Optical enhancement with magnification versus white-light endoscopy for detecting gastric intestinal metaplasia and neoplasia: a randomized controlled trial.

[BACKGROUND AND AIMS] The detection of gastric intestinal metaplasia (GIM), gastric intraepithelial neoplasia (GIN), and early gastric cancer (EGC) using white-light endoscopy (WLE) is unsatisfactory. This study assessed the value of optical enhancement combined with magnification endoscopy (ME-OE) versus WLE for detecting GIM, GIN, and EGC.

[METHODS] Patients at high risk for gastric cancer were randomly assigned to the ME-OE group or WLE group at a 1:1 ratio. Suspicious GIM, GIN, and EGC lesions underwent targeted biopsy sampling in both groups. The diagnostic yield, diagnostic efficacy, and agreement of suspicious lesions were compared between the 2 groups.

[RESULTS] Two hundred eighty-five patients were finally analyzed. The per-patient diagnostic yields of GIM, GIN, and EGC were 36.6% in the ME-OE group and 23.8% in the WLE group (P = .018). The per-lesion diagnostic yield of GIM, GIN, and EGC in the ME-OE group was higher than that in the WLE group (66.7% vs 48.7%, P = .017). Sensitivity (82.8% vs 54.3%, P = .003), specificity (84.2% vs 81.4%, P = .738), positive predictive value (88.9% vs 70.4%, P = .038), negative predictive value (76.2% vs 68.6%, P = .419), and accuracy (83.3% vs 69.2%, P = .028) for GIM were compared between the 2 groups. The intraobserver agreements of experienced endoscopists were excellent for ME-OE (κ = 0.81 and κ = 0.83) and good for WLE (κ = 0.63 and κ = 0.62). The interobserver agreements of experienced endoscopists were good for both groups (κ = 0.75 and κ = 0.61, respectively).

[CONCLUSIONS] ME-OE showed better performance for detecting GIM than WLE in high-risk populations. (Clinical trial registration number: NCT04411589.).