Reducing Colonoscopies in Patients Without Significant Bowel Disease

Overview

Investigating people with bowel symptoms uses a test that detects traces of blood in the stools, the FIT test. There are many possible reasons for positive tests. A few people have cancer. However, most participants with symptoms don't have any serious bowel disease but have benign problems such as piles or irritable bowel syndrome (IBS). It is very difficult to diagnose on symptoms alone, those participants who have serious bowel disease and those who do not. After a positive test, people are invited for colonoscopy - a sort of articulated tube that is passed up the bowel. Most people invited for colonoscopy don't have cancer. Only about 5% of those with positive FIT tests have cancer. About 25% have other bowel diseases, but most have nothing serious wrong at all. So they have the inconvenience and discomfort of colonoscopy but don't get any benefit from it. The investigators want to try adding another test, the volatile organic compound (VOC) test, to see if the investigators can separate those with positive FIT tests who do have something wrong, from those who don't. The VOC test uses a urine sample. Using both tests might also be better for detecting cancer. FIT alone misses about 20%. So the investigators think that using both tests might not only be better for detecting cancer, but also might mean that a lot of people will avoid having to have colonoscopy. This study will recruit 1,819 participants with bowel symptoms from NHS trusts in the UK. They will provide stool samples for FIT and urine for VOC analysis. They will have colonoscopy to get a definite diagnosis. Then the investigators will look at their FIT and VOC test results to see if in future, people with both tests negative.

There is currently disparity between demand and available resources for colonoscopy. At present around 300,000 participants (and rising) are being referred annually to NHS trusts suspected of colorectal cancer (CRC). These participants are offered invasive colonic examinations (colonoscopy or CT colonography) but only 30% will have significant bowel disease. Significant bowel disease includes neoplasia (cancer and benign tumours) and significant treatable benign conditions such as inflammatory bowel disease and microscopic colitis. Of the remaining 70%, 40% have completely normal colonic investigations and 30% have functional bowel conditions such as irritable bowel syndrome or diverticular disease. Set against this there is and will remain for the foreseeable future a capacity shortfall for colonoscopy. This limits the ability of the NHS to extend colorectal cancer detection within the Bowel Cancer Screening Programme (BCSP) or to target those participants that present for the first time through the Emergency Department (25% of all colorectal cancer diagnoses). The increasing demand, limited capacity and lack of a triage tests have left NHS trusts with a conundrum of how best to stratify those with symptoms and at risk of SBD including CRC. The National Institute for Health and Care Excellence (NICE) has recommended a stool test (faecal immunochemical testing for haemoglobin, known as FIT) in the assessment of those suspected of CRC. NICE have recommended 10 μgHb/g faeces as the cut off for investigation of people with low risk symptoms who, account for only 10% of those referred with suspected CRC. When applied to high risk symptom groups, FIT will miss a significant number of participants with CRC (\~10%) if used on its own at the threshold recommended by NICE (10 μgHb/g faeces). FIT will also miss a large number of significant potentially pre-cancerous polyps (\~40%). Early detection and removal of such polyps will reduce risk of CRC. The Bowel Cancer Screening Programme (BCSP) has set a FIT cut off of \>120 μgHb/g faeces, compared to the NICE threshold of 10 μgHb/g faeces, but even at the lower threshold recommended by NICE some SBDs will be missed. Whilst a lower threshold might improve detection of SBD, it will increase the number of colonoscopies that find no abnormality. Consequently, the investigators have been investigating a urine test (in addition to FIT) to improve detection of participants with SBD with a view to reducing unnecessary colonoscopies. The urine test analyses volatile organic compounds (VOCs) that originate from the body and provides a chemical 'fingerprint' that is disease specific. Stool FIT and urine VOCs identify different biological characteristics of SBD - haemoglobin (as a marker of excess blood loss) versus metabolic response to inflammation. In a preliminary study of 562 participants, stool FIT on its own (at a threshold determined from the data) detected 80% of those with colorectal cancer. However, the addition of urine chemical testing improved this to 97%. The number of CRC cases missed by combined FIT and urine chemical testing is similar to that of colonoscopy, the current gold standard.