Application of transnasal capsule esophagoscope (E.G scan) in patients with upper GI bleeding: A pilot trial of feasibility, safety, and utility to predict high risk endoscopic lesions

Introduction: The optimal timing of urgent endoscopy (within 24 hours) for patients with upper GI bleeding (UGIB) remains controversial. Emergency endoscopy (EE) (within 4 hours) is indicated in patients with high risk laseions (Forrest I-IIA), but risk stratification (triage) based on only the clinical criteria can be insufficient to properly select these patients. A novel disposable transnasal capsule esophagoscope, the E.G.Scan (EGSC) allows direct visualization of the esophageal and gastric content. The aim of the present study was to determine whether the application of EGSC in patients with acute UGIB could be useful to improve current pre-endoscopic risk stratification.

Patients and methods: We prospectively investigated 22 patients admitted with UGIB into our emergency department. UGIB was defined as a history of vomiting fresh or altered blood with or without melaena in the preceding 7 days. After hemodynamic stabilization, both nasogastric tube aspiration (NGA) and EGSC obtained before EE. The first generation EGSC system (Intromedic) was applied successfully without complications in all patients with a disposable ultrathin transnasal probe with camera capsule at the tip. Esophageal and gastric content was defined as clear, coffee ground material and bloody. Endoscopic findings were defined as high risk (HRL) (Forrest I, IIA-B) and low risk (LRL) (Forrest IIC, III) lesions. Finally results of NGA and EGSC were compared to the diagnosis EE.

Results: In all but one patient there was an excellent correlation in the assessment of esophageal and gastric content between the EGSC and EE, in contrast in 8 out of 22 patients NGA and endoscopy demonstrated discordant results. Ongoing bleeding (bloody content) was demonstrated in 5 out of 6 patients with EGSC. More importantly, compared to EE Forrest HRL and LRL were differentiated with EGSC better than with NGA: PPV: 78% vs. 65%,; NPV: 87% vs. 100%, sensitivity: 78% vs. 100%, specificity: 85% vs. 46% and accuracy: 82% vs. 68%, respectively. 100% of patients with bleeding esophageal varices were also identified on EGSC.

Conclusions: EGSC is feasible, safe and valuable method to assess gastric and esophageal content in the emergency room setting. Our pilot study demonstrated that EGSC can be useful to predict high-risk endoscopic lesions in patients with UGIB. Further, randomized prospective study is needed to determine the diagnostic value of EGSC as to able to identify patients who would benefit from EE.

The study was supported by the Intromedic co.