OPTICAL DIAGNOSIS OF POLYPS < 10MM AND IMPACT ON SURVEILLANCE INTERVAL (DISCARD3)

Introduction Optical diagnosis (OD) of diminutive (1–5mm) and small (6–9mm) polyps can be performed with advanced endoscopic imaging. When coupled with a ‘resect and discard’ strategy, for high confidence diagnoses, significant histopathology time and cost savings are possible. The threshold for implementation in practice is ≥90% OD-histology surveillance interval concordance. The aim of this study was to assess feasibility and implementation of ≤5mm and <10mm polyp OD in a bowel cancer screening setting. Methods This prospective feasibility study involved 8 bowel cancer screening endoscopists completing a validated OD training module and performing procedures. All <10mm consecutive polyps were optically assessed using white-light and narrow-band images and given high or low confidence diagnoses until 120 high confidence ≤5mm polyp diagnoses had been made. All polyps had standard histology. Where high confidence OD did not match histology deeper histology levels were analysed. Histology-derived and OD-derived surveillance intervals were calculated using BSG, ESGE and US multi-society task force guidelines. Results 565 patients were invited with 40 exclusions. Of the 525 patients included, 1771 polyps were identified of which 1752 were resected and retrieved. Of these, 1560 were <10mm (1329 ≤5mm and 231 6–9mm). Most <10mm polyps (74.9%, 1169/1560) were diagnosed with high confidence. There were no <10mm polyp cancers. High confidence OD was accurate in 81.5% of ≤5mm and 92.8% of 6–9mm polyps; overall accuracy (<10mm polyps) was 83.5%. The sensitivity for OD of a ≤5mm adenoma was 93.0% with a PPV of 90.8%. The sensitivity for OD of a ≤5mm serrated polyp was 80.2% with a PPV of 73.1%. Surveillance interval concordance, according to BSG, ESGE, and US multi-society task force guidelines, was assessed in 229 patients, who had only <10mm polyps and high confidence diagnoses (see table 1). Conclusions On a per-polyp basis OD accuracy was greater for 6–9mm polyps compared to diminutive polyps and there were no <10mm polyp cancers. A ‘resect and discard’ strategy for high confidence ≤5mm polyp OD, in a group of bowel cancer screening colonoscopists, is feasible and safe with performance exceeding the 90% surveillance interval concordance (BSG, ESGE, US guidelines) required for implementation in clinical practice. For high confidence OD of polyps <10mm, performance exceeds the 90% surveillance interval concordance threshold for BSG and ESGE guidelines but not US guidelines.