Validation of the Modified Location-based Resect-and-discard Strategy Requiring Pathology Examination of Sigmoid...

Recently, the location-based resect-and-discard (LBRD) strategy, which does not depend on optical diagnosis, was developed and demonstrated different surveillance interval agreement with the pathology-based reference in several researches. We aimed to evaluate the performance of LBRD in our first-time colonoscopy cohort, and improve the LBRD.

The first-time colonoscopy with complete pathologic information were enrolled. The accuracy of LBRD strategy applied in diminutive polyps in different colonic segments was used as indicator to develop modified LBRD (mLBRD) strategy. Surveillance interval agreement with pathology-based reference was compared between LBRD and mLBRD. The ≥ 90% agreement with pathology was used as benchmark.

The polyps in sigmoid colon were significantly associated with higher proportion of neoplastic compared with polyps in rectum. The accuracy of LBRD applied in polyps in sigmoid colon were only 53.5%, which was significantly lower than that applied in polyps in other colonic segments. Thus, we hypothesized that mLBRD requiring pathology examination of diminutive polyps in sigmoid colon was more efficient in clinical use. The mLBRD significantly outperformed LBRD in surveillance interval agreement with pathology-based reference (90.2% vs. 83.4%, P <0.001), had lower proportion of patients assigned a longer surveillance interval (3.6% vs. 10.5%, P <0.001) and reached the benchmark, although the proportion of patients with an immediate surveillance interval recommendations and pathology examination avoided decreased.

The mLBRD, but not LBRD, achieved sufficient surveillance interval agreement with pathology-based surveillance interval assignment and reduced over 30% of pathology examinations.