Volume, distribution and acidity of gastric secretion on and off proton pump inhibitor treatment: a randomized double-blind controlled study in patients with gastro-esophageal reflux disease (GERD) and healthy subjects

Abstract

Background: Postprandial accumulation of gastric secretions in the proximal stomach above the meal adjacent to the esophagogastric junction (EGJ), referred to as the 'acid pocket', has been proposed as a pathophysiological factor in gastro-esophageal reflux disease (GERD) and as a target for GERD treatment. This study assessed the effect of proton pump inhibitor (PPI) therapy on the volume, distribution and acidity of gastric secretions in GERD and healthy subjects (HS).

Methods: A randomized, double blind, cross-over study in 12 HS and 12 GERD patients pre-treated with 40 mg pantoprazole (PPI) or placebo b.i.d. was performed. Postprandial secretion volume (SV), formation of a secretion layer and contact between the layer and the EGJ were quantified by Magnetic Resonance Imaging (MRI). Multi-channel pH-monitoring assessed intragastric pH.

Results: A distinct layer of undiluted acid secretion was present on top of gastric contents in almost all participants on and off high-dose acid suppression. PPI reduced SV (193 ml to 100 ml, in HS, 227 ml to 94 ml in GERD; p < 0.01) and thickness of the acid layer (26 mm to 7 mm, 36 mm to 9 mm respectively, p < 0.01). No differences in secretion volume or layer thickness were observed between groups; however, off treatment, contact time between the secretion layer and EGJ was 2.6 times longer in GERD compared to HS (p = 0.012). This was not the case on PPI.

Conclusions: MRI can visualize and quantify the volume and distribution dynamics of gastric secretions that form a layer in the proximal stomach after ingestion of a liquid meal. The secretion volume and the secretion layer on top of gastric contents is similar in GERD patients and HS; however contact between the layer of undiluted secretion and the EGJ is prolonged in patients. High dose PPI reduced secretion volume by about 50% and reduced contact time between secretion and EGJ towards normal levels.

Trial registration: NCT01212614.

Figures

Fig. 1

Flow chart of the double blind, crossover placebo controlled study

Fig. 2

Definition of the EGJ position and 3D reconstruction of gastric content. a Transverse abdominal MR images at time points t = 10 and 80 min. EGJ is marked by a white arrow. Stomach contours are depicted as yellow lines. b The 3D reconstructed contours of stomach (black closed lines) and esophagus (dark blue closed lines) together with the color-coded gastric content (blue to red) of the same subject at the same time points. Gastric content was color-coded to highlight the formation of the secretion layer, here depicted in red. A non-linear color-coding of the contrast optimized grayscale values in the MR images was applied. The location of the EGJ is marked by the black dot at the end of the esophagus

Fig. 3

Schematic of post-processing steps for the quantification of acid layer thickness and volume from MR images. Gastric secretion accumulated predominantly on the meal surface and a secretion concentration gradient was observed from the surface of the meal into the test meal along the direction of gravity (i.e. the right side of the subject lying in the right lateral position) [10, 13]. The x-axis of the image data was set anti-parallel to the direction of gravity. Gastric content was sliced along the x-axis using a slice thickness of 1 mm and corresponding %secretion values were averaged along the other two Cartesian coordinates. This resulted in a 1D projection of mean %secretion values along the x-axis that was computed for each gastric secretion scan and stacked together over time to allow visualization of the formation of the gastric secretion layer (‘layer-graphs’). In the layer-graphs, the layer thickness was defined as the distance from the meal surface to the x coordinates having a threshold value of ≥70 % secretion. Layer volume (LV) was calculated by summing all pixels above this threshold

Fig. 4

Grouped meal and secretion volume curves. Meal Volume (MV) and Secretion Volume (SV) dynamics over 120 min for GERD patients and Healthy Subjects (HS), under placebo (black dots/lines) and PPI (grey triangles/lines). The approximately linear emptying pattern of the test meal is typical for a high calorie, viscous liquid meal

Fig. 5

Formation of individual secretion layers. Layer thickness over time is depicted for all healthy subjects (HS) and GERD patients under placebo (black dots/lines) and PPI therapy (grey triangles/lines)

Fig. 6

Correlation of layer and secretion volume under placebo and PPI therapy. GERD and HS data are depicted as triangles and circles, respectively; placebo volumes (black), PPI volumes (grey). The offset in the correlation indicates that layer formation started when approximately 50 ml secretion was present in the stomach. For placebo, the regression coefficient was R2 = 0.7 with regression slope and offset of 1.04 ± 0.04 and -52 ± 7 ml, respectively. On pantoprazole therapy, the regression coefficient was reduced to R2 = 0.3 with regression slope and offset of 0.56 ± 0.06 and offset -18 ± 5, respectively

Fig. 7

Layer graphs including EGJ position and time to Layer formation. Layer graphs of a healthy subject (left) and a GERD patient (right) under placebo (top) and PPI therapy (bottom). The vertical axis represents the MR image x-axis, which is aligned along the direction of gravity. The colours code the average %secretion. The threshold level of the layer (i.e. ≥70 % secretion) and the position of the EGJ are indicated as white dashed and black dotted line, respectively