Abstract
Colorectal cancer is the third most common cancer, and the second leading cause of cancer death in the UK, with 35,000 diagnoses and 16,000 deaths per year. It costs the National Health Service (NHS) expenditure of more than £300 million pounds in surgical, adjuvant and palliative treatment. The disease is uniquely placed for screening strategies, as 70–90% of the tumours arise from premalignant, adenomatous polyps. This transformation may take 10 years or more and provides us with a window of opportunity to detect early tumours (Dukes' A) and also prevent cancer development by excision and removal of premalignant polyps during colonoscopy. The NHS Bowel Screening Programme in the UK is based on the principle of secondary prevention whereby detection and treatment of existing preneoplastic and early neoplastic lesions would result in reduction of prevalence and improvement in survival of patients with colorectal cancer.
Background
Colorectal cancer (CRC) is the third most common cancer, and the second leading cause of cancer death in the UK with 35,000 diagnoses and 16,000 deaths per year. 1 It costs the National Health Service (NHS) expenditure of more than £300 million in surgical, adjuvant and palliative treatment. 2
Fifty-five percent of colorectal cancers diagnosed in the UK have advanced disease in the form of lymph node (Dukes C) or distant metastases (Dukes D) (Cancer Research UK bowel cancer statistics) and have five-year survival rates of 38% and 3%, respectively. There are increased cost implications as a result of increased treatments required such as adjuvant or palliative therapies.
Alternatively, patients who present with cancer confined to the bowel wall and its surrounding tissues (Dukes' A and B) have 64–83% five-year survival rates, respectively, and can be successfully treated with surgery alone. Many patients remain asymptomatic, and so the only method to detect early cancers is through a population-based screening process. The NHS Bowel Cancer Screening Programme (NHSBCSP) was introduced in England in 2006 to screen men and women aged between 60–69 years in order to detect these early cancers. The aim was to improve survival in patients with CRC and reduce the overall burden of treating advanced disease.
Colorectal cancer is uniquely placed for screening strategies, as 70–90% of the tumours arise from premalignant, adenomatous polyps. 3 This transformation may take 10 years or more and provides us with a window of opportunity to detect early tumours (Dukes' A) and also prevent cancer development by excision and removal of premalignant polyps during colonoscopy. 4
Screening tests
The screening modalities available for the early detection of CRC include faecal occult blood testing (FOBT), colonoscopy, flexible sigmoidoscopy (FS) and radiological investigations.
Faecal blood tests (FOBT)
These are qualitative chromogen tests, which convert a colourless compound to a coloured one due to the pseudoperoxidase activity of haemoglobin. The most widely available one is the Haemoccult® test employing guaiac-impregnated paper which detects the presence of haem in the faeces. The patients are advised to avoid any foods containing large amounts of haemoglobin or myoglobin as well as peroxidase containing vegetables such as turnips, broccoli, horseradish, cantaloupe melon, parsnips, radishes and tomatoes to reduce the false-positive cases. The usual gFOBT protocol consists of collecting two samples from each of three consecutive bowel movements at home. 5
The effectiveness of these tests in detecting occult blood can also vary with the degree of faecal hydration. 6 In the Nottingham study, 2% of those screened required further investigation but only 11% of these were found to have colorectal cancer and 37% polyps. 7 The estimated sensitivity of the Haemoccult test for colorectal cancer varied from 55% to 57% if the slides were not re-hydrated. 7,8 However, if the slides were re-hydrated (i.e. adding a drop of water to the slide window before processing) as in the Minnesota trial, the sensitivity increased to 92%. 9 In spite of this it should be noted that the positive predictive value (PPV) of the Haemoccult test for colorectal cancer was fairly low. It reported that nearly 80% of all positive tests were false-positives. These cases would then need further diagnostic interventions with a small chance of adverse events. However, the encouraging aspect is that the PPV for detecting adenomas ≥1 cm was higher which might prevent the development of CRC in the future. 10
The introduction of immunochemical FOTB (iFOBt) introduced in 2001 detect the presence of globin instead of the haem component of the haemoglobin (Hb). They are based on an antigen-antibody reaction that specifically detects human Hb and hence are not affected by dietary restrictions. 11 As globin is degraded by digestive enzymes in the upper gastrointestinal tract, iFOBt also are more specific for lower gastrointestinal bleeding, thus improving their specificity for CRC. Also a recent randomized controlled trial showed that, in patients with a positive iFOBt who underwent colonoscopy the number of detected cancers and advanced adenomas was 2.6 times and 3.5 times higher as compared to those with a positive guaic-based test, respectively. 12 Another randomized controlled trial from the Netherlands also reported a positivity rate difference of 3.1% in favour of iFOBt with a statistically higher number of cancers and advanced adenomas detected. 13 A further cost-effective analysis from the same group demonstrated that after one round of screening the iFOBt would save the healthcare system €72 as opposed to €27 by the gFOBt compared with no screening. 14
The evidence of the effectiveness of FOBT in reducing CRC mortality is based on four large randomized control trials from participants in Denmark, Sweden, USA and the UK involving a total of 327,043 participants. 7,8,10,15,16 All these trials performed biennial screening and reported a reduction in the relative risk for CRC mortality from 13% to 33%. A meta-analysis combining the four randomized trials showed that screening resulted in a statistically significant relative reduction in CRC mortality of 16% (fixed and random effects models: RR 0.84, CI 0.78–0.90). When adjusted for screening attendance in the individual studies, there was a 25% relative risk reduction in mortality (RR 0.75, CI 0.66–0.84) for those attending at least one round of screening using FOBT. 10 As expected, all the trials reported a higher number of Dukes A cancers. The details of the trials are illustrated in Table 1.
Effectiveness of FOBTs in reducing mortality in colorectal cancer – the trials
The main hurdle in screening for CRC is the uptake. Studies across Europe and the US have reported compliance rates between 40–54%. The main reasons for patient non-compliance are embarrassment and worries that tests would be unpleasant or painful. 17 Even in trial settings the participation does not exceed 60%. 7 A study performed has suggested that GP recommendations of screening could improve patient compliance. A cross-sectional postal survey of GPs in the UK reported that GPs from practices located in more deprived locations were also more likely to have positive attitudes towards FOBT and its recommendation to patients. They also identified certain system-related barriers such as lack of trained healthcare providers to conduct screening and investigate the FOBT. 18
Colonoscopy
Colonoscopy as the sole screening tool is advocated by numerous societies in North America. The American Cancer Society and the US Multisociety Task Force on Colorectal Cancer have both approved its use and advocated a schedule with screening colonoscopy every 10 years from 50 years of age. 19 However, the evidence for using colonoscopy as a screening tool is extrapolated from a case-control study which used FS as the initial investigation, and if polyps were found a full colonoscopy with polypectomy was performed. 20 The screened group in this study showed an 80% incidence reduction in CRC. Another case-control study of colonoscopy performed on symptomatic patients in the USA reported a 50% reduction in mortality. 21 The latest study from Canada conducted a natural experiment involving a 14-year follow-up of a cohort of all men and women 50–90 years of age exposed to different intensities of colonoscopy use. They calculated that for every 1% increase in complete colonoscopy rate, the hazard of death decreased by 3%. 22 Another rationale for using colonoscopy as a screening tool comes from the evidence of reduction in incidence of colon cancer in patients who have had polypectomy after baseline colonoscopy. The National Polyp Study reported a 76% to 90% reduction in the incidence of CRC after clearing colonoscopy when compared with reference populations. 23 Another Italian study also demonstrated similar incidence reduction in CRC of around 80% in patients who had their colon cleared of ≥5 mm adenomas as compared to a reference population. 24
Even though it seems obvious that colonoscopy could be the ideal investigation for CRC as it visualizes the whole colon, the evidence appears compelling its uptake is limited by the resources available. Also, its effectiveness in a community setting remains to be proven. A population-based study discovered that nearly 6% of newly diagnosed right-sided CRC had a colonoscopy 6 months to 3 years prior to diagnosis suggesting a substantial miss rate. 25 Another population-based, case-control study from Canada suggested that the reduction in CRC mortality seen with screening colonoscopy was restricted to left-sided tumours, while it failed to reduce mortality rates in people with right-sided tumours. 26 Another consideration should be given to the occurrence of interval cancers which in a study evaluating occurrence of tumours in patients with adenomatous polyps was found to be as high as 5% in patients who had undergone colonoscopy and polypectomy in the last 5 years. 27
Even though patient surveys indicate that more than 50% of those surveyed preferred colonoscopy as the investigation of choice for screening CRC screening, 28 another study which compared the attendance of subjects invited for screening had the lowest uptake for colonoscopy as compared to other modalities (iFOBT and FS). 29 Colonoscopy does require a more extensive work-up which includes bowel preparation as well as post-procedure care due to the sedation used during the procedure. Colonoscopy is a technically demanding procedure and is operator skill-dependent. The likelihood of a successfully completed colonoscopy with visualization of the entire colon is 85–95% with an experienced colonoscopist but falls to as low as 50% with a less experienced colonoscopist. 30 Different screening centres report varying rates of adenoma detection which is an important quality indicator and on multivariate analysis an independent predictor of the risk of interval cancer. 31 All these factors put a tremendous pressure on resources as colonoscopy for positive FOBT itself would generate around 10,000 colonoscopy sessions (at six cases per session) per annum in the UK.
Flexible sigmoidoscopy (FS)
FS (every 5 years) either alone or combined with FOBT (3 years) has been recommended by the US Preventive Services Task Force, American Cancer Society, US Multisociety Task Force, American College of Radiology Joint as a screening tool for CRC. 6 As nearly two-thirds of colorectal cancers and adenomas are found in the rectum and sigmoid, these can be examined by a screening FS. Also, screening colonoscopy has failed to demonstrate any reduction in CRC mortality for right-sided cancers. 26 It is relatively easier to perform as compared to colonoscopy and can be performed safely by a nurse endoscopist with adequate training. 32,33
Numerous case-control studies as well as randomized controlled trials have shown that FS was associated with a 60–80% reduction in CRC mortality for the area of the colon within its reach, with a protective effect lasting for 10 years or more. 34–36 A large randomized controlled trials conducted by the UK Flexible Sigmoidoscopy Trial investigators have recently published their findings examining the role of once only sigmoidoscopy screening undertaken between the ages of 55 and 64 years. The participants underwent screening FS and were referred to colonoscopy if they had polyps with high-risk criteria: 1 cm or larger; three or more adenomas; tubulovillous or villous histology; severe dysplasia or malignant disease. They had an uptake rate of 71% and a median follow-up of 11.2 years. They detected a 33% reduction in incidence and 43% reduction in mortality due to CRC in people attending screening. The incidence of distal CRC (rectum and sigmoid) was reduced by 50%. The reduction in mortality achieved by once only FS is higher than that achieved by FOBT screening (Table 1) methods. This suggests that this approach could be applied to a nationwide screening programme if an adequately trained workforce is made available. 37
Other possible screening modalities
Radiological investigations such as virtual colonoscopy include the use of computed tomography (CT), magnetic resonance (MR) imaging or positron emission tomography (PET) as a means of imaging the colon. The advantages of radiological examination are mainly that they are non-invasive procedures. Disadvantages include the radiation burden with CT, missed lesions such as flat adenomas and adenocarcinomas. 38 MR colography does not carry the risks of radiation, however, require the patient to be scanned in claustrophobic scanners and cannot be performed on patients with metal implants and cardiac pacemakers. MR colography does not have any advantage in sensitivity or specificity when compared with other modalities. 39 There are no data in the use of PET in screening at present, but it has been shown to be very useful in detecting metastatic cancer.
Other modalities that are becoming of interest are capsule endoscopy, the use of detecting molecular abnormalities as a basis for screening and faecal DNA screening tests.
Increased risk due to family history of CRC
The screening for CRC in high-risk individuals may need to be tailored as an individual with a single first-degree relative with CRC has an increased risk of approximately 2.25 times that of the general population. Individuals with more than one first-degree relative with colorectal cancer have an increased risk of approximately 4.25 times that of the general population. There is evidence to suggest that individuals with a first-degree relative with colorectal adenoma also have an increased risk of CRC of approximately twice that of the general population.
40
As the prevalence of CRC is high at least 9.8% of patients will have a first-degree relative affected by colon cancer.
41
The British Society of Gastroenterology (BSG) and Association of Coloproctology for Great Britain and Ireland (ACPGBI) in their recent update have stratified risk according to family history and made recommendations for colonoscopy surveillance:
42
High to moderate risk: These are people with three or more first-degree relatives with CRC (none less than 50 years); two first-degree relatives with CRC with mean age less than 60 years. Colonoscopy every 5 years from age 50–75 years;
Low to moderate risk: These are people with only one affected relative less than 50 years or only two affected first-degree relatives aged 60 years or older. Once-only colonoscopy recommended at 55 years.
In all other patients who are known to have a relative with CRC there is no evidence for any colonoscopic surveillance. The high-risk groups with genetic susceptibilities are not discussed in this review.
The NHS Bowel Cancer Screening Programme
The NHS Bowel Screening Programme in the UK is based on the principle of secondary prevention whereby detection and treatment of existing pre-neoplastic and early neoplastic lesions would result in reduction of prevalence and improvement in survival. The programme was rolled out in July 2006 and has almost achieved nationwide coverage to date. It offers screening with gFBOT every two years to all men and women aged between 60–69 years. Patients over the age of 70 years can request a kit to be sent to them.
As described above the detection of traces of blood due to small bleeds from gastrointestinal tumours in the faeces forms the mainstay of the screening process. This can be detected by inexpensive tests (supplied in kit form) which are relatively simple to distribute and are performed by the patients at home. However, the presence of blood in the faeces can be due to any cause in the GI tract. Therefore, the tests lack specificity and positive results have to be investigated by a definitive diagnostic colonoscopy.
All the eligible participants are sent the FOBT kit along with an information leaflet. They are advised to omit red meat or vitamin C supplements from their diet for two days before starting to complete the kit.
The kit comprises six tiles onto which samples of stool are smeared. Three separate stool samples are required. The samples are then sent by the patient in a prepaid package.
If zero out of six samples are positive, a further kit is sent out at two years. If one to four tests are positive, a further kit is immediately sent out. If the second test is unclear then the patient is reviewed in a nurse-led clinic at a designated screening centre. All patients with a positive test (five out of six slides positive) are then invited for a screening colonoscopy.
Prior to this, they are reviewed in the nurse-led clinic and counselled as regards their eligibility for colonoscopy, the procedure of colonoscopy itself and its possible outcomes and complications. These include incomplete examination, missed lesion, bleeding perforation and a mortality risk of 1 in 10,000. All colonoscopies are performed by an accredited colonoscopist as per the standards set out by the Joint Advisory group on GI Endoscopy (JAG).
Anyone who has had a full bowel investigation (colonoscopy or a barium enema and flexible sigmoidoscopy) within the previous four years is excluded. Also, people on a bowel polyp surveillance programme, those currently being treated for bowel cancer or awaiting bowel investigations arranged by their GP, those who have had had their bowel removed and people suffering from an inflammatory bowel disease, such as Crohn's disease or ulcerative colitis are also excluded.
Of every 1000 people sent a FOBT kit, approximately 20 will give a positive result and be offered colonoscopy. Sixteen of these 20 will undergo the colonoscopy and of these eight (50%) will demonstrate a normal colon, six will reveal colonic polyps and two (1%) will have a screening detected cancer. Any incomplete colonoscopy will require a barium enema or virtual colonoscopy in order to assess the bowel.
The final report on the second round of the CRC screening published by the Department of Health in 2006 reported an uptake rate of 51.9% as opposed to 58.5% in the first round. The sensitivity of the FOBT used in the Pilot (Hemascreen) was very similar to that observed in the Nottingham trial and ranged from 57.7–64.4%. The PPVs for both cancer and neoplasia (any polyps) were 5.6% and 33.5%, respectively.
The uptake of colonoscopy in patients with a positive test was very high (96.9%).The rate of cancers detection was 0.91 for every 1000 colonoscopies. Of the cancers detected 19.4% were Dukes A and 32.3% had Dukes B cancer.
Summary
Colorectal cancer is one of the most common cancers diagnosed in the UK. In 2007 the Government committed to reducing cancer deaths in the under 75s by at least 20% by this year. The current NHSBCSP, implemented in 2006, using the gFOBT method, has reduced morbidity and mortality from CRC. The use of gFOBT has been deemed an affordable test for the programme and has been shown to be cost-effective in reducing mortality rates. Costs related to the screening programme have to be measured against cost savings due to decreases in the use of adjuvant, neoadjuvent treatments and palliative care services. There are also possible cost savings related to the detection of early CRC being more amenable to laparoscopic resection than larger advanced tumours found at clinical presentation. These savings are a result of reduced hospital stays and reduced complications relating to minimally-invasive surgery.
The current CRC screening programme seems to be a pragmatic solution for a significant disease. This is especially true when considering the current NHS financial constraints. Future tests may well be combined tests that hopefully will reduce the number of false-negative results but may be more costly. A major challenge will be to increase the uptake of the tests. This may be achieved with greater public awareness education and a concerted effort by healthcare professionals in both primary and secondary care.