Kisiel, John B.

Background & Aims: Multi-target stool DNA (mt-sDNA) is an established stool-based colorectal cancer screening test, yet long-term performance is not well-described. We evaluated follow-up colonoscopy rates & quality, neoplasia yield, and repeat-testing adherence over 10 years with an integrated screening program. Methods: We assembled a retrospective cohort of mt-sDNA tests (September 2014–July 2025) within a multi-site health system. A validated large-language-model pipeline extracted colonoscopy, pathology, and quality endpoints from free-text reports. Detection outcomes (adenoma detection rate, advanced neoplasia, modified advanced adenoma, sessile serrated lesion (SSL), and negative exams) were calculated among high-quality exams; adenocarcinoma, bowel preparation adequacy, and cecal intubation rate (CIR) used all screening exams. We assessed colonoscopy completion ≤1 year after positive mt-sDNA, follow-up colonoscopy findings, findings by time-to-colonoscopy, and adherence to repeat mt-sDNA after negative result. Results: Among 148,624 tests in 110,410 individuals, positivity was 13.6%. Of 19,506 positives, 65.1% completed colonoscopy within 12 months (median 82 days). Colonoscopy after a positive mt-sDNA had high yield: advanced adenoma diagnosis was 30.7%; adenocarcinoma was 1.3%. Of patients who had colonoscopy within 3 years of a negative mt-sDNA test, 45.2% had adenoma/serrated lesion, 9.3% had advanced adenoma and 0.5% had colon cancer. Adenoma detection rate after positive mt-sDNA (>50%) exceeded specialty society quality benchmarks and increased over time; Longer intervals (>6 months) to colonoscopy after a positive test were associated with higher advanced adenoma (trend P=0.001).After an initial negative mt-sDNA, only 35.3% were screened again by mt-sDNA testing within 3 years. Conclusions: Over a decade,mt-sDNA screening achieved high colonoscopy qualitybut revealed critical gaps in the care cascade. One-third of positive tests lacked timely follow-up and nearly two-thirds of negative tests were not repeated on schedule, suggesting opportunities for system-level interventions.

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.

ABSTRACT Background and Objectives Occult intra‐abdominal metastases are identified in approximately 20% of patients with radiographically localized pancreatic ductal adenocarcinoma (PDAC) through staging laparoscopy, as peritoneal cytology is limited in sensitivity. Methylated DNA markers (MDMs) have demonstrated high diagnostic accuracy for PDAC in tumor tissue, blood, and pancreatic secretions. This study assesses the feasibility and diagnostic performance of MDMs in peritoneal lavage fluid collected during staging laparoscopy, comparing their performance to mutant KRAS (m KRAS ) detection. Methods DNA from peritoneal lavage fluid of PDAC patients was analyzed for m KRAS mutations (codons 12, 13, and 61) using droplet‐digital PCR and for 13 MDMs and a reference gene (B3GALT6) using quantitative methylation‐specific PCR. An MDM score was generated from the average signal of the 13 MDMs and evaluated for its ability to detect intra‐abdominal metastases via receiver operator characteristic (ROC) analysis. Results Among 48 patients, 15 (31%) had intra‐abdominal metastases. The MDM score showed 87% sensitivity and 90% specificity (AUC = 0.98), outperforming m KRAS , which showed 40% sensitivity and 85% specificity (AUC = 0.62; p < 0.0001). Conclusions MDMs in peritoneal lavage fluid may enable highly accurate molecular staging of PDAC and warrant validation in larger studies.

e15647 Background: Colorectal cancer (CRC) screening offers an opportunity to prevent invasive disease by detecting and removing precancerous lesions. Colonoscopy-based detection of adenomas and serrated polyps is strongly associated with reduced risk of interval CRC. While the performance of noninvasive CRC screening tests is well described, comparative evidence on adenoma detection across screening colonoscopy and stool-based strategies with follow-up colonoscopy remains limited. We conducted a systematic review to compare adenoma detection across CRC screening modalities. Methods: MEDLINE (PubMed) and Embase were searched from January 1, 2000, through July 10, 2025, for peer-reviewed, U.S.-based studies reporting colonoscopy among average-risk adults aged ≥45 years, who underwent screening colonoscopy or follow-up colonoscopy after a positive stool-based screening test [fecal immunochemical test (FIT) or multi-target stool DNA (mt-sDNA) test]. The primary outcome was the detection of one or more adenomas at colonoscopy. High-risk or surveillance cohorts, AI-assisted colonoscopy, diagnostic accuracy trials, and non-U.S. studies were excluded. Results: Among 7,710 de-duplicated records, 147 studies underwent full-text review, and 47 studies were included in quantitative synthesis, with reported adenoma detection for screening colonoscopy alone (n = 31), follow-up colonoscopy after a positive FIT alone (n = 4), follow-up colonoscopy after a positive mt-sDNA test alone (n = 4), and for more than one of the screening strategies of interest (n = 8). Across included studies, overall adenoma detection differed by screening pathway. Follow-up colonoscopy after a positive mt-sDNA test demonstrated the highest adenoma detection (66.3%; 95% CI, 61.2-71.1), with lower adenoma detection observed for colonoscopy after a positive FIT (50.6%; 95% CI, 43.4-57.8) and screening colonoscopy alone (33.8%; 95% CI, 30.9-36.8) (Table 1). Conclusions: Detection of adenomas was highest following colonoscopy performed after a positive mt-sDNA test compared with FIT-positive follow-up or screening colonoscopy alone. These findings highlight differences in downstream detection across CRC screening tests and suggest that, in settings where colonoscopy resources are constrained, initial mt-sDNA screening with timely follow-up colonoscopy may help prioritize patients most likely to benefit. CRC screening modality No. of Studies Adenoma detected Study population Proportion, % (95% CI) Heterogeneity, I² (%) Between-study variance (τ 2 ) P-value* mt-sDNA 8 17,436 29,098 66.3 (61.2-71.1) 94.6 0.0924 <0.0001 FIT 11 17,588 32,630 50.6 (43.4-57.8) 94.0 0.2187 <0.0001 Screening colonoscopy 38 2,351,801 6,378,891 33.8 (30.9-36.8) 99.8 0.1710 NA *P-value represents adenoma detection vs. screening colonoscopy adenoma detection.

5624 Background: Endometrial cancer (EC) is the most common gynecologic cancer and abnormal uterine bleeding (AUB) is the predominant presenting symptom. While <10% of AUB represents underlying EC, endometrial sampling is currently required to rule out EC. To develop an alternative, non-invasive rule-out strategy, we aimed to refine a previously reported EC-associated methylated DNA marker (MDM) panel in vaginal swabs. Methods: Patients ≥18 years old with AUB who met ACOG recommendations for endometrial sampling and those undergoing a hysterectomy for biopsy-proven EC or atypical hyperplasia (AH) were prospectively enrolled in a multicenter cohort study. A vaginal fluid specimen was collected via swab by a provider during a speculum exam. DNA was extracted from samples and bisulfite treated, followed by a target and signal amplification method to detect 8 previously identified MDMs. The cross-validated EC classification model used the mean of standardized MDMs. Clinical classification of benign endometrium (BE), EC, and AH was based on central pathology review. EC diagnoses were based on hysterectomy specimens. Endometrioid and mucinous constituted type 1 EC; serous, carcinosarcoma, clear cell, and mixed histology ECs were collectively analyzed as type 2. Results: Of 693 participants enrolled, 670 were evaluable with valid results: 524 in the AUB cohort and 146 in the suspected EC cohort. Within the AUB cohort, 16 had EC or AH. The final analysis cohort included 508 BE, 156 EC, and 6 AH. Of the evaluable participants 20% identified as Black, 3% Asian or Pacific Islander, 75% White, and 3% Other; 7% reported Hispanic ethnicity. Median age was 66 and 50 years for EC and BE AUB, respectively; 80% of those with EC and 43% with benign AUB reported being postmenopausal. Among participants with EC, 102 (65%), 17 (11%), 30 (19%), 6 (4%) were stage I, II, III, and IV, respectively; one EC was unstaged. Type 2 EC accounted for 42 (27%) of all EC cases. A 4-MDM panel yielded an AUC of 0.97 (0.96-0.99, 95% CI) for EC; models with 5-8 MDMs gave identical performance. All-stage EC sensitivity was 97, 97, 96, 95, and 92% at specificities ranging from 75, 80, 85, 90, and 95%, respectively. At the 85% (82-88%) specificity threshold, sensitivities across EC stage, type, and age were consistently high (Table). Sensitivity for AH was 33% (10-70%). Conclusions: An MDM panel demonstrated high sensitivity for EC across stage and EC type, supporting the potential for a non-invasive vaginal swab-based test to rule out EC. Future studies are necessary to finalize the panel and algorithm, prior to validating the test. Clinical trial information: NCT06294886 . EC Sensitivity (%, 95% CI) at 85% Specificity All Stages (n=156)* 96 (92-98) Stage I (n=102) 94 (88 - 97) II (n=17) 100 (82 - 100) III (n=30) 100 (89 - 100) IV (n=6) 100 (61 - 100) Type 1 (n=114) 95 (89 - 98) 2 (n=42) 100 (92 -100) Age ≤ 66 yrs (n=86) 95 (89 - 98) > 66 (n=70) 97 (90 - 99) *n=1 EC was un-staged.

Abstract Background: The Alaska Tribal Health System is working to increase colorectal cancer (CRC) screening among Alaska Native peoples. It is unknown why people who initially agree to screening might not follow through to complete their screening. Methods: As part of a larger intervention investigating multi-target stool DNA (mt-sDNA) test use in rural Alaska communities, we interviewed Alaska Native people who had agreed to scheduling a colonoscopy or having a mt-sDNA kit sent to their home but who did not complete their selected test. Results: Of 531 Alaska Native people who initially agreed but did not complete screening, 299 (56%) were able to be contacted for a follow-up survey. Of those, 256 (86%) answered the survey; 187/224 (83%) who had chosen mt-sDNA, and 69/75 (93%) who had chosen colonoscopy. Reported reasons for non-completion of stool DNA test kits included being busy or forgetting, being uncomfortable doing a stool test, and not knowing how to do the test or concerns over how to return the kit for processing. Reasons for non-completion of colonoscopy included scheduling issues, physical health issues, travel, and housing barriers, as well as cost, fear, and being too busy. Of note, many respondents in both groups indicated they would still be interested in CRC screening in the future or were currently pursuing screening. Conclusions: This was the first exploration of reasons for non-adherence to CRC screening after initial test request by Alaska Native people. Many of the identified themes are amenable to health care system changes for future interventions to improve screening outreach and uptake. Citation Format: Diana Redwood, Christie A. Flanagan, Lauren Jeffries, Melissa Toffolon, Joseph Klejka, Lila J. Finney Rutten, Judith S. Kaur, John B. Kisiel. Non-completion of colorectal cancer screening among Alaska Native people in an intervention trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2474.

Abstract INTRODUCTION: Colorectal cancer (CRC) is the 2nd most fatal cancer in the U.S.; most benefits of CRC screening are from detection of advanced pre-cancerous lesions (APLs). With a goal to improve non-invasive stool test APL sensitivity, we sought novel APL signatures in stool metaproteomes (host- and gut microbiome-derived proteins) among cancer-free controls (CTRL), APL, and CRC patients. Since dysbiosis in gut microbiota is associated with CRC initiation, we also examined the microbial source of differentially expressed proteins (DEPs). METHODS: Mass spectrometry was performed on archival stool supernatants (n=63 patients; 15 CTRL, 32 APL [16 adenomas & 16 sessile serrated lesions], 16 CRC). A sectioning and database-enrichment strategy was applied for peptide-spectrum matching. Microbial and human-derived proteins were searched against the Integrated Gene Catalogue (IGC) and UniProt human reference proteome. DEPs were identified using thresholds of P &lt; 0.05 and fold-change ≥ 1.5. Pathway annotation of human DEPs was performed with the REVIGO webserver, while gut microbial DEPs were functionally categorized using the KEGG Automatic Annotation Server (KAAS). To assess the taxonomic distribution of microbial DEPs, sequences were aligned against the NCBI non-redundant database using BLASTP. RESULTS: Mass spectrometry detected 879 human and 32,514 microbiome-derived proteins. When limiting detection to proteins present in ≥ 75% of patients, only 117 human and 204 microbial proteins remained. There were 17 (4,875), 288 (9,638), and 63 (3,053) unique human (microbial) proteins detected in CTRL, APL, and CRC groups, respectively. When the protein presence cutoff was set at &gt;30% of patients, there were 52 (13 human [11 upregulated, 2 downregulated], 39 microbial [28 upregulated, 11 downregulated]) DEPs detected for the CRC + APL vs CTRL comparison. Some pathways associated with upregulated human DEPs included: inflammatory response, cellular response to iron, and endothelial cell migration. The top upregulated microbial DEPs were derived from the Phocaeicola, Lachnospiraceae, and Oscillospiraceae genera. The top downregulated microbial DEPs belonged to the Bacteroides, Eubacterium, Agathobacter, Coprococcus, and Blautia genera. CONCLUSION: This work highlights the biological diversity and molecular heterogeneity of the stool proteome. We identified several host and microbial protein markers that differentiate APLs and early CRC from controls. These results provide a foundation for future validation studies, which are necessary to determine the clinical utility of these protein biomarkers prior to their integration into a noninvasive stool test. Citation Format: Emmalee J. Northrop-Albrecht, Vinod K. Gupta, William R. Taylor, Jason P. Sinnwell, Doug W. Mahoney, Patrick H. Foote, Kelli N. Burger, Jaeyun Sung, John B. Kisiel, . Comprehensive investigation of stool metaproteomes among patients with precancerous lesions and colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7656.

Abstract Esophageal squamous cell carcinoma (ESCC) remains a leading cause of cancer-related mortality worldwide, and there are limited molecular screening options. Following a prior discovery and validation studies in subjects from the United States, China, and Iran, we aimed to validate methylated DNA markers (MDM) for ESCC detection in Malawi, a country with one of the highest ESCC incidence. Fourteen MDMs were tested on 177 tissue samples. Ten MDMs showed excellent performance (AUCs ≥ 0.85); a random forest model achieved an AUC of 0.97. The consistent cross-population performance suggests potential for universal application of these MDMs and further justifies exploring nonendoscopic sampling methods, such as swallowed cell collection devices in combination with MDMs, for efficient and cost-effective screening in high-incidence populations. Prevention Relevance: This study validates MDMs for early detection of ESCC in Malawi, supporting their potential role in noninvasive, population-based screening strategies in Africa to enable early detection and timely intervention for cancer interception, potentially reducing cancer burden in high-incidence, resource-limited regions.

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.