How do I interpret narrow band imaging (NBI) findings?

How to Interpret Narrow Band Imaging (NBI)

Interpret NBI findings by systematically evaluating microvascular patterns and mucosal surface architecture, where dysplastic and neoplastic lesions appear brown at standard magnification due to increased vessel density, and specific irregular patterns at high magnification indicate malignancy. 1

Fundamental Principles of NBI Technology

NBI uses narrow wavelength light filters (415 nm blue and 540 nm green) that optimize hemoglobin absorption, creating enhanced visualization of superficial mucosal and submucosal microvessels without requiring dye application. 1 The technology generates false-color images where abnormal tissue displays distinct color and pattern changes compared to normal mucosa. 1

Standard Magnification Interpretation (30x)

Color Assessment

• Brown appearance: Indicates higher vessel density characteristic of dysplastic and neoplastic lesions in both squamous and columnar epithelium 1
• Normal pink/pale appearance: Suggests benign tissue with normal vessel density 1

Clinical Application

At standard magnification, focus on identifying brownish areas that contrast with surrounding normal mucosa, as these represent areas requiring targeted high-magnification examination or biopsy. 1

High Magnification Interpretation (up to 100x)

Esophageal Squamous Mucosa

Intraepithelial Papillary Capillary Loops (IPCL) Assessment:

• Normal pattern: Regular, uniform IPCL with consistent configuration, density, and size 1
• Dysplastic/neoplastic pattern: Dilated, irregular-shaped IPCL with abnormal density and configuration 1
• Brown dots on magnification: Represent abnormal IPCL indicating squamous cell dysplasia or carcinoma 1

The configuration, density, and size of IPCL change progressively with dysplasia and neoplasia, allowing depth prediction. 1

Barrett's Esophagus and Columnar Epithelium

Three Key Abnormalities to Identify: 2

1. Irregular/disrupted mucosal patterns: Loss of normal villous/gyrus-forming architecture
2. Irregular vascular patterns: Disruption of normal regular vessel networks
3. Abnormal blood vessels: Presence of atypical vessel morphology

Diagnostic Algorithm:

• No abnormalities present: 98% negative predictive value for high-grade intraepithelial neoplasia (HGIN) 2
• One abnormality: Suspicious for dysplasia, requires targeted biopsy
• Two or more abnormalities: 85% of HGIN cases display this pattern; 94% sensitivity for HGIN 2

Normal Barrett's Pattern:

• Villous/gyrus-forming patterns (80% of cases) with regular vascular networks 2
• Flat mucosa with regular long branching vessels (20% of cases) 2

Gastric Lesions

ME-NBI Classification System (Three Pattern Types): 3

Type A Pattern:

• Regular surface pattern with regular microvascular architecture
• Mainly characteristic of noncancerous lesions 3
• Specificity: 84.4% for excluding cancer 3

Type B Pattern:

• Irregular surface pattern with irregular microvascular architecture
• Good indicator of differentiated adenocarcinoma 3
• Indicates intramucosal or superficially invasive cancers 3
• Oval/tubular microsurface pattern with fine network or loop microvascular pattern 4

Type C Pattern:

• Absent surface pattern with severely irregular/corkscrew microvascular architecture 3
• Indicative of undifferentiated adenocarcinoma or differentiated cancer with deep submucosal invasion 3
• Sensitivity: 92.3% for identifying undifferentiated adenocarcinoma 3

Depth of Invasion Prediction:

• Destructive microsurface patterns correlate with submucosal invasion, particularly in differentiated tumors 4
• ME-NBI accurately predicted invasion depth in 95% of differentiated adenocarcinomas 3

Colorectal Lesions

The microvascular network surrounding colonic glands changes in metaplasia, dysplasia, and neoplasia. 1 Higher vessel density and irregular patterns indicate neoplastic transformation. 1

Practical Implementation Strategy

Pre-Examination Preparation

• Ensure adequate mucosal preparation with aspiration of fluid/debris and defoaming agents 5
• Maximize acid suppression in Barrett's esophagus patients (twice-daily PPI) before surveillance 1

Systematic Examination Approach

1. Initial white-light survey: Complete standard examination first 5
2. NBI activation: Switch to NBI mode for detailed inspection 1
3. Adequate inspection time: Longer examination correlates with higher detection rates 5
4. Targeted evaluation: Focus on suspicious areas identified on white-light or brownish areas on standard NBI 1
5. High magnification: Apply to areas requiring detailed characterization 1, 3

Biopsy Strategy

• Target biopsies to areas with abnormal NBI patterns rather than random sampling 1, 2
• NBI guidance reduces total biopsy number while maintaining diagnostic yield 1, 5

Clinical Performance Metrics

Overall Diagnostic Accuracy: 6

• Sensitivity: 96.2% for neoplastic lesions
• Specificity: 91.0%
• Positive predictive value: 96.2%
• Negative predictive value: 97.2%

Barrett's Esophagus Specific: 6

• Sensitivity: 95.4%
• Specificity: 90.7%
• Negative predictive value: 90.7%

Gastric Neoplasia: 6

• Sensitivity: 96.1%
• Specificity: 91.04%

Colorectal Neoplasia: 6

• Sensitivity: 96.7%
• Specificity: 91.3%

Common Pitfalls and Caveats

Training Requirements

Adequate training programs are essential for implementation, as NBI interpretation requires experience and familiarity to reduce missed lesion rates. 5, 7 The technology is operator-dependent and performs best in expert hands. 7

Technical Limitations

• Active inflammation: Avoid surveillance biopsies during active esophagitis (Los Angeles grade C/D), as inflammation causes over-diagnosis 1
• Inadequate preparation: Poor mucosal visualization compromises diagnostic accuracy 5
• Standard vs. high-definition systems: Combine NBI with high-definition endoscopy systems for maximum benefit 5

Clinical Context Limitations

• Inflammatory bowel disease: NBI is NOT recommended for routine colitis surveillance, as it shows no benefit over white-light endoscopy and is inferior to chromoendoscopy 5
• Missed lesions: 25.3% of early adenocarcinomas are missed at initial endoscopy in Barrett's patients, emphasizing need for meticulous technique 1

Interpretation Challenges

While NBI increases dysplasia/cancer detection by 34% compared to white-light endoscopy alone 1, 5, standardized classification systems that include low-grade dysplasia do not yet exist. 1 Use the patterns described above as frameworks rather than absolute diagnostic criteria, and always confirm suspicious findings with histopathology. 7, 6