Abstract
Background and study aims Post-polypectomy surveillance interval (SI) is determined based on the number, size,
and histology of colorectal polyps. Electronic chromoendoscopy in association with
magnifying imaging colonoscopy allows “in vivo” polyp histology prediction. Colorectal
polyps ≤ 5 mm can be resected and discarded without pathologic assessment if the endoscopic
technology when used with high confidence provides ≥ 90 % agreement between the post-polypectomy
SI and the SI based on pathological assessment. The aim of this study was to evaluate
the agreement between the post-polypectomy SI based on flexible spectral color imaging
enhancement (FICE) chromoendoscopy in association with magnified imaging and the pathology-based
SI.
Patients and methods Each diagnosed colorectal polyp received a histology prediction (neoplastic or non-neoplastic)
based on the FICE capillary-vessel pattern classification. Each prediction was classified
as high or low confidence. SI based on the FICE prediction was compared to the pathology-based
SI recommendation according to the US Multi-Society Task Force on Colorectal Cancer
guideline. Sensitivity, specificity and accuracy of FICE in diagnosing neoplastic
lesions were compared with the pathology assessment. Interobserver and intraobserver
agreement for FICE-based SI predictions was evaluated using the kappa coefficient.
Results A total of 267 polyps had histology prediction assessed with high confidence in 136
patients. Sensitivity of FICE was 98.7 % (95 % CI: 93.5 – 99.3) and specificity was
62.5 % (95 % CI: 43.6 – 78.9). Prediction accuracy was 94.4 % (95 % CI: 88.6 – 96 – 1)
in differentiating between neoplastic and non-neoplastic lesions. Therefore, magnifying
FICE colonoscopy-based SI recommendation was consistent with pathological assessment
in 88.3 % of general cases (95 % CI: 82.1 – 92.6) and in 89.7 % (95 % CI: 83 – 94.5)
of the high-confidence evaluation cases. The intraobserver agreement value for FICE-based
SI predictions was 0.87 (high-confidence evaluations), and the interobserver agreement
values were 0.78 (high- and low-confidence evaluations) and 0.82 (high-confidence
evaluations) (95 % CI: 0.79 – 0.95).
Conclusions FICE-based SI demonstrated 89.7 % concordance with the pathology-based SI.
