eP131 - A Retrospective, Single Center Study of the Positive Predictive Value of Non-Invasive Stool-Based DNA Test in Diagnosing Colorectal Cancer

Purpose/Background: Colorectal cancer (CRC) incidence has increased, and early detection is crucial for improving outcomes. Non-invasive stool-based DNA testing has emerged for CRC screening. Currently, Cologuard® is the only available non-invasive stool-based DNA test. It has a reported sensitivity of 92% of detecting CRC, but limited studies have evaluated the positive predictive value (PPV) in diagnosing CRC. This study aims to investigate the PPV of a positive non-invasive stool-based DNA test in diagnosing CRC in a single community-based colorectal practice.Methods/Interventions: A retrospective analysis of a single colorectal practice from January 2018 to August 2023 was performed. Inclusion criteria is patients aged 45 years or older who had a positive non-invasive stool-based DNA test (Cologuard®) and underwent a diagnostic colonoscopy at our institution. Exclusion criteria was a personal history of CRC. Data collected included demographics (age, sex, race/ethnicity, insurance status, family history of CRC, and history of prior colonoscopies before the positive stool-based DNA test), clinical symptoms (if any) at the time of the positive stool-based DNA test, colonoscopy findings/pathology, and stage of CRC at time of diagnosis. The primary outcome was the PPV of a non-invasive stool-based DNA test in detecting CRC. Secondary outcomes were the PPV of other colorectal findings, such as adenomas and diverticulosis.Results/Outcomes: A total of 102 patients with a positive non-invasive stool-based DNA test met our inclusion criteria. The mean age was 65.7 years old. Males (61.8%) and Caucasians (82.4%) were predominant. Two patients were diagnosed with CRC after a positive stool-based DNA test. The overall PPV of a positive test for CRC was 1.96%. Both patients diagnosed with CRC were African-Americans. Race was a significant risk factor in diagnosing CRC from a stool-based DNA test CRC (p value 0.007). The PPV for advanced adenomas (defined as a polyp with high-grade dysplasia, polyp ≥ 1 cm, or a sessile serrated polyp) was 6.9% and non-advanced adenomas was 52.9%. Patients aged 65 or older were significantly more likely to have non-advanced adenomas (p value 0.004) and diverticulosis (p value 0.002) on colonoscopy with a positive test compared to younger aged cohorts. Non-advanced adenomas found after a positive stool-based DNA test had a significance in patients who never had a prior colonoscopy compared to patients who had a prior colonoscopy (p value 0.03).Conclusion/

Discussion: This retrospective single practice study resulted in a PPV of 1.96% for CRC after a positive non-invasive stool-based DNA test. This study may highlight some limitations of a positive non-invasive stool-based DNA test leading to a diagnosis of CRC even though it has limitations of its own. A low PPV may lead to unnecessary patient anxiety with the fear of having colorectal cancer solely based on a positive stool-based DNA test. A low PPV may overburden healthcare providers who perform colonoscopies with unnecessary procedures and could lead to an increase in costs for both patients and healthcare systems. However, with a PPV of 59.8% for identifying non-advanced or advanced adenomas, it may lead to certain patients undergoing colonoscopy who would otherwise not have without a positive screening test. This study emphasizes the importance of optimizing individual risk factors, clinical presentation, and diagnostic modalities when using screening tests. Further research is needed to refine the utility of a non-invasive stool-based DNA test in CRC screening to enhance its PPV.