Intended for healthcare professionals


Colorectal adenocarcinoma: risks, prevention and diagnosis

BMJ 2016; 354 doi: (Published 14 July 2016) Cite this as: BMJ 2016;354:i3590
  1. Sri G Thrumurthy, honorary research fellow1,
  2. Sasha S D Thrumurthy, medical officer2,
  3. Catherine E Gilbert, foundation year doctor3,
  4. Paul Ross, consultant medical oncologist4,
  5. Amyn Haji, consultant colorectal surgeon5
  1. 1Department of Surgery, University College Hospital, London, UK
  2. 2Department of General Medicine, Tan Tock Seng Hospital, Singapore
  3. 3Department of Surgery, Queen Elizabeth Hospital, London, UK
  4. 4Department of Medical Oncology, Guy's & St Thomas' Hospital, London, UK
  5. 5Department of Surgery, King's College Hospital, London, UK
  1. Corresponding author: A Haji amynhaji{at}

What you need to know

  • Colorectal cancer is the third most common cancer in men and the second most common cancer in women worldwide

  • It is rare below 40 years of age

  • Increasing age, male gender, and a family history of colorectal cancer are the greatest risk factors for the disease

  • Patients presenting with suspicious symptoms and signs should be referred and investigated urgently in a specialised unit

Colorectal cancer is the fourth most common cause of cancer related mortality globally, with 1.4 million new cases and 700 000 deaths annually.1

What is colorectal cancer?

Colorectal cancer refers to tumours of the rectum or large bowel (including the appendix) that arise from the colorectal mucosa (fig 1). Adenocarcinoma is the most common form of colorectal cancer (>95%). Rarer subtypes include carcinoid tumour, sarcoma, and lymphoma; these present differently from adenocarcinoma1 and will not be discussed in this review.


Fig 1 Distribution of bowel cancer by anatomical site, UK (2007-09) (adapted from Cancer Research UK, bowel cancer incidence statistics,

How does colorectal adenocarcinoma develop?

Colorectal cancer typically develops from adenomatous polyps that undergo dysplastic changes to become cancerous (fig 2).2 Tumours can occur sporadically, but there are some inherited colorectal cancer syndromes (see table 1). Several risk factors are also recognised (see box 1).


Fig 2 Progression from colorectal polyp to cancer (adapted from Johns Hopkins Colon Cancer Center. Polyps 101.

Table 1

Inherited syndromes predisposing to colorectal cancer

View this table:

Box 1: Risk factors for colorectal cancer

Sociodemographic factors
  • Older age

    • In the UK from 2009 to 2011, 43% of colorectal cancer was diagnosed in people aged ≥75 years, and 95% in those aged ≥50 years7

    • Age-specific incidence rates increase steeply after age 50, with the highest rates above age 85 years

  • Male sex

    • A meta-analysis of 17 studies comprising 924 932 patients showed men were at significantly greater risk than women (relative risk 1.83 (95% CI 1.69 to 1.97)) at developing advanced colorectal cancer across all age groups8

Lifestyle factors
  • Red meat and processed meat consumption

    • A meta-analysis of 24 prospective studies found the relative risk of colorectal cancer for the highest versus the lowest intake of red or processed meat in the pooled cohort was 1.22 (95% CI 1.11 to 1.34)9

    • Relative risk for every 100 g/day increase in intake was 1.14 (1.04 to 1.24)

    • Cancer risk increased linearly with increasing meat intake up to 140 g/day

  • Obesity

    • 13% of bowel cancers in the UK have been linked to obesity10

    • A meta-analysis of 43 studies including nearly 9 000 000 individuals worldwide found a relative risk of 1.33 (1.25 to 1.42) in obese individuals (body mass index >30) compared with those with normal BMI11

    • Association with BMI is stronger in men than women,12 and in women it may vary with menopausal status and use of hormone replacement therapy13

  • Alcohol

    • 11% of bowel cancers in the UK have been linked to excessive alcohol consumption10

    • A meta-analysis of 61 cohort and case-control studies found that, compared with infrequent or non-drinkers, those who consumed 1.6-6.2 British alcohol units daily had a 21% higher risk (95% CI 1.13 to 1.28) of bowel cancer, and those who drank over 6.2 units daily had a 52% higher risk (95% CI 1.27 to 1.81)14

    • Overall, bowel cancer risk increased by 7% per unit of alcohol consumed daily

  • Tobacco smoking

    • 8% of bowel cancers in the UK have been linked to tobacco smoking10

    • A meta-analysis of 28 prospective studies and 1 463 796 subjects suggests current cigarette smokers have a 20% higher risk of disease (95% CI 1.10 to 1.30) than people who have never smoked15

    • A meta-analysis of 106 observational studies suggests that former smokers have a 18% higher risk of bowel cancer (95% CI 1.11 to 1.25) than never smokers; and that bowel cancer risk rises by 7-11% per 10 cigarettes smoked daily16

Medical factors
  • Family history

    • 20% of bowel cancers are associated with hereditary factors other than familial adenomatous polyposis (FAP) and hereditary non-polyposis colorectal cancer (HNPCC)6

    • A meta-analysis of 59 studies suggested that bowel cancer risk was 80% higher in people with an affected first degree relative.17 A population lifetime risk of 1.8% for a 50 year old increased to 3.4% (95% CI 2.8 to 4.0%) with at least one affected relative or 6.9% (4.5 to 10.4%) with two or more relatives

    • A meta-analysis of 27 studies suggested a relative risk of 3.87 (2.40 to 6.22) for patients with an affected relative who was diagnosed before the age of 45 years18

  • Colorectal adenomas or polyps

    • A pooled analysis of 8 prospective studies comprising 9167 patients revealed that 1% of patients with colorectal adenomas >20 mm diameter or with high grade dysplasia develop cancer within 4 years of adenoma removal19

    • A meta-analysis of 7 studies comprising 11 387 patients revealed that patients with low risk polyps detected at first colonoscopy had a 80% higher risk of advanced cancer than those with no polyps detected20

  • Inflammatory bowel disease (IBD)

    • A meta-analysis of population based cohort studies found that patients with IBD (ulcerative colitis or Crohn's colitis) have a 70% higher risk (95% CI 1.2 to 2.2) of developing colorectal cancer than the general population21

    • This risk increases with the duration and extent of disease: those with IBD for >20 years have a 5% risk of developing bowel cancer21

  • Diabetes

    • A meta-analysis of 15 studies including 2 593 935 patients found that diabetes was associated with a 30% higher risk of colorectal cancer (relative risk 1.30 (1.20 to 1.40))22

    • Diabetic patients also suffered greater mortality from bowel cancer (relative risk 1.26 (1.05 to 1.50)), but there was evidence for heterogeneity between studies (P=0.04)

Who gets colorectal cancer?

The incidence of colorectal cancer strongly increases after 50 years of age (fig 3), and median age at diagnosis is about 70 years in developed regions.23 In 2012, age standardised incidence rates were highest across Oceania (41.0 and 29.2 per 100 000 in men and women respectively), North America, and Europe, and lowest across western Africa (4.5 and 3.8 per 100 000) and south-central Asia (fig 4).8 Previously low risk regions (such as Spain and several countries in Eastern Europe and East Asia) have seen rapid rises in incidence of colorectal cancer. This has been attributed to the adoption of high fat diets heavy in red and processed meats, physical inactivity, excessive alcohol consumption, and smoking. The US and other high income countries have seen a plateau or drop in disease incidence, possibly due to an increased use of diagnostic endoscopy and polypectomy.24 25


Fig 3 Age specific incidence rates of colorectal cancer per 100 000 population, UK (2009-11) (adapted from Cancer Research UK, bowel cancer incidence statistics,


Fig 4 Estimated worldwide age standardised (per 100 000) incidence of colorectal cancer in men in 2012 (adapted from GLOBOCAN 20121)

What are the risk factors for colorectal cancer?

Most colorectal cancer cannot be attributed to any single risk factor, although increasing age and male sex have consistently shown strong associations with disease incidence in epidemiological studies.23 The evidence for the other major risk factors for colorectal cancer is summarised in box 1.

The key genetic syndromes associated with colorectal cancer are summarised in table 1; and the current UK recommendations for screening and surveillance of asymptomatic patients in moderate-risk family groups are summarised in table 2.26

Table 2

Summary of recommendations for colorectal cancer (CRC) screening and surveillance in moderate risk family groups (adapted from Cairns et al26)

View this table:

How do patients present, and who should be referred?

The commonest presenting features of colorectal cancer are abdominal pain, change in bowel habit, rectal bleeding, and microcytic anaemia, although these commonly feature in other gastrointestinal diseases. Left sided colorectal tumours typically present with altered bowel habit (such as loose stools, increased frequency, and intestinal obstruction secondary to progressive luminal narrowing), rectal bleeding or mucus, or tenesmus. Right sided lesions may present more insidiously with weight loss, abdominal pain or mass, or iron deficiency anaemia.27 Urgently investigate and refer men and non-menstruating women with iron deficiency anaemia, as 10% of such patients will have colorectal cancer.28 29

A UK population based case-control study of 2093 patients aimed to quantify the pre-diagnostic features of colorectal cancer (see table 3 for details).30 The 2015 NICE guidelines for the urgent referral of suspected colorectal cancer are summarised in table 4.31

Table 3

Pre-diagnostic features of colorectal cancer (adapted from Hamilton et al30)

View this table:
Table 4

Urgent referral criteria for suspected colorectal cancer for an appointment within two weeks (from National Institute for Health and Care Excellence31)

View this table:

How to investigate suspected colorectal cancer

Advise patients that more than one investigation may be necessary to confirm or exclude a diagnosis of colorectal cancer.32

Colonoscopy is the recommended first line imaging for suspected colon cancer in patients without major comorbidity.32 Bowel preparation with oral laxatives improves the diagnostic yield of pan-colonic imaging. Biopsies are taken from any suspicious lesions unless contraindicated (for example, in coagulopathy) to confirm the diagnosis and degree of tumour differentiation (well, moderately, or poorly differentiated) to guide treatment. Patients who have undergone incomplete colonoscopy (due to intra-procedural discomfort or poor bowel preparation) may be offered repeat colonoscopy, computed tomographic (CT) colonography, or barium enema. Patients with major comorbidity and frail or elderly patients with poor mobility and poor tolerance to bowel preparation may be offered alternative imaging such as CT colonography or flexible sigmoidoscopy in the first instance, followed by biopsy of suspicious lesions.32


Colonoscopy is operator dependent and requires full bowel preparation. Completion rates (that is, passage of colonoscope to caecum) vary widely because of technical challenges, and experienced endoscopists typically achieve completion rates of 98%. A 90% rate is considered acceptable.33 Histological confirmation of malignancy requires multiple biopsies. Overtly malignant lesions may be injected with dye contrast to mark the area and facilitate subsequent surgical resection. Procedural risks are mainly related to sedation and bowel perforation (table 5).34 35 Prospective comparative studies have suggested that patients are as comfortable with conventional colonoscopy as CT colonography. Both of these investigations are preferred over barium enema.36 37 Poor bowel preparation and incomplete colonoscopy are the two main variables contributing to false negative tumour detection.33

Table 5

Common methods of colonic imaging

View this table:

Box 2 lists advice for people referred for colonoscopy. If there is a clear contraindication or more than one relative contraindication to bowel preparation, offer admission on the day before colonoscopy for inpatient preparation.

Box 2: General advice for patients referred for colonoscopy

  • All patients referred for colonoscopy should be sent comprehensive written instructions by their endoscopy unit, including the following:

    • Deadline for stopping iron preparations, anti-diarrhoeal and codeine-containing preparations, and fibre-containing laxatives beforehand

    • Consuming a specific low fibre diet and increasing fluid intake in the week before endoscopy

    • Avoiding all solid food 24 hours before endoscopy, and taking an oral bowel cleansing preparation (such as macrogols or sodium picosulfate with magnesium citrate) at fixed times during this period

  • The main risks of bowel preparations are dehydration and electrolyte imbalance

  • Absolute contraindications to the use of bowel cleansing preparations include allergies to these agents, bowel obstruction, perforation or ileus, gastric emptying disorders, toxic megacolon, and certain drug-specific contraindications (such as macrogols for patients with phenylketonuria or glucose-6-phosphate dehydrogenase deficiency)

  • Relative contraindications include chronic renal failure, heart failure, dehydration, impaired swallow or increased risk of aspiration, acute inflammatory bowel disease, patients who are frail or not self caring, and the use of any medication that may affect fluid or electrolyte balance or are dangerous with hypokalaemia (such as diuretics, steroids, antidepressants, digoxin, carbamazepine, lithium)

  • It is vital to consider drugs such as antiepileptics, where decreased absorption may cause clinical deterioration

CT colonography

CT colonography (“virtual colonoscopy”) may be indicated after failed colonoscopy, in light of patient preference, or when a patient is deemed unsuitable for colonoscopy. Unlike conventional colonoscopy, CT colonography allows intraluminal visualisation of the colourectum without the need for sedation, while it offers a similar sensitivity for cancer detection but a lower specificity for polyp and tumour detection (table 5).38 CT colonography may detect extraluminal pathology in patients referred for weight loss or abdominal pain. Intravenous contrast and oral faecal tagging agents (such as Gastrografin, which also has a laxative effect) are used to differentiate stools and mass lesions (polyps and cancers). CT colonography is not suitable for pregnant patients. It is contraindicated in patients with iodine allergies and may not be possible for those with impaired swallowing or high aspiration risk. Tissue biopsy and polyp removal are not possible with CT colonography, so any lesions detected will require endoscopic evaluation to confirm a histological diagnosis.

As with colonoscopy, patients must be given comprehensive information about stopping certain medications and adopting a low residue diet in the days before the test. Special instructions are provided for diabetic patients. Depending on hospital policy, patients will be requested to take oral Gastrografin (with or without an accompanying bowel cleansing agent) at set times for one or two days before the procedure. The procedure lasts about 30 minutes, during which a muscle relaxant (such as buscopan) may be administered with intravenous contrast, while the colon is gently insufflated with carbon dioxide gas via the rectum, and scans obtained in various patient positions.

Double contrast barium enema

Double contrast barium enema is well tolerated and safe with excellent completion rates, but its relative impracticality and lower diagnostic yields compared with the above investigations have resulted in its declining use across most UK centres.

Other tests

Testing for faecal occult blood (FOB) and serum tumour markers (such as carcinoembryonic antigen) are not useful in the investigation of suspected colorectal cancer. While FOB testing is effective for population screening in asymptomatic cohorts, it is too insensitive to guide the investigation of patients with colorectal symptoms. Similarly, tumour markers lack sensitivity and specificity but are useful in the follow-up of treated patients. A negative FOB test or normal serum tumour markers should not delay the referral of symptomatic patients.

What can be done to prevent colorectal cancer?

Primary prevention


A meta-analysis of 25 prospective studies revealed a 10% risk reduction in developing colorectal cancer for an extra 10 g of total dietary fibre ingested daily. In particular, cereal fibre and whole grains were associated with a reduction in colorectal cancer risk.39

A meta-analysis of 19 cohort studies suggested that, for every 400 g of dairy products consumed daily, the risk of developing colorectal cancer was significantly reduced (relative risk 0.83 (95% CI 0.78 to 0.88). Similarly, the daily consumption of 200 g of milk or 50 g of cheese was associated with a lower risk of developing colorectal cancer.40

Another meta-analysis of 15 studies including 12 305 patients revealed that every 300 mg of daily calcium intake (up to 1900 mg/day) reduced the risk of developing colorectal cancer.41

Physical activity

It is estimated that 5% of colon cancers (but not rectal cancers) in the UK are linked to inadequate physical activity.42 43 Meta-analyses of cohort studies have revealed a 17-24% risk reduction in colon cancer from the most to the least physically active people.43 44 45


Two large trials in the 1980s, designed to evaluate the prevention of vascular events by aspirin, revealed a 37% risk reduction of colorectal cancer in patients with a daily intake of 300 mg aspirin for at least five years.46 Various observational studies support the use of long term aspirin in the chemoprevention of colorectal cancer,47 and data from a large randomised controlled trial revealed that taking 600 mg aspirin daily for two years led to a risk reduction (hazard ratio 0.41) in colorectal cancer incidence among patients with HNPCC.48 Non-steroidal anti-inflammatory drugs (NSAIDs) have also been shown to reduce colorectal cancer risk in various cohort studies and case-control studies,49 and randomised controlled trials have shown that COX-2 inhibitors reduce adenoma incidence (relative risk 0.72 (0.68 to 0.77)), potentially reducing subsequent cancer risk.49 A large, randomised, double-blind trial revealed that daily supplementation with 1200 mg calcium reduced colorectal adenoma recurrence (adjusted risk ratio 0.85 (0.74 to 0.98), P=0.03).50

Despite insufficient evidence to recommend routine use of most of the above agents for prevention of colorectal cancer, the US Preventive Services Task Force recently recommended use of low dose aspirin for the primary prevention of cardiovascular disease and colorectal cancer in adults aged 50-59 years who have a ≥10% 10-year risk of cardiovascular disease, are not at increased risk for bleeding, have a life expectancy of ≥10 years, and are willing to take low dose aspirin daily for ≥10 years.51 No similar recommendations exist in current British or European guidelines.

Secondary prevention (screening)

Faecal occult blood (FOB) testing is based on the tendency of cancer and polyps to bleed.52 A meta-analysis of randomised controlled trials suggested that FOB screening reduced colorectal cancer mortality by 25% (relative risk 0.75 (0.66 to 0.84)) and estimated that FOB screening prevented approximately 1 in 6 deaths from colorectal cancer.53 The NHS bowel cancer screening programme currently offers FOB testing to patients aged 60-74 years, of whom those with a positive FOB test are offered colonoscopy.54

In a large randomised controlled trial across 14 UK centres, a single flexible sigmoidoscopy between the ages of 55 and 64 years reduced colorectal cancer incidence by 23% (hazard ratio 0.77 (0.70 to 0.84)) and mortality by 31% (hazard ratio 0.69 (0.59 to 0.82)).55 Flexible sigmoidoscopy is expected to be fully integrated into the NHS Bowel Cancer Screening Programme by 2018.

US guidelines recommend various other screening strategies for average risk patients from 50 years of age, including five-yearly flexible sigmoidoscopy (with or without FOB testing), double contrast barium enema, or 10-yearly colonoscopy,56 all of which reduce mortality compared with no screening. Specific colonoscopy based screening protocols exist for patients at increased risk due to syndromes such as FAP, HNPCC, and MYH-associated polyposis (MAP)56 57; a strong family history of colorectal cancer (a single first degree relative with colorectal cancer or advanced adenoma diagnosed at <60 years old, or two first degree relatives with colorectal cancer or advanced adenomas)26 56; inflammatory bowel disease (ulcerative or Crohn's colitis)56; and acromegaly.58 59

How patients were involved in the creation of this article

No patients were involved in the creation of this article.


  • Contributors: SGT, SSDT, and CEG conceived the review, extracted evidence, and drafted the manuscript. All authors helped write the article (including article direction, interpreting the literature, and editing the manuscript). AH is guarantor and has approved the final article.

  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and have no relevant interests to declare.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract