Analysis of the meal-dependent intragastric performance of a gastric-retentive tablet assessed by magnetic resonance imaging


Dipl.-Ing. A. Steingoetter, Department of Gastroenterology, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland.


Background : Modern medical imaging modalities can trace labelled oral drug dosage forms in the gastrointestinal tract, and thus represent important tools for the evaluation of their in vivo performance. The application of gastric-retentive drug delivery systems to improve bioavailability and to avoid unwanted plasma peak concentrations of orally administered drugs is of special interest in clinical and pharmaceutical research.

Aim : To determine the influence of meal composition and timing of tablet administration on the intragastric performance of a gastric-retentive floating tablet using magnetic resonance imaging in the sitting position.

Methods : A tablet formulation was labelled with iron oxide particles as negative magnetic resonance contrast marker to allow the monitoring of the tablet position in the food-filled human stomach. Labelled tablet was administered, together with three different solid meals, to volunteers seated in a 0.5-T open-configuration magnetic resonance system. Volunteers were followed over a 4-h period.

Results : Labelled tablet was detectable in all subjects throughout the entire study. The tablet showed persistent good intragastric floating performance independent of meal composition. Unfavourable timing of tablet administration had a minor effect on the intragastric tablet residence time and floating performance.

Conclusion : Magnetic resonance imaging can reliably monitor and analyse the in vivo performance of labelled gastric-retentive tablets in the human stomach.


The bioavailability and effectiveness of orally administered drugs can be controlled and optimized by integrating the drugs in gastric-retentive drug delivery systems, based on muco-adhesive, floating, high-density and swelling principles.1–4 This is especially important for drugs that act locally in the stomach,5, 6 have a short absorption window in the small intestine7 or must be released at the site of macronutrient digestion, as is the case in pancreatic insufficiency and cystic fibrosis.8 Medical imaging modalities, such as γ-scintigraphy, are reliable tools for tracing labelled oral drug dosage forms in the animal and human gastrointestinal tract, and therefore represent important methods for the evaluation of the in vivo performance of such drug delivery systems.9–15 However, most of these techniques suffer from poor temporal and three-dimensional spatial resolution. In addition, scintigraphic methods expose patients or volunteers to ionizing radiation that precludes repetitive study designs.

In a recent study, the feasibility of magnetic resonance imaging (MRI) to monitor the intragastric course of a labelled and orally administered gastric-retentive tablet has been demonstrated.16 The use of MRI for this purpose has increased in importance due to the feasibility of labelling tablets with superparamagnetic iron oxide (Fe3O4) particles. Fe3O4, which is paramagnetic, induces a strong local reduction of the T2* relaxation time, causing a signal void (so-called susceptibility artefact) in the resulting magnetic resonance images.17–21 Consequently, the intragastric position of an ingested labelled tablet can be traced by the susceptibility artefact created in the image.

In addition, the use of MRI has increased in significance due to the development of open-configuration magnetic resonance systems with a gap between the magnetic rings, permitting scanning in the sitting position. This technique takes into account the effect of gravity, which is of importance when evaluating the in vivo performance of oral drug delivery systems. Currently, experimental data are lacking on the impact of meal composition and intake time on the intragastric performance of gastric-retentive tablets. Therefore, in this study, we used an open-configuration magnetic resonance system to evaluate the in vivo performance of an Fe3O4-labelled gastric-retentive tablet in relation to different meals in the stomachs of asymptomatic volunteers, with special emphasis on intragastric tablet position and residence time.

Materials and methods

This study was approved by the Institutional Review Board and informed consent was obtained from all volunteers who participated in the study.

Study population

Twenty healthy male volunteers (age, 19–38 years; body mass index, 17.9–26.3 kg/m2) participated in the study. Volunteers were instructed to arrive fasted on the study day or to have a light and fat-free breakfast at least 6 h before the onset of the measurements. Fifteen of these volunteers were randomly divided into three ‘meal groups’, i.e. hamburger meal, cheese meal and pasta meal, to study the influence of the test meal composition on the intragastric tablet position. The composition of each test meal is listed in Table 1. Within each meal group, the volunteers consumed the same meal. The remaining five volunteers were assigned to a ‘challenge group’ to study the influence of the timing of tablet administration, referred to as the ‘challenge test’. These five volunteers consumed the hamburger meal under the same conditions as above, but with the tablet ingested at a different time point (described below).

Table 1.  Composition of the test meals
 Of portion
Weight (g)Fat (g)kcal
Hamburger meal
 French fries16020.8448
 Green beans120042
 Orange juice100045
 Fruit salad2000136
Cheese meal
 Curd cheese9014.4189
 Whipped cream204.44451.4
 Orange juice100045
 Fruit salad2000136
Pasta meal
 Minced meat12014.4231.6
 Olive oil2020176.8
 Orange juice100045
 Fruit salad2000136

Tablet formulation

A floating tablet formulation was labelled with Fe3O4 crystallites as ‘negative’ magnetic resonance contrast marker (i.e. creating a signal void in the image). Tablets were prepared in house at the pharmacy of the University Hospital Zurich. The tablet composition was as follows: Fe3O4 crystallites (1%), citric acid monohydrate (2.5%), NaHCO3 (2.5%), polyvinylpyrrolidone K30 (1%), metolose 90SH-4000SR (82%), magnesium stearate (1%) and lactose (10%). Fe3O4 crystallites were thoroughly mixed with the tablet compounds before drying of the tablet powder and final tablet formation. Tablets were formed with a hydraulic press using a tool size of 15 mm × 6.9 mm and a compression force of 3 kN for 5 s. The tablet weight was 400 g.

Influence of test meal composition

Volunteers were asked to eat the meal within 15 min. Immediately after completion of the meal, the marked floating tablet was ingested together with 50 mL of water. Subsequently, each volunteer was placed in a sitting position in the 0.5-T open-configuration magnetic resonance system (Signa SP, GE Medical Systems, Milwaukee, WI, USA), as shown in Figure 1. A standard send–receive surface coil for abdominal MRI was wrapped around the upper abdomen of the volunteer for signal reception. Magnetic resonance measurements were taken at 30-min intervals for at least 180 min and at most up to 240 min. At each time point, a stack of 24 sagittal magnetic resonance images, covering the total gastric region (volume scan), was acquired within three breath holds of 25 s using a T1-weighted fast spoiled gradient echo sequence (repetition time, 150 ms; echo time, 8 ms; flip angle, 60°; field of view, 280 mm; slice thickness, 8 mm; interslice gap, 0 mm; matrix, 256 × 160). The volunteers were not allowed to drink or eat during the 4-h study period. However, they were allowed to move freely between the measurements.

Figure 1.

Seated volunteer with fixed abdominal send–receive surface coil in the 0.5-T open-configuration magnetic resonance system.

Challenge test (influence of the timing of administration)

To assess the effect of the timing of tablet administration, i.e. to investigate tablet floating performance under sub-optimal conditions, five volunteers were asked to ingest the tablet after completion of the first quarter of the total meal. Ingestion of the tablet during meal intake instead of after meal intake is potentially challenging for the tablet, as it has to float up past the incoming meal in order to retain good intragastric floating performance. After meal intake, magnetic resonance imaging in the sitting position was performed in the same way and at the same time intervals as described above.


Image analysis

The gastric residence time of the tablet Tmax was determined and defined as the time between the first and last detection of the tablet in the stomach. To analyse gastric emptying of the meal, the gastric contents were outlined in every image of a volume scan acquired during the first 180 min. The area of the outlined gastric contents was calculated and multiplied by the slice thickness to obtain the volume of gastric contents, which was then plotted over time. To determine the tablet floating performance, the location of the tablet within the stomach was analysed at every time point based on the acquired volume scan. Intragastric tablet location was calculated relative to the meal level. Floating performance was expressed as a percentage and was defined as the ability of the tablet to stay afloat on the meal surface. A value of 100% corresponded to a tablet floating at the meal surface and a value of 0% corresponded to a tablet positioned at the bottom of the stomach (Figure 2). Analysis was performed using the medical image processing software eFilm Workstation 1.5.3 (eFilm Medical Inc., Toronto, Ont., Canada) and a home-built software package implemented in IDL 5.4 (Research Systems Inc., Boulder, CO, USA).

Figure 2.

Sketch of the coronal view of the stomach illustrating the calculation of the tablet floating performance from the intragastric tablet position. Sagittal image planes (1–24) are depicted as thin broken lines. The gastric incisura dividing the proximal and distal stomach volume is indicated by a bold dotted line.

Statistical analysis

The gastric residence time and floating performance of the tablet and the gastric emptying of the different meals (hamburger, cheese, pasta) were compared using analysis of variance (anova) with repeated measures. The tablet floating performance observed in the challenge test was compared with the corresponding data in the hamburger group. Data were expressed as the median (interquartile range). P < 0.05 was considered to be statistically significant.

Three-dimensional visualization

Three-dimensional visualization of the stomach or meal volume, together with the tablet, was performed for descriptive analysis of the intragastric tablet performance. A surface triangulation based on the outlined contours of the stomach or meal volume and the tablet was applied using a freely available software tool called IsoSurf v1.5d (http://svr-, Dr Graham Treece). The resulting three-dimensional surface points were then visualized with a three-dimensional CAD/CAM Viewer, 3Space Assistant, from TGS Europe (Merignac Cedex, France).


Influence of meal composition

Tablet administration was well tolerated by all volunteers in the three meal groups (hamburger, cheese and pasta). MRI was performed at all time points for all volunteers. The susceptibility artefact originating from the Fe3O4-labelled tablet allowed the detection of the intragastric tablet position throughout the entire study period. The average diameter of the circular artefact area, as detected in the sagittal magnetic resonance images, was 40 mm. Figure 3 shows typical sagittal magnetic resonance images depicting the gastric contents and tablet artefact of a volume scan at time t = 90 min for one volunteer for each test meal. Figure 4 depicts the three-dimensional visualization of the meal volume and tablet in a volunteer (hamburger meal) at times t = 60, 120 and 180 min.

Figure 3.

Sagittal magnetic resonance images showing the outlined gastric volume and tablet artefact from each of the meals at time t = 90 min. The size of the artefact is different from the size of the tablet, depending on the Fe3O4 concentration in the tablet. Stomach wall, tablet artefact and anatomical landmarks are indicated by capital letters: H, heart; K, kidney; L, lung; S, stomach; T, tablet.

Figure 4.

Sagittal magnetic resonance images of volume scans and corresponding three-dimensional visualization of the meal volume and tablet in a volunteer from the hamburger group at times t = 60, 120 and 180 min. The tablet floating position on the meal surface in the proximal gastric volume is well depicted.

The gastric emptying curves of the different meals are plotted in Figure 5(a). Gastric emptying was not statistically different (P = 0.71) between the three meal groups. The half-times of gastric emptying (T50%) were as follows: hamburger, 89.1 min (85.1–99.2 min); cheese, 88.0 min (78.7–98.6 min); pasta, 104.8 min (81.3–121.7 min).

Figure 5.

(a) Gastric emptying curves of the meal groups over 180 min. Data plotted as the median (interquartile range). (b) Tablet floating performance as a percentage over 240 min for the three meal groups (hamburger, cheese, pasta).

No difference (P = 0.77) in the tablet residence time Tmax was detected between the three test meals: hamburger, 240 min (240–240 min); cheese, 240 min (240–240 min); pasta, 240 min (240–240 min). In one subject from each meal group, the floating tablet emptied before the end of the study period (hamburger, 90 min; cheese, 180 min; pasta, 180 min).

The tablet floating performances for the three meals are plotted over time in Figure 5(b). No difference (P = 0.72) in tablet floating performance was determined between the meals: hamburger, 94.8% (80.0–97.7%); cheese, 96.0% (92.4–98.5%); pasta, 97.7% (87.9–99.2%).

Challenge test

The half-time of gastric emptying (T50%) for the challenge group was 118.6 min (112.1–126.1 min). The gastric residence time of the tablet Tmax was 240 min (240–240 min). Also, in one subject from this group, the floating tablet emptied before the end of the study period, at time t = 120 min. The floating performance of the tablet in the challenge group is plotted over time in Figure 6. The variation of the tablet floating performance was greater in the challenge group than in the other analysed groups. No difference (P = 0.36) in tablet floating performance was determined between the challenge group and the hamburger group: challenge, 92.6% (51.8–97.9%); hamburger, 94.8% (80.0–97.7%).

Figure 6.

Tablet floating performance as a percentage over 240 min for the challenge group (full line) and the corresponding hamburger group (broken line). Data from the challenge group are plotted as the median (interquartile range).


The oral delivery of drugs to the human systematic circulation is of major interest, because it represents the most convenient and popular method of drug administration for patients and physicians. Gastric-retentive oral drug delivery systems were developed to optimize the time course of intestinal delivery for specific drugs, e.g. drugs that require interaction with specific macronutrients or have a short intestinal absorption window. In combination with timed release, these delivery systems can improve the bioavailability of drugs and lower toxicity due to the reduction in unwanted drug plasma peaks.

In this study, the intragastric floating performance and residence time of a gastric-retentive tablet, administered together with different semi-solid meals, were determined by MRI in the sitting position. In all subjects, the tablets showed good persistent intragastric floating performance. The tablet floating performance and residence time showed no dependence on the meal composition or time of administration. The use of an open-configuration magnetic resonance system provides the unique opportunity to evaluate the tablets intragastrically in a sitting position.

The experimental set-up permitted volunteers to move freely in between measurements, which is essential for a 4-h study period. In comparison with more established imaging techniques, such as γ-scintigraphy, X-rays and ultrasound, MRI has the advantage of imaging the complete three-dimensional gastric volume at high spatial and temporal resolution and adjustable image contrast. In addition, MRI does not involve exposure to ionizing radiation and allows the performance of repeated studies in healthy volunteers. Finally, MRI allows the assessment of complementary physiological data of gastric function, as shown in recent studies by Kunz et al.22 and Marciani et al.,23 using whole-body magnetic resonance systems.

The intragastric floating performance and residence time of the ingested Fe3O4-labelled tablet were determined by localizing and analysing the susceptibility artefact in the magnetic resonance images. The susceptibility artefact was clearly detectable in all magnetic resonance images and was distinguishable from other intragastric signal voids, such as air or solid pieces of meal, thus permitting a reliable calculation of tablet performance. The three-dimensional visualization of the magnetic resonance data better illustrated the intragastric position of the tablet relative to the gastric contents and proved helpful for the understanding of the tablet's intragastric pathway.

The results of this study showed that there were no differences in gastric meal emptying, tablet floating performance or tablet residence time between the three test meals. This may not be surprising given the small number of analysed subjects (n = 5). However, the plotted data (Figure 5) and the medians and interquartile ranges listed in the ‘Results’ section do not indicate the emergence of any trend. This suggests that the non-significant results are due to the small differences in these parameters between the meals rather than to the small sample size.

As mentioned above, no difference in gastric meal emptying was determined. Nevertheless, for each meal group, noticeable inter-subject variations were observed. The observed inter-individual and intra-individual variations for gastric emptying are consistent with those reported in the literature.24

The administered tablet showed persistent good floating performance in all volunteers, independent of the meal composition. Nevertheless, for each meal, one subject emptied the tablet before the end of the study period. For this subject in the hamburger meal group, the tablet showed a poor floating ability immediately after intake and was emptied after only t = 90 min. This effect may be explained by the particular gastric anatomy of this particular volunteer. More meal volume was stored in the distal stomach than in the gastric fundus in comparison with that typically observed in the other volunteers. We assume that, in this subject, the cardia was positioned very low, explaining the low initial intragastric position of the ingested tablet in relation to the meal surface. For the subjects from the cheese meal and pasta meal groups with early gastric emptying, the tablet started to lose floating ability after time t = 180 min, and was emptied afterwards. Interestingly, these two volunteers also had the lowest values for the half-time of gastric emptying in their respective meal groups, suggesting an influence of the gastric emptying pattern on the intragastric tablet performance. However, the similarity of the gastric emptying characteristics for the different meals did not allow us to analyse the influence of gastric emptying on the tablet floating performance in this study. The observations described highlight the potential problems for the development of an optimal gastric-retentive tablet formulation. To ensure a constant and complete intragastric release of the drug in the stomach, the release characteristics must be adjusted according to the intragastric residence time of the tablet. This parameter, however, may be affected by individual differences in gastric anatomy and gastric emptying.

The tablet analysed in the challenge group (the floating tablet was administered during the meal instead of afterwards) also showed a persistent but more variable floating performance. As in the other groups, the tablet emptied earlier in one subject. This result indicates that the floating performance of the tablet is not significantly affected by its ingestion during or after the meal.

Today, the efficacy of most pharmaceuticals is evaluated by analysing their pharmacokinetics using repeatedly collected blood or saliva samples from a large number of subjects.25 In such studies, the influence of the drug delivery system on the bioavailability of the drug remains unspecified. Studies conducted to investigate the effects of newly developed sophisticated delivery systems are usually performed in vitro or in animals.

From the results derived in this study, we conclude that in vitro or animal studies alone will not be sufficient to reliably predict the intragastric fate and dynamics of oral drug delivery systems or administered drugs in the human gastrointestinal tract. More extensive human studies are needed to obtain a comprehensive knowledge of the fate of different drug delivery systems and applied drugs in the human body. This will enable the development of highly sophisticated and specified systems that can further improve and control the bioavailability and effectiveness of administered drugs. In the future, MRI, as a non-invasive and radiation-free imaging tool, might attain equal importance to γ-scintigraphy in this field of research.


We acknowledge the technical and organizational support of Karl Treiber and Bernadette Stutz. The study was supported by F. Hoffmann-LaRoche, Basel, Switzerland and the Swiss National Science Foundation (SNF grants 32-54056.98 and 31-55932.98).