Background: It is theorized that repeated gastric distention leads to dilatation of the cardia and the development of GERD. We hypothesize that cardia circumference correlates with the presence and the severity of GERD, and we developed software to measure cardia circumference from static endoscopic images. Our aims were to validate the software and to quantify cardia circumference along the spectrum of GERD.
Methods: Software-based measurements were compared with actual measurements in animal and mechanical models. A retrospective review of an endoscopic database and patient charts produced 273 subjects, grouped as follows: controls, GERD, < or = 3-cm Barrett's esophagus, or > 3-cm Barrett's esophagus. A blinded abstractor measured cardia circumference by using images from the database.
Results: Software and actual measurements correlated closely and were reproducible among observers. Median cardia circumference for each group was the following: control, 31.8 mm; GERD, 37.8 mm; < or = 3-cm Barrett's esophagus, 38.4 mm; and > 3-cm Barrett's esophagus, 45.0 mm (p < 0.001). By using 34.3 mm as a cutoff, cardia circumference was 85.3% sensitive and 89.6% specific for the diagnosis of GERD.
Conclusions: There was a direct relationship between cardia circumference and the presence of GERD. This finding augments our understanding of the anatomic contributions of the esophagogastric junction in the pathogenesis of GERD. Cardia measurement may prove to be a useful diagnostic tool.