Colorectal cancer

How common it is in the UK. Cancer Research UK counts 48,213 new bowel cancers a year (2019, 2021 to 2022), more than 130 a day, which makes it the 4th commonest cancer and 12 percent of all new cancers: about 26,800 in men and about 21,400 in women. It is the 2nd commonest cause of cancer death, with about 17,700 deaths a year (2022 to 2024), 48 a day. More than two-fifths (43 percent) of new cases are in people aged 75 and over, rates are highest at 85 to 89, and 59 percent of deaths are in the over-75s. The lifetime risk of a diagnosis is 1 in 20 for women and 1 in 17 for men born in 1961. The trends run in opposite directions: incidence rates are up about 6 percent since the early 1990s but down 3 percent in the last decade, and are projected to fall a further 7 percent by 2038 to 2040 (to about 47,700 cases a year), while mortality rates have almost halved (down 46 percent) since the early 1970s and fell 6 percent in the last decade. The rectum is the commonest single site (2016 to 2018). Deprivation shows through: mortality rates for lower gastrointestinal cancer are 22 percent higher in the most deprived fifth than the least for women and 27 percent higher for men, about 1,700 deaths a year in the UK.

The rise in early-onset disease. Colorectal cancer is becoming a young person's disease at the same time as it retreats in the old. The American Cancer Society's 2026 report finds overall incidence falling 0.9 percent a year over 2013 to 2022, driven by a 2.5 percent annual fall in people aged 65 and over, while incidence rose 3 percent a year in adults aged 20 to 49 and 0.4 percent a year at 50 to 64, dominated by tumours of the distal colon and rectum; rectal cancer incidence overall turned upwards by 1 percent a year from 2018 to 2022 after decades of decline and now accounts for 32 percent of colorectal cancers, up from 27 percent in the mid-2000s. Mortality under 50 has risen 1 percent a year since 2004 and at 50 to 64 by 1 percent a year since 2019, while it fell 2.3 percent a year in the over-65s from 2012. The earlier 2023 report put the same shift another way: the share of cases in people under 55 went from 11 percent in 1995 to 20 percent in 2019, and 60 percent of new cases were advanced in 2019 against 52 percent in the mid-2000s. The pattern is not American. A study of registry data from 36 countries found incidence in adults under 50 rising in 19 of them and, in nine spanning three continents (Australia, Canada, Denmark, Germany, New Zealand, Slovenia, Sweden, the UK and the USA), rising in the young while stable or falling in older adults; the uptick began in the mid-1990s in most high-income countries, and the steepest rises were in Korea and New Zealand (Siegel 2019). No cause has been established. Diet, obesity, type 2 diabetes, inactivity, alcohol, smoking and the microbiome are the candidates, and the generational pattern points at an exposure acting early in life.

What raises the risk, at population level. These are the factors that shape how many bowel cancers occur in a country, not an account of why any one person got theirs. Age comes first, and then a list that is unusually modifiable: Cancer Research UK judges 54 percent of UK bowel cancers preventable. Processed meat causes 13 percent of UK cases (risk 18 percent higher per 50 g a day; colon cancer risk 22 percent higher per 100 g a day of red meat, with no association for rectal cancer); too little fibre causes 28 percent (risk 10 percent lower per 10 g a day of total or cereal fibre, 20 percent lower per 90 g a day of whole grains); overweight and obesity cause 11 percent (colon cancer risk 30 percent higher in men and 12 percent higher in women per 5 units of body mass index, 25 percent higher per 10 cm of waist, and adenoma risk 47 percent higher in obesity); alcohol causes 6 percent (risk 7 percent higher per unit a day, 33 percent higher above 6 units a day); smoking causes 7 percent (17 to 21 percent higher risk in current smokers, adenomatous polyp risk about doubled, serrated polyp risk more than doubled); too little physical activity causes 5 percent (colon cancer risk 19 percent lower in the most active); and ionising radiation 2 percent. Medical conditions matter too: type 2 diabetes raises risk 22 to 30 percent, and inflammatory bowel disease (ulcerative or Crohn's colitis) raises it 70 percent, with a 5 percent risk of bowel cancer after 20 years of disease. Family history more than doubles risk with one affected first-degree relative. Inherited syndromes account for a minority but a decisive one: mismatch repair variants (Lynch syndrome) are carried by about 1 in 279 people and cause 1 to 4 percent of colon cancers, familial adenomatous polyposis fewer than 1 percent, and about 1 in 45 people carry a MUTYH variant (Win 2017; Cancer Research UK). Aspirin runs the other way: ever-use is associated with 17 percent lower risk, and longer use with lower risk still.

How it shows itself, and when to refer. The NHS lists the symptoms as changes in poo (looser, harder or going more or less often than usual), blood in the poo or bleeding from the bottom, a persistent feeling of needing to poo, tummy pain, a lump in the tummy, bloating, unexplained weight loss, and tiredness or breathlessness from anaemia; it advises seeing a GP for any of them, an urgent appointment or NHS 111 for black or dark red poo or bloody diarrhoea, and 999 or A&E for non-stop bleeding or large clots. None of these symptoms is specific, which is why England now puts a test between the symptom and the referral. NICE NG12 (1.3.1, adapted from the HealthTech guidance on quantitative faecal immunochemical testing in primary care) tells GPs to offer a quantitative faecal immunochemical test to adults with an abdominal mass, a change in bowel habit or iron-deficiency anaemia; aged 40 and over with unexplained weight loss and abdominal pain; under 50 with rectal bleeding plus abdominal pain or weight loss; 50 and over with unexplained rectal bleeding, abdominal pain or weight loss; and 60 and over with anaemia even without iron deficiency. A result of at least 10 micrograms of haemoglobin per gram of faeces triggers a suspected cancer pathway referral (1.3.2); below that, safety netting applies and strong clinical concern still overrides the test (1.3.3). A rectal mass goes straight to referral, and people with a rectal mass, an unexplained anal mass or anal ulceration do not need the test first. A previous negative screening result does not exempt a symptomatic person. In England in 2019, 38 percent of bowel cancers were diagnosed through an urgent suspected cancer referral, 22 percent after an emergency presentation and 12 percent through screening; 47 percent of staged cases in 2022 were stage I or II (Cancer Research UK).

Screening: what is offered and what it is worth. England invites everyone aged 50 to 74 every two years for a faecal immunochemical test done at home, with a colonoscopy for an abnormal result; the kit replaced the older guaiac test in June 2019, results come about two weeks after the laboratory receives the sample, and people aged 75 and over can request a kit every two years on the programme helpline (NHS; GOV.UK). The threshold that decides who is called for colonoscopy is a rationing decision as much as a clinical one: the English programme uses 120 micrograms of haemoglobin per gram of faeces, a figure no programme publication states but which the modelling of the English pilot calls the current threshold (Br J Cancer 2022; the same paper's estimates of what it buys are in the glossary). The symptomatic threshold in primary care is twelve times lower, at 10 micrograms per gram. The United States starts earlier: the US Preventive Services Task Force recommends screening from 50 to 75 (grade A) and from 45 to 49 (grade B), with selective screening from 76 to 85, by colonoscopy, faecal testing, stool DNA or computed tomography colonography. The evidence that screening saves lives is randomised. Annual faecal occult blood testing cut 13-year colorectal cancer mortality by 33 percent in 46,551 people in Minnesota (Mandel 1993). A single flexible sigmoidoscopy offered once to people aged 55 to 64 cut incidence by 26 percent and mortality by 30 percent over 17 years, and by 35 and 41 percent in those who actually attended (Atkin 2017). NordICC, the first randomised trial of screening colonoscopy, found a 10-year risk of colorectal cancer of 0.98 percent in the invited group against 1.20 percent in usual care, an 18 percent reduction, with 455 people needing to be invited to prevent one cancer; only 42 percent of those invited attended, and a later instrumental-variable analysis of the same trial estimated a 35 to 41 percent reduction in incidence among those screened (Bretthauer 2022).

How it is diagnosed and staged. Colonoscopy with biopsy is the test that makes the diagnosis and removes the precursor at the same visit; computed tomography colonography is the alternative when colonoscopy is not possible. Staging is by computed tomography of the chest, abdomen and pelvis, with magnetic resonance imaging of the pelvis for every rectal cancer. The eighth edition of the TNM system is the language: Tis is carcinoma in situ in the mucosa; T1 is into the inner layer, T2 into the muscle, T3 through the muscle into the surrounding tissue, T4a through the outer lining into the peritoneum and T4b into a nearby organ; N1a, N1b and N1c mean one node, two or three nodes, or tumour deposits without node involvement, and N2a and N2b mean four to six or seven or more nodes; M1a, M1b and M1c mean one distant site, two or more, or any spread to the peritoneum (Cancer Research UK). Rectal cancer adds two questions computed tomography cannot answer: how close the tumour comes to the plane the surgeon will cut along (the mesorectal fascia, which becomes the circumferential resection margin in the specimen), and whether tumour has spread into the veins outside the bowel wall. High-resolution magnetic resonance imaging answers the first with 92 percent specificity for predicting a clear margin, and in the 349 patients it predicted clear, 327 were clear at surgery (MERCURY, 408 patients, BMJ 2006). Every metastatic tumour is tested for RAS and BRAF V600E mutations before systemic therapy (NICE NG151 1.4.1), mismatch repair status is tested at diagnosis, and where the tumour started, the right colon or the left colon and rectum, predicts both biology and the benefit of EGFR antibodies. In the United States 34 percent of cases are found localised, 37 percent regional, 23 percent distant and 6 percent unstaged (SEER).

Pathology, the two pathways, and prevention. Almost all bowel cancers are adenocarcinomas of the gland cells that line the bowel; mucinous and signet ring cell tumours are rarer forms of adenocarcinoma, and squamous, neuroendocrine, sarcomatous, lymphoid and melanocytic tumours of the bowel are separate diseases with their own records (Cancer Research UK). Grade runs 1 to 4, from cells that look almost normal to cells that look nothing like them. The cancers arise along two routes. The conventional adenoma-carcinoma sequence, mapped in 172 specimens in 1988 and formalised as a model in 1990, runs from normal mucosa through adenoma to carcinoma as mutations accumulate: loss of the chromosome 5 region that carries the polyposis gene, RAS mutation (found in 58 percent of adenomas over 1 cm but only 9 percent of smaller ones), loss of chromosome 18 (73 percent of carcinomas) and loss of chromosome 17p (75 percent of carcinomas) (Vogelstein 1988; Fearon and Vogelstein 1990). The serrated pathway accounts for about 30 percent of colorectal carcinomas and runs instead through hyperplastic polyps, sessile serrated lesions and traditional serrated adenomas, with methylation of promoter islands and BRAF or KRAS mutation, and it produces most of the sporadic mismatch repair-deficient cancers (Bettington 2013). Prevention follows from the first route: in the National Polyp Study, removing adenomas at colonoscopy left 1,418 patients with 76 to 90 percent fewer cancers than expected, and at a median of 15.8 years mortality from colorectal cancer among 2,602 patients whose adenomas were removed was 53 percent lower than the general population would predict (Winawer 1993; Zauber 2012). Aspirin is the drug: in the CAPP2 trial, 600 mg of aspirin daily for at least two years cut colorectal cancer in Lynch syndrome carriers over ten years (hazard ratio 0.65 in the intention-to-treat analysis, 0.56 in those who completed two years), and NICE NG151 (1.1.1) now says to consider daily aspirin for more than two years in Lynch syndrome (Burn 2020).