Multichannel Intraluminal Impedance in the Evaluation of Patients with Persistent Globus on Proton Pump Inhibitor Therapy

Abstract

Objectives: Studies suggest an association between globus sensation (“globus”) and gastroesophageal reflux. Although globus often persists despite proton pump inhibitor (PPI) therapy, the role of nonacid reflux has never been determined. Our goal was to use 24-hour multichannel intraluminal impedance (pH/MII) to determine 1) reflux characteristics in patients with globus despite PPI therapy; 2) whether there are differences in reflux patterns between patients with globus and patients with heartburn alone; and 3) whether there are differences in the number of positive symptom indices when standard pH testing is compared to pH/MII. Methods: Twenty-one adult patients with globus and 12 with heartburn symptoms alone underwent 24-hour pH/MII on PPI therapy. Reflux frequencies and characteristics were compared in both groups. For patients with globus, the symptom index and symptom sensitivity index were calculated. Logistic regression was performed to determine which reflux characteristics correlated with symptoms. Results: With pH/MII, we detected 1,160 reflux episodes; 64.7% were nonacid, and 55% reached the proximal sensor. In identifying patients with positive symptom indices, pH/MII increased the yield of standard pH testing by 27.8%. Proximal reflux was a significant predictor of globus symptoms (p = .04). Nonacid reflux approached significance in predicting globus (p = .08). Compared to patients with heartburn alone, those with globus had a higher mean frequency of proximal reflux (27.76% versus 15.63%; p = .04). Conclusions: In adults on PPI therapy with refractory globus, acid reflux does not predict globus symptoms. Through detection of nonacid reflux, pH/MII increased the yield of standard pH testing in identifying positive symptom indices. Additionally, pH/MII provided important information regarding reflux-symptom correlation. Proximal reflux may be predictive of globus symptoms.