An Approach to Comprehensive Gastric Emptying Study

Objectives Although it has been reported that gastric emptying is closely tied to antral motility, as assessed by scintigraphy motility correlates poorly with scintigraphic gastric emptying times. To help to localize the cause of delayed emptying we explored the relative influence on gastric emptying of antral motility and fundal accommodation.

Methods Forty-four gastric emptying studies (GES) were retrospectively selected from our database according to their percent emptying at 2 hours: emptying diagnostic categories (eDxC) = 1)rapid, 2)normal, 3)borderline-delayed and 4)delayed. Regional gastric emptying was evaluated using geometric mean time-activity curves derived from manually drawn fundal regions-of-interest (ROI) and from computer generated proximal ROIs applied to GES images at successive 15-minute time points. Fundal ROIs were generally defined by the initial food distribution and modified to maintain a fixed position relative to the most inferior aspect of the stomach as food redistributed during processing and emptying. Using a similar strategy, proximal ROIs were automatically derived from the oral (proximal) 50% of the total-stomach GES ROIs because it is generally accepted that the fundus and proximal body act as a unit from the point of view of accommodation. Percent of regional emptying at 1 and 2 hours was also characterized by eDxCs as rapid, normal, borderline-delayed or delayed. Antral motility was characterized by motility diagnostic categories (mDxC): 1)hyper, 2)normal, 3)borderline-hypo and 4)hypo from anterior dynamic antral scintigraphy (DAS) at the time of prominent antral activity. Discordance between total-stomach GES and DAS was evaluated by noting the number of studies with differences of 1, 2, or 3 between their respective DxCs. Total-stomach and regional half-emptying times were also compared, as were 2-hour total-stomach and regional 1 and 2-hour percent emptying. Total-stomach/regional emptying discordance was evaluated from their eDxC differences.

Results In the 39 studies with DAS, total gastric emptying and antral motility (evaluated at 68±27 min) had differences of 1, 2, or 3 in their respective DxCs in 11 (28%), 9 (23%) and 7 (18%) studies, respectively. Of the 7 studies with a difference of 3 DxCs, 6 patients had delayed total-stomach emptying. All of these had either delayed or borderline-delayed proximal emptying at 1 and 2 hours. The single rapid emptying study with a difference of 3 DxCs had normal 1-hour proximal and rapid 2-hour proximal emptying. Percent of 2-hour total-stomach emptying was well correlated with fundal emptying at 1 and 2 hours (r=0.85, r=0.84, respectively). Total-stomach/fundal emptying discordance by 1, 2, and 3 eDxCs was 36%, 2% and 0% at 1 hour and 34%, 2% and 0% at 2 hours. Fundal and/or proximal half-emptying time was greater than 2 hours in 13 of 44 patient studies. Mean half-emptying times were not significantly different in the remaining 31 (59±29 vs 65±25 min, p = 0.15), although they were poorly correlated (r=0.68). In the 44 patient studies, fundal and proximal percent emptying differed significantly (38±30% vs 32±27%, p < 0.05 at 1 hour; 67±30% vs 62±31%, p < 0.002 at 2 hours) although they were well correlated (r=0.80 and r=0.95, respectively.)

Conclusions Total-stomach gastric emptying is coupled more to fundal/proximal emptying than to antral motility, as demonstrated by the high correlation between regional and total-stomach emptying and by their concordance, generally within one eDxC. Total-stomach gastric emptying/motility discordance is thus a demonstration of the relatively independent function of these different regions of the stomach.