Non-invasive blood and stool CRC screening tests: Available modalities and their clinical application

Introduction

Colorectal cancer (CRC) screening significantly reduces CRC incidence and mortality, but only 65% of eligible individuals report being up-to-date with screening (1). Colonoscopy is the most widely used opportunistic screening method in the United States and is associated with many barriers to uptake. Providing patients a choice of colonoscopy and/or stool-based tests, improves screening adherence in randomized controlled trials (2,3). Non-invasive screening options have expanded from stool occult blood and multi-target DNA tests, to multi-target stool RNA tests, and novel blood-based tests, the latter only U.S. Food and Drug Administration (FDA) approved for patients who refuse colonoscopy and stool-based tests. This review summarizes the test characteristics of stool and blood CRC screening options and provides guidance on clinical implications of their use.

Stool occult blood tests

Guaiac-based fecal occult blood testing (gFOBT) significantly reduces CRC mortality by 33%-35% when implemented on an annual or biennial basis (4,5). Fecal immunochemical testing (FIT) has supplanted gFOBT with advantages including independence from dietary restriction and medication-related interference, use of antibodies specific to human globin, and the need for only a single stool sample.

The most common threshold for a positive FIT in the U.S. is ≥ 20 micrograms (μg) of hemoglobin per gram (g) of stool. FIT is approved by the FDA as a qualitative positive or negative result based on a threshold value (6). A meta-analysis summarized test characteristics of commercially available FITs at various detection thresholds (7). The CRC sensitivity and specificity was 75% and 95% for ≥ 20 ug hemoglobin/g stool, and 91% and 90% for 10 ug hemoglobin/g stool, respectively. The sensitivity for advanced adenomas ranged from 25% at 20 μg/g to 40% at a 10 μg/g. Programmatic use of FIT in adults ages ≥ 50 years at 20 ug/g of stool, in cohort and case control studies, has been shown to significantly reduce CRC mortality by 33%-40% and advanced stage CRC by 34% (8,9).

More than 57,000 average-risk individuals ages 50–69 years were randomized to biennial FIT or one-time colonoscopy and followed for 10 years (10). CRC mortality and incidence was similar between the groups: 0.22% with FIT vs. 0.24% with colonoscopy and 1.13% with FIT vs. 1.22% with colonoscopy, respectively. Thus, confirming biennial FIT screening is non-inferior to one-time colonoscopy in important CRC-related outcomes.

Multi-target stool tests

Two multitarget stool DNA tests (mt-sDNA) known as Cologuard™ and Cologuard Plus™ have been approved by the FDA. Both tests include a FIT (with a positivity threshold of 20 μg hemoglobin per gram of stool) combined with DNA methylation markers. The test result is qualitative, reported as a positive or negative. Cologuard markers include methylated BMP3, NDRG4, and mutant KRAS while Cologuard Plus assesses methylated LASS4, LRRC4, and PPP2R5C. The respective mt-sDNA tests were studied in 9989 of 12,776 and 20,176 of 26,758 average-risk individuals undergoing colonoscopy and the results were compared to a commercially available FIT (with a positivity threshold of 20 μg hemoglobin/gram of stool) (11, 12). In both trials, the sensitivity for CRC and advanced precancerous lesions was higher with the mt-sDNA tests compared to FIT but had a significantly lower specificity for advanced precancerous lesions versus FIT (Table 1). An age-related decline in specificity was noted in both trials with mt-sDNA, a trend not observed with FIT. This reduction may be attributed to age-related DNA methylation.

Multi-target stool RNA test

A multi-target stool RNA test (mt-sRNA) commercially available as ColoSense™ is FDA-approved. It combines FIT (at a positivity threshold of 20 μg hemoglobin/gram of stool) with RNA-based stool markers. The combined results of the RNA markers, FIT, and smoking status provide a qualitative single test result. In the trial, 8,920 adults aged ≥45 underwent the mt-sRNA test and FIT followed by colonoscopy (13).  The mt-sRNA showed higher sensitivity for CRC than FIT (94.4% versus 77.8%) and advanced adenomas (45.9% versus 28.9%) but lower CRC specificity (84.7% vs 94.7%) (Table 1). Unlike mt-sDNA-based tests, mt-sRNA showed consistent performance across age groups, addressing concerns about age-related declines in specificity attributed to DNA methylation.

Blood-based tests

In 2014, the first blood-based (BBT) CRC screening test known as Epi proColon™ was FDA but not Centers for Medicare & Medicaid Services (CMS) approved for average-risk adults ≥50 years of age who are offered and refused other U.S Preventive Services Task Force (USPSTF) endorsed CRC screening tests. It is a qualitative test for detection of circulating methylated Septin 9 (mSeptin9). The accuracy of mSeptin9 to detect CRC was assessed in a subset of 7941 asymptomatic average risk adults undergoing screening colonoscopy (14).  The sensitivity and specificity for CRC were 48% and 91.5%, respectively. The sensitivity for advanced adenomas was 11.2%. An increase in sensitivity to 63.9% and reduction in specificity to 88.4% for CRC was demonstrated in a sub-analysis of available samples where an additional (third) polymerase chain replicate was performed. Epi proColon™ is not currently reimbursed by Medicare and not endorsed in the latest USPSTF guidelines.

Technologic advancements have improved the detection of circulating tumor markers in the blood. The Shield™ BBT approved by the FDA in 2024 for average risk adults ≥ 45 years integrates three types of cfDNA data (epigenetic changes resulting in the aberrant methylation or fragmentation patterns, and genomic changes resulting in somatic mutations) into a positive or negative test result. In the trial, 22,877 average-risk, asymptomatic individuals ages 45–84 were enrolled and clinical validation was performed in 7,861 of the participants (15). The sensitivity for CRC was 83.1% which decreased to 55% for stage I tumors (Table 1). CRC specificity was 89.6% and the sensitivity for advanced adenomas and large sessile serrated lesions was 13.2%.

Another BBT SimpleScreen™, which is not yet FDA-approved, analyzed circulating, cell-free DNA methylation patterns in 27,010 evaluable average-risk, asymptomatic adults ages 45–85 years undergoing screening colonoscopy (16).  The sensitivity and specificity for CRC was 79.2% and 91.5%, respectively. Similar to Shield, the sensitivity for stage I CRC was low at 57.1%. The sensitivity for advanced precancerous lesions, a secondary endpoint, was 12.5% which did not meet the prespecified study criteria. 

Effectiveness and cost-effectiveness

Modeling studies have evaluated novel noninvasive CRC screening tests compared to FIT and colonoscopy (17-20). One compared a hypothetical BBT performed every 3 years that meets the minimum CMS threshold CRC sensitivity and specificity of 74% and 90%, respectively, to other established CRC screening tests beginning at age 45 (17).  Every 3-year BBT reduced CRC incidence and mortality by 40% and 52%, respectively compared to no screening. However, the reductions were much lower than yearly FIT (72% and 76%, respectively), every 10 year colonoscopy (79% and 81%, respectively), and triennial mt-sDNA (68% and 73%, respectively). The BBT resulted in fewer quality-adjusted life-years per person compared to the alternatives. Additionally, FIT, colonoscopy, and mt-sDNA were less costly and more effective. Advanced precancerous lesion detection was a key measure for a test’s effectiveness. BBT characteristics would require a CRC sensitivity and specificity of >90% and 90%, respectively, and 80% sensitivity for advanced precancerous lesions at a cost of ≤$120–$140 to be cost-effective compared to FIT at comparable participation rates.

Another analysis simulated colorectal neoplasia progression and compared clinical effectiveness and cost between annual FIT, every 3 year stool mt-sRNA, every 3 year stool mt-sDNA tests, every 3 year stool Shield™; these outcomes were compared to colonoscopy every 10 years and no screening in adults ≥ age 45 over different adherence rates (19). At real-world adherence rates of 60%, colonoscopy prevented most CRC cases and associated deaths. FIT was the most cost-effective strategy at all adherence levels. Between the multi-target stool tests and Shield™, mt-sRNA was the most cost-effective. Compared to FIT, mt-sRNA reduced CRC cases and deaths by 1% and 14%.

The third study evaluated CRC incidence and mortality, quality-adjusted life-years and costs with annual FIT, colonoscopy every 10 years, mt- sDNA tests, mt-sRNA test, and BBTs (20).  The latest mt-sDNA (Colguard plus™) and mt-sRNA achieved benefits approaching FIT but the Shield™ test was substantially less effective. The authors hypothesized that if 15% of the population substituted Shield™ for current effective CRC screening strategies, an increase in CRC deaths would occur and require 9-10% of the unscreened population to uptake screening with Shield to avert the increases in CRC deaths due to the substitution effect.

Clinical implications

The effectiveness of non-invasive screening strategies depends on their diagnostic performance, adherence, and ensuring a timely colonoscopy after a positive test. Two claims-based studies found 47.9% and 49% of patients underwent follow-up colonoscopy within 6 months of an abnormal stool or BBT CRC screening test, respectively (21-22).

Conclusions

Non-invasive stool mt-sDNA and mt-sRNA have higher effectiveness than the new BBTs. BBTs can lead to increased CRC mortality if substituted for the FDA and CMS-approved, USPSTF-endorsed, CRC screening modalities. If future BBTs increase their sensitivity for CRC (including early-stage CRC) and advanced precancerous lesions and decrease their cost, they may prove to have similar cost-effectiveness to stool-based tests. Currently, BBTs are not a substitute for colonoscopy or other stool tests and should be offered to patients who refuse other CRC screening modalities.  A personalized, risk-adapted approach, paired with improved adherence and follow-up are essential to optimize the population-level impact of CRC screening and ensure equitable, effective cancer prevention.