Gastric Emptying

The radionuclide gastric-emptying study is the accepted standard methodology used to measure gastric transit. It is physiological, quantitative, accurate, and reproducible. An upper gastrointestinal contrast study can detect a gross delay in gastric emptying, but it is not sensitive for the detection of less severe but symptomatic gastroparesis. Other methodologies have been proposed and investigated, most recently the wireless motility capsule; however, all have issues, and none has found routine clinical use. However, the radionuclide study cannot differentiate a severe functional delay from anatomical obstruction, such as a tumor or pyloric channel ulcer. Endoscopy or contrast barium radiography is required for that purpose.

Patients with delayed gastric emptying (gastroparesis) present clinically with symptoms of postprandial nausea, vomiting, and abdominal pain. Some of the clinical indications for the radionuclide gastric emptying study are listed ( Box 10.1 ). Rapid gastric emptying is less common than delayed. Patients may present with similar symptoms, although more commonly postprandial abdominal cramps, diarrhea, flushing, and tachycardia (dumping syndrome).

Anatomy and Physiology

The regions are designated from proximal to distal as cardia, fundus, body, antrum, and pylorus ( Fig. 10.2 ). Gastric mucosal glands secrete hydrochloric acid and digestive enzymes. Gastric motility is controlled by both neuromuscular gastric activity and small intestinal neuroendocrine feedback.

Gastric anatomy. The proximal stomach (fundus) undergoes receptive relaxation and accommodation of ingested food. Its tonic contraction is responsible for liquid emptying. The distal stomach (antrum) is responsible for solid emptying. It has phasic contractions that mix and grind food into small enough particles to pass through the pylorus.

The gastric fundus and antrum have distinct functions. The more proximal fundus acts as a reservoir, accepting large meals with only a minimal increase in pressure (receptive relaxation and accommodation). Fundal tonic contractions produce a constant pressure gradient between the stomach and pylorus, resulting in liquid emptying. The more distal antrum has phasic contractions initiated by a neural pacemaker. Muscular contractions sweep down the antrum in a ring-like pattern, squeezing food toward the pylorus. Larger food particles are not allowed to pass and are retropelled. The solid material is converted into chyme through contact with acid and peptic enzymes and mechanical grinding. Food particles must be broken down until they are small enough to pass through the pyloric sphincter (1–2 mm). The antrum is responsible for solid emptying. The pylorus, at the junction of the antrum and duodenal bulb, acts as a sieve, regulating gastric outflow ( Fig. 10.3 ).

Solid gastric emptying. Four selected images, from left to right, at time 0, 15 minutes, 30 minutes, and 60 minutes. Initially, the solid meal resides in the fundus; then it moves into the antrum, where it is ground up to a small enough size to pass through the pylorus into the small intestines.

The pattern of emptying for solid and liquid meals is different. Solids have a delay before emptying begins (lag phase) lasting 5 to 25 minutes. This is the time required to grind the food into small enough particles so that they can pass through the pylorus. Following the lag phase, solids empty in a relatively linear pattern ( Fig. 10.4 ). The rate of emptying depends on the size and contents of the meal. Meals with greater volume, weight, carbohydrates, protein, or fat empty slower ( Box 10.2 ). Liquids have no delay before emptying begins. Clear liquids empty in a monoexponential pattern ( Fig. 10.5A and B ). Full liquids and clear liquids ingested simultaneously with a solid meal empty in a slower multiexponential pattern.

Biphasic pattern of solid emptying. (A) Sequential solid gastric-emptying images over initial 60 minutes. The ingested meal moves normally from the gastric fundus to the antrum and then begins to clear into the small intestines. (B) Time–activity curve for the patient generated by a region of interest (ROI) around the stomach shows a biphasic pattern of solid gastric emptying. An initial delay (lag phase) of 9 minutes is seen before emptying begins. Emptying clears the stomach in a linear pattern for the remainder of the 60-minute study.

(A) Normal liquid emptying. Sequential 1-minute images of a 30-minute study after ingestion of 300 mL water with 100 μCi Tc-99m diethylenetriaminepentaacetic acid (DTPA). Distribution is uniform in the stomach throughout the study. Hot spot in lower portion of the stomach in second row of images is due to overlap with small bowel activity. (B) Time–activity curve was generated by drawing a whole-stomach region of interest (ROI). Emptying begins immediately. The clearance pattern is monoexponential, with a half-emptying time of 15 minutes (normal <25).

Gastric Stasis Syndromes

The majority of patients with chronic gastroparesis have a functional cause; that is, there is no known pathological etiology. An exception is diabetic gastroenteropathy, which occurs in patients with long-standing insulin-dependent diabetes. Here gastroparesis is caused by vagal nerve damage as part of a generalized autonomic neuropathy. In addition to producing disagreeable postprandial symptoms, the delay in emptying may exacerbate the problem of diabetic glucose control because the timing of the insulin dose with food ingestion and absorption may be unpredictable. A common gastric disorder is nonulcer dyspepsia, characterized by ulcer-like or dyspeptic symptoms. It is reported that gastroparesis occurs in 20% to 40% of these patients.

Most patients with gastroparesis have a chronic problem with delayed emptying. However, some have acute and potentially reversible causes (e.g., due to viral gastroenteritis, trauma, metabolic derangements). Common causes of and associations with chronic and acute gastric stasis syndromes are listed in Box 10.3 . Rapid gastric emptying is most commonly seen in patients who have had gastric surgery (e.g., pyloroplasty or gastrectomy), in patients with hyperthyroidism, or in those with a gastrinoma producing large amounts of the hormone gastrin (Zollinger–Ellison syndrome). A minority of patients with insulin-dependent diabetes have rapid gastric emptying ( Box 10.4 ), with symptoms of palpitations, diaphoresis, weakness, and diarrhea.

Patient Preparation

Hyperglycemia per se, independent of diabetic gastroenteropathy, may cause a delay in gastric emptying. Thus, gastric-emptying studies should be performed when patients are under good diabetic control, with a fasting blood preferably <250 mg/dL. Also, many commonly used nongastric therapeutic drugs can delay gastric emptying ( Table 10.1 ). In our clinic, we allow the referring physician to decide whether the patient should continue to take his or her medications before the study. This will depend on the study indication (e.g., whether to make the diagnosis of gastroparesis or to determine whether a specific therapeutic drug has been effective). When indicated, the drugs should be stopped 48 to 72 hours before the study. Medications taken should be considered when interpreting the study.

Therapy of Gastroparesis

Metoclopramide (Reglan) is the most common therapeutic drug prescribed for delayed gastric emptying. It is not effective in all patients, symptom relief is not always accompanied by an improvement in gastric emptying, and serious side effects may occur in some patients (e.g., tardive dyskinesia seen in 10% of patients taking the drug for longer than 3 months). Domperidone (Motilium), used in Europe, is not approved for clinical use in the United States. Cisapride (Propulsid) was removed from the U.S. market due to serious arrhythmias. Erythromycin, a motilin agonist, improves gastric motility; however, it has a high incidence of nausea and vomiting. Analogs of erythromycin as well as other potential therapeutic drugs are in development. Surgical intervention is sometimes used to treat refractory gastroparesis. One approach is implantation of an electrical gastric stimulator. In patients who are not responsive to standard therapies, gastric resection may be indicated.

Radiopharmaceuticals

For accurate and reproducible quantification of solid gastric emptying, the radionuclide must be tightly bound to the solid component of the meal. Elution of the radiolabel results in a part-solid, part-liquid labeled mixture and may be quantitatively unreliable. Radiolabeled eggs are most commonly used for solid gastric-emptying studies. Tc-99m sulfur colloid (Tc-99m SC) binds to albumen in egg white during cooking. Although whole eggs work well for gastric emptying, egg whites have higher percent binding. Tc-99m SC remains stable in the acidic stomach and is not absorbed by the gastrointestinal (GI) tract.

For liquid meals, the radiotracer must equilibrate rapidly within the liquid and be nonabsorbable in the gastrointestinal tract. Tc-99m- or In-111-labeled diethylenetriaminepentaacetic acid (DTPA) and Tc-99m SC meet these criteria. Dual-phase solid-liquid studies use one radiotracer for the solid meal and another for the liquid phase, for example, In-111 DTPA as the liquid marker (171, 247 keV) and Tc-99m SC (140 keV) as the solid marker ( Fig. 10.6 ). They are differentiated by their separate photopeaks.

Dual-isotope, dual-phase solid–liquid emptying. Simultaneously ingested solid and liquid meals. The solid egg meal is labeled with Tc-99m sulfur colloid (SC), and the liquid (water) meal is mixed with In-111 diethylenetriaminepentaacetic acid (DTPA). One-minute images were acquired. Time–activity curves were generated by drawing regions of interest (ROIs) around the stomach. The solid and liquid meals are distinguished by their different photopeaks. This study shows very delayed solid emptying but normal liquid emptying.

Standardization of Solid Gastric-Emptying Scintigraphy

Various methodologies have been used over the years for solid gastric-emptying studies, including different meals, patient positioning, instrumentation, framing rate, and study length as well as quantitative methods. All of these factors may affect normal values. Thus, normal values must be well validated for the specific meal and methodology used. In the past, gastroenterologists have expressed concern about the different methodologies and different normal values. They advocated for a standardized gastric emptying protocol.

Standardized Solid Gastric-Emptying Study

In 2008, an expert panel of gastroenterologists and nuclear medicine physicians published Consensus Recommendations for Radionuclide Gastric Emptying. The recommendations were based on a protocol published by Tougas and colleagues in 2000 and are described in detail in Box 10.5 . The protocol was selected because it simplified the procedure and because the normal values were considered valid based on the large number of normal subjects studied (123). The meal consists of an egg-substitute sandwich (4 oz egg white, equivalent to 2 whole eggs, 2 slices of bread), strawberry jam (30 g), and water (120 mL). The study requires 1-minute anterior and posterior images acquired at four time points (immediately after meal ingestion and at 1, 2, and 4 hours) and calculation of the percent gastric retention at each time point ( Figs. 10.7–10.9 ).

Box 10.5

Standardized Gastric-Emptying Scintigraphy: Summary Protocol

Patient Preparation

• 
  

  The referring physician should decide which medications are to be continued/discontinued before the study.

  
  
• 
  

  For diabetics, a fasting blood sugar (FBS) is recommended before the study. The study should not be performed if the FBS is >250 mg/dL.

  
  
• 
  

  Begin the study in the morning after an overnight fast. The patient should be NPO for at least 6 hours before study.

Radiopharmaceutical and Meal

• 
  

  Tc-99m SC 1 mCi (37 MBq) bound to 4 oz egg substitute (Egg-Beaters or generic egg-white equivalent) by microwaving or scrambling after mixing egg with Tc-99m SC. Stir once or twice while cooking until mixture reaches consistency of omelet. Meal also consists of two slices of white bread, strawberry jam (30 g), and water (120 mL).

  
  
• 
  

  Meal should be ingested within 10 to 15 minutes.

Instrumentation

• 
  

  Gamma camera: Large-field-of-view, dual-detector camera

  
  
• 
  

  Collimator: Low energy, parallel hole, high resolution or general purpose

  
  
• 
  

  Tc-99m photopeak with a 20% window around 140 keV photopeak

  
  
• 
  

  Computer setup: 1-minute frames (128 × 128 word mode matrix)

Patient Position

• 
  

  Position patient either standing upright or lying supine, with camera heads anterior and posterior. A single-head camera can be used, acquiring an image first anteriorly, then posteriorly.

Imaging Procedure

• 
  

  Acquire 1-minute frames at time 0 (immediately after ingestion) and 1, 2, and 4 hours; 3-hour images are optional.

Processing

• 
  

  Draw region of interest for anterior and posterior stomach views.

  
  
• 
  

  The geometrical mean of the anterior and posterior views is determined at each time point. The percent retention or gastric emptying is calculated at each imaging time point.

  
  
• 
  

  Decay correction is mandatory.

Interpretation

• 
  

  Delayed gastric emptying:

  
  
  
  
  • 
    

    1 hour <10% emptying (>90% retention)

    
    
  • 
    

    2 hours <40% emptying (>60% retention)

    
    
  • 
    

    4 hours <90% emptying (>10% retention)

  
  
  
  
• 
  

  Rapid gastric emptying:

  
  
  
  
  • 
    

    >70% emptying at 1 hour

    
    
  • 
    

    >90% emptying at 2 hours

    
    
  • 
    

    These values apply to the entire meal and are not valid for different meals or incomplete ingestion of the standard meal.

  
  

NPO, Nil per os; SC, sulfur colloid.

Normal solid gastric emptying using the standardized Tougas and colleagues protocol. (Left) Anterior (above) and posterior (lower) images acquired simultaneously at time 0, 1, 2, 3, and 4 hours. Three-hour images are not required. Regions of interest are drawn around the stomach. (Right) Time–activity curve and results (below). The percent geometrical-mean-corrected gastric emptying result at 2 hours (112 minutes) is 44% (normal >40%), and at 4 hours (233 minutes), it is 90% (normal >90%).

Delayed solid gastric emptying, using the standardized protocol. At 2 hours (119 minutes), the percent gastric emptying is 22% (normal >40%), and it is 60% at 4 hours (240 minutes; normal >90%).

Rapid solid gastric emptying. At 1 hour (67 minutes), solid gastric emptying is 88% (abnormal >70%). We also consider >90% emptying at 2 hours to be rapid. The initial (time 0) region of interest (ROI) includes abdominal activity. Because of the rapid emptying, there was transit to the small bowel by the first image after ingestion. Thus, for accurate calculation of the total ingested radiolabeled meal, the large ROI is required for the initial image.

The Tougas and colleagues publication found that infrequent imaging (each hour) was as accurate as more frequent imaging (e.g., every 10 minutes). The rationale for a 4-hour study length was based on several publications reporting a higher rate of detection of gastroparesis at 4 hours compared with 2 hours. One investigation reported that by extending the study from 2 to 4 hours, the number of patients diagnosed with gastroparesis increased by 30%. Widespread use of this protocol ensures that results are comparable between institutions. Although a 4-hour study length might at first seem demanding of clinic logistics, the standardized protocol can actually improve patient flow. Multiple patients can be studied on one camera in a single morning because total imaging time per patient is short.

Liquid Gastric Emptying

Accepted teaching had long been that liquid gastric-emptying studies were less sensitive for the detection of gastroparesis than solid studies, that liquids become abnormal only in late stages of gastroparesis, and that only a solid gastric-emptying study is necessary for clinical purposes. However, no data had been published to support this, and it has recently proven to be incorrect.

Two publications of 140 patients have compared clear-liquid (water)-only emptying with the standardized solid gastric-emptying meal in patients referred for suspected gastroparesis. More patients had abnormal liquid than solid emptying. The most important finding was that 25% of patients who had a normal solid study had delayed liquid emptying. The pathophysiologic explanation for these findings is unclear; however, it may help explain the patient’s symptoms. Many of our patients are referred for both studies.

The methodology for clear-liquid-only studies is detailed in Box 10.6 . Because water empties rapidly, only a 30-minute study is required (1-minute framing rate). Imaging begins immediately after ingestion of the 300 mL water containing 1 mCi Tc-99m DTPA or Tc-99m SC. Alternatively, the liquid and solid studies can be performed sequentially on the same day, the clear liquid first with In-111 DTPA, followed by the solid study labeled with Tc-99m SC. Simultaneous evaluation of liquid and solid gastric emptying is possible and done at some institutions using a dual-isotope acquisition technique ( Fig. 10.10 ). Liquids radiolabeled with In-111 and ingested simultaneously with the standardized Tc-99m-labeled solid meal empty slower than liquid-only meals. The results of liquid-only and simultaneously ingested liquids and solids correlate poorly and likely provide different pathophysiologic information, although at present the reason is uncertain. However, if abnormal, either may explain the patient’s symptoms.

Geometrical mean correction for attenuation. The anterior detector time–activity curve (TAC) shows a rising curve before emptying begins; however, all the food is in the stomach at time 0. The posterior detector curve shows decreasing counts from time 0. The geometrical-mean-corrected curve has a normal pattern with an initial lag phase, then linear emptying.

Quantification of Gastric Emptying

Various methods have been used over the years to quantify gastric-emptying studies. However, investigations have found no clinical advantage to more frequent imaging, other methods of quantification, or calculation of the lag phase. Percent emptying is more accurate than a rate of emptying (T _½ ) with the standardized methodology because there are only 4 to 5 data points. For the standardized solid gastric-emptying study, a gastric region of interest (ROI) is drawn for all anterior and posterior views at each time point (see Figs. 10.7–10.9 ). Decay and attenuation correction is required for accurate quantification. The percent emptying is calculated at each time point.

Percent emptying = initial meal counts minus meal counts at 1, 2, and 4 hours, each divided by initial meal counts, and all corrected for attenuation and decay.

Attenuation Correction

Attenuation on gastric-emptying studies varies as the ingested meal transits the stomach because stomach contents move from the relatively posterior fundus to the more anterior antrum. With a single gamma camera detector positioned anteriorly, the counts increase during the early portion of the study in spite of the fact that all the food is already in the stomach from time zero. This results in a quantitative error. Attenuation correction is mandatory for correct results (see Fig. 10.10 ). Otherwise, there may be an underestimation of emptying, by as much as 10% to 30%. This is most pronounced in obese patients; however, it also occurs in nonobese patients and is not predictable between patients. The commonly used method for attenuation correction is mathematical and involves calculating the geometrical mean (square result of the product of the anterior and posterior counts) at each time point.

The left anterior oblique projection acquisition method is an alternative approach that minimizes the effect of attenuation. The study is acquired in this one projection. It compensates for attenuation because the head of the gamma camera is roughly parallel to the movement of the stomach contents, and thus the effect of attenuation is minimized ( Fig. 10.11 ). The advantages of this method are that it requires only a single-headed camera, and no mathematical correction is needed. Results correlate well with those of the geometrical mean method, although the geometrical mean is more accurate and preferable whenever possible ( Fig. 10.12 ).

Left anterior oblique (LAO) acquisition—diagram of transverse view. When the camera is placed in the LAO projection, rather than the anterior view, the stomach contents move roughly parallel to the head of the gamma camera, compensating for the effect of varying attenuation. No mathematical correction for attenuation is required. ANT, Anterior.

Comparison of lag phase (LAG) and percent emptying (%GE) in two patients (A [left] and B [right] ) using anterior (ANT)-only camera head, geometrical mean (GM), and left anterior oblique (LAO) attenuation correction, from top to bottom. (A) ANT-acquisition time–activity curve (TAC) shows a modest rise in counts over 15 minutes before declining. GM correction of anterior and posterior views (middle) results in a shortened lag phase of 8 minutes and greater %GE than the ANT view. LAO uncorrected method (bottom) results are very similar to GM method. (B) Different patient. ANT acquisition shows very long lag phase of 60 minutes and slowly rising TAC. GM correction shortens the lag phase considerably, although there is still some initial rise, suggesting incomplete correction (middle). The %GE is improved greatly. The LAO results are similar, although not identical, to the GM.

Correction for radioactive decay is mandatory for solid-meal studies because of the short half-life of Tc-99m (6 hours) and the relatively long duration of the study (4 hours). Down-scatter (In-111 into the Tc-99m window) and up-scatter (Tc-99m into the In-111 171 keV window) have been shown to be inconsequential if the administered dose ratio of Tc-99m to In-111 is at least 4:1. Therefore, with standard doses of 1 mCi Tc-99m and 100 to 200 μCi In-111, no correction is necessary.

Liquid Gastric Emptying Quantification

Attenuation correction is not necessary for liquid gastric emptying because of its rapid transit. A gastric ROI is drawn, and a half-time of emptying (time in minutes when counts become half of peak counts) and/or an exponential mathematical fit T _½ is calculated ( Figs. 10.13 and 10.14 ). If small bowel activity moves into the ROI, it should be excluded to correct for potential artificial delay. Normal values for a clear-liquid (water) study alone have been well validated, with a half-emptying time of <25 minutes. Adding salt, sugar, and so forth to the clear liquid meal slows the rate of emptying.

Normal liquid-only (water) gastric emptying. Only the first 8 of 30 1-minute images are shown. A region of interest (ROI) was drawn for the stomach, and a time–activity curve was generated. The half-time of emptying was 12.84 min (normal <25 minutes). The exponential T _½ fit (yellow line) was 8.17 minutes. Expon, Exponential

Delayed liquid-only (water) gastric emptying. Only the first 8 of 30 1-minute images are shown. T _½ of >40 minutes is delayed emptying (normal <25). Expon, Exponential.