Ahead of the anticipated USPSTF update on colorectal cancer screening guidelines that are likely to recommend initiation of colorectal cancer screening at age 45 for "average risk" individuals, we thought we’d take this opportunity to review data and practical considerations for the available screening methods.
Colonoscopy is the gold standard for colorectal cancer screening and has the rare advantage of being able to potentially simultaneously detect cancer/precancerous lesions (with an 89-98% sensitivity) and prevent disease. Most organizations recommend colonoscopy once every ten years, although follow-up time may vary depending on pathology or other risk factors. Colonoscopy is expensive, relatively inconvenient (often requiring anesthesia and a day taken off work), requires a cleanout, and is invasive, leading to a small but real procedural risk. These and other systemic factors have contributed to significant racial disparities in colonoscopy provision. When a colonoscopy happens, the results can be impressive: a 2018 case-control study from Kaiser looking at death from colon cancer found that primary screening colonoscopy in normal-risk patients was associated with decreased colorectal cancer mortality (adjusted odds ratio 0.33, 95% CI 0.21-0.52).
gFOTB and FIT Testing
Stool testing for occult blood has long been recommended for colorectal cancer screening and includes guaiac-based fecal occult blood testing (gFOBT) and fecal immunochemical tests (FIT testing). Both tests measure hemoglobin in the stool. gFOBT detects the peroxidase activity of heme, an enzyme that breaks down hydrogen peroxide and is not specific to human blood, making false positives possible due to foods which also have peroxidase activity, such as red meat. There are lots of reasons for blood in the stool, and a positive result requires a follow-up colonoscopy. FIT testing measures antibodies specific for human hemoglobin and may be qualitative or quantitative, resulting in a lower false-positive rate than gFOBT. Most guidelines recommend annual high-sensitivity gFOBT or FIT testing, but FIT testing is preferred because it only requires one stool sample instead of three for gFOBT, resulting in higher adherence with greater sensitivity (74%) and a similar specificity (94%). Both tests are associated with lower colorectal cancer specific mortality. Five randomized trials with over 400,000 participants found that repeated rounds of biennial screening with gFOBT decreased colorectal cancer specific mortality when compared with no screening (risk ratio [RR] 0.91, 95% CI 0.84-0.98 at 19.5 years to RR 0.78, 95% CI 0.65-0.93 at 30 years). One prospective study with over five million participants found that one-three rounds of screening with a biennial FIT test was associated with lower colorectal cancer mortality when compared with no screening (RR 0.90, 95% CI 0.84-0.95). In general, mortality reductions are similar for both men and women, and though some studies have included adults younger than 50 years, none stratified their results by that age group. In general, gFOBT or FIT testing are reasonable and more affordable alternatives to colonoscopy with minimal risk of harm other than missed cancers or adverse events resulting from follow-up diagnostic colonoscopy.
Multi-target stool DNA testing combines multiple molecular assays with a hemoglobin immunoassay. The molecular assay components look for aberrantly methylated BMP3, NDRG, and NDRG4 or mutated K-ras and beta-Actin found in cells shed from colonic neoplasms. The hemoglobin assay is similar to FIT testing alone. Stool may be collected at home. Stool DNA testing has higher sensitivity than FIT testing for detecting both colorectal cancer (92.3% vs 73.8%) and advanced adenomas and sessile serrated polyps (42.4% vs. 23.8%). It is also much better than FIT testing for detecting serrated sessile polyps larger than one centimeter (42.4% vs. 5.1%). There are no data for mortality outcomes. Stool DNA testing is, however, associated with more false positives than FIT testing and a positive DNA test requires a follow-up colonoscopy. If the colonoscopy is normal, close follow-up is required to make sure that a lesion was not missed. For those choosing stool DNA testing as their colorectal cancer screening option, screening is recommended every three years.
Computed Tomography Colonography (CTC)
CTC, also called virtual colonoscopy, is an oral contrast CT protocol study of the abdomen and pelvis focused on the colon and is recommended every five years when performed. The USPSTF reviewed seven heterogeneous studies evaluating the accuracy of CTC in 5,328 patients and reported 86%-100% sensitivity for colorectal cancer detection, and 67%-94% sensitivity and 86%-98% specificity for detection of adenomas at least 10 millimeters in size. Benefits of CTC compared with colonoscopy include relatively lower invasiveness, no need for sedation or analgesia, ability to detect potentially harmful extracolonic findings, and an extremely low rate of adverse events. But like colonoscopy, there is a lack of randomized trial evidence supporting its efficacy for the reduction of colorectal cancer incidence or mortality (Obaro et al., 2018). Another downside is the radiation exposure that can accumulate to meaningful levels over time and repeat scans. CTC may be preferable to colonoscopy for patients seeking to reduce discomfort and can be used for patients who are on anticoagulants that cannot be stopped for the procedure, but it does not allow for the ability to obtain biopsy samples or remove colorectal polyps (Kato et al., 2020). CTC appears to be more often used among patients aged 45-49 compared with patients over 65, among patients of color, and among patients with lower socioeconomic status or public insurance (Hong et al., 2020).
Never even heard of blood or urine testing for colon cancer screening? Join the club. Both appear to be potentially promising options for patients unable or unwilling to undergo colonoscopy or stool-based testing, but evidence for mortality reduction is not yet available. The serum test Epi proColon looks for the biomarker septin 9 and was found to have a sensitivity of 68% and a specificity of 79% to detect colorectal cancer based on one nested case-control study (n=6,845), although a second generation of the test may have better performance characteristics. Similar to stool-based testing, the low sensitivity for detection of advanced adenomas with the serum test (22%) makes ruling out adenomas unrealistic. Specificity of the serum test for detecting adenomas was 79% in the same small trial. Similarly, data from the one available small study (n=228) evaluating metabolomic urine testing (PolypDx) suggest that this has potential to be an option for detecting adenomas (sensitivity 22%, specificity 79%); further studies suggest improved sensitivity is possible with refined criteria. Again, any positive blood or urine screen would need to be followed up with colonoscopy. No consistent data on cost of blood or urine testing are available, but there may be issues with insurance coverage based on lack of robust data for either of these methods.
Bottom line: Efficacy and practicality differ among methods for colon cancer screening. The bottom line is that any screening is better than no screening at all. Perhaps the “best” method is the one that the patient in front of you is willing to undergo.
For more information, see the topic Colorectal Cancer Screening in DynaMed.