INTRODUCTION:Multitarget stool DNA (mt-sDNA) is a sensitive stool-based test for colorectal cancer (CRC) screening. Limited referral-center data suggest low rates of incident aerodigestive cancers after a positive mt-sDNA and negative colonoscopy; community-based outcomes are unknown. METHODS:Retrospective cohort study of adults referred for colonoscopy after positive mt-sDNA in a large community practice (2017-2021). Among patients with negative colonoscopy, we manually recorded post-colonoscopy recommendations, findings if additional colonoscopies had been performed, aerodigestive cancer diagnosis, and last clinical follow-up. Observed cancer incidence was compared with Surveillance, Epidemiology, and End Results (SEER) data. RESULTS:Of 1,176 patients completing colonoscopy after positive mt-sDNA, 328 (28%) had negative findings. Median cohort age was 67 (IQR 59-72) years. Colonoscopy was performed at a median of 53 days after mt-sDNA sample collection. Post-colonoscopy recommendations aligned with multi-society task force (MSTF) guidelines in 59% of patients. Thirty-two patients underwent repeat colonoscopy with a median of 3.5 (1.9-4.4) years later; 2 (6.3%) had advanced serrated lesions, and no CRC was detected. Median clinical follow-up was 4.4 (3.4-5.6) years. Seven patients developed aerodigestive cancers (2 pancreatic, 5 lung), consistent with expected age- and sex-adjusted SEER incidence (n=7.62) (p=0.823). DISCUSSION:In this community cohort, when a positive mt-sDNA was followed by negative colonoscopy, rates of subsequent advanced colorectal neoplasia and aerodigestive cancer were low. Most patients received recommendations for subsequent colonoscopy that were consistent with MSTF guidelines. Findings further support the current MSTF recommendation of no additional evaluation for extra-colonic malignancy in asymptomatic patients after high quality follow-up colonoscopy.

INTRODUCTION:Multi-target stool DNA (mt-sDNA) testing is a noninvasive colorectal cancer (CRC) screening test with higher sensitivity and real-world adherence than fecal immunochemical testing (FIT) or guaiac fecal occult blood testing (FOBT). However, data on adherence and follow-up colonoscopy (FU-CY) within individual health systems are limited. This study evaluated mt-sDNA adherence, FU-CY completion, and predictors within Massachusetts General Hospital (MGH). METHODS:We conducted a retrospective cohort study of 23,665 patients aged 45-75 years with mt-sDNA test orders in the MGH health system (June 2016-May 2023). Adherence was defined as return of a valid mt-sDNA kit. FU-CY was assessed within 12 months after a positive result. Multivariable logistic regression models identified patient- and provider-level predictors of adherence and FU-CY completion. RESULTS:Overall adherence to mt-sDNA screening was 78.4%, with a positivity rate of 10.5%. FU-CY completion within one year was 74.2%, with 91% completed within six months. Adherence was highest among adults aged 65-75 (80.1%), Asian or Pacific Islander patients (83.0%), Medicare beneficiaries (84.0%), rural residents (82.4%), and those with prior mt-sDNA testing (89.3%). Lower adherence was observed among Black (73.6%) and Hispanic (75.7%) patients, Medicaid beneficiaries (60.0%), and orders placed by nurse practitioners/physician assistants (71.7%). Multivariable analyses identified independent predictors of adherence and FU-CY completion, including race/ethnicity, payer type, income, provider specialty, geography, and patient return history. Age and sex were significant predictors of FU-CY completion. CONCLUSIONS:Within a large health system, mt-sDNA screening demonstrated high overall adherence and timely FU-CY completion, exceeding those historically reported for FIT/FOBT.

INTRODUCTION:Colorectal cancer remains a leading cause of cancer associated death in the United States and colonoscopy the primary screening strategy for prevention. Rates of adenomatous and serrated neoplasia detection are inversely associated with postcolonoscopy colorectal cancer. This crucial quality metric depends on accurate ascertainment of colorectal neoplasia findings from both endoscopy and histopathology records. We aimed to assess the feasibility of a random forest machine learning model to rapidly and accurately categorize colorectal neoplasia from electronic health record data.METHODS:A retrospective cohort study compared neoplasia detection rates among individuals undergoing colonoscopy at a large academic institution to develop a rule-based algorithm to categorize colorectal neoplasia from endoscopy reports and pathology systematized nomenclature of medicine - clinical terms. This cohort provided a large training set to develop a natural language processing system using a random forest approach to automatically classify unstructured pathology findings into adenoma, serrated, or advanced neoplasms. This system was manually validated through an independent holdout set.RESULTS:The training set comprised 35,953 unstructured pathology reports with matched systematized nomenclature of medicine - clinical terms from 95,188 unstructured colonoscopy reports. The final model was assessed on an independent holdout set of 337 manually annotated procedures obtaining an area under the receiver operating characteristic curve of 0.997 (confidence interval [CI] 0.994-1), 0.99 (CI 0.98-1), and 0.99 (CI 0.98-0.99) for prediction of adenoma, serrated, and advanced lesions, respectively.DISCUSSION:The random forest-based hybrid natural language processing system for classification of colonoscopy results was both accurate and explainable. NLP combined with effective machine learning algorithms can provide a scalable strategy for colonoscopy quality monitoring.

Adherence to colorectal cancer (CRC) re-screening is essential to maximize screening effectiveness. This study assessed adherence to a multi-target stool DNA (mt-sDNA) test among previous users in the USA across different payer types. Data from Exact Sciences Laboratories LLC (01/01/2023–12/31/2023) were used. Insured patients (45–85 years) who were shipped an mt-sDNA test during the data coverage period and had previously completed mt-sDNA screening with a negative result ≥ 2.5 years prior were included. Mt-sDNA re-screening adherence rate and mean time to test return were compared across payer types, and their associations with patient characteristics were assessed using multivariable regression models. Of 793,567 patients (50–75 years: 89.0

BACKGROUND:Atrophic gastritis (AG), particularly in the context of autoimmune mechanisms or chronic Helicobacter pylori infection, is a recognized precursor to gastric neoplasia. However, surveillance strategies remain inconsistent. This study aimed to determine the cumulative incidence and risk factors for gastric neuroendocrine tumor (gNET) and adenocarcinoma (AC) in a large United States-based cohort of patients with AG. METHODS:A retrospective cohort study of 224 patients diagnosed with AG at a single tertiary center between 2010 and 2023 was conducted. Patients with prevalent gNET (n = 25) or AC (n = 7) were excluded from respective survival analyses. Kaplan-Meier methods estimated 5- and 10-year cumulative incidence, and univariate Cox regression identified risk factors for neoplastic progression. RESULTS:Among 199 patients at risk of gNET, the 5- and 10-year cumulative incidence rates were 6% and 18%, respectively. Hypergastrinemia and alcohol use disorder were associated with increased gNET risk. For the 217 patients at risk of AC, the 5- and 10-year cumulative incidence rates were 3% and 9%, respectively. Severe dysplasia and a history of peptic ulcer disease were the strongest predictors of AC. Histologic severity alone was not a reliable predictor of neoplastic progression. CONCLUSION:Patients with AG face significantly elevated risks of both gNET and AC compared with the general population. Risk stratification based on clinical and biochemical factors - such as hypergastrinemia, alcohol use, and dysplasia - may better inform surveillance strategies than histology alone. These findings support more personalized endoscopic surveillance protocols in high-risk populations with AG.

BACKGROUND AND AIMS:An irregular Z-line is characterized by a squamocolumnar junction (SCJ) that extends proximally above the gastroesophageal junction by <1 cm, whereas Barrett's esophagus is defined as a columnar-lined esophagus (CLE) that extends proximally by ≥1 cm with the presence of specialized intestinal metaplasia on biopsy sampling. Measurement of the CLE is most accurate for lengths ≥1 cm, and, as such, guidelines do not recommend biopsy sampling of an irregular Z-line when seen on endoscopy. However, a CLE is often estimated by visual inspection rather than direct measurement, making this characterization imprecise. In this study, we present methodology to standardize the characterization of the SCJ, hypothesizing that the shape of the Z-line can be used as a surrogate classifier. We present a computer-generated algorithm capable of automated segmentation and shape complexity quantification of the Z-line. METHODS:We selected and manually segmented 849 images of the Z-line. We used the nnUNet framework (Nature Methods, Heidelberg, Germany) to train a model to segment the Z-line. An additional dataset of 58 videos containing the Z-line were obtained from the Mayo Clinic Endoscopy video library. A high-quality image containing the Z-line was selected from each video. Ten gastroenterologists (5 esophageal experts) rated each of the 58 video-image pairs containing the Z-line as "regular" or "irregular," including their degree of confidence. Fleiss κ statistics were used to determine interobserver variability. The "ground truth" classification was determined by the esophageal expert majority vote. A wavelet decomposition model was then used to determine the threshold of irregularity based on the ground truth. Heat maps were generated for each Z-line to determine localized areas of complexity. RESULTS:Fair agreement, with a Fleiss κ of .39, was observed among the 10 endoscopists when rating the Z-line as regular versus irregular using this dataset. Moderate agreement was observed among the 5 esophageal experts with a Fleiss κ statistic of .42, and fair agreement was observed among the 5 nonesophageal experts with a Fleiss κ statistic of .31. The wavelet energy coefficient optimal threshold to classify an SCJ as irregular was determined to be 1.53 × 107 with an accuracy of 78%. CONCLUSIONS:Our computer-generated model was capable of automatic segmentation and classification of the Z-line. We established a threshold of complexity using the wavelet energy coefficient to standardize the classification of the SCJ.

89 Background: With estimates of over 150,000 incident and 50,000 fatal colorectal cancer (CRC) cases in 2024, efforts are ongoing to improve the national screening performance and engagement in the US. The Centers for Medicare & Medicaid Services (CMS) have established criteria for covering new blood-based CRC screening tests if minimum CRC sensitivity and specificity thresholds are met. However, the current blood-based tests are challenged by their relative inability to detect advanced precancerous lesions (APLs). At present, the importance of APL versus CRC sensitivity on screening outcomes remains undetermined. To address this knowledge gap, we conducted a simulation study to compare the screening effectiveness and outcomes of a hypothetical blood-based test across a range of sensitivities for CRC and APL detection. Methods: Outcomes were simulated using the CRC-AIM microsimulation model, which has been previously calibrated and validated. Predicted outcomes were calculated for average-risk individuals screened triennially with a blood-based test between ages 45-75 years, assuming perfect adherence. Base-case screening sensitivity and specificity inputs were based on CMS minimum thresholds (74% CRC sensitivity, 90% specificity) with assumed 10% APL sensitivity. Additional scenarios considered an increase in either APL or CRC sensitivity. Outcomes of interest included life years-gained (LYG), CRC incidence, and mortality reductions. Results: Compared to the CMS minimum performance (Table), increasing APL sensitivity by 1% (from 10% to 11%) resulted in a similar increase in LYG as compared to increasing CRC sensitivity by 11% (from 74% to 85%). Additionally, increasing APL sensitivity by 1% resulted in greater reductions in CRC cases and deaths (3% decrease) as compared to increasing CRC sensitivity by 11% (0% and 1% reductions, respectively). Conclusions: Data from this study show that increased APL sensitivity is a key determinant of screening-related outcomes, such as LYG, derived from a blood-based strategy, and that similar results would require much larger increases in CRC sensitivity. With the current CMS minimum performance threshold, blood-based tests that are designed to have high APL sensitivity can outperform tests with higher CRC sensitivity and little to no APL sensitivity. Comparison of CRC screening health outcomes with CMS proposed minimum threshold blood-test versus hypothetical tests with either increased APL and CRC sensitivity. APL Sensitivity CRC Sensitivity CRC Specificity Follow-up Colonoscopies CRC Cases (% reduction) CRC Deaths (% reduction) Life-Years Gained (LYG) (% increase) 10% 74% 90% 809 50 (ref) 18 (ref) 220 (ref) 11% 74% 90% 813 48 (-3%) 17 (-3%) 226 (3%) 10% 85% 90% 809 50 (0%) 18 (-1%) 227 (3%) APL: Advanced precancerous lesions; CRC: Colorectal cancer.

Background and Aims:A positive stool-based test (SBT) result requires timely follow-up colonoscopy (FU-CY) to realize the full benefits of screening recommendations. The current investigation examined the real-world rate of, and time to, FU-CY adherence among patients aged 45-49 years who were screened with either a fecal immunochemical test (FIT)/fecal occult blood test (FOBT) or a multitarget stool DNA test (mt-sDNA). Methods:This retrospective study utilized a large integrated national multipayer claims database. Patients were aged 45-49 years and had a positive SBT result between January 1, 2017, and June 30, 2022 (study period); the first screening test result served as the index date. The primary outcome was adherence to FU-CY, defined as evidence of a colonoscopy (based on Current Procedural Terminology codes) within 12 months of a positive SBT. Results:A total of 13,921 patients with a positive SBT result were identified during the study period (FIT/FOBT, n = 6429; mt-sDNA, n = 7492). Among those with a positive SBT, FU-CY adherence was significantly higher among patients screened with mt-sDNA compared to those with FIT/FOBT (85.0% vs 35.2%, respectively; P < .01). Within the first 90 days, 62.9% of mt-sDNA users completed FU-CY, compared to only 22.8% of FIT/FOBT users. The mean time to FU-CY was 74.5 days for those screened with mt-sDNA (median, 53 days), compared to 90.8 days for FIT/FOBT (median, 63 days). Conclusion:FU-CY adherence rates were significantly higher in patients screened with mt-sDNA compared to those screened with FIT/FOBT. These results suggest that in younger patient populations, mt-sDNA may be a promising tool to improve FU-CY adherence rates and enhance the quality of colorectal cancer screening.

Objective: To provide contemporary data on subsequent screening after negative-result multitarget stool DNA (mt-sDNA) tests and follow-up colonoscopy after positive-result mt-sDNA tests. Patients and Methods: Negative-result mt-sDNA tests (for patients aged 50-72 years) and positive-result mt-sDNA tests (for patients aged 50-75 years) were identified among average risk patients from a 9-county region in Southeast Minnesota from January 1, 2016 to December 31, 2022. Competing risks models of time to subsequent colorectal cancer (CRC) screening were modeled separately for the negative mt-sDNA and positive mt-sDNA cohorts. Multistate Cox proportional hazards models compared rates of CRC screening modality by patient demographic characteristics. Results: At 3.5 years after a negative-result mt-sDNA test (n=18,739 tests), 55.0% (95% CI, 53.9%-56.3%) of patients were rescreened, which increased to 81.0% (95% CI, 80.0%-82.1%) at 5 years. Most tests were repeat mt-sDNA tests (48.3% at 3.5 years; 95% CI, 47.2%-49.5%). Rescreening with any modality was more likely with older age and among females and less likely among Black persons, Asian persons, and those with other or mixed race. After a positive-result mt-sDNA test (n=2863 tests), 80.9% (95% CI, 79.6%-82.6%) and 84.4% (95% CI, 83.2%-86.0%) of patients completed follow-up colonoscopy by 6 months and 1 year, respectively. Those of other or mixed race had lower rates of follow-up colonoscopy compared with White persons. Conclusion: Although rates of overall rescreening after a negative-result mt-sDNA test and follow-up colonoscopy after positive-result mt-sDNA tests were high, racial disparities were apparent. Targeted interventions are needed to improve equity in CRC screening adherence and follow-up care across diverse patient populations.

88 Background: There are an estimated 60 million individuals in the US who are not up-to-date with average-risk colorectal cancer (CRC) screening, as screening rates and adherence have remained stubbornly below the national target of 80%. Efforts are ongoing to improve CRC screening performance and participation, including the development of new blood-based tests. Despite high expectations for these tests, performance remains lower than other guideline-recommended strategies, particularly with respect to advanced precancerous lesion (APL) sensitivity. We conducted a simulation study of estimated clinical and economical outcomes for CRC screening with a blood-based test (at perfect adherence) compared to the multi-target stool-DNA (mt-sDNA) test (at published adherence rates). Methods: Utilizing CRC-AIM, a calibrated and validated microsimulation model, CRC screening outcomes were calculated for 1 million average-risk individuals screened between ages 45-75 years with triennial blood-based versus mt-sDNA testing. CRC and APL sensitivity and specificity inputs were derived from two large clinical validation studies for these screening modalities: ECLIPSE (NCT04136002) and DeeP-C (NCT01397747), respectively. To demonstrate the maximum benefits and burdens of blood-based screening, adherence to initial screening and follow-up colonoscopy after a positive blood test was modeled at 100%, while real-world adherence estimates of 65.6% were used for the mt-sDNA test. Outcomes of interest included life years-gained (LYG) and CRC cases missed by the blood-based test compared to the stool-based test, additional treatment costs imposed by theblood-based test, and additional CRC-related deaths per 1 million screened. Results: Compared to triennial screening with mt-sDNA at real-world adherence, blood-based screening at perfect adherence resulted in a higher number of incident (n=18,464) and fatal (n=6,483) CRC cases that would been avoided with the mt-sDNA test. Over a lifetime, screening with blood-based tests required 1,276,310 more tests (21% more) and 52,942 additional follow-up colonoscopies (7% more) compared to mt-sDNA screening. This increased testing burden did not generate additional benefits and instead reduced life-years gained by 67,645 per 1 million screened, resulting in an additional $1.6 billion in treatment costs compared to the mt-sDNA strategy. Conclusions: Data from this CRC-AIM modeling study show that even with perfect adherence, blood-based screening yields inferior clinical and economic benefits compared to mt-sDNA screening, due to suboptimal APL detection with the former test.