Immunoassay-based screening for colorectal cancer. More colonoscopies performed and more cancers detected than with Haemoccult.

In adults with no specific risk factors, colorectal cancer screening based on detection of occult blood in stools with a guaiac test (Haemoccult) has several technical limitations. Sensitivity is low, only about 50% for colorectal cancer. Immunochemical tests for faecal occult blood are now also available. Do they improve the performance of colorectal cancer screening? To answer this question, we reviewed the available literature using the standard Prescrire methodology. These immunochemical tests are specific for human haemoglobin, are read automatically, and provide quantitative results. Their sensitivity in detecting lesions, and the proportion of patients who subsequently require colonoscopy, depends on the chosen cutoff. In 9 screening studies including a total of about 86 000 participants, immunochemical tests had a sensitivity of between 61% and 89% for colorectal cancer, depending on the test modalities and positivity cutoffs. In France, immunochemical tests have been compared with Haemoccult for mass screening in a total of 85 149 persons. They detected nearly twice as many colorectal cancers as Haemoccult, but led to twice as many colonoscopies. Both methods detected about 1 cancer per 14 colonoscopies. The main risks of screening are those associated with colonoscopy, which results in about 3 serious complications (including perforation and bleeding) per 1000 procedures. We found no studies of screening-related complications in France. In practice, immunoassay-based tests appear to increase the number of colorectal cancers detected but are associated with a proportionate increase in the number of patients who undergo colonoscopy. In France, better assessment and supervision of colonoscopies resulting from mass screening are needed.