Narrow band imaging for characterization of high grade dysplasia and specialized intestinal metaplasia in Barrett's esophagus: a meta-analysis

Endoscopy. 2010 May;42(5):351-9. doi: 10.1055/s-0029-1243949. Epub 2010 Mar 3.

Abstract

Background and study aim: Narrow band imaging (NBI), a novel endoscopic technique that highlights mucosal surface structures and microvasculature is increasingly advocated as a tool to detect and characterize neoplasia and intestinal metaplasia in patients with Barrett's esophagus. We aimed to assess the diagnostic accuracy of NBI with magnification for the diagnosis of high grade dysplasia (HGD) and specialized intestinal metaplasia (SIM) in patients with Barrett's esophagus.

Methods: We performed a meta-analysis of studies which compared NBI-based diagnosis of HGD and SIM with histopathology as the gold standard.

Results: Eight studies including 446 patients with 2194 lesions met the inclusion criteria. For diagnosing HGD, the pooled sensitivity, specificity, diagnostic odds ratio (DOR), and area under the curve (AUC) were 0.96 (95 % confidence interval [CI] 0.93-0.99), 0.94 (95 %CI 0.84-1.0), 342.49 (95 %CI 40.49 - 2896.89) and 0.99 (SE 0.01) on a per-lesion analysis with similar results on per-patient analysis.. For the characterization of SIM, the pooled sensitivity, specificity, DOR, and AUC were 0.95 (95 %CI 0.87-1.0), 0.65 (95 %CI 0.52-0.78), 37.53 (95 %CI 6.50-217.62) and 0.88 (SE 0.08) on a per-lesion analysis.

Conclusion: NBI with magnification is accurate with high diagnostic precision for diagnosis of HGD in Barrett's esophagus on the basis of irregular mucosal pit patterns and/or irregular microvasculature. NBI has high sensitivity but poor specificity for characterizing SIM.

Publication types

  • Comparative Study
  • Meta-Analysis
  • Review

MeSH terms

  • Barrett Esophagus / pathology*
  • Endoscopy, Gastrointestinal / methods*
  • Esophagoscopy / methods
  • Humans
  • Image Enhancement / methods*
  • Intestinal Mucosa / pathology*
  • Metaplasia
  • Observer Variation
  • Precancerous Conditions / pathology*
  • Reproducibility of Results
  • Sensitivity and Specificity