Light blue crest (blue fringe): endoscopic diagnosis of pathology

I have been greatly interested in the meaningful data from Tahara et al. presented in response to our reported finding using magnifying narrow band imaging (NBI). The authors suggested that a finding of the light blue crest by magnifying NBI, especially in the gastric corpus, could be a good indicator of high risk for the development of gastric cancer. I think their data highlight two important points about endoscopic diagnosis of intestinal metaplasia in the stomach.

First, the authors indicated that endoscopy could diagnose and show the distribution of intestinal metaplasia, which is associated with a risk of developing gastric cancer. When we investigated, in a case–control study, the relationship between risk of gastric cancer and histological findings of biopsy specimens according to the Sydney system, the most important risk factor was the presence of intestinal metaplasia at the lesser curvature of the corpus: odds ratio of 15.1 (95 %CI 4.3 – 52.6, P < 0.001) [1]. Therefore, it is important to evaluate not only the presence of intestinal metaplasia, but also its distribution when assessing the risk for development of gastric cancer.

Currently, the diagnosis of gastritis is basically evaluated by histology of biopsy specimens from certain points on the gastric mucosa, as stated by the Sydney system. However, ”blind biopsy” cannot evaluate the precise distribution of gastritis and this method potentially causes sampling error. To landscape the whole stomach, chromoendoscopy with methylene blue is more advantageous [2], but it demands preparation with a mucolytic agent and vigorous washing away of mucus or excessive dye. Although the NBI image is dark in the stomach and color contrast is lower compared with that of methylene blue chromoendoscopy, at least the presence or absence of the light blue crest is evaluable when observing magnified areas, so it can target a biopsy site. Furthermore, the area of the light blue crest appears as light-blue–whitish patches in nonmagnifying NBI images, so subjective semiquantitative assessment may be possible. Nevertheless, one of the main virtues of NBI is convenience. The effect of NBI can be achieved just by pressing a small button on the control head of the video endoscope. We expect this easier method to facilitate its widespread use in clinical practice.

Secondly, the data of Tahara et al. suggested that new endoscopic imaging could be an alternative to biopsy. According to our investigation, the appearance of the light blue crest corresponded closely with the presence of histological intestinal metaplasia. We speculated that the light blue crest originated from dense reflection of short-wavelength light at ciliated structures on the epithelial surface of intestinal metaplasia (i. e. the brush border), so the light blue crest is a light image and not a surface pattern. When we first noticed the appearance of the light blue crest, we named it the ”blue fringe” [3] because it looked like a pale blue flashing light fringing on the crest of the epithelial surface or gyri; during submission it was recommended to revise the term to ”light blue crest.” Sometimes thin white lines that mimic the light blue crest were observed, but these were probably due to a tangential view of the epithelial surface; a typical light blue crest is relatively thick, more lightly blueish, and more reflecting.

The appearance of the light blue crest is clearly visible at maximum zoom level. Taking an excellent magnified image is key to evaluating magnified NBI findings correctly. Tips for magnifying observation include using a cap and fine movement of the zoom lever. In our previous report, we used a transparent plastic cap but we now prefer a black rubber cap (MB-162 or MB-45, Olympus Medical Systems, Inc., Tokyo, Japan) because it is softer and can avoid contact bleeding during close observation. Sometimes the zoom lever is extended all the way down at once, but we use it more like a ”microscope”, focusing and adjusting the amount of zoom with a fine movement. NBI with zoom can relate gross endoscopic appearance to its histological findings very well; therefore we think it could accelerate our understanding of the pathogenesis of diseases in the digestive tract by endoscopy. Elucidation of the pathogenesis would result in improved diagnosis and treatment methods, and finally benefit patients. To achieve this, we think that every attempt should be made to clarify and systematize magnified NBI findings as fully as possible in accordance with histology. We do not yet know how many answers can be found, but at least through the process, we could improve endoscopic diagnosis and treatment of gastrointestinal diseases.

Competing interests: None

References

• 1 Shiotani A, Iishi H, Uedo N. et al . Histologic and serum risk markers for noncardia early gastric cancer.  Int J Cancer. 2005;  115 463-469
• 2 Areia M, Amaro P, Dinis-Ribeiro M. et al . External validation of a classification for methylene blue magnification chromoendoscopy in premalignant gastric lesions.  Gastrointest Endosc. 2008;  67 1011-1018
• 3 Uedo N, Ishihara R, Iishi H. et al . New diagnostic modality of gastric intestinal metaplasia: narrow band imaging system with magnifying endoscopy.  Gastrointest Endosc. 2005;  61 AB184