Magnifying narrow-band imaging, when used in addition to conventional white-light imaging, greatly improves accuracy, sensitivity, and specificity in the detection of gastric mucosal cancers, Dr. Yasumasa Ezoe and colleagues reported in the December issue of Gastroenterology.
"This has enormous significance in clinical practice, because the examination with high positive predictive value and high negative predictive value might enable the clinician to make appropriate judgments as to which lesion needs pathology to confirm the diagnosis," the authors wrote (Gastroenterology 2011 Dec. 1 [10.1053/j.gastro.2011.08.007]).
Indeed, when combined, the two imaging modalities "might be the best approach for making accurate diagnoses of small gastric cancers," offering a so-called "optic biopsy" for gastric mucosal cancers, the authors suggested.
Dr. Ezoe of Kyoto University and colleagues studied patients aged 20 years or older seen at multiple institutions in Japan who had either untreated gastric cancers or a history of gastric cancer. Patients with prior surgical stomach resection were excluded from the study, although prior minimally invasive procedures, such as endoscopic mucosal resection and endoscopic submucosal dissection, were allowed.
All 353 patients included in the study initially underwent conventional white-light imaging (C-WLI). When a newly detected, undiagnosed, small (10 mm or less), depressed gastric lesion was detected, patients were immediately randomized to undergo detailed examination of the lesion using either C-WLI (n = 176) or magnifying narrow-band imaging (M-NBI) in a 1:1 ratio.
All lesions initially evaluated with C-WLI were subsequently evaluated with M-NBI, to ascertain the predictive value of both modalities together.
In the case of multiple lesions, only the first lesion was included in the study. Small, depressed lesions with apparent erosion or ulceration were also not evaluated, "as it is difficult to visualize surface changes in these lesions," wrote the authors.
After all the target lesions were examined, at least one biopsy specimen was collected and the revised Vienna classification system was used to diagnose C4 (mucosal high-grade neoplasia) or C5 (submucosal invasion by neoplasia) specimens as cancerous.
Overall, 20 patients in each group had a newly diagnosed gastric cancer (13% for both groups).
According to the authors, when compared with biopsy results, the diagnostic accuracy of M-NBI was significantly greater than that of C-WLI (90.4% versus 64.8%, respectively; P less than .001). M-NBI also beat C-WLI on specificity (94.3% vs. 67.9%, respectively; P less than .001).
However, in terms of sensitivity, the two techniques were similar, with M-NBI at 60.0% and C-WLI at 40.0%, respectively (P = .34).
The authors then looked at M-NBI plus C-WLI, and found that the former "significantly enhanced the diagnostic performance of the latter," with accuracy increasing to 96.6% when both were used together, specificity increasing to 96.8%, and sensitivity to 95%, with P less than .001 for all values when compared with C-WLI alone.
Similarly, "C-WLI followed by M-NBI dramatically improved the positive predictive value from 13.8% to 79.2%" compared with C-WLI alone (P less than .001). The negative predictive value also increased, from 89.8% with C-WLI alone to 99.3% when both techniques were used together (P less than .001).
The authors conceded that their sample size was small, and that larger studies will be needed to confirm the diagnostic utility of each modality. Additionally, they did not compare these modalities to dye-based imaging methods.
However, they added, dyes "are only used in a few countries and institutes, and then, the standard worldwide endoscopic method to diagnose early gastric cancer is still C-WLI without any dye use," they added.
The authors disclosed no personal conflicts of interest. The study was sponsored by a grant from the Ministry of Health, Labor, and Welfare of Japan.