Authors: Henry P. Parkman, Reena Anand, Alexandra C. Barrett, Rona Cooper, Simindokht Dadparvar, Alan H. Maurer
Categories: Original Article, Gastric emptying scintigraphy, Gastroparesis, Early satiety, Nausea
Source: Digestive Diseases and Sciences
Authors: Henry P. Parkman, Reena Anand, Alexandra C. Barrett, Rona Cooper, Simindokht Dadparvar, Alan H. Maurer
Patients with symptoms of gastroparesis are assessed with gastric emptying scintigraphy (GES). However, patients may not eat the entire meal due to their symptoms. The aim of this study was to quantify in normal subjects the effect of different ingested meal sizes and how these results could affect interpretation of patients’ GES.
Normal subjects underwent three GES tests ingesting the currently established Tc-99 m radiolabeled liquid egg white sandwich served as a full (100%), half (50%), or quarter (25%) meal. Percent gastric retention was calculated for each imaging time (0, 0.5, 1, 2, 3, 4 h). T1/2 for GE was calculated using power exponential curve fit.
24 normal subjects completed three GES tests. Percent gastric retention at 0.5, 1, 2, 3 h postprandially were significantly less for half and quarter meal compared to the full meal. Mean + 2SD gastric retention values at 2 h for 50% meal and 25% meal were 41.6% and 59.5%. Compared to T1/2 for the full-size meal (94 ± 21 (SD) min), T1/2 for half meal (67 ± 19 min; p < 0.01) and quarter meal (69 ± 28 min; p < 0.01) were significantly shorter. Mean + 2SD T1/2 emptying values for 50% and 25% meal ingestion were 105.2 and 125.8 min.
Gastric emptying is affected by amount of the meal ingested. T1/2 was shorter, and percent retention at 0.5, 1, 2, 3 h postprandially were less for half- and quarter-size meals compared to the full meal. Based on our results, if a patient ingests 50% of the standard GES meal, > 41.6% retention at 2 h or T1/2 is > 105.2 min is consistent with delayed GE.
Patients with symptoms of gastroparesis are typically assessed with gastric emptying scintigraphy (GES) to determine if they have delayed gastric emptying (gastroparesis) [1]. Because of their symptoms of early satiety, nausea and vomiting, many patients undergoing GES are not be able to eat the entire standardized egg sandwich meal. The rate of gastric emptying is dependent on total calories and meal content ingested [2]. How ingestion of less than the full standard meal impacts the final reported interpretation of the GES test result is not standardized. Can one use the current normal control values which were obtained in normal subjects who consumed the entire meal? Or should final GES study interpretation be based on normative values based on limited standardized meal ingestion size? Currently, if the GES results are not delayed for a small meal ingested, the GES results are often interpreted as non-diagnostic. A recent study by Shah et al. in patients undergoing GES used cluster analysis and showed that the normal values established for the current standardized GES meal could be applied to a subset of patients who ingested at least 50% of the standard meal [3].
The aim of this study was to measure the effect of different ingested meal sizes on gastric emptying in normal subjects and how those results could aid in the interpretation of patients with limited meal ingestion. This was performed utilizing normal subjects who ingested the full (100%), half (50%), and a quarter (25%) of the currently accepted GES sandwich meal [4, 5]. In addition, we also assessed the effect of limited meal ingestion on the intragastric meal distribution immediately postprandially (IMD-0) [6].
Normal volunteer subjects were recruited for this study. The study was performed after institutional IRB approval. After obtaining informed consent, a medical history was taken from each subject to ensure they were healthy with no GI symptoms or history of GI surgery. This was confirmed also with patients filling out a questionnaire capturing medical and surgical history and medications. Medications that subjects were taking were reviewed. Subjects could not be taking medications known to delay gastric emptying, such as opiates, anticholinergic agents, and GLP-1 receptor agonists. The questionnaire included a modified form of the Patient Assessment of Upper GI Symptoms (PAGI-SYM) which asks about symptoms of gastroparesis, dyspepsia, GERD, constipation [7]. Menstruating females were asked about the timing of their menstrual cycle and the gastric emptying studies were scheduled to be performed during the first 10 days of their menstrual cycle. Study subjects underwent three gastric emptying tests, at least 6 days between tests.
Study subjects reported for GES fasting after midnight. The currently consensus recommended radioactive egg sandwich meal with plain water was utilized [4, 5]. The full (100%) meal consisted of 4 oz. (120 g, equal to two large eggs) liquid egg white; two slices of white bread (120 kcal), strawberry jam (30 g, 74 kcal), and water (120 mL). The total caloric value of the meal is 255 kcal (72% carbohydrate, 24% protein, 2% fat, and 2% fiber). The egg white was radiolabeled with 18.5 MBq (0.5 mCi) of Tc-99 m sulfur colloid and microwaved until firm consistency of scrambled eggs. The radiolabeled egg white was prepared and then eaten for the full meal as a sandwich with two slices of white bread and jelly and ingested with 120 mL water.
In randomized order, the meal given to the patient was either the full egg sandwich meal (100%), half sandwich meal (50%), or quarter sandwich meal (25%). A concordant reduced proportion of the water (8 oz, 4 oz, and 2 oz) was also ingested. The dose of Tc-99 m sulfur colloid was similarly reduced to 9.25 MBq (0.25 mCi) for 50% of the meal and 4.62 MBq (0.125 mCi) for 25% of the meal. All subjects were required to complete the meal ingestion within 10 min.
Following ingestion of the test meal, imaging in the anterior and posterior projections was performed at 0, 0.5, 1, 2, 3, and 4 h with patient standing upright in front of the gamma imaging camera. In between images, the subject was sitting or standing.
The percentage of gastric meal retention for each time period was calculated using the geometric mean of the decay corrected counts. We also calculated the overall T1/2 for gastric emptying using power exponential curve fit analysis. In addition, we calculated the intragastric meal distribution immediately postprandially (IMD-0). IMD-0 is used as a measure of proximal stomach accommodation and was calculated as the percent radioactivity in the proximal stomach divided by the amount of radioactivity in the entire stomach, as described by Orthey et al. [6].
We compared the percent retention at 1, 2, 4 h for each of the three meal size groups using ANOVA followed students t test with Bonferroni correction. Comparisons of the T1/2 (time to empty half the meal) and the intragastric meal distribution immediately postprandially (IMD-0) were performed.
In our past studies at Temple University, gastric emptying has been performed in 31 normal control subjects, the percent retention at 2 h is 58.6% with SD of 19.8%. To detect a change in emptying of the meal by 15% from 58.6 to 43.6% retention, using an alpha level of 0.05 and power of 0.8, the number needed is 14. We will be comparing two different meal sizes to the normal meal size. Using Bonferroni correction for multiple comparisons, the number needed is 14 × SQRT (2) = 20. For this study, an extra five were enrolled to cover potential dropouts. We planned to enroll 25 normal subjects for this study.
There were 24 normal subjects (11 males, 13 females; average age 26 years; average weight 167 pounds) who completed all three GES tests. The 25th patient only had one gastric emptying test performed and was not included.
With the full (100%) size meal, the % gastric retention at 1 h was 74.5 ± 9.7%, at 2 h was 30.5 ± 16.5%, and at 4 h 2.2 ± 2.7%. These values compare well with the original Tougas study with median values for percent gastric retention at 60, 120, and 240 min were 69%, 24% and 1.2%, respectively (Tougas).
Figure 1 shows an example of the three different size meals on the images for gastric emptying scintigraphy for one individual. For the entire group of 24 normal subjects, compared to percent retention for the full sandwich meal, percent gastric retention at 0.5, 1, 2, 3 h postprandially were significantly less for the half sandwich meal and the quarter sandwich meal (Table 1). At 4 h postprandially, there was no significant difference for the percent gastric retention for the full meal sandwich (2.2 ± 2.7), the half meal sandwich (1.9 ± 1.9), and the quarter meal sandwich (3.2 ± 2.7). At 2 h postprandially, the percent retention for the full-size meal was 30.5 ± 16.5%, but was 14.6 ± 13.5 for the half-size meal and 18.3 ± 20.6 for the quarter-size meal (p < 0.01). The GE T1/2 for the full sandwich meal was 94.1 ± 21.3 (SD) min) with the T1/2 for gastric emptying for the half sandwich meal equal to 66.6 ± 19.3 min (p < 0.01) and the quarter sandwich meal was 68.6 ± 28.6 min (p < 0.01). Using these normal subject results, our findings suggest that if a patient can ingest at least 50% of the standard GES meal, if > 41.6% retention is present at 2 h or T1/2 is > 105.2 min, the results would be consistent with delayed GE.Fig. 1Images from one normal subject for the three different meal size GES studies. Top Row: Full-size meal (100%). Middle Row: Half-size meal (50%). Bottom Row: Quarter-size meal (25%)Table 1Effect of ingested meal size on gastric emptying scintigraphy in 24 normal subjectsFull mealHalf mealQuarter meal% retention—0100 ± 0100 ± 0100 ± 0% retention—0.5 h90.1 ± 5.0285.5 ± 8.181.0 ± 11.7**% retention—1 h74.5 ± 9.753.0 ± 19.052.1 ± 21.3% retention—2 h30.5 ± 16.514.6 ± 13.518.3 ± 20.6% retention—3 h9.5 ± 9.33.3 ± 3.14.6 ± 4.4% retention—4 h2.2 ± 2.71.9 ± 1.93.2 ± 2.7Upper limit of normal for % retention at 2 h30.5 + 2(16.5) = 63.514.6 + 2(13.5) = 41.618.3 + 2(20.6) = 59.5Upper limit of normal for % retention at 4 h2.3 + 2(2.7) = 7.71.9 + 2(1.9) = 5.73.2 + 2(2.7) = 8.6t-1/2 Power exp (min94.1 ± 21.366.6 ± 19.368.6 ± 28.5IMD-00.78 ± 0.090.78 ± 0.090.80 ± 0.09Results expressed as mean ± SD**p < 0.05. **p < 0.01 versus full meal
In contrast, the intragastric meal distribution immediately postprandially (IMD-0) was not affected by alterations in the amount of meal ingested (Table 1).
This study was designed to obtain normal gastric emptying values for normal subjects who ingested less than 100% of the current standardized GES meal. Our study confirms that gastric emptying is affected by the amount of the meal ingested for GES [2]. The T1/2 was significantly shorter, and the percent retention at 0.5, 1, 2, 3 h postprandially were significantly less for half- and quarter-size meals compared to the full complete sandwich meal. Translating this to patients undergoing GES, patients ingesting less than the full meal may have a normal gastric emptying result, when they in fact may have delayed gastric emptying with a full meal. Thus, caution is needed in interpreting GES in patients ingesting less than the full sandwich meal and the gastric emptying result is normal when using the normal values established for the full meal.
Gastric emptying depends on the meal’s total content (caloric and volume) and the amounts of fat, protein, and carbohydrate [2, 8]. Older studies suggested gastric emptying is regulated to obtain constant energy delivery into the small intestine [2]. Using the gastric emptying breath test, gastric emptying time increased when meal size was increased from 1 to 2 MJ to 3 MJ [9]. Using the muffin-based [13C] octanoate breath test, T1/2 for the 350-kcal meal was significantly longer than for a 250-kcal meal [10]. Thus, increasing meal size prolongs gastric emptying. On the other hand, decreasing meal size speeds gastric emptying. The average half-emptying times for solid food, with 1692-, 900-, and 300-g meals, were 277, 146, and 77 min, respectively [8].
Currently, for proper interpretation of GES, patients need to ingest the entire radiolabeled meal, for comparison to the established normal values based the entire meal [5]. Normal results for gastric emptying based on the Tougas meal are < 60% retention at 2 h and < 10% retention at 4 h [5]. However, it is not unusual for patients with symptoms of nausea and early satiety, to not be able to eat the entire meal. In a 5 year study of 798 patients undergoing GES, 259 patients (32%) ingested less than 100% of the standard meal but greater than or equal to 50% of the meal and only 30 (4%) ingested less than 50% [3]. According to current clinical practice guidelines, studies with less than 100% meal ingestion would be deemed non-diagnostic and non-interpretable [4], potentially adversely affecting diagnosis and treatment management decisions. Using cluster analysis, Shah et al. concluded that Tougas meal normative standards for gastric emptying at 4 h could be applied for differentiating normal and delayed populations if at least 50% of the standard meal was ingested [4]. Our results support that there appears to be a minimum threshold requirement of 50% meal ingestion for GES result interpretation. That is, if a patient can ingest at least 50% of the standard GES meal and if > 41.6% retention is present at 2 h or T1/2 is > 105.2 min, the results are consistent with delayed GE.
Gastric accommodation can be evaluated by assessing the intragastric meal distribution [6]. Usually most of the meal is in the proximal stomach on the initial imaging of the gastric emptying scintigraphy test. Impaired gastric accommodation results in more distal intragastric meal distribution. Of note, in this study, the gastric accommodation response as assessed by IMD-0 was not significantly affected by meal size.
We recognize several technical limitations to our study. First, the number of normal subjects included is limited and our results ideally would need to be expanded by others to include a larger number of subjects. Our study enrolled 25 normal subjects, with three completed studies in 24 of the subjects. This was the appropriate number by our sample size calculation. The original Tougas study with the egg sandwich meal studied 123 normal subjects in a multicenter study [5]. Also, because of concerns to reduce radiation exposure to normal volunteers, we utilized a relatively low dose of Tc-99 m sulfur colloid (18.5 MBq (0.5 mCi)) for the full (100%) meal which was reduced further by one half and one quarter for the smaller meal sizes. This resulted in very low counts in the images for the reduced meal sizes particularly after 2 h (Fig. 1). The resulting low image count statistics for the residual activity in the stomach may explain why the mean % retention and T1/2 values for the 25% meals actually increased at 2 and 4 h compared to the 50% meal (Table 1). The upper limit of normal for 4 h retention were not significantly different for the different size 7.7% for the full meal, 5.7% for the half meal, and 8.6% for the quarter meal. While the radioactive counts available for the smaller meals becomes low at 4-hour suggestion is that similar to the full-size meal, using the conventional value of > 10% gastric retention at 4 h for smaller meals suggesting delayed gastric emptying [11].
Both the two hour retention and four hour retention are important to classify a patient for potential gastroparesis when the entire meal has been consumed [12, 13]. This also appears true for individuals consuming a minimum of 25% of the meal for GES. With the full meal, there can be discordance between the 2 h and 4 h gastric retention values in assessing for gastroparesis. When such discordance is present, the Society of Nuclear Medicine procedure guideline on GES recommends reporting that gastric emptying is delayed when either the 2 h or 4 h upper limit of normal value is exceeded [14]. Clinically, the 4 h gastric retention values are thought to be more sensitive than 2 h data. It is interesting that the 2 h retention, but not the 4 h retention, was different among the different meal sizes. As expected, the smaller meals show a reduction in the amount retained primarily at 2 h. Appropriate caution is recommended in the interpretation of GES studies with limited meal consumption, based primarily with the 2 h retention values. Our study, however, now provides revised cutoffs for determining delayed gastric emptying based on the 2 h retention values. Similar to the results reported by Shah et al., our results also support the finding that if the patient has ingested at least 50% of the total standard meal, a 4 h gastric retention of > 10% can still be used to support a diagnosis of delayed gastric emptying.
Overall, our study confirms that gastric emptying is affected by the amount of the meal ingested. The T1/2 and the percent retention at 0.5, 1, 2, 3 h postprandially were significantly less for half- and quarter-size meals compared to the full sandwich meal. Translating our results to clinical evaluation of patients being studied for possible delayed gastric emptying brings up two important points. First, patients ingesting less than the full meal may have an incorrect normal gastric emptying results if their results are compared to normative values for the full meal. Thus, caution is needed in interpreting GES in patients ingesting less than the full meal and the gastric emptying result shows no evidence for a delay in gastric emptying. Second, if the patient ingests 50% of the standard GES meal, gastric retention > 41.6% at 2 h and/or > 10% at 4 h would suggest delayed GE. From our study, we suggest continuing the gastric emptying test for patients that have consumed at least half the egg sandwich meal, recording in the GES report how much of the meal was ingested, so that if needed alternative normal values, as reported in this manuscript, can be used to interpret the gastric emptying test. This will facilitate the ability to reach a diagnosis of delayed gastric emptying for a significant number of patients with limited meal ingestion.