USA to promote screening for asymptomatic women.
In the 1970s, Walton concluded that squamous-cell carcinoma of the cervix lent itself to control by cervical cytology screening, and that a reduction in the incidence of the disease was related directly to the proportion of women screened (Walton et al, 1976). Although screening had been carried out on a localised basis in the UK for some time, it was not until the national screening programme was set up in 1988 that every health authority became obliged to provide a cervical cancer screening service.
The programme we are familiar with today aims to reduce the number of women who die of cervical cancer by encouraging regular screening. The screen test detects cancers and precancerous conditions, known as cervical intraepithelial neoplasia (CIN) or cervical glandular intraepithelial neoplasia. These conditions can often be treated if they are identified early enough.
Is screening effective?
Three years after the programme started, mortality rates among women began to decrease and have continued to fall at an increasing rate. This could reflect improvements in the programme in terms of tightening up quality checks and laboratory procedures, often in response to the occasional slip-ups that have occurred over the years. Apparently, deaths are now falling at 7% a year. From 1979 to 1995, deaths from cervical cancer fell by 40% (www.cancerscreening.nhs.uk/cervical/risks.html).
The biggest risk factor for cervical cancer is not having a smear test, with almost half of the new cases of the disease being diagnosed among inpatients who have never had a smear test. Screening is not 100% effective, but in results published for screening in British Columbia from 1955 to 1985, the death rate from cervical cancer per 100,000 women fell from 11.4 in 1958 to 3.1 in 1985. Meanwhile, the incidence of cervical cancer fell from 28.4 in 1955 to 6.4 in 1985 (Anderson et al, 1988).
According to the Imperial Cancer Research Fund, cervical screening prevents between 1,100 and 3,900 cases of cervical cancer a year (Sasieni et al, 1996). Among those who are screened, about 19,000 women in England and Wales will have CIN3, the most serious type of abnormality. This can be detected only by screening and requires treatment to prevent the development of invasive cancer.
Is the programme successful?
The effectiveness of the programme may be judged by coverage. In 1988-1989, screening among the target age group of 25-64 was 45%. By 1999-2000 this had increased to 83.7%. These figures seem to point to the effectiveness of the programme.
It is worth noting that before 1994-1995, the calculation was based on a five-and-a-half year period to allow for delays between the invitation and the actual test. Figures are now based on a strict five-year period. In the long term, according to the National Cervical Screening Programme, coverage of 80% among the target age group would produce a reduction in the death rate of 65-70%.
Although the screening programme is open to women aged 20 to 64, those aged 20 to 25 are not included when calculating rates of coverage. It is hoped that this will provide a more accurate picture because women are not necessarily invited for screening as soon as they reach their 20th birthday.
Screening period discrepancies
National Office for Cervical Screening guidelines recommend that screening should be carried out at intervals of three to five years, and it is up to individual health authorities to interpret and implement this. This has resulted in a situation in which health authorities in some areas offer three-yearly screening (about 70%), while others offer tests every five years. There is evidence to suggest that three-yearly screening is more effective than five-yearly screening (IARC Working Group, 1986). This is shown in Table 1.
The staffing of screening laboratories may be a factor in whether smears are offered at three or five-yearly intervals. Recruitment and retention rates for laboratory staff are poor because of low pay and high stress levels. Cytoscreeners need a minimum of two years’ intensive training, while biomedical scientists have to have an honours degree for state registration, and an MSc or equivalent for higher-grade positions that carry considerable responsibility. These positions are becoming more difficult to fill because they offer little financial reward for the responsibility involved.
The recall-period debate
Does the discrepancy in recall periods amount to a postcode lottery? Perhaps it does, but health authorities that use a five-year recall system may have very good reasons for doing so. They may be spending limited money wisely in other areas of health care. A study by Waugh and Robertson (1996) shows that reducing screening from five to three years greatly increases the costs but saves few extra lives.
The Department of Health has said: ’Cervical cancer is a very slowly developing condition and, in terms of saving as many lives as possible, it is far more important to screen as many eligible women as possible than to offer more frequent screening. Our priorities are to ensure that as many women as possible have access to screening and attend on an informed-choice basis.
‘We do understand that some women being screened once every three years and some being screened every five years because of where they happen to live, is an inequality. That is why the Advisory Committee on Cervical Screening has undertaken work looking into this, which it is hoped will inform future discussion. This was commissioned from Cancer Research UK, and the ACCS will discuss the findings at its next meeting scheduled for June.’
Taking a smear
The aim of the person taking the smear is to obtain a representative sample of cells from the surface of the transformation zone, the area of the cervix that is susceptible to precancerous changes, as it undergoes constant physiological changes. The cervix should not be swabbed before the smear is taken. It should be fully visualised and scraped firmly but not roughly. The smear is taken with a spatula or brush.
If a spatula is used, the material should be spread on the non-frosted part of the slide in a linear motion along the longest part of the slide (two sweeps are recommended). If a brush is used, a series of curling loops travelling the length of the slide is preferable. The cells need to be evenly spread on the slide and promptly fixed with an alcohol-based fixative. This will preserve the morphological detail of the cells and their constituents in as life-like a state as possible, while allowing subsequent staining and microscopic examination. Delayed fixation will result in the cells drying out and flattening, making interpretation more difficult.
Several specialist terms are used to describe cervical smear results and nurses should be familiar with their meaning. Patients may have questions about their test results and the implications for their health.
Dyskaryosis refers to abnormal changes in the cell nuclei. All abnormal epithelial cells that look as though they derive from CIN or invasive cancer of the cervix are described as being dyskaryotic. The abnormalities seen in these nuclei consist of any combination of:
- Disproportionate enlargement of nuclei;
- Hyperchromasia (increased staining intensity) and multinucleation;
- Irregularity in form and outline;
- Irregular distribution of chromatin, which may appear as stippling, clumps or strands beneath the nuclear membrane, producing apparent irregularities in the thickness of the nuclear membrane;
- Abnormalities in the number, size and form of nucleoli.
(Nucleoli present in abnormal squamous cells suggest either widespread CIN3, microinvasive disease or invasive disease).
Cervical intraepithelial neoplasia
This is a histological term used to describe the precancerous changes seen in biopsies. The main features are described in Box 1. There are three grades of CIN:
- CIN 1 affects the basal third of the squamous epithelium;
- CIN 2 affects up to the middle third of it;
- CIN 3 neoplastic cells extend into the uppermost third of the epithelium and often replace the entire thickness.
With each increase in grade of severity there is a corresponding generalised worsening of some features of dyskaryosis. As a rule, the abnormalities are graded by assessing the nuclear/cytoplasmic area ratio. In mild dyskaryosis, the nucleus is enlarged but occupies less than half the total area of cytoplasm. In moderate dyskaryosis, the nuclear/cytoplasmic ratio is greater, with the nucleus occupying no more than two-thirds of the cytoplasm. In severe dyskaryosis, the nuclear/cytoplasmic ratio is greater again, with the nucleus often surrounded by only a narrow band of cytoplasm. In addition, there may be abnormal maturation of cytoplasm, including keratinisation and bizarre shaped cells.
Cytology that may indicate invasive disease
Carcinoma cannot be reliably diagnosed from a smear, but there may be strong indications of its presence, including:
- Smears usually show very large numbers of dyskaryotic cells, reflecting the widespread CIN 3 that is generally present with early invasion;
- There is variation in the size and shape of dyskaryotic cell nuclei, more so than that associated with CIN 3, often including very small cells;
- There is hyperchromasia with such coarse aggregation of nuclear chromatin that translucent areas appear between the aggregates (nuclear windowing);
- There are tissue fragments of small, severely dyskaryotic cells;
- There is cytoplasmic keratinisation, including thick anucleate fragments in keratinising squamous-cell carcinoma;
- Large nucleoli are present: these are sometimes irregular and multiple, particularly in large non-keratinising squamous-cell carcinoma;
- The smear background contains necrotic debris, leucocytes and blood (tumour diathesis).
One of the latest techniques in cervical screening is liquid-based cytology, which is widely used in the USA and has been given the go-ahead in Scotland. With LBC, a special cervical brush is used to take the smear. The end of the brush is then dropped into a pot containing fluid, and the pot is then sealed and sent off to the laboratory for analysis. It is hoped that LBC will reduce the number of inadequate smears and the number of false negative results.
The National Institute for Clinical Excellence has considered the evidence on LBC, including a systematic review of the literature commissioned, as part of the health technology assessment programme. Pilot studies have been set up by NICE to evaluate LBC before its possible controlled introduction into the NHS.
Routine testing for the human papillomavirus is being evaluated as certain genotypes of HPV are associated with a high risk of cervical cancer.
It makes sense to link the LBC and HPV trials, as HPV testing can be carried out on the same sample. For this reason, the two methods are being piloted at the same sites. There are three sites - Norfolk and Norwich University Hospital, Royal Victoria Infirmary, Newcastle, and Southmead Hospital, Bristol - each of which has converted to using LBC. They are evaluating one of two systems available in the UK, namely, ThinPrep and Autocyte Prep.
According to the NHS screening programme, the pilots are designed to evaluate all the effects, costs and practical implications of introducing LBC technology, such as:
- The effect on test results (proportions of tests classified as inadequate, negative, borderline/mild dyskaryosis, moderate dyskaryosis, severe dyskaryosis or worse) and the consequent need for repeat screening, recall in less than three years and additional diagnostic investigation;
- The extent to which productivity improvements in cytology laboratories are realised in routine practice, the acceptability of LBC to laboratory staff and their need for training, and the identification of quality-assurance guidance before full implementation;
- The impact in the primary care setting with regard to the training of screening personnel, avoidance of repeat visits and ease of implementation;
- The logistical implications of using LBC, including the transport of specimens, storage, waste disposal and laboratory throughput.
The HPV pilots aim to determine whether this testing technique can be used to prioritise smears that show borderline and mild dyskaryosis. The pilots are evaluating the following issues:
- The extent to which HPV testing in women with low-grade cytological changes reduces the need for colposcopy;
- The positive predictive value of the HPV test in women with low-grade smear abnormalities and the negative predictive value for women with persistent mild dyskaryosis;
- Public acceptability of HPV testing as part of the screening programme. This needs to be seen as a component of an improved cervical screening test and not a test for a sexually transmitted disease;
- The anxiety of patients returned to normal recall after a negative HPV test despite an earlier abnormal smear is being monitored, as well as the anxiety of those whose HPV test is positive and is followed by immediate referral for colposcopy;
- HPV prevalence with low-grade abnormalities in the UK and the impact on a laboratory of starting these tests.
The HPV pilots essentially replace the assessment of low-grade abnormalities by colposcopy with assessment by HPV testing. Patients whose initial smear shows borderline nuclear change or mild dyskaryosis also have material tested for the high-risk HPV genotypes. If they test positive for HPV, they are referred immediately for colposcopy.
Women who test negative for HPV will have a repeat smear in six months’ time, as now, and a second HPV test. If an abnormality has persisted as mild dyskaryosis or progressed to a smear that is moderate, severe or worse, or if a woman is positive to HPV, she is referred for colposcopy. If the abnormality has regressed or is no worse than borderline nuclear change and the woman tests negative for HPV on both occasions, she will be discharged back to routine screening.
Most smear-takers will be pleased to hear that if LBC becomes a reality their job should become easier. This is because the cells would no longer need to be spread on a microscope slide but would instead be placed in a small container for transport to the laboratory. It also means that whatever cells are sampled are then sent to the laboratory rather than being left on the spatula. This will standardise preparation and improve quality control. Although many smear-takers produce quality smears, less experienced staff may have higher rates of inadequate smears.
Many people think an automated screening process is a good idea, but we should be cautious. Some years ago an automated system that picked out the cells it considered to be ‘not normal’ was tested. It then needed a human to interpret the images, a skill that required a lot of training.
Guidelines require each technician to do no fewer then 3,000 smears a year to maintain competence (the rationale being that you can compare many negative smears with the smaller proportion of positive smears, thus helping to maintain reporting standards). Paradoxically, if an automated system was used for primary screening the expertise needed to interpret the images would be lost. This would inevitably mean a change in training methods.
The automated system was never implemented in the UK, but other systems are being developed. Technology improves with time and, provided the correct training is given to maintain quality, anything that improves service provision should be considered.
Free guidelines on all aspects of the screening programme are available from the national office for cervical screening, The Manor House, 260 Ecclesall Road South, Sheffield S11 9PS. Tel: 0114 271 1060; fax: 0114 271 1089; or log on to: www.cancerscreening.nhs.uk