大腸病変検出における画像強調内視鏡技術の比較：ランダム化比較試験のネットワークメタアナリシス

Different image-enhanced endoscopy (IEE) methods have shown promise in improving colorectal lesion detection, but direct comparisons remain limited. This network meta-analysis compared the detection and diagnostic performance of various IEE modalities for colorectal lesions.

We searched PubMed, Embase, and Cochrane Library for randomized controlled trials (RCTs) up to May 31, 2025. Primary outcomes included adenoma (adenoma detection rate [ADR]) and advanced ADR; secondary outcomes included polyp (polyp detection rate), sessile serrated lesion detection rate, and mean number of colonic lesions detected per patient. Surface Under the Cumulative Ranking Curve (SUCRA) and pairwise risk ratios (RRs) with 95% confidence intervals (CI) were used. Lesion characterization was evaluated by the summary receiver operating characteristic area under the curve (AUC).

In total, 54 RCTs with 28,663 participants were included. Regarding ADR, texture and color enhancement imaging (TXI) ranked highest (SUCRA: 88.4; RR: 1.48, 95% CI: 1.09-2.02), followed by linked color imaging (LCI) when compared with high-definition white light imaging (HD-WLI). For advanced adenoma detection, both TXI (RR: 1.42) and LCI (RR: 1.23) showed superiority over HD-WLI. New-generation narrow band imaging (NBI) ranked highest for the number of adenomas detected (SUCRA: 82.3%), and LCI led in polyps (SUCRA: 88.4%). For lesion characterization, new-generation NBI (AUC: 0.880) and blue light imaging (AUC: 0.812) effectively distinguished adenomas from non-neoplastic lesions. However, this finding is based on a limited number of RCTs and should be interpreted cautiously.

Our findings indicate TXI and LCI may provide improved detection performance compared with conventional imaging. Blue light imaging and new-generation NBI seem to enhance lesion characterization. Further comparative studies are needed to guide optimal IEE modality selection for colorectal neoplasia.