Gastric scintigraphy with a liquid–solid radiolabelled meal: performances of solid and liquid parameters

Abstract

To assess the clinical performance of parameters of liquid-solid gastric emptying (GE) scintigraphy. Fifty-three controls and 476 patients underwent GE scintigraphy using a liquid-solid test meal (non-ulcer dyspepsia, n=180; gastro-oesophageal reflux disease, n=123; dyspepsia after anti-reflux surgery, n=29; diabetes mellitus, n=96; cystic fibrosis prior to heart-lung transplantation, n=48). Time-activity curves were fitted by a power exponential function and half-emptying times (T1/2) were computed. The lag phase (Tlag) and constant emptying (TRE) times of solid emptying were also calculated using a mathematical method (maximum slope tangent method). TRE and T1/2 of solids were higher in each subgroup of patients vs. controls (P=0.0001) and in cystic fibrosis patients vs. gastro-oesophageal reflux patients (P=0.0001). Tlag was significantly higher only in non-ulcer dyspepsia patients vs. controls (P=0.001). There was no significant difference for liquid parameters. Using the mean+/-1.96 SD of the solid and liquid T1/2 values obtained in controls, GE was normal (n=251; 53%), delayed (n=183; 38%), accelerated (n=33; 7%) or mixed (n=9; 2%). Delayed solid T1/2 was the most prominent alteration (n=189), and alterations of liquid GE alone were present in only 24 (5%) patients. A good correlation was found between solid T1/2 and TRE (r=0.88), but no correlation between Tlag and TRE, suggesting that these estimates represent independent phases of GE. In 26 patients, all GE parameters of solids and liquids were normal except Tlag (n=8) or TRE (n=18). The lack of significant differences between the different patient subgroups did not allow emptying profiles to be drawn according to patient pathology. Liquid GE scintigraphy provided poor and unreliable information in terms of patient discrimination and the drawing of pathophysiological profiles of abnormal GE. Tlag and TRE may confirm GE alteration, especially when solid T1/2 values are at the superior limit of normality, and may improve the performance of GE scintigraphy, rather than using liquid parameters.