Model-Based Evaluation of Colorectal Cancer Screening Effectiveness: Three Rounds of Multitarget Stool DNA Testing Versus One Colonoscopy

Using a validated microsimulation model, this study demonstrates that triennial next-generation multitarget stool DNA testing (ng mt-sDNA) yields superior 10-year cumulative effectiveness in detecting colorectal cancer and precancerous lesions, reducing mortality, and gaining life-years compared to a single decennial colonoscopy, primarily due to higher real-world adherence.

The Gist

Imagine your body is a vast, complex city, and Colorectal Cancer (CRC) is a dangerous, slow-growing weed that can take root in the city's plumbing (your colon). If left alone, this weed can grow into a massive, destructive vine that ruins the whole neighborhood.

The goal of cancer screening is to find and pull out these weeds before they become a disaster. But there's a catch: you have to actually do the screening.

This paper compares two different ways of keeping the city clean over a 10-year period:

The Two Strategies

1. The "Big Inspection" (Colonoscopy)

• How it works: You hire a team of expert inspectors to come in, shut down the pipes, and do a full, deep dive inspection of the entire plumbing system. They can see everything and pull out weeds immediately.
• The Catch: It's a huge production. You have to fast, take strong laxatives to clean the pipes, get sedated (put to sleep), and take a day off work. Because it's so invasive and inconvenient, many people simply skip it or delay it.
• The Plan: You only do this once every 10 years.

2. The "Smart Mailbox Check" (ng mt-sDNA Test)

• How it works: This is a new, high-tech test you do at home. You just collect a tiny sample of your stool (like checking your mailbox for a specific type of letter) and mail it back. It looks for DNA signs that the weeds are growing.
• The Catch: It's not as "deep" as the big inspection in one single moment.
• The Plan: Because it's easy, non-invasive, and requires no prep, people actually do it. The plan is to do this three times over the 10-year period (every 3 years).

The Big Experiment

The researchers built a super-computer simulation (a digital "what-if" game) involving 1 million virtual people. They wanted to see: If we follow the real-world rules (where people often skip the hard stuff), which strategy saves more lives over 10 years?

They didn't just look at how good the tests are in a lab; they looked at how good they are when real people are involved.

The Results: The "Mailbox Check" Wins

Here is what the simulation found, using simple analogies:

• Finding the Weeds: The "Big Inspection" (Colonoscopy) found fewer weeds because so many people skipped the appointment. The "Smart Mailbox Check" (done 3 times) found 13% more precancerous weeds and 11% more early cancers.
  • Analogy: It's like having a security guard check the front door once every 10 years vs. having a friendly neighborhood watch check the windows every 3 years. Even if the guard has a better flashlight, if they aren't there, the burglars get in. The frequent, easy checks caught more trouble.
• Saving Lives: The "Smart Mailbox Check" reduced cancer deaths by 33%, while the "Big Inspection" only reduced them by 20%.
• Extra Years of Life: People in the "Mailbox Check" group gained 62% more life-years (time alive) compared to the "Big Inspection" group.

Why Did the "Easy" Option Win?

The paper argues that adherence (people actually doing the test) is the secret sauce.

• Colonoscopy: Only about 38% of people actually show up for their 10-year checkup.
• Stool Test: About 72% of people actually do the test because it's easy and done at home.

Even though the colonoscopy is a "perfect" tool in theory, it fails in practice because people avoid it. The stool test isn't perfect in a single shot, but because people do it three times, the cumulative effect is much stronger. It's like taking three small steps forward versus trying to take one giant leap and falling short.

The "Perfect World" Scenario

The researchers also asked: What if everyone did the Colonoscopy?
In that "perfect world" where 100% of people get the big inspection, the Colonoscopy actually does slightly better at finding cancer. But the authors point out that perfection doesn't exist in the real world. Since we can't force everyone to get a colonoscopy, the strategy that gets the most people tested (the stool test) is the winner for saving lives.

The Bottom Line

Think of it like changing the oil in your car.

• Colonoscopy is like taking the car to a master mechanic for a full engine overhaul. It's great, but it's expensive, takes all day, and many people put it off until it's too late.
• The Stool Test is like checking the oil dipstick yourself every few months. It's quick, easy, and you actually do it.

This study suggests that for the average person, doing the easy test three times is a better way to keep the city (your body) safe than waiting for the one big, scary inspection that many people will never show up for. By making screening easier, we catch the "weeds" earlier and save more lives.

Technical Summary

1. Problem Statement

Colorectal cancer (CRC) remains a leading cause of cancer-related mortality in the United States. While screening is highly effective, current guidelines offer multiple modalities with differing intervals and adherence rates, primarily decennial colonoscopy (every 10 years) and triennial multitarget stool DNA (mt-sDNA) testing (every 3 years).

The core problem addressed is that single-round test performance metrics often fail to reflect cumulative effectiveness over time. Colonoscopy is invasive, resource-intensive, and suffers from suboptimal real-world adherence (estimated at 38%). Conversely, noninvasive stool tests like mt-sDNA have higher adherence rates (72%) but are often dismissed due to perceived lower sensitivity in single-round comparisons. This study aims to determine which strategy yields superior population-level outcomes over a 10-year horizon when accounting for real-world adherence, screening intervals, and follow-up compliance.

2. Methodology

The study utilized a microsimulation-based approach using the validated Colorectal Cancer and Adenoma Incidence and Mortality (CRC-AIM) model.

• Model Structure: CRC-AIM is a continuous-time agent-based model that simulates the natural history of CRC (adenoma initiation, growth, progression to preclinical cancer, and clinical detection). It has been rigorously validated against SEER incidence/mortality data and CISNET models.
• Cohort: The simulation modeled a cohort of 1 million average-risk individuals in the U.S., stratified by age (45, 55, 65) to represent commercially insured and Medicare populations. Results were weighted using U.S. Census data.
• Strategies Compared:
  1. Triennial ng mt-sDNA: Next-generation multitarget stool DNA testing (Cologuard Plus™) offered every 3 years for 9 years (3 rounds total).
  2. Decennial Colonoscopy: One-time screening colonoscopy offered at the start of the 10-year period.
• Key Inputs:
  • Test Performance: Sensitivity and specificity derived from the BLUE-C clinical trial (for ng mt-sDNA) and published literature (for colonoscopy). Sensitivity was stratified by lesion size (<6mm, 6–10mm, ≥10mm) and disease stage.
  • Adherence Rates (Real-World):
    • Initial screening adherence: 72% for ng mt-sDNA vs. 38% for colonoscopy.
    • Follow-up colonoscopy adherence (after positive stool test): 77%.
  • Time Horizon: 10 years.
• Outcomes Measured:
  • Precancerous lesions detected.
  • CRC cases detected (split by screening detection vs. symptomatic detection).
  • CRC mortality reduction (vs. no screening).
  • Life-years gained (LYG).
• Sensitivity Analyses: The study varied adherence rates by ±10% and ±20% and performed threshold analyses to determine the adherence levels required for colonoscopy to match ng mt-sDNA outcomes.

3. Key Contributions

• Cumulative vs. Single-Round Analysis: The study shifts the focus from single-test sensitivity to longitudinal cumulative effectiveness, demonstrating how repeated testing with high adherence can outperform a single high-sensitivity invasive test with low adherence.
• Real-World Adherence Integration: Unlike theoretical models assuming perfect adherence, this analysis incorporates empirically derived adherence rates, highlighting the critical impact of patient participation on screening efficacy.
• Next-Generation Test Data: It utilizes performance data from the BLUE-C study for the next-generation mt-sDNA test (ng mt-sDNA), which shows improved sensitivity over previous iterations.
• Policy Relevance: The findings provide a quantitative basis for addressing colonoscopy capacity constraints by validating stool-based strategies as a primary screening tool for the general population.

4. Key Results

Over a 10-year period per 1,000 individuals offered screening:

• Detection Rates:
  • Precancerous Lesions: ng mt-sDNA detected 13% more lesions (102 vs. 90) than colonoscopy.
  • CRC Cases: ng mt-sDNA detected 11% more CRC cases (10.8 vs. 9.7).
  • Screening vs. Symptomatic Detection: A critical finding was the mode of detection. Under ng mt-sDNA, 69% of CRCs were detected via screening, compared to only 18% for colonoscopy. This represents a >275% increase in pre-symptomatic detection.
• Mortality and Survival:
  • CRC Mortality Reduction: ng mt-sDNA achieved a 33% reduction in CRC mortality compared to no screening, versus 20% for colonoscopy.
  • Life-Years Gained (LYG): ng mt-sDNA resulted in 62% more LYG (5.1 vs. 3.1 per 1,000).
  • Population Subgroups: The advantage was consistent across commercial (64% higher LYG) and Medicare (61% higher LYG) populations.
• Sensitivity Analysis:
  • Even with adherence variations (±10% to ±20%), ng mt-sDNA maintained a 56–70% higher LYG advantage.
  • Threshold Analysis: For colonoscopy to match ng mt-sDNA in precancerous lesion detection, adherence would need to rise to 43.4%. To match LYG, adherence would need to reach 63.1%, and for mortality reduction, 64.0%.
  • Perfect Adherence Scenario: In a theoretical scenario where both strategies had 100% adherence, colonoscopy showed slightly higher mortality reduction (52% vs. 43%). This confirms that the superiority of ng mt-sDNA in the base case is driven primarily by real-world adherence gaps in colonoscopy, not inherent test failure.

5. Significance and Implications

• Clinical Practice: The study supports the expanded use of noninvasive stool-based screening (ng mt-sDNA) as a first-line strategy. It suggests that the "gold standard" of colonoscopy may be less effective in practice than repeated stool testing due to low participation rates.
• Health System Capacity: With over 60 million eligible adults and limited colonoscopy capacity (exacerbated by the pandemic and age guideline changes), shifting to high-adherence stool testing can alleviate system bottlenecks. It allows colonoscopy resources to be prioritized for diagnostic follow-up of positive stool tests rather than primary screening.
• Public Health Impact: By increasing the proportion of cancers detected at pre-symptomatic stages, ng mt-sDNA offers a pathway to significantly reduce CRC mortality at the population level.
• Limitations & Future Work: The authors note limitations including the exclusion of the sessile serrated pathway (SSP) in the model, the lack of formal cost-effectiveness analysis (which requires longer time horizons), and the assumption of static adherence. Future research should focus on long-term cost-benefit analyses and strategies to further optimize follow-up adherence.

Conclusion:
The modeling study concludes that triennial ng mt-sDNA screening demonstrates superior cumulative effectiveness compared to decennial colonoscopy over a 10-year period. This superiority is driven by higher real-world adherence and favorable longitudinal performance, resulting in greater detection of precancerous lesions and early-stage cancers, reduced mortality, and increased life-years gained.