Screening for colorectal cancer: the role of clinical laboratories

Abstract

Background:

Colorectal cancer (CRC) is a leading cause of cancer incidence and mortality. Screening can result in reductions in incidence and mortality, but there are many challenges to uptake and follow up.

Content:

Here, we will review the changing epidemiology of CRC, including increasing trends for early and later onset CRC; evidence to support current and emerging screening strategies, including non-invasive stool and blood based tests; key challenges to ensuring uptake and high quality screening; and the critical role that clinical laboratories can have in supporting health system and public health efforts to reduce the burden of CRC on the population.

Summary:

Clinical laboratories have the opportunity to play a seminal role in optimizing early detection and prevention of CRC.

1.0. Introduction

Colorectal cancer (CRC) is the second most common cause of cancer death in the United States, and a leading cause of cancer incidence and mortality worldwide. Incidence and mortality from CRC can be reduced through screening. Several challenges to optimizing CRC prevention and screening exist, including changing epidemiology that includes increasing incidence among younger age groups, as well as suboptimal participation in screening and guideline appropriate follow up. This review provides an overview of these challenges, including the potential roles that clinical laboratories can play in partnering with health systems and public health authorities to reduce the burden of CRC.

2.0. Epidemiology of colorectal cancer

2.1. CRC Burden and Inequities

Colorectal cancer (CRC) is common worldwide, and in the United States specifically it is the third most common incident cancer, second leading cause of cancer death, and is expected to account for 153,020 new cancer cases and 52,550 deaths in 2023 (1). CRC incidence and mortality vary across demographic groups in the US, with higher incidence and mortality observed for Alaska Native, American Indian, and non-Hispanic Black individuals compared to non-Hispanic White individuals. CRC presents at a more advanced stage in non-Hispanic Black, Alaska Native, American Indian, and Hispanic people compared to non-Hispanic White people (1). Five year CRC survival is closely related to stage, at over 90% for localized cancer and less than 15% for distant stage disease, underscoring the key role of early detection in optimizing outcomes (2).

2.2. Changing CRC Trends By Age, Anatomical Location, and Stage

In the US, overall CRC incidence (and mortality) have been decreasing over time (Figure 1, panel A). This decrease is noted across all racial/ethnic populations with incidence decreasing by over 30% in Black and White individuals from 2000 to 2017. However, there are several established and emerging trends of concern. Early onset CRC incidence under age 50 years has been rising for over 20 years (Figure 1, panel B), to the point that incidence among individuals age 45 today is similar to historic incidence among 50 year-olds. This trend is expected to continue, with CRC expected to become the leading cause of cancer death among individuals age 20–49 years by 2030 (3). Furthermore, disparities in early onset CRC mortality exist, with improvements in survival over the last 20 years seen among White individuals, but not among Black, Asian or Hispanic individuals (4). These trends may be linked to social determinants of health such as lack of access to healthcare resources, subsequently leading to delayed CRC diagnosis and advanced disease stages at presentation (5).

Figure 1:

Age- and delay-adjusted trends in Colorectal cancer (CRC) incidence among men and women 2000–2020, Surveillance Epidemiology and End Results Program 22 for all ages combined (Panel A), and stratified by age <50, 50–64, and >64 years (Panels B, C, and D). The data demonstrate that CRC incidence is rising among individuals <50, and that incidence among individuals age 50–64 has been rising since 2013. Data accessed May 30 2023 at https://seer.cancer.gov/statistics-network/explorer/

Worrisome trends regarding later onset CRC above age 50 are also emerging, with an apparent flattening in incidence over time for 50 to 64 year olds (Figure 1, panel C) and some reports showing increases in CRC incidence among 50 to 59 year olds (4–6). Notably, increases in incidence have been mainly attributable to rectal, and to a lesser extent, distal cancer with prominent increases in advanced stage disease (1). Observed trends in the US are also being reported from around the world, with increases in early onset CRC incidence ranging from 1.8% – 3.1% in Denmark, New Zealand, Australia and the United Kingdom (7).

It has been postulated that these trends are attributable to either a birth cohort or time period exposure effect, whereby persons born 1950 and later experience substantially higher CRC risk compared to those born prior to 1950 (1). Features of the postulated birth cohort CRC phenomenon include: increased risk for early onset CRC under age 50; emerging increased risk for CRC among individuals 50 and older; higher risk for rectal cancer, and elevated risk for more advanced stage at diagnosis (Figure 2) (1). Hypothesized reasons for increasing risk include perinatal exposures, such as sulfonamide antibiotics and maternal obesity, exposures during childhood and adolescence, like sugar sweetened beverages and obesity, and adult exposures, such as alcohol, but the full range of risk factors and mechanisms have yet to be defined (Table 1) (8–13).

Figure 2:

Birth cohort colorectal cancer. The phenomenon of birth cohort Colorectal cancer (CRC) can be conceptualized as including, among persons born 1950 and later: 1) increased risk for early onset CRC before age 50; 2) emerging increased risk for CRC age 50 and older; 3) increased risk for rectal cancer and advanced stage at diagnosis.

Table 1:

Hypothesized and Established Risk Factors for Birth Cohort and Early Onset CRC

Established Associated Factors	Hypothesized Risk Factors
Obesity	Dysbiosis
Sedentary lifestyle	Lack of breast feeding
Hyperlipidemia	Dietary patterns especially with increases in processed foods,
Diabetes	sugar, fat, and refined grains
Red or processed meats	High fructose corn syrup
Alcohol	Food additives (e.g. synthetic dyes, monosodium glutamate,
Tobacco smoking	titanium dioxide)
Sugar-sweetened beverages	Food insecurity
Maternal weight and weight gain during pregnancy	Elevated birth weight/childhood obesity
In utero exposures:	Excessive exposure to insulin and growth hormone in utero
Long-acting sulfonamide antibiotic exposure	Environmental toxins (e.g. pesticides, benzene)
Maternal obesity	Cesarean delivery
Bendectin (doxylamine; pyridoxine; dicyclomine)	Pediatric onset inflammatory bowel disease
	Periodontal disease
	Other antibiotic use:
	Ampicillin
	Amoxicillin
	Gentamycin

Associated and hypothesized factors pooled from multiple sources (8–13).

The changing epidemiology of CRC highlights the rationale for recent guideline recommendation changes by the US Preventive Services Task Force (USPSTF), the American Cancer Society, and others to initiate screening at age 45 instead of age 50 (14–17). Notably, some groups, such as the American Association of Family Medicine and the American College of Physicians continue to endorse initiation of screening at 50, citing a need for more evidence to support a lower age and concerns regarding the resources required to screen 45 to 49 year old individuals at a population level (18, 19). Whether initiating screening at age 45 or 50, trends towards increasing CRC risk underscore the importance of ensuring individuals eligible for screening have every opportunity to access either invasive tests such as colonoscopy, or non-invasive, lab-based screening tests such as the fecal immunochemical test (FIT) or stool DNA-FIT (sDNA-FIT).

3.0. Screening for CRC

3.1. Rationale

CRC screening meets widely utilized criteria for appraising screening strategies based on:

1. Frequency of CRC in the population;

2. Well understood pathogenesis, including a long lead time between progression from normal mucosa to precancerous polyps and subsequent CRC during which screening can be used for early detection and prevention;

3. Availability of effective interventions to prevent and treat CRC;

4. Wide availability of multiple acceptable test strategies for screening; and

5. Established cost-effectiveness (2, 20).

Randomized trials, observational studies, and modeling analyses all support the potential for screening to reduce incidence and mortality, and screening is widely endorsed by the USPSTF and others.

3.2. CRC Screening Options

USPSTF recommends a menu of strategies for screening, including annual guaiac fecal occult blood (gFOBT), annual fecal immunochemical test (FIT), stool DNA (sDNA)-FIT (Cologuard) every 1–3 years, computed tomographic colonography (CTC) every 5 years, sigmoidoscopy every 5 years (or every 10 years when combined with annual FIT), and colonoscopy every 10 years for average risk individuals. Additional Food and Drug Administration-approved tests include the colon capsule (Pill-Cam COLON 2; Medtronic Minneapolis, Minnesota, USA), as well as methylated serum septin 9 (Epi proColon, Epigenomics Inc., San Diego, California, USA), but these have not been USPSTF-endorsed, and Epigenomics recently stopped sales of the methylated serum septin 9 test in the US. It should be noted that outside the United States, the most common test offered for population screening is FIT, and that sigmoidoscopy may be offered as a once-only or an every 10 year screen (21).

3.2.a. Options: lab-based tests

The three currently available non-invasive tests take different approaches for screening, all utilizing stool samples taken at home.

3.2.a.i. gFOBT

The gFOBT evaluates for the presence of a peroxidase reaction with heme within a stool sample, and requires 3 samples with manual development and interpretation. Kits may be distributed at point-of-care or by mail, and returned in person or by mail. Challenges with gFOBT include the need for multiple samples, and patient instructions that recommend dietary and medication restrictions such as avoidance of red meat and non-steroidal anti-inflammatory medications, despite mixed evidence regarding the importance of such restrictions (22–24).

3.2.a.ii. FIT

FITs utilize an antibody against the globin moiety of heme to detect bleeding from colorectal neoplasia, and generally require a single stool sample. Available FITs in the US include quantitative FITs as well as qualitative FITs, though all available tests are only currently approved for reporting a qualitative (abnormal/normal) result in the US. Kits may be distributed at point-of-care or by mail, and returned in person or by mail. Quantitative FITs can be batch machine processed, but qualitative FITs require manual development and interpretation. Advantages of FIT over gFOBT include ability to use a single instead of multiple stool samples, manufacturer instructions that do not specify a need for dietary or medication restrictions, and superior test characteristics, to the point that expert societies recommend use of FIT over gFOBT.(25) Substantial within-brand variation in FIT performance has been observed, and thus far the USPSTF has only formally recommended the Polymedco OC Sensor quantitative FIT, and OC Light qualitative FIT for screening (14, 26).

3.2.a.iii. sDNA-FIT

The sDNA-FIT is a multi-marker assay that includes assessment for Kirsten rat sarcoma virus (KRAS) mutations, abnormal Neuregulin 4 (NDRG4) and bone morphogenetic protein 3 (BMP3) methylation, and hemoglobin concentration (27). Using these measurements, a proprietary algorithm is used to generate a qualitative positive/negative result. A single sDNA-FIT is currently available on the market (Cologuard). Test orders are sent to the company, which follows up with patients to mail the testing kit, promote test completion, and assist with test return to a central for processing. The test requires patients to collect a whole spontaneous bowel movement, take a sample for FIT, mix a preservative buffer into the remaining sample, and package all contents for return by a mail courier. Ordering tests and viewing results can either be done entirely through clinical providers accessing a web-based ordering portal, or by working with the company to embed the ordering interface into electronic health record interfaces such as Epic.

3.2.b. Evidence to support use and comparative performance

3.2.b.i. Test Characteristics

Sensitivity, specificity, and test positivity vary widely across tests (Table 2). Sensitivity for CRC and advanced neoplasia (adenomas with villous features, high grade dysplasia, size ≥1cm or CRC) are between 50 – 75% and 7 – 21% for gFOBT, 74 – 81% and 25 – 27% for FIT; 93% and 47% for sDNA-FIT, using colonoscopy as a reference standard. Sensitivity for CRC and advanced neoplasia are estimated to be 95% and 95% for sigmoidoscopy, and 95% and 95% for colonoscopy, respectively (28). Specificity for CRC and advanced neoplasia is between 96 – 98% and 96 – 99% for gFOBT, 93 – 94% and 95 – 96% for FIT; 85 – 89% for sDNA-FIT, and between 86 – 89% for colonoscopy, respectively (2, 26). Specificity for colonoscopy was downgraded for detection and removal of non-adenomatous polyps, and was calculated via the formula:

Table 2.

Characteristics of USPSTF-recommended CRC screening strategies and impact on CRC incidence and mortality

Test Strategy	Approach/Assay	CRC Sensitivity	CRC Specificity	AN Sensitivity	AN Specificity	Positivity Rate	Relative Impact on CRC Incidence vs No Screening	Relative Impact on CRC Mortality vs No Screening	Comment
Guaiac FOBT	Peroxidase reaction with heme in stool	50 – 75%	96 – 98%	7 – 21%	96 – 99%	10.1%	0 to ↓ 20%*	↓ 9–22%	Risk reduction estimates based on RCT data
FIT	Antibody against globin moiety of heme	74 – 81%	93 – 94%	25 – 27%	95 – 96%	6.4%	↓ 10%	↓ 10%	Risk reduction estimates based on observational data
sDNA-FIT	Algorithm using results of testing for 7 DNA markers in occult blood	93%	85%	47%	89%	13.5%	NR	NR	--
CT Colonography	CT scan of colon requiring bowel prep and oral contrast	86 – 100%	75%	89%	94%	13.4%	NR	NR	--
Sigmoidoscopy	Flexible endoscope used to evaluate rectum and distal colon	95%	87%	95%	87%	17.3 – 23.4%	↓ 22%	↓ 26%	Risk reduction estimates based on RCT data
Colonoscopy	Flexible endoscope used to evaluate entire colon and rectum	95%	86 – 89%	95%	86 – 89%	35.3%%	↓ 47 – 69%	↓ 50 – 68%	Risk reduction estimates based on observational data and a single RCT

Data regarding sensitivity, specificity, positivity rate, and relative impact on CRC incidence/mortality pooled from several sources (2, 26, 35, 77–89)

Abbreviations: CRC, colorectal cancer; AN, advanced neoplasia; RR, relative risk; IRR, incidence rate ratio; NR, not reported; HR, hazard ratio; *1 out of 4 large RCTs of gFOBT showed incidence reduction

(Patients correctly identified as not having CRC on colonoscopy) ÷ [(Patients correctly identified as not having CRC) + (patients without CRC incorrectly identified as having CRC on colonoscopy)]. Generally lower performance for sensitivity among non-invasive tests, particularly for advanced neoplasia, is the fundamental reason why non-invasive tests need to be repeated relatively frequently.

3.2.b.ii. Acceptability

Randomized trials have shown that participation rates are substantially higher when non-invasive tests such as FIT, or a choice of tests such as FIT or colonoscopy, are offered compared with offering only colonoscopy (29–32). The highest rates of up-to-date screening have been shown in health systems consistently offering a combination of test strategies, highlighting the important role of lab-based non-invasive tests in optimizing screening participation rates for health systems and at the population level (33).

3.2.b.iii. Impact on Incidence and Mortality (Table 2)

Randomized trials utilizing no screening as a comparator have shown that exposure to gFOBT can reduce mortality, and in one trial, incidence (21, 34). RCTs have shown exposure to sigmoidoscopy substantially reduces CRC incidence and mortality through the detection and removal of precancerous polyps (21). A recently reported RCT has shown that inviting individuals to colonoscopy versus no invitation can reduce CRC incidence; a reduction in CRC mortality was shown on per-protocol analyses among those participating in screening but not on intent-to-screen analysis (35). Observational studies have shown exposure to FIT is associated with reduced risk for incidence and mortality, but RCTs results are not yet available (36). No RCTs or observational studies linking sDNA-FIT or CT colonography to reduced incidence and mortality have been reported. Head-to-head comparisons of various strategies such as FIT, sigmoidoscopy, and colonoscopy are underway, but results comparing the relative impact on cumulative CRC incidence and mortality are not yet available.

Given that the sensitivity and specificity of all available tests are well understood, modeling studies have been utilized to fill evidence gaps regarding the potential comparative effectiveness of available tests on incidence and mortality. These have suggested that similar rates of cancer prevented and deaths averted can be achieved with all strategies currently recommended by the USPSTF, albeit assuming perfect participation in screening and follow up, and optimal quality of testing. While perfect participation is an aspirational goal, the best test is one that is done, and the USPSTF has recommended a “menu of options” to optimize participation (14).

3.2.c. Emerging Lab-Based Non-Invasive Test Strategies

Advances in translational science, as well as novel healthcare policies, promise to usher in a new wave of lab-based, non-invasive strategies for CRC screening. Novel molecular approaches include blood testing for aberrant epigenetic and genetic changes in circulating tumor DNA (ctDNA) and proteomic (including glycoproteomic) patterns, as well as stool testing for epigenetic and genetic changes in tumor and polyp derived DNA, micro-RNA patterns, and protein markers (37). Importantly, in 2021 the Centers for Medicare and Medicaid Services issued a National Coverage Decision in the US supporting coverage of any novel blood-based test that garners FDA approval and has at least 74% sensitivity, and 90% specificity for CRC. This was done in order to provide “a pathway for Medicare coverage to support innovation and to accelerate access to emerging blood-based biomarker screening tests as soon as patient and test criteria are met” (38). This coverage decision has provided key target metrics that are expected to result in development of a new wave of blood-based tests for CRC screening. It may also suggest the Centers for Medicare and Medicaid services will set a similar standard for coverage decisions surrounding novel stool-based CRC screening tests. As a result of evolving science for biomarker detection and recent policy decisions, results of 4-large scale, prospective studies of novel, lab-based screening tests which are near completion among individuals at average risk for CRC undergoing usual care screening colonoscopy are highly anticipated (Table 3). These include one study of a blood-based assay analyzing ctDNA patterns, one study of a blood based assay utilizing ctDNA and proteomic markers, a study of a stool-based test evaluating abnormal expression of 8 RNA biomarkers combined with FIT (sRNA-FIT), and one study of an updated version of the currently approved sDNA-FIT test. Notably, the study of the updated sDNA-FIT also includes assessment of a blood-based assay for genetic and epigenetic alterations. Additional earlier stage studies leveraging ctDNA, proteomic, glycoprotein, and epithelial markers in the blood, as well as DNA, RNA, proteins, and microbiome patterns in the stool, have also been reported and show promise for CRC screening, but performance requires verification in large scale prospective studies (Table 3).

Table 3.

Emerging Non-Invasive, Lab-Based Strategies for CRC Screening

Later phase, large-scale prospective studies
Study	Specimen Source	Assay/Approach	Study Design	Enrolled subjects (n)	Test Characteristics
Evaluation of the ctDNA LUNAR Test in an Average Patient Screening Episode (ECLIPSE)(90)	Blood	Measurement of ctDNA	Observational study of average risk patients ages 45–84 years old, undergoing routine CRC screening	22,877	CRC sensitivity: 83%, CRC Specificity: 90%, AA Sensitivity: 13%*
Prevention of Colorectal Cancer Through Multiomics Blood Testing (PREEMPT CRC)(92)	Blood	Measurement of tumor and non-tumor derived signals from ctDNA, epigenetic and protein biomarkers	Observational study of average risk patients ages 45–85	>30,000	Not published yet, prior data from earlier trial shows CRC sensitivity: 94%, CRC Specificity: 94%*
Colorectal Cancer and Pre-Cancerous Adenoma Non-Invasive Detection Test Study (CRC-PREVENT)(40)	Stool	Multi-target stool RNA test plus a FIT	Interventional study of average risk individuals >45 years of age	8,289	CRC sensitivity: 94%, AA sensitivity: 45%, Specificity: 88%
Clinical Validation of An Optimized Multi-target Stool DNA (Mt-sDNA 2.0) Test, for Colorectal Cancer Screening “BLUE-C”(93)	Stool	Mt-sDNA 2.0 screening test	Observational study including patients age 40 and older who are enrolled in screening colonoscopy and will also complete mt-sDNA 2.0 test and FIT	23,494	CRC sensitivity: 94%, CRC Specificity: 91%, AA Sensitivity: 43%*
Early phase studies
Sample Collection Study for the CellMax Life Circulating Tumor Cell and Circulating Tumor DNA Platforms for the Early Detection of Colorectal Cancer and Adenomas (91)	Blood	Evaluation of aberrations in ctDNA via NGS, detection of circulating epithelial cells	Observational study of average risk patients ages 45–80	1,038	CRC sensitivity: 92.1%, AA sensitivity: 54.5%, AA specificity: 91%*
Collection of Samples USOPTIVAL Study (96)	Blood	Evaluation of cell-free DNA methylation and fragmentation characteristics, tumor-derived signal deduction and machine learning algorithm	Observational study of patients between the ages of 45 and 84 who were either average risk or had suspected AA or newly diagnosed CRC that has not been resected	997	CRC sensitivity: 93%, AA sensitivity 54%, AA specificity: 92%*
Non-invasive Identification of Colorectal Cancer and Adenomas in Early Stages (NICE)(95)	Blood	InterVenn Glycoprotein test	Prospective, multi-site study using glycoproteomic testing for early detection of advanced adenoma and colorectal cancer for average-risk patients undergoing routine screening colonoscopy	575	Not available

^*Data available from abstracts only

Data from the above clinical trials were cited from multiple sources(90–96)

CRC, colorectal cancer; ctDNA, Circulating tumor DNA ;FIT, fecal immunochemical test; MTs-DNA, Multitarget stool DNA; AA, advanced adenoma; NGS, next generation sequencing.

3.2.c.i. Early results from large trials of non-invasive tests

Results for the Guardant ctDNA blood-based Shield test have been reported in abstract form for a study that included 65 CRC cases among an age/sex stratified random sample of 7,796 cancer-free controls (39). The test was reported to have 83% sensitivity for CRC (54/65, 95% CI 72–90%), 13% sensitivity for advanced precancerous lesions (147/1,116; 95% CI 11–15%), and 90% specificity for the absence of CRC or advanced precancerous lesions (5,982/6,680; 95% CI 89–90%). Whereby advanced adenoma is defined as any adenoma with greater than or equal to 10 mm in size or with greater than 25% villous histology or high-grade dysplasia. Results for the Geneoscopy sRNA-FIT study report 94% sensitivity for CRC, 45% sensitivity for advanced adenoma, and 88% specificity for their cohort of 8,289 enrolled subjects (40). Also in press release form, the study of a second generation sDNA-FIT has reported 94% sensitivity for CRC, 43% sensitivity for advanced precancerous lesions, and 91% specificity for their cohort of over 20,000 subjects (https://www.exactsciences.com/newsroom/press-releases/next-generation-cologuard-test-demonstrates-94-percent-sensitivity, accessed October 2023). Results from the other large prospective studies are expected to be made available within the next two years. Available results may herald an opportunity to make these and other non-invasive blood and stool-based CRC screening tests available for clinical use in the near future. Novel FDA approved, CMS-covered tests could have the potential to optimize screening through better acceptability, availability, detection of neoplasia, or all 3 relative to 1 or more currently available tests.

4.0. Opportunities for clinical laboratories to address current challenges to CRC screening

While the rationale and evidence to support CRC screening are robust, outcomes of CRC screening remain suboptimal due to multiple challenges. Clinical labs can play a crucial role in helping optimize CRC screening and prevention by taking advantage of opportunities to address many current challenges (Table 4).

Table 4.

Clinical Lab Opportunities for Optimizing Colorectal Cancer Screening and Outcomes

Opportunity	Challenge	Intervention
Optimize screening participation	Suboptimal screening rates	Establish infrastructure to implement mailed outreach programs
		Implement embedded EHR tools to allow for standing orders at point of lab care for screening tests such as FIT for patients not up to date (“Lab-FIT”)
Optimize screening effectiveness	Subpar patient instructions	Distribute instructions following best practices for literacy, with consideration for wordless instructions, multiple translations, and links to instructional videos
	Distribution of low performance stool tests	Phase out gFOBT and more broadly adopt FIT
		Select FIT brands with consistently good performance characteristics
	Discarding of irreplaceable FITs due to missing collection dates	Result FITs missing collection dates or out of window from collection to processing and report abnormal results, and advise repeat testing for normal results
	Non-reporting of quantitative FIT values that correlate closely with probability of neoplasia	Support implementation of quantitative FIT reporting Report quantitative fecal hemoglobin values
	Low rates of colonoscopy follow up after an abnormal non-invasive CRC screening test	Flag abnormal results as critical values, provide education on test result interpretation, partner to create registries of patients with abnormal lab-based CRC screening test results
	Socioeconomic variability in access	Promote access to FDA/CMS approved testing
Reduce inappropriate screening test exposure	High rates of FIT overuse/misuse in emergency department and inpatient settings	Eliminate or restrict emergency department and inpatient use of CRC screening tests such as FIT and guaiac FOBT

4.1. Optimize screening participation

Screening participation rates remain suboptimal, and as of 2021 just 59% of the US population was up to date in terms of a CRC screening test, far short of the 80% target set by the National Colorectal Cancer Round Table (1, 41). A complicating factor is that screening suffered a setback as a result of the COVID-19 pandemic and estimates suggest the negative impacts on CRC-incidence and CRC-mortality could last for years (42). Modeling studies suggest pandemic-related disruptions to screening can be mitigated through expansion of stool-based testing (43). Improving overall screening rates will likely rely heavily on lab-based stool testing, and potentially in the near future blood tests, underscoring the role that clinical labs will have in facilitating optimal participation.

4.1.a. Partner to support mailed FIT outreach

Mailed outreach is one of several evidence-based strategies for increasing CRC screening completion along with visit-based FIT distribution, patient navigation, patient reminders and patient education (44, 45). Among all strategies for promoting CRC screening completion, mailed outreach appears to be the most effective, based on indirect comparisons. Meta-analyses have shown that offering mailed outreach can result in an absolute 28% increase in screening completion compared to usual, health-system visit-based offers for screening (46). Real world examples of implementation have shown impressive results. Kaiser Permanente established a mailed-FIT program for nearly 700,000 eligible screening-aged adults in 2008 that led to dramatic increases in proportion of patients up to date from 38.9% in 2000 to 82.7% in 2015, along with concurrent decreases in CRC incidence and mortality, and elimination of disparities in CRC incidence and mortality between Black and Non-Hispanic White individuals (33, 47). Mailed FIT approach is the standard approach to population-based CRC screening in over 15 countries around the world (21). Clinical laboratories can partner with providers to establish necessary infrastructure to successfully implement a mailed outreach program and significantly increase rates of up to date CRC screening utilizing guides available from the Centers for Disease Control, National Association of Chronic Disease Directors, National Colorectal Cancer Roundtable, and others (48–50).

4.1.b. Collaborate to allow for standing orders for FIT

A policy of having standing orders for FIT that can be acted on by different members of the healthcare team at time of usual healthcare visits, including primary care, specialty or nurse visits, has been shown to facilitate screening completion (51). We have observed some centers leverage this approach to a concept known as “Lab-FIT”, in which lab personnel, as part of patient interaction during lab visits, make note of open/overdue healthcare reminders, and, in the case of patients not-up-to-date with CRC screening, offer to distribute FIT kits to promote screening completion. Implementing Lab-FIT, and perhaps in the future even allowing for standing orders for blood-based CRC screening tests to patients not up-to-date at the time of usual care blood draws, could be one opportunity for clinical labs to help promote screening participation. Lab-FIT could lead to inappropriate testing among those with severe comorbidities. As such any implementation of this strategy should include multidisciplinary discussion with primary care clinicians and other stakeholders to set up clear parameters for consideration for this strategy.

4.2. Optimize screening effectiveness

Screening effectiveness can be impacted by problems with high quality test completion, selection of tests, and systems barriers to test ordering and reporting as well as follow up to colonoscopy completion after an abnormal non-invasive test. Clinical laboratories can play a large role in overcoming these barriers.

4.2.a. Distribute Optimal Patient Instructions for Stool-Based Tests

Reliance on patients for sample collection and handling naturally leads to increased variability in sample quality, and highlights the importance of effective patient instruction. Instructions deemed complex and burdensome by patients are a barrier to screening adherence, and patient literacy and language must be addressed to enhance comprehension (52). Effective interventions include utilization of low-literacy instructions, which led to decreased rates of mishandled FIT samples, and wordless instructions, which were unanimously preferred over written instructions in one cohort (53, 54). Manufacturer instructions for FIT often need to be supplemented with additional instructions to optimize proper test completion, for example by using strategies recommended at https://research.kpchr.org/mailed-fit/ (date accessed October, 2023). Resources available include wordless instructions, multiple translations, and QR codes with links to videos on test completion. By ensuring testing instructions are clear and easy to comprehend, clinical laboratories can help improve screening participation across the populations they serve, and improve lab efficiency by reducing work burden of following up on unsatisfactory samples.

4.2.b. Replace gFOBT with FIT

With a single sample, FIT is more sensitive and similarly specific for colonic bleeding due to neoplasia, can be done effectively with one sample instead of three samples as required by gFOBT, and, in the case of some FITs, be performed with automated reading (25). Replacing FIT with gFOBT has also been shown to increase screening completion (55). Clinical laboratories can work to improve overall screening effectiveness by phasing out gFOBT as a screening test and more broadly adopting FIT.

4.2.c. Use a High Quality FIT

Commercially available FITs vary in test performance, number of samples required, and cutoff of hemoglobin concentration to determine an abnormal/positive result. Among the nearly 65 FITs commercially available in the US, significant heterogeneity in terms of sensitivity and specificity exist, with sensitivity for advanced neoplasia and CRC ranging from 2–66% and 50–97% respectively, and specificity for AN ranging between 60–99% (26). In their evidence review of available FIT tests in the United States, the USPSTF found sufficient evidence to recommend the Polymedco OC-Sensor and OC-Light tests due to their consistently good performance characteristics, and notably did not find sufficient evidence to support use of many of the other commercially available FIT tests in the United States (26). A comparative effectiveness study comparing 5 different brands of FIT is underway, and may help to clarify relative performance of FIT brands (56). In the meantime, clinical labs have an opportunity to optimize detection and prevention of CRC by careful selection of which FIT brand is offered to their patients.

4.2.d. Reduce Rates of Ineffective FIT

Another challenge seen with FIT is user error and mishandling, which can lead to invalid tests. In routine practice, patients have been observed to provide FITs with missing collection dates, too much or too little fecal sample, a sample taken with a kit that is past the manufacturer expiration date, or kits returned at a time outside of the recommended window between sample collection and lab processing. In a study by Wang et al., among 1,871 patients, 19.8% of tests were mishandled, most commonly due to missing collection dates (53). In many instances these tests would be discarded, but as their study notes, rates of positivity were similar between the correctly labeled group and mishandled group. Time at room temperature, as well as variability in ambient air temperature can impact rates of hemoglobin degradation within FIT samples, as well as positivity rates, and refrigeration can mitigate degradation (57–60).

These data highlight the importance of promoting prompt return and proper storage of samples, underscoring the importance of research showing that providing a deadline can promote timely FIT return (61). Importance of time from sample collection to processing also supports current laboratory practice of monitoring collection dates, and provides a rationale for why some labs routinely discard FITs for which the collection date cannot be established. However, we view the practice of discarding FITs with missing collection dates or out of window with regard to collection date as a major missed opportunity to support patients when stool specimens are otherwise adequate. As has been discussed, screening participation in the population is suboptimal, and patients often collect and submit stool samples with great effort. Further, cancers intermittently bleed, such that a positive test on one date might be negative if repeated even at a short interval on another date. As such, these samples should be viewed as irreplaceable.

A patient and public-health centric approach would allow for a standard lab policy of processing FITs missing collection dates or out of the time window from collection. Results with an abnormal or positive result can be reliably interpreted as true positive results, while normal/negative results can be interpreted as uncertain, with a recommendation to repeat the test. This strategy could reduce the risk of missing the opportunity to identify a patient with an abnormal, mishandled sample who might not return a repeat test or might have a repeat test that is normal due to the intermittent nature of bleeding from colonic neoplasia. This scenario has important clinical implications, since 1 in 20 with abnormal FIT have CRC, and failure to follow up with colonoscopy after an abnormal FIT confers 2.5 fold increased risk for CRC-related mortality (62–64). As such, many clinical labs may be discarding irreplaceable abnormal FITs where patients may not submit a repeat sample or submit a subsequent negative sample, and thus miss an opportunity for early detection of CRC that could come from reporting results for the original mishandled sample.

4.2.e. Report Quantitative FIT Results

For patients who have had a quantitative FIT with an abnormal result, the absolute hemoglobin concentration is measurable, and research has definitively shown that increasing hemoglobin concentration is directly proportional to the positive predictive value for finding advanced adenomas and CRC at follow up colonoscopy. For example, the most commonly used quantitative FIT in the US is reported as abnormal/normal based on reaching a hemoglobin concentration of 100 ng hemoglobin/mL buffer, but the positive predictive value for advanced neoplasia increases from 25.9% to 31.0% for a 100 vs 150 ng hemoglobin/mL buffer threshold (65). Just as absolute values of white blood cell counts and other continuous lab measures are used to guide urgency of clinical care, absolute hemoglobin concentrations among patients with an abnormal FIT could also be used to guide urgency of colonoscopy follow up, particularly in settings with long wait times for colonoscopy. This issue is particularly salient at this moment given worldwide backlogs in colonoscopy associated with the COVID-19 pandemic (66). As such, clinical labs could have a role in advocating for reporting of qualitative and quantitative FIT results as a strategy for improving clinical care.

4.2.f. Optimize Result Reporting and Messaging

Non-invasive CRC screening can only be effective if individuals with abnormal tests complete the screening process with a colonoscopy. Yet, colonoscopy follow up rates for abnormal FIT tests range from 18–65%, and are estimated at 66% after an abnormal sDNA-FIT (67). Furthermore, colonoscopy completion rates are lower among people of lower socioeconomic station and minority groups (68, 69). In identifying reasons for poor follow up, common reasons include lack of patient knowledge of the result, lack of knowledge of the significance of the result, and provider failure to respond to a result (70).

We postulate clinical laboratories can play a vital role in optimizing messaging for both patients and providers. Abnormal results should be flagged as critical values which may optimize recognition by both patients and providers, potentially increasing rates of follow up. Furthermore, laboratories may be able to optimize patient understanding of results by providing information regarding risk of cancer and advanced neoplasia with positive results, and also reminding patients and providers that cancers intermediately bleed, making colonoscopy referral conditional on a second abnormal result clinically inappropriate. For example, an abnormal FIT result can be accompanied by a result comment, which we have employed at UC San Diego:

“Follow up colonoscopy is recommended to exclude colorectal cancer. Among individuals with abnormal FIT, 1 in 20 have colorectal cancer and 1 in 4 have an advanced polyp. Repeat testing after an abnormal result is not recommended because cancers can intermittently bleed.” (62).

Clinical labs can also potentially spearhead quality improvement projects aimed at increasing colonoscopy follow up after an abnormal test, with one way being through creation of weekly lists or a registry of patients with abnormal results. Leveraging lists and registries of abnormal test patients has been successfully implemented and proven to improve colonoscopy follow up in several health systems (71). We suggest clinical labs, in partnership with primary clinicians and patient navigators, can work towards improving identification of patients who require follow up, and ultimately improving outreach to those patients.

4.2.g. Improve Screening Access

Screening rates vary across the US population, with 51% of Hispanic individuals, 48% of Asian individuals, 52% of Native American/Alaska Native individuals up-to-date compared to 60% non-Hispanic White individuals, and lower screening rates among those with lower socioeconomic status, Medicaid insurance, and the uninsured (1). Non-invasive tests such as FIT have been shown to result in higher screening uptake among underserved populations (29). As such, addressing screening inequities will rely heavily on offering non-invasive stool tests. Clinical labs can play a role in addressing screening inequities by promoting access to all FDA and CMS approved testing for people of all socioeconomic and racial backgrounds.

4.3. Reduce inappropriate screening test exposure

While suboptimal CRC screening rates remain a main challenge, overuse and misuse of non-invasive lab-based tests are also a challenge with clinical and resource implications. Examples of overuse include FIT after exposure to prior colonoscopy, and misuse includes repeating FIT after a prior abnormal result and utilization of FIT for non-screening indications such as in the inpatient or emergency room setting among patients presenting with signs or symptoms of GI disease.

Inappropriate use is common, with rates ranging from 26 – 51% of ordered FITs (72, 73). Overuse contributed significantly, with one study finding that 28% of inappropriate FIT tests ordered were within the recommended surveillance interval from previous colonoscopy (72). One way clinical laboratories could reduce overuse is through an EHR intervention whereby orders considered inappropriate would be flagged and prompt a best practice alert or trigger a clinical decision support system (CDSS). Implementation of CDSSs have been shown to increase guideline adherence among ordering providers, as well as a reduce unnecessary testing (74).

Another significant contributor to inappropriate ordering is misuse, particularly in emergency department (ED) and inpatient settings. Bhatti et al conducted a study of FIT use specifically in the inpatient and ED setting and found that 99.5% of tests ordered were inappropriate, with the vast majority being for indications such as anemia or overt GI bleeding (75). This study led to FITs being removed from inpatient and emergency department ordering menus, a process termed “disinvesting”. Disinvesting from inpatient stool-test use has been adopted by several institutions, and in one health system led to a 98% reduction in inpatient use and cost-savings of approximately $40,000 per year (76). Clinical labs can consider similar methods of disinvestment in inpatient FIT by flagging or restricting use in the inpatient and ED settings, which can ultimately have downstream consequences such as delivery of higher-value care and significant cost-savings.

6.0. Conclusion

CRC is a highly prevalent disease, with increasing incidence among persons born 1950 and later due to a birth cohort risk effect. System-level strategies for preventing inappropriate use of FIT for screening purposes will need to be carefully managed in context of emerging enthusiasm for use of quantitative FIT (in combination with age) to help triage patients with GI symptoms such as change in bowel habit to fast track evaluation, as recommended by UK NICE guidelines (97). Screening for CRC can reduce incidence and mortality. Multiple options for screening exist, including invasive tests such as colonoscopy, and non-invasive, lab-based tests such as FIT and sDNA-FIT, with more non-invasive lab options, including blood-based options, on the horizon. Challenges to optimizing CRC screening outcomes include suboptimal screening rates (with variation by race/ethnicity and socioeconomic status), suboptimal quality of testing due to process issues, and low rates of follow up of abnormal non-invasive test results. Clinical labs can have a major role in improving CRC outcomes by partnering with stakeholders, including patients, clinicians, and health systems, to improve access to non-invasive tests, implement strategies to promote completion such as mailed FIT outreach, put in place processes and infrastructure that promote high quality non-invasive screening and abnormal test follow up. As such, clinical laboratories are poised to play a significant role in decreasing CRC morbidity and mortality throughout the populations they serve.

List of abbreviations:

CRC

colorectal cancer

US

United States

USPSTF

United States Preventative Services Task Force

FIT

fecal immunochemical test

sDNA-FIT

stool DNA-FIT

gFOBT

guiac fecal occult blood test

CTC

computed tomographic colonography

RCT

randomized control trial

ctDNA

circulating tumor DNA

CMS

center for Medicare & Medicaid Services

FDA

food and drug administration

COVID-19

coronavirus disease 19

AN

advanced neoplasia

EHR

electronic health record

CDSS

clinical decision support system

ED

emergency department

GI

gastrointestinal

RR

relative risk

IRR

incidence rate ratio

NR

not reported

HR

hazard ratio