Complementary use of white-light endoscopy and narrow-band imaging for post-EMR scar assessment: a systematic review and diagnostic meta-analysis.

BACKGROUND AND AIMS: EMR is a well-established treatment for large laterally spreading colorectal lesions. However, recurrence remains a key limitation. Accurate endoscopic evaluation of the post-EMR scar is essential, yet white-light endoscopy (WLE) alone may miss subtle residual neoplasia, contributing to routine biopsies and increased costs. Enhanced imaging modalities, particularly narrow-band imaging (NBI), improve visualization of mucosal and vascular patterns relevant to the detection of recurrence. Despite these advances, the incremental benefit of NBI over WLE for post-EMR scars remains uncertain. We aim to evaluate the diagnostic performance of NBI for the assessment of post-EMR scars, comparing its performance with WLE.

METHODS

We performed a systematic review and meta-analysis using the PubMed, Embase, and Cochrane databases. We searched for randomized and nonrandomized studies comparing WLE with NBI in patients submitted to surveillance colonoscopy after colorectal EMR and reporting at least one of the outcomes of interest. The outcomes assessed were the area under the curve (AUC), sensitivity, and specificity of both endoscopic techniques. Outcomes were reported with their respective 95% CI.

RESULTS

This study included 5 studies with 848 scars evaluated. The pooled sensitivity and specificity of WLE for recurrence detection were 0.799 (95% CI, 0.705-0.868) and 0.950 (95% CI, 0.875-0.981), respectively. The pooled sensitivity and specificity of NBI for recurrence detection were 0.904 (95% CI, 0.843-0.942) and 0.936 (95% CI, 0.876-0.968), respectively. The comparison between these AUCs results in a P value of .461.

CONCLUSIONS

Across 5 studies of 848 post-EMR scars, both WLE and NBI demonstrated high diagnostic performance for scar assessment. When applied in a complementary and sequential approach, with NBI performed after careful white-light evaluation, diagnostic sensitivity and confidence were enhanced, improving the ability to differentiate postresection scars from residual or recurrent neoplasia. This real-world strategy supports a selective and targeted biopsy approach, with the potential to avoid unnecessary biopsies and improve cost-effectiveness in post-EMR surveillance.