Applied Radiology

Dr. Rehm is an Associate Professor of Radiology and the Director of Nuclear Medicine at the University of Virginia Health System, Charlottesville, VA. She is also a member of the Editorial Board of this journal.

Nuclear medicine evaluation of gastric motility in adults is an extremely useful but often underutilized procedure. Underutilization undoubtedly results from a number of factors and misunderstandings. The purpose of this article is to change attitudes and make gastric emptying studies user-friendly for patients and radiologists. The images, lacking the anatomic detail so appealing to radiologists, could be considered relatively "ugly" and featureless even when compared with other scintigraphic examinations, such as bone or hepatobiliary scans. Perhaps equally unappealing to radiologists and patients is the idea of a breakfast meal of radioactive liver eaten after an overnight fast. It is possible that the number of imaging protocols and variety of radiolabeled meals appear challenging. The different protocols have different "normal values" that are dependent on the scan technique and meal, which can make it difficult to compare results between institutions. In the case of gastric emptying studies, too many options can be a disadvantage. ^1-5

Evaluation of gastric emptying becomes much simpler when one looks at it as a diagnostic study that is clinically useful for diagnosis and for monitoring the effects of therapy. The goal is to provide an accurate and reproducible quantification of gastric motility. The examination is noninvasive, accepted by patients, and performed with equipment that is widely available. From this standpoint, much of the variation in methodology and all the esoteric processing is relegated to the realm of research. A gamma camera with a computer and a radiolabeled meal are all that is required. From the scintigraphic images, a time-activity curve is generated that displays the amount of radioisotope in the stomach (defined by a region of interest [ROI]) over time (Figure 1).

A variety of techniques have been used to evaluate gastric emptying, but none fills the niche occupied by scintigraphy. Mechanical obstruction should be excluded by contrast radiography or esophagogastroduodenoscopy first. However, these procedures are relatively insensitive for detection of gastric dysmotility unless it is severe. Radiographic studies are associated with significantly greater radiation than the nuclear examination, and do not allow for quantification. The radiation dose to the patient for a gastric emptying study varies with the specific radionuclide, the dose administered, and bowel transit-time, but it is small. Whole body exposure from a 1.0 mCi (37.0 MBq) Tc-99m sulfur colloid solid meal is approximately equal to 22 days of environmental background radiation in Charlottesville, VA.

The procedure is useful to establish the presence of dysmotility but does not determine an etiology. It is the appropriate study when chronic dysmotility is suspected (Table 1). The clinical diagnosis is difficult because the patient's presenting clinical signs and symptoms are often vague and may overlap with other conditions. Most patients referred for the examination have diabetes mellitus and are being evaluated for gastroparesis diabeticorum, with autonomic neuropathy as the suspected etiology. However, there are numerous, less common etiologies of gastroparesis (Table 2). The study also can be performed to evaluate for rapid gastric emptying, which most commonly occurs after vagotomy. Rapid emptying may be associated with other conditions, and occasionally with diabetes (Table 3). ^1-4 After the diagnosis of dysmotility is established, the examination is also useful to evaluate the effectiveness of medical therapy. Scintigraphic gastric emptying studies are not usually employed in the evaluation of acute gastric obstruction for which other diagnostic studies are more appropriate.

Anatomy and physiology

For purposes of gastric emptying, one can view gastric anatomy and physiology in the simplest fashion. The proximal stomach is represented by the cardia and gastric body and serves as a reservoir. Following a meal, neural refluxes allow it to distend to contain the ingested volume without a significant increase in intra-gastric pressure. When this neural reflex pathway is disrupted (for example, after vagotomy), the ingested meal results in increased intragastric pressure. The dumping syndrome, characterized by sweating, tachycardia, dizziness, and weakness, may result. The distal stomach, composed of the antrum, serves as the grinder of the ingested contents through cycles of rhythmic contractions at a rate of 3 per minute initiated by a gastric pacemaker located on the greater curvature at the junction of the proximal and distal stomach. The pylorus controls outflow to particles <1 mm. Solids begin to empty after a short period during which no emptying occurs, referred to as the lag phase. During the lag phase, the contents of the stomach are broken down to small particles that can pass through the pylorus. The duration of the lag phase may vary in length (Figure 2). Compared with the overall results of the gastric emptying study, no clinical significance has been determined for the lag phase alone. ⁴ Between meals, nondigestible materials (for example, a swallowed foreign body) are cleared by strong antral contractions.

Factors affecting gastric emptying

The rate of gastric emptying is dependent on what is consumed. Liquid and solid contents empty from the stomach at different rates. Non-nutrient liquids, such as water, empty in an exponential fashion in 10 to 20 minutes. Nutrient liquids empty more slowly. Liquids ingested in conjunction with solids empty more slowly than liquids ingested alone. All solids empty in a linear fashion. However, the emptying rate is influenced by the contents of the meal, such as the amount of carbohydrate, fat, and protein as well as the total caloric content and volume of the meal. Patient position and patient activity also affect the rate of gastric emptying. Emptying is faster when the patient is sitting than when lying supine. Emptying also increases if the patient is allowed to ambulate during breaks in the scanning procedure. ^1-4 Therefore, it is important that the entire gastric emptying procedure be standardized, not just the radiolabeled meal.

Radiolabeled meal

Measurement of gastric emptying using a liquid phase marker is appropriate for infants, but is a topic beyond the scope of this article. An examination using a solid test meal is appropriate for adults. Although simultaneous measurement of liquid and solid emptying can be performed in a single, dual-isotope gastric emptying study, in the vast majority of cases assessment of solid emptying alone is adequate and preferred. Gastric emptying evaluation using a solid meal is more sensitive for detecting abnormal motility than a liquid meal. In normal patients, both liquid emptying and solid emptying will be normal. In patients with gastroparesis, liquid emptying may be delayed or normal. Because the test's sensitivity for detecting abnormalities lies in the assessment of solid emptying, simultaneous liquid and solid gastric emptying studies (using In-111 DTPA and Tc-99m sulfur colloid, respectively, as labels) are rarely necessary in clinical practice. A number of authors have recommended that each clinic validate its own procedure and establish its own standard values. However, a more practical approach is to rigorously adopt a validated procedure from an academic center and "do it like the experts."

The only things required for the study are a gamma camera with computer capability and a radionuclide-labeled meal. It is critical that the radionuclide remains attached to the solid component of the meal to reflect the physiology of solid emptying accurately. If the radionuclide dissociates from the solid, it will leave the stomach in the liquid phase; the resulting scan and calculations will result in misinformation and an overestimation of gastric emptying. A variety of food substances of varying palatability and availability (oatmeal, fiber, etc) have been labeled with a number of radiopharmaceuticals. However, Tc-99m sulfur colloid labeling of chicken liver or egg has been found superior. The egg sandwich and the chicken-liver stew meals both contain fat, carbohydrate, and protein and mimic a normal meal. However, there is no consensus as to the optimal meal. ^1-4

Both the Tc-99m sulfur colloid chicken-liver stew meal and the egg sandwich meal are relatively easy to prepare. Historically, in vivo labeling of the chicken liver was achieved by injecting Tc-99m sulfur colloid intravenously into a chicken that was subsequently slaughtered. The in-vivo labeled chicken liver then was cooked and incorporated in a can of prepared beef stew. Fortunately, it was subsequently shown that in-vivo labeling of the liver was not necessary. Further investigations proved that in-vitro labeling of the liver or in-vitro labeling of eggs or egg whites performed as well as in-vivo labeling of liver, with no significant elution of the Tc-99m sulfur colloid from the solid component. Because of the ready availability of eggs and the general patient preference of eggs over liver, a labeled egg meal is widely used (Tables 4 and 5). A standard meal is two large eggs or 3 egg whites beaten and mixed with 0.5 to 1.0 mCi (18.5 to 37.0 MBq) of Tc-99m sulfur colloid. The eggs or egg whites are cooked in omelet fashion in a nonstick frying pan and placed between two slices of standard white bread. Using an egg meal provides similar information regarding gastric emptying as using a liver meal. However, purists will note that an egg or egg-white meal is correctly considered a semisolid (rather than solid) meal. Because the egg meal has a shorter emptying time than a stew meal, a shorter examination is possible, which is an advantage in a busy clinic. Because the meal must consist of an adequate volume to reflect a typical meal, the egg is typically administered as an egg sandwich with fluid to help with ingestion. Patients are occasionally allergic to eggs, which may require utilization of a liver meal. It is important that appropriate normal values be applied for each standardized meal.

The patient should eat the entire meal within 10 minutes. If a longer period is allowed, some gastric emptying may occur before the patient has finished the meal, and before the imaging and data acquisition has begun. Any standardized technique must take into account the type of meal, patient position during imaging, specifics of the scan acquisition, and methods of quantitation.

Imaging technique

The scanning period should be at least as long as the time needed for half of the meal to empty, based on an average normal subject. A meal of stew with liver empties more slowly than eggs, so a 120-minute scan is recommended. A 90-minute study is appropriate for an egg-sandwich meal.

Most people eat in the sitting position and remain in the sitting or standing position following a meal. Patient positioning in the semi-upright or upright position mimics the normal physiologic state. Therefore, imaging in these positions would seem preferable. However, gamma camera logistics may determine patient positioning and scan protocol. The supine position can be used, and may be necessary for some dual-headed cameras to utilize the geometric mean technique. Patient comfort is always critical and at times may be the limiting factor for positioning.

More frequent image acquisition will result in more data points and a more representative time-activity curve. Modern nuclear medicine camera-computer systems allow frequent images to be obtained at regular intervals with minimal technologist effort. However, some clinics employ infrequent static images, rather than continuous dynamic imaging as the basis for generation of the TAC.

Quantitative analysis: Corrections for background, decay and attenuation

On the computer, an ROI is drawn around the stomach using either an individual image or a summed image. Counts in the ROIs must be corrected for radioactive decay to allow comparison with the start of the study. If decay correction is not done, gastric emptying will be overestimated. For example, assume a patient has no gastric emptying. Decay alone will result in a decrease in counts in the gastric ROI, indicating apparent emptying of 16% at 90 minutes and 21% at 120 minutes if decay correction is not performed.

Attenuation due to tissue between the camera and the stomach is also a factor affecting quantitation. From the perspective of a gamma camera placed anterior to the patient, as the radionuclide moves from the more posterior gastric fundus to the more anterior gastric body and antrum, the thickness of tissues between the radioisotope and the camera gradually decreases. Therefore, the attenuation of photons arising from the labeled meal decreases . Due to this geometry, the gamma camera will record counts from the gastric ROI that increase as the meal moves from the fundus to the antrum. Imaging in the anterior or posterior view only can result in errors up to 35% in determination of emptying if attenuation is not taken into account. ⁴

Direct measurement of attenuation is the most accurate method to correct for attenuation, as is done using a transmission scan in positron emission tomography (PET). However, that complex technique is impractical and unnecessary for gastric emptying studies. An alternative method to compensate for attenuation incorporates a mathematical correction known as the geometric mean method (Figure 3). ⁶ This technique requires two opposed views of the area of interest, typically anterior and posterior, closely in time. Both views should be obtained nearly simultaneously if a single-headed camera is used, or can be obtained simultaneously if a dual-headed camera is used. The geometric mean value for the gastric ROI counts is the square root of the product of the counts in the stomach ROI on the anterior and on the posterior views [geometric mean = (anterior * posterior) ^1/2 ]. The geometric mean technique compensates for the variable attenuation of the labeled meal during movement from the fundus to the pylorus. The geometric mean value for gastric counts is decay-corrected and then recorded on the TAC for each time point.

An alternative method that allows for attenuation compensation employs the left anterior oblique (LAO) method of acquisition. ^7,8 It is particularly valuable when a dual-headed camera is not available, and is well suited to using a single-headed camera. It eliminates the need for the technologist to switch a single camera head back and forth between the anterior and posterior positions at each time point, as needed for the geometric mean method. By positioning a single camera head roughly parallel to the axis of the stomach, the effect of attenuation is relatively constant and there is no need for a mathematical correction for attenuation. The LAO value for gastric counts is decay-corrected and then recorded on the TAC for each time point. In the LAO projection, there may be superimposition of the antrum on the first portion of the duodenum. Nevertheless, the method is adequate for clinical purposes, is technologist friendly, and is more accurate than using the anterior or posterior view only (Figures 4 and 5).

Patient movement during the scan should be discouraged. Occasionally, patient motion can be a problem and lead to spurious results. Basic gastric emptying software supplied with the gamma camera commonly applies the same ROI to every image in a dynamic scan. If there is significant patient motion, the stomach will not remain within the ROI and the emptying determination will be incorrect. Patient motion is easily identified if the images are viewed in cinematic mode as a movie. Remedies include motion correction, which is available on some cameras, or manual placement of the ROI around the stomach on each frame to adjust for the change in patient position during the scan.

Reporting

Reporting can be done using a variety of parameters that include the percent emptying at the end of the study, the emptying rate (%/minute) or the half time of emptying (T1/2). Because reporting a T1/2 value implies an exponential rate of emptying, that parameter is not truly appropriate, although it is often used (incorrectly). Either the percent emptying at the end of the study or the emptying rate are more appropriate values to report. Normal values must be based on the meal used as well as the specific imaging and processing protocol. In the vast majority of clinics, adoption of a standardized protocol that has been documented in the literature and application of its normal values is preferable to establishing values for the normal range on-site.

Conclusion

Gastric emptying studies using either Tm-99m sulfur-colloid­labeled egg or stew meals can provide accurate and reproducible quantification of gastric motility. This clinically useful test can be performed using a single- or dual-headed gamma camera equipped with a computer, but the meal and scan procedure should be standardized. Factors critical for accurate determination include decay correction and compensation for attenuation, using either the left anterior oblique view or the geometric mean technique. In some cases, correction for patient motion is necessary. Quantitative determination of normal and abnormal emptying values must be appropriate to the labeled meal used as well as the specific imaging and processing protocol.