S254 How Well Do the U.S. and European Surveillance Guidelines Discriminate High- and Low-Risk Groups: Data From the New Hampshire Colonoscopy Registry

Abstract

Introduction: Surveillance colonoscopy guidelines were updated by the European Society of Gastrointestinal Endoscopy (ESGE) and the United States Multi-Society Task Force (USMSTF) for Colorectal Cancer Recommendations in 2020. Our goal was to use the New Hampshire Colonoscopy Registry (NHCR) to evaluate the ability of the USMSTF (3 year) and ESGE guidelines to discriminate high versus low risk groups for metachronous colorectal neoplasia. We also examined the impact of adding polyp specific groups from the USMSTF 5-year recommendations to the USMSTF 3-year on performance. Methods: We risk stratified NHCR patients who had an index and follow up colonoscopy at least 12 months later as per USMSTF and ESGE guidelines based on index findings. We excluded examinations that were incomplete, had poor bowel preparation, baseline CRC, IBD, or polyposis syndromes. The outcome was the metachronous risk for significant colorectal neoplasia (advanced adenomas, CRC, and large serrated polyps). The sensitivity, specificity, and c-statistic were calculated for each guideline recommendation group. The c-statistics were calculated using probabilities from a logistic regression adjusting for age, sex, and time to follow up. Results: We included 20972 patients (average age 58.8 ± 8.6, 48.1% male). The addition of villous histology to ESGE resulted in an increase in sensitivity but decrease in specificity. The addition of other polyp groups into USMSTF 3 years yielded c-statistics ranging from 0.600 to 0.611. The addition of 1-2 small SSPs into USMSTF 3 years resulted in a 4.1% increase in sensitivity, with only a 1.9% decrease in specificity (Table 1). Conclusion: The ESGE and USMSTF guidelines had similar modest c-statistics, suggesting similar discriminating ability to predicting metachronous neoplasia. This is further suggested by the similar sensitivities and specificities with the addition of villous histology into the ESGE category. Additionally, the modest c-statistics may indicate the need for further adjustment of the guidelines or use of other markers. Further examination of the USMSTF 3-year guidelines were notable for increasing sensitivities with the addition of 1-2 SSPs, suggesting that SSPs may be important predictors of future neoplasia and therefore may need further study for modifications of the guidelines. For example, perhaps 1-2 SSPs should have a 3-year interval.Table 1.: Comparison of ESGE and USMSTF guidelines for metachronous colorectal neoplasia