Aims:
Post-polypectomy surveillance intervals are determined taking pathology results into consideration. Study aim was to compare a novel polyp-based resect-and-discard strategy (PBRD) with optical polyp diagnosis and pathology based surveillance interval assignment.
Methods:
The PBRD strategy (Table 1) and optical polyp diagnosis, using the NICE classification with iScan or Optivista imaging, were applied in real time during colonoscopies for small polyps (< 10 mm) in a prospective study. The primary outcome was the agreement in surveillance interval assignment between the polyp-based and the optical diagnosis based resect-and-discard strategies in reference to the pathology-based reference-standard. Secondary outcomes were the overall reduction in required pathology exams with the different strategies, and the proportion of patients that can be provided with surveillance recommendations immediately following the colonoscopy.
Results:
553 patients (mean age 62, 48% male, 567 small polyps) were enrolled into the study. In applying the PBRD (Table 1), the surveillance interval agreement with pathology was of 90% (95% CI:87 – 92). When applied by the endoscopist immediately after colonoscopy, the agreement was of 81% (95% CI:69 – 76). The optical diagnosis strategy achieved a surveillance interval agreement of 91% (95% CI:89 – 94). With respect to histopathology, the polyp-based approach would save 56% of all histopathology exams, while 69% could be omitted with the optical diagnosis strategy. The polyp-based strategy would provide 84% of patients with immediate surveillance recommendations, while 78% and 46% of patients could be provided with immediate recommendations following the optical and pathology based strategies, respectively.
Conclusions:
The PBRD reaches the 90% PIVI benchmark when applied according to the developed algorithm. However, when used in real time by clinicians, the surveillance interval agreement with pathology is less accurate, mainly because of Shorter surveillance intervals deliberately chosen by clinicians. Optical diagnosis using NICE in combination with iScan and Optivista achieved similar results as the polyp-based strategy surpassing the 90% PIVI benchmark.
