S273 Risk Factors Impacting Advanced Adenoma Detection Rate Following Negative Multitarget Stool DNA Testing

Abstract

Introduction: Colonoscopy remains the gold standard for colorectal cancer (CRC) screening, but less invasive screening modalities have been employed more recently, including the multitarget Stool DNA (MT-sDNA or Cologuard) test, which combines detection of blood products with genetic markers in the stool. Data regarding the false-negative rate of the MT-sDNA test in real-world clinical practice is limited. Our primary aim was to determine the rate of false-negative MT-sDNA testing and evaluate for factors associated with higher false-negative rates within our health system. Methods: Adults (≥18 years old) with a negative MT-sDNA test between 2017 and 2022 and subsequent colonoscopy within three years of the MT-sDNA test, regardless of colonoscopy indication were included. Our primary outcome of interest was advanced adenoma (AA) detection rate, defined as adenoma with villous features, size ≥ 1.0 cm, high-grade dysplasia, or early invasive cancer. Demographic and procedural variables including age, sex, race, BMI, colonoscopy indication, polyp size, and polyp location were manually extracted from patient charts and compared between the two groups (AA vs. no AA) using chi-squared analysis. Results: A total of 370 patients met the inclusion criteria, of which 31 (8.4%) were found to have AA and 3 (0.81%) were found to have CRC on colonoscopy within 3 years of negative MT-sDNA test. There were no demographic differences between the two groups. AA detection rate was significantly higher in patients who underwent colonoscopy for GI bleeding (32.3% vs 14.2%, p=0.008) as opposed to other indications. Among patients who had polyps (N=148), AA detection was associated with more numerous polyps (2 [IQR 1-4] vs 1 [IQR 1-2], p < 0.001), and larger polyp size (14 [SD 5.1] vs 5.1 [SD 2.2], p < 0.001). AAs were also significantly more frequently found in the hepatic flexure (6.5% vs 0.3%, p=0.050) and transverse colon (41.9% vs 5.6%, p=0.002) compared to other locations (Table). Conclusion: The results of this study validate the 8% quoted false-negative rate for MT-sDNA testing shown in prior literature. Large polyps in the transverse colon and hepatic flexure are more likely to result in a false negative MT-sDNA test and therefore these locations should be examined in more detail during endoscope withdrawal. Finally, a negative MT-sDNA test result should be interpreted with caution and gastroenterologists should have a low threshold to perform a colonoscopy if otherwise clinically indicated. Table 1. - Baseline demographics and procedural variables in patients with negative MT-sDNA testing and advanced adenoma on colonoscopy (n=31) as compared to patients without advanced adenoma (n=339) Advanced Adenoma Non-Advanced Adenoma p value n=31 n=339 Male Sex, n(%) n(%) 16 51.6% 126 37.2% 0.114 Age – years Median (IQR) 67 59-74 66 59-71 0.157 Race n(%) 0.168 White 28 90.3% 272 80.2% Black 3 9.7% 59 17.4% Other 0 0.0% 8 2.4% BMI – kg/m2 Median (IQR) 30.7 25.6-35.2 30.1 26-34 0.807 Personal History of Colon Cancer n(%) 0 0.0% 3 0.9% 0.600 Family History of Colon Cancer n(%) 0 0.0% 32 9.4% 0.074 Personal History of IBD n(%) 0 0.0% 3 0.9% 0.600 Indication for Colonoscopy n(%) GI Bleeding 10 32.3% 48 14.2% 0.008 Iron deficiency anemia 4 12.9% 28 8.3% 0.380 Diarrhea 1 3.2% 34 10.0% 0.216 Constipation 0 0.0% 6 1.8% 0.457 Weight loss 0 0.0% 5 1.5% 0.497 Abnormal Imaging 2 6.5% 11 3.2% 0.355 IBD 0 0.0% 0 0.0% Screening 14 45.2% 193 56.9% 0.207 Other 0 0.0% 14 4.1% 0.250 Prep Quality n(%) Poor 2 6.5% 24 7.1% 0.896 Inadequate 0 0.0% 2 0.6% 0.669 Fair 10 32.3% 79 23.3% 0.265 Adequate 2 6.5% 35 10.3% 0.493 Good 17 54.8% 187 55.2% 0.972 Excellent 0 0.0% 12 3.5% 0.288