Radionuclide Gastroesophageal Motor Studies*

Oropharyngeal Motor Function

The normal esophagus propels a bolus of solid or liquid from the mouth to the stomach. Liquids require approximately 2 s to traverse the esophagus, whereas solids can take about 9 s. The process begins with swallowing. Of the 6 phases that constitute swallowing, only the first phase is totally voluntary: positioning of the bolus on the tongue’s central groove and propulsion of the bolus toward and through the faucial arches (oral phase) (7). The subsequent phases are involuntary and are initiated by elevation and retraction of soft palate, with closure of the nasopharynx associated with simultaneous opening of the posterior mouth sphincter (oropharyngeal phase). The next 2 steps (proximal and distal pharyngeal phases) take the bolus into the hypopharynx (8). The most important and delicate task of these 2 phases is avoiding passage of food through the larynx into the respiratory tree (9). Aspiration is avoided through a complex, multistep process involving a reflex mechanism consisting of sensitive afferences and articulated motor efferences (10). The pharyngoesophageal phase is marked by progression of the bolus through the striated cricopharyngeus muscle, or upper esophageal sphincter, whose function is to avoid reflux of the esophageal content back into the pharynx (11,12). Although the first phases of swallowing are concentrated in a relatively short space immediately posterior and inferior to the mouth, their physiologic coordination requires complex interaction of muscular and nervous structures. Coordinating this process requires participation of the IV, V, IX, X, and XII cranial nerves and of sympathetic fibers originating mostly at T4–T6 (Fig. 1). Oropharyngeal dysphagia can be caused by degenerative disorders of the central nervous system (Parkinson’s disease, multiple sclerosis, and supranuclear palsy), of the spinal motor neuron (amyotrophic lateral sclerosis), of the neuromuscular junction (myasthenia gravis), or of striated muscle per se (polymyositis and muscular dystrophy) (13,14).

Esophageal Motor Function

The last step of swallowing (the esophageal phase) involves the traveling of a much longer distance by the bolus yet involves a much simpler peristaltic mechanism aided by gravity. The primary peristaltic pump originating in the pharynx propagates through the inner (circular) and the outer (longitudinal) muscular layers of the esophagus at about 4 cm/s. Coordination of the primary peristaltic pump with postswallowing relaxation of the lower esophageal sphincter (LES) lets the bolus enter the stomach (15). Disorders of peristalsis can be caused by derangement of motor innervation provided by the vagus nerve for the striated muscle fibers, derangement of the sympathetic and parasympathetic innervation provided for the smooth muscle, and derangement of the myenteric plexus (Table 1) (16).

Functional esophageal disorders are a heterogeneous group of chronic disturbances often complicated by association with psychiatric or psychologic disorders, as determined on the basis of a lack of identifiable structural or metabolic damage (Table 2). After irritable bowel syndrome, esophageal disorders are the second most frequent functional gastrointestinal disorder and are usually not associated with GERD or other structural, metabolic, or motor disorders (17,18).

Achalasia is a primary esophageal motility disorder of unknown cause. The disease results in partial or absence of relaxation of the LES and loss of esophageal-body peristalsis. It is characterized by progressive neuronal degeneration in the myenteric plexus, with a paucity of ganglion cells, the presence of neural fibrosis, and some degree of chronic inflammation at histology. There is also a significant decrease in the synthesis of nitric oxide, a mediator of relaxation in the LES. Dysphagia and varying degrees of regurgitation, weight loss, and chest pain are the most common clinical presentation. Achalasia may present at any time, with the highest incidence occurring between 20 and 50 y of age. Most patients seen today in a surgical practice have had dilation or botulinum toxin injections before. Nevertheless, having undergone previous therapies should not be regarded as proof that a patient has achalasia. In fact, one can confuse the clinical manifestations of achalasia with those of other esophageal disorders; thus, patients must have, in addition to dysphagia and regurgitation, the radiologic and manometric findings consistent with achalasia.

GERD is common among the populations of industrialized countries (19). Heartburn, the most common symptom of gastroesophageal reflux, occurs daily in approximately 10% of the population and is the major reason for the consumption of antacids in our society. GERD can be overlooked for several reasons: First, the most common symptoms, heartburn and regurgitation, occur in only half of these patients; second, because of the natural history of the disease and the frequency of its spontaneous remission, many patients will not seek medical advice; and third, there is no diagnostic standard for GERD.

Symptoms of GERD are divided into 2 categories: esophageal, which includes typical esophageal reflux complaints (heartburn, regurgitation, and dysphagia) and noncardiac chest pain, and extraesophageal, which includes pulmonary (asthma and recurrent aspiration pneumonia), otolaryngologic (hoarseness, chronic dry cough, chronic sore throat, and globus sensation), and dyspeptic symptoms (upper abdominal pain, nausea, vomiting, and bloating).

The work-up of patients with gastroesophageal symptoms should include confirmation of the presence of pathologic gastroesophageal reflux and exclusion of other motility disorders or lesions of the esophagus and stomach, quantification of the severity of reflux, and clear definition of the anatomy of the esophagus and gastroesophageal junction.

Gastric Motility and Emptying

Upper Gastrointestinal Series

Upper gastrointestinal radiography with barium contrast medium allows examination of the esophagus, stomach, and duodenum, whereas special protocols have been developed to evaluate the oropharyngeal phase of swallowing. This is the procedure most patients who complain of dysphagia still undergo before any other testing. On the other hand, current diagnostic protocols call for esophagogastroduodenoscopy as a first-line approach for patients who complain of heartburn or regurgitation, chest pain, or epigastric pain or have unexplained vomiting or severe indigestion. This leaves a more limited role than in the past for an upper gastrointestinal series for these symptoms. Barium radiography is now considered a second-line diagnostic test whenever endoscopy reveals no obvious abnormalities explaining these symptoms.

Different barium suspensions (solid, semisolid, or liquid) can be used to characterize the parameters of swallowing. Although capable of identifying the major categories of dysfunction in patients with pharyngoesophageal dysphagia (42), the examination cannot estimate muscle fatigue, measure pharyngeal contractile forces, or estimate the intrabolus pressure during swallowing (12).

The esophagogram should include upright double-contrast views with a high-density barium suspension to assess mucosal disease, and prone single-contrast views with a low-density barium suspension to assess distensibility and motility. Barium swallows allow a dynamic evaluation of esophageal motility and transit of the barium boluses from the mouth to the stomach through the entire esophagus. Cine- and videoradiography (possibly recorded as digital videofluorography) help in evaluating functional disorders of the pharyngoesophageal and the esophageal phases of swallowing. Particular attention is paid to evaluation of the gastroesophageal junction and hiatus through varying of the patient position (e.g., oblique, standing, and supine).

An upper gastrointestinal series should always include evaluation of the stomach and duodenum, even if symptoms suggest a primarily esophageal disorder.

Ultrasonography

The oral phase of swallowing can be evaluated by positioning the ultrasonography probe under the chin to the hyoid region, using transverse and longitudinal scans to visualize the tongue and mouth floor both at rest and during swallowing of a bolus (43,44). This procedure can be used to identify various abnormalities, such as inability to keep the bolus in the mouth, lack of backward propulsion of the bolus, and asymmetric contraction of the tongue.

Although a simple ultrasonography examination can give useful information on visceral wall thickness, a functional evaluation is needed for patients with upper gastrointestinal disorders to gain data on esophageal and gastric motility. As with other diagnostic ultrasonographic applications, the test is highly dependent on the operator’s skill; in addition, the test is relatively time consuming because it requires repeated and prolonged observations (45,46).

The proximal (cervical) tract of the esophagus and its distal tract (gastroesophageal junction) can be explored in children. Ultrasonography can detect gastroesophageal reflux in children up to 5 y of age with 100% sensitivity and 87.5% specificity (47).

Even though air in the gastrointestinal tract and a thick abdominal wall can interfere with visualization of the fundic and antral regions, ultrasonography can evaluate gastric volume and emptying and transpyloric flow (45,48). A high correlation has been found between gastric-emptying values evaluated by ultrasonographic procedures and those derived by radionuclide-based procedures (49). In addition, ultrasonography can measure gastric area and volume and detects gastric contraction and distension (50).

Functional ultrasonography of the stomach is indicated mainly for evaluation of patients with dysmotility-like dyspepsia (45), evaluation of both dyspeptic and nondyspeptic patients with chronic disease potentially causing delayed gastric emptying (e.g., diabetes mellitus, systemic sclerosis, and myotonic dystrophy) (51,52), monitoring of the effects of pharmacologic and nonpharmacologic agents potentially affecting gastrointestinal motility (53), and evaluation of newborns with suspected hypertrophic pyloric stenosis (54).

Breath Test for Gastric Emptying

The use of breath tests in gastroenterology has become increasingly popular since stable isotopes such as ¹³C were introduced. Mass spectrometers are required to measure the concentration of these nuclides in expired air. These instruments are now widely available (at relatively low cost) and easy to operate. This technologic evolution has caused the β-emitter ¹⁴C to largely be replaced by the mass isotope ¹³C. This change avoids the need for extensive isotope inventories and record keeping, eliminates radiation exposure for both patients and personnel, and allows gastroenterologists to perform these procedures in their own environment.

Combined with different food substrates, octanoic acid (solid meal) or acetate (liquid meal) labeled with ¹³C are used to assess gastric emptying. The underlying concept is that these compounds pass unabsorbed through the stomach to the duodenum, where they are quickly absorbed. After absorption in the duodenum, portal circulation transports the ¹³C-labeled substrate to the liver, where fast metabolic degradation produces ¹³CO₂, which is excreted with exhaled air. Breath is therefore tested for enrichment with¹³CO₂ at regular intervals for up to 6 h, thus deriving the basic parameters of ¹³CO₂ appearance in the breath (beginning of gastric emptying) and time-related enrichment (a rising curve whose slope is related to the gastric-emptying rate) (55,56).

The test is safe and noninvasive, can be performed on children and during pregnancy, and can be repeated whenever necessary, as when monitoring the efficacy of therapy or assessing the effects of drugs on gastric emptying (56). Because the breath test is not an imaging technique, information is not provided about intragastric distribution of the different phases of the meal. Moreover, considering gastric emptying as the sole limiting step of the delivery of ¹³CO₂ to the breath can be misleading (thus reducing the reliability of the test) in certain conditions such as malabsorption; in diseases of the pancreas, liver, or lungs; or in the presence of visceral hemodynamic changes (e.g., physical exercise) (46,56).

Fiberoptic Endoscopy

Small, dedicated fiberoptic endoscopes are available to directly visualize all mucosal surfaces of the nasopharynx, pharynx, and larynx. Nasoendoscopy is minimally invasive, repeatable, and easy to perform and allows bedside evaluation for nonambulatory patients. Although this technique can evaluate both the motor and the sensorial components of swallowing, it can examine only the pharyngeal phase, with the additional limitation of a swallowing blackout (57).

A recent variant of the technique uses endoscopic delivery of pulsated air on the laryngeal mucosa to estimate the adductional reflex contraction of vocal folds or the patient’s ability to discriminate increasing pressure (58).

Esophagogastroduodenoscopy

This is the procedure that most gastroenterologists and surgeons choose when evaluating patients with gastroesophageal disorders, as it permits direct visualization of the esophageal, gastric, and duodenal mucosa and permits tissue biopsies of suggestive lesions. Endoscopy is indicated for any patient complaining of dysphagia or dyspepsia, generally as a first-line diagnostic procedure. More rarely (depending on the choice of the general practitioner and local logistics), endoscopy is performed after an upper gastrointestinal series revealing no obvious abnormalities.

The gastroesophageal flap valve is easily seen during esophagogastroduodenoscopy, thus revealing esophagitis and abnormal hiatal and paraesophageal hernias. Esophageal diverticula can also be visualized during endoscopy. This invasive technique is not always well accepted by patients and requires a relatively long learning curve for the operator. A consensus has been reached on the grading of mucosal damage and esophagitis, thus minimizing interobserver variability and subjectivity in reporting results (59). However, about 55% of patients with typical GERD-related symptoms do not have gross esophagitis, although careful endoscopic examination may detect, in about 12% of patients with GERD, the presence of intestinal metaplasia (Barrett’s esophagus, a condition associated with a 40-fold increased probability of development of esophageal adenocarcinoma).

Pharyngeal Manometry

A transnasally positioned manometric probe permits estimation of the rate of upper esophageal sphincter relaxation and the strength of pharyngeal contraction while also measuring the duration of these 2 events (60). Manometry can be performed at the time of videofluoroscopy (manofluorography) for better positioning of the pressure sensor and for distinguishing between recordings of intrabolus pressure and recordings from within a closed lumen (8).

Pharyngeal manometry is more difficult to perform than the esophageal test because of the extreme longitudinal and radial asymmetry of intraluminal pressures recorded from within the pharynx during deglutition (61). Moreover, movements during the pharyngeal phase of swallowing frequently displace the pressure sensor, thus making the technique difficult to standardize (62).

Stationary Esophageal Manometry

Esophageal manometry is usually performed on patients with dysphagia in whom endoscopy or a barium swallow study has excluded obvious structural abnormalities. The dynamic pressure measurement is especially useful in the diagnosis of primary esophageal motility disorders such as achalasia, diffuse esophageal spasm, nutcracker esophagus, or hypertensive LES. It is also useful in the characterization of esophageal disorders secondary to systemic diseases such as scleroderma, dermatomyositis, and polymyositis.

Although esophageal manometry is rarely indicated in patients with typical GERD-like symptoms, it is crucial for correct placement of the probe for esophageal pH testing in patients with atypical symptoms or unresponsive to proper medical therapy and in patients being considered for antireflux surgery. Esophageal manometry may identify a defective LES, suggesting the diagnosis of GERD, and provides valuable information on peristaltic function (thus allowing selection of the most appropriate antireflux procedure).

Ambulatory pH Monitoring

A pH electrode at the end of a catheter is placed in the esophagus 5 cm above the upper limit of the LES for prolonged monitoring of esophageal pH. The pH electrode is attached to a battery-operated ambulatory device that can record data for up to 24–48 h. This device records pH over a circadian cycle to identify the frequency and duration of esophageal mucosa exposure to acid and the ability of the esophagus to clear gastric reflux and allows correlation of the intensity and duration of symptoms with reflux episodes.

Esophageal exposure to gastric juice is evaluated as 6 major components contributing to a combined score: the percentage of the total time, upright time, and supine time that esophageal pH drops below 4; the total number of reflux episodes per day; the number of episodes > 5 min; and the duration of the longest episode (63,64).

Biliary Reflux

The combination of bile and hydrochloric acid has a noxious effect on esophageal mucosa. An ambulatory bile-reflux monitor (Bilitec 2000; Medtronic) has been developed to detect the presence of bile reflux in the esophagus (65,66). Using bilirubin as a marker for bile, this spectrophotometric system records the frequency and duration of bile exposure in either the stomach or the esophagus over a 24-h period. Combined with 24-h pH testing, the bile-reflux detector gives a more complete profile of a patient’s reflux, thus identifying patients at greater risk for developing complications such as Barrett’s esophagus. Nevertheless, reflux alone of bile in the esophagus cannot be considered the main risk factor for such complications.

Reflux of duodenal content into the stomach, possibly causing alkaline gastritis and intestinal metaplasia in the gastric mucosa, occurs in up to 40%–50% of patients who undergo cholecystectomy (Fig. 2). This condition can also be secondary to gastric surgery that alters the function of the pylorus. As already emphasized, reflux of duodenal and gastric contents into the esophagus can play a role in the pathogenesis of Barrett’s esophagus.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Delayed abdominal hepatobiliary scan (45 min after injection, when radioactivity has almost cleared from hepatocellular compartment) obtained with a ^99mTc-iminodiacetic acid analog for a patient with persisting dyspepsia after endoscopic cholecystectomy. Reflux of radioactive bile from duodenum into stomach is obvious (arrows), possibly explaining persistence of symptoms.

Multichannel Intraluminal Impedance (MII)

In this new technique for evaluating esophageal function and gastroesophageal reflux, a change in resistance to alternating current between 2 metal electrodes is produced by the presence of a bolus inside the esophageal lumen (67,68). Combined MII and esophageal manometry provide simultaneous information on intraluminal pressure changes and bolus movement, whereas combined MII and pH monitoring allow detection of reflux episodes irrespective of their pH values (i.e., acid vs. nonacid reflux). Combined MII and pH testing shift the gastroesophageal reflux testing paradigm, in that reflux events are no longer detected solely by pH changes. In fact, reflux presence, distribution, and clearance are detected primarily by MII and are characterized simply as acid or nonacid on the basis of pH change and as liquid, gas, or mixed on the basis of MII.

On the other hand, combined MII and esophageal manometry identify those patients with abnormal esophageal motility on manometry who actually have a clinically important defect in esophageal function. In fact, whereas manometry is an indirect measure of esophageal function, impedance makes it possible to follow the bolus movement.

Impedance technology has great potential for identifying the relative roles of acid and nonacid reflux, such as in patients who continue to have reflux-related symptoms even though their acid is being successfully suppressed or controlled with traditional acid-suppressant therapy. This technique can also help in assessing patients with reflux-related respiratory problems and in understanding dysphagia and its manometric and clinical correlates.

Electrogastrography

By using mucosal, serosal, or cutaneous electrodes, this technique evaluates the fasting and postprandial myoelectrical activity generated by the gastric antrum (69). Cutaneous electrogastrography is most frequently used clinically through the positioning of 3 or 4 electrodes on the epigastrium (46). The normal gastric pacesetter potentials range from 2.5 to 3.75 cycles per minute (cpm). Either bradygastria (<2.5 cpm), tachygastria (3.75–10 cpm), or bradytachygastric arrhythmia (mixed pattern) is detectable (69). Such gastric dysrhythmias can be observed in dysmotilitylike dyspepsia, idiopathic gastroparesis, diabetes mellitus, nausea of pregnancy, and motion sickness and after drug intake or gastric surgery (69), whereas increased wave amplitude has been reported in gastroparesis due to mechanical obstruction (70). Electrogastrography has a complementary role in evaluating patients with dyspeptic symptoms, in predicting gastroparesis, and in assessing gastric motor function in patients with GERD or with chronic constipation or atypical upper-gastrointestinal symptoms (46).

Scintigraphy Procedure

Image Processing

Few studies have evaluated patients with oropharyngeal dysphagia with a radiolabeled swallow test; most of these studies concerned patients with neuromuscular dysphagia, and the analysis was based on the oral and pharyngeal transit times and their corresponding indices (99,100). Because of the high interobserver variability in the calculation of transit times (G. Mariani et al., unpublished data, 2003), we believe that simple retention indices obtained after single swallows are a reliable means of assessing oropharyngeal dysphagia.

Quantitative Parameters.

The main goal of quantitative analysis is to quantify retention and measure the rate of esophageal transit. Most quantitative approaches use multiple-swallow techniques to overcome problems related to the intraindividual variability of a single wet swallow. Usually 3 regions of interest (ROIs) are drawn, encompassing the upper, middle, and lower thirds of the esophagus. The data can be summed for a total esophageal measurement. The gastric fundus should be identified and excluded from the lower third region, as oscillations linked to respiratory movements interfere with the analysis (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Typical time–activity curves obtained for oropharyngoesophageal radionuclide transit study (10-mL liquid bolus, upright, anterior view) of healthy subject. Shown are curves for mouth (A), pharynx (B), and whole esophagus (C) and for upper third (D), middle third (E), and lower third (F) of esophagus. Summed image at center defines the 3 esophageal ROIs.

Based on 1 wet (15 mL of water labeled with ^99mTc-sulfur colloid) and 3 successive dry swallows, Klein and Wald calculated the fourth-swallow residual activity fraction. A fraction > 19.8% in the posterior view (or 13.1% in the anterior view) is considered abnormal. They also proposed a mathematic 2-compartment model to calculate the mean transit time, by separating the time–activity curve of the first swallow into a rapid component (major fraction of the bolus that has traveled the esophagus) and a slow component (residual fraction of the bolus clearing with subsequent swallows). The mean transit time is calculated as the ratio of the area under the fast component to the maximum height. Although this index is elegant, it did not prove to be diagnostically meaningful (97).

An alternative approach is based on multiple independent swallows (6 swallows are adequate for a reliable scintigraphic study). The individual swallows of a given esophageal scintigraphy study are normalized to their corresponding starting points, arranged consecutively, and then condensed in a sum image. Curves are generated by plotting the counting-rate columns assembled in each image. The following indices have been suggested (75,84,98):

• Transit time: lag time from the starting point until activity falls to ≤10% of peak activity;

• Mean transit time: (∑cts)/cts_max;

• Mean time: (∑cts(t)x t)/∑cts (t);

• Esophageal emptying (EE) at 10 s after T_max as fraction of peak activity: EE_{T max+10 s}; and

• Esophageal emptying at 12 s after swallow as fraction of peak activity: EE_12 s.

Mean transit times and mean times do not reflect true values, since the esophageal time–activity curves never actually drop to zero. Calculation of esophageal emptying should take into account residual activity after each swallow; the following formula attempts to correct for this residual background activity (usually expressed as percentages): esophageal emptying = (cts_max – cts_(sec))/(cts_max – cts_mra), where cts_mra is the mean residual activity calculated on the first 5 swallows. However, this procedure overcorrects for background, sometimes yielding emptying rates > 100%. Tatsch demonstrated a high discriminating capacity for EE_12 s (95% sensitivity, 96% specificity), followed by mean time, by EE_{T max+10 s}, and lastly by transit time (similar specificities but lower sensitivity) (79,84).

Quantitative analysis of esophageal scintigraphy can also be performed with the time–activity curves obtained separately on the 3 ROIs: upper, middle, and lower third of the esophagus. In this case, one should keep in mind the delay in clearance at the distal third that is due to the time lag required for relaxation of the LES (76). This procedure can constitute a second-line approach reserved for patients in whom cine-mode visual analysis or condensed-dynamic-image analysis has detected abnormalities in the overall esophagus, so that the specific abnormal segments can be identified.

Regional-transit-time topography has recently been developed. The method, which is based on visualization of the esophageal bolus transit as profiles connecting pixels containing the same number of counts, produces a multicontour plot and calculates relative local transit times along the esophagus. The counting rate measured at a certain cross-section is inversely proportional to the velocity of flow and directly proportional to transit time. The relative local transit times so obtained concisely describe the local kinetics of bolus transit along the esophagus, with a longitudinal resolution of about 25% of the esophagus (103,104).

General Clinical Application

The wide range in sensitivity and specificity reported for esophageal scintigraphy can at least in part be attributed to the different disorders evaluated. The potential clinical role of esophageal scintigraphy (including the oropharyngeal phase of swallowing) in patient management is better appreciated in light of the pathophysiologic changes underlying esophageal motility disorders (Figs. 4 and 5).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Examples of radionuclide swallow studies of patients with oropharyngeal dysphagia. (A) Time–activity curve for ROI drawn over oral region of patient with amyotrophic lateral sclerosis after patient swallowed liquid bolus (10 mL, upright, anterior view) shows significant retention and markedly delayed clearance of activity from mouth, with double swallow (piecemeal deglutition because entire bolus could not be swallowed at once). (B) Static image of patient with severe oropharyngeal dysphagia after total thyroidectomy (damage of upper left laryngeal nerve). Dynamic recording of the swallowing of a liquid bolus (10 mL, upright) was impossible because of wide movements due to coughing. Aspiration in trachea is obvious. (C) Semisolid bolus (10 mL, upright, anterior view) did not cause gross aspiration, thus permitting dynamic recording. Time–activity curve obtained for ROI drawn over pharyngeal region on completion of semisolid transit study shows markedly delayed, irregular, and incomplete clearance of radioactivity from pharynx.

Diffuse Esophageal Spasm.

Diffuse esophageal spasm is characterized by uncoordinated esophageal contractions that can interfere with esophageal clearance (Fig. 6). Different manometric criteria have been proposed for diagnosing this condition, including either the presence of spontaneous and repetitive contractions or the presence of high-amplitude or prolonged contractions. The key diagnostic criterion is the presence of simultaneous contractions induced by wet swallows. Besides the discrepancies in manometric criteria, diffuse esophageal spasm is an intermittent phenomenon that is evident manometrically in >10% (but ≪100%) of wet swallows (106,112). Therefore, in patients with this condition, esophageal scintigraphy should include multiple wet swallows, visual analysis of bolus dynamics (showing chaotic bolus movements), and definition of the time–activity curves (showing multiple peaks after a single swallow).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Radionuclide esophageal transit study (15-mL semisolid bolus, upright, anterior view) of patient with amyotrophic lateral sclerosis. Shown are time–activity curves for upper third (A), middle third (B), and lower third (C) of esophagus. Summed image at left defines the 3 esophageal ROIs. All 3 portions of the esophagus clear initially with a normal-appearing pattern, but there is marked retention of radioactivity, mainly in mid esophagus but also somewhat in proximal third, and slow subsequent passage of radioactive bolus to lower third. A similar scintigraphic pattern with delayed clearance of esophageal radioactivity is also seen in patients with achalasia or with scleroderma, but with retention in distal rather than middle portion of esophagus (Fig. 7).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Radionuclide esophageal transit study (10-mL liquid bolus, upright, anterior view) of patient with esophageal dysphagia. (A) One-second frames from dynamic recording show irregular movements of radioactive bolus up and down the esophagus. (B) Time–activity curve for ROI drawn over pharynx shows the normal efficiency of a single swallow without regurgitation. (C) Time–activity curve for ROI drawn over entire esophagus shows incoordinate clearance of radioactivity, the typical pattern for a patient with diffuse esophageal spasm. Review of the dynamic recording in cine mode excluded gastroesophageal reflux as an alternative cause of the irregular, multiple-peak pattern in the esophageal time–activity curve.

Achalasia.

Classic achalasia is caused by degeneration of neurons in the wall of the esophagus. These neurons are involved in the production of nitric oxide, which relaxes esophageal smooth muscle, therefore causing the basal LES pressure to rise (106,113–116). The result is impaired esophageal clearance and delayed transit times. Esophageal scintigraphy readily detects the delay in transit (Fig. 7), with a sensitivity close to 100% (74). Distal esophageal retention does not clear after the patient assumes an upright position or drinks a glass of water. Scintigraphic evaluation of esophageal clearance is widely accepted as the method of choice in the follow-up of patients with achalasia, for monitoring the efficacy of treatment (myotomy, pneumatic dilation, or botulinum toxin injections). To optimize sensitivity for this condition, the patient should be upright when studied and should swallow a bolus of unlabeled water after the radiolabeled swallow (117–122).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Radionuclide esophageal transit study (15-mL liquid bolus, supine, posterior view, 1 wet swallow followed 30 s later by 3 dry swallows at 15-s intervals) of a patient with achalasia. (A) Summed image of whole esophagus, with ROI. (B) Time–activity curve shows marked retention of radioactivity in the esophagus. Marked esophageal retention of radioactivity remained virtually unchanged after patient stood upright for some minutes. In contrast, esophageal scintigraphy of patients with scleroderma shows similar pattern of radioactivity retention, which, however, clears after they stand.

Application to Pediatric GERD

Regurgitation and vomiting are common symptoms in infants with GERD, probably because of the limited capacity of the esophagus in this age group. In infants, it is difficult to discriminate pathologic from physiologic reflux. If the event is within physiologic levels, the infant thrives well and usually responds to simple measures such as nursing in an upright position or thickening of the food.

Most infants with severe reflux have symptoms by the age of 2 mo; the condition usually has a benign course, and about 60% of these babies are free of symptoms by the age of 18 mo. However, about 30% have persistent symptoms until approximately 4 y of age, 5% develop esophageal strictures, and another 5% die if they do not receive adequate treatment.

Clinical presentation of severe gastroesophageal reflux in children includes regurgitation or vomiting and failure to thrive. The spectrum of complications includes recurrent infections due to aspiration, asthma, apnea, cough, and stridor (137,138), along with the sequelae of esophagitis, including anemia, esophageal stricture, esophageal dysmotility (139), Sandifer’s syndrome, and even sudden infant death. Primary respiratory disorders may predispose to gastroesophageal reflux as a result of increased intraabdominal pressure caused by coughing. Contrary to common beliefs, crying does not exacerbate gastroesophageal reflux (140).

Most conventional tests for diagnosing gastroesophageal reflux in adults are not suitable in infants and children because of the radiation dose or because they require substantial cooperation from the patient (141). In particular, a barium upper-gastrointestinal series (with either intermittent imaging or cineradiography) uses a nonphysiologic test meal and requires provocative measures such as abdominal compression. Other diagnostic tests are more or less invasive, such as the acid infusion test (during which the patient is asked to communicate symptoms), intraluminal pH monitoring, esophageal manometry, upper digestive endoscopy, and tracheal aspiration (searching for lipid-laden macrophages to confirm aspiration of milk) (142).

The radionuclide method that has been adapted for use in infants and children (143) is performed at the time of scheduled feeding. ^99mTc-Sulfur colloid (3.7–18.5 MBq [100–500 μCi]) mixed with milk or formula is used to label the liquid meal at a concentration of <1.85 MBq/mL (<50 μCi/mL). About half the regular meal is labeled, leaving the remainder unlabeled for completing the feeding study. If possible, the scintigraphic study includes the initial swallowing phase and esophageal transit. After the radiolabeled half of the meal has been given, the remaining, unlabeled, portion is given to wash radioactivity from the mouth, pharynx, and esophagus. The baby is then burped and placed supine for imaging of the chest and gastric area (reflux in infants is more likely to occur in the supine than in the prone position) (144). According to age and cooperation, the child can be positioned either lying on the imaging table or lying on the γ-camera head. A high-sensitivity collimator should be used to maximize radioactive counts, because the activity during an episode of reflux can be low or in a small volume.

Sequential images should be recorded continuously for about 5–10 s per image for at least 60 min, since about 25% of reflux episodes can be missed when limiting the acquisition to 30 min. Although dynamic acquisition at 20, 30, or even 60 s per frame (64 × 64 matrix) is usually adequate to identify gastroesophageal reflux, using shorter framing (e.g., 5 s per frame) can help in detecting brief reflux episodes and in evaluating clearance of reflux from the esophagus. At the end of dynamic acquisition, 5-min static images of the anterior and posterior lung fields are obtained, taking care to exclude the high gastric radioactivity. An additional procedure normally applied only to adults is to progressively increase abdominal pressure, as is done during a conventional barium radiography study. If aspiration is suspected, static imaging should be repeated at about 2 and 24 h.

For data analysis, ROIs are manually defined over the lower, middle, and upper esophagus and the oropharynx to generate time–activity curves, in which reflux episodes are characterized as sharp spikes. The dynamic acquisition should be reviewed in the cine mode to identify movements of the body, which can simulate reflux spikes in the time–activity curve.

Gastroesophageal reflux can be quantitated using different indices, which usually consider the volume of each episode, the frequency of the episodes, and the rate of reflux clearance from the esophagus. A simple approach is to express esophageal activity (either in selected frames from dynamic recording or in the static images) as a fraction of gastric activity (145). This approach clearly discriminates adult patients with gastroesophageal reflux from those without reflux (11.7% ± 1.8% vs. 2.7% ± 0.3%). When the mean value of esophageal activity was expressed as a fraction of initial gastric activity, patients with peptic esophagitis had significantly different values from those of controls (3.66% ± 0.81% vs. 0.66% ± 0.12%) (146). One can also calculate an overall reflux index based on the ratio of summed activities in each episode to gastric activity over the entire 60 min of the dynamic study (recorded at 20 s per frame) (147). If higher than 5%, this index denotes the presence of reflux.

The sensitivity of scintigraphy in detecting gastroesophageal reflux has been reported to be 75%–100%, depending on the protocol used (148). These values are consistently higher than those of conventional barium studies or manometry.

Scintigraphy Procedure

Patients fast for at least 8 h before a gastric-emptying study. In addition, it is preferable to study women during the first 10 d of the menstrual cycle, to avoid possible hormonal effects on gastrointestinal motility. Any medication that potentially interferes with gastric motility (narcotic analgesics, anticholinergics, antidepressants, calcium channel blockers, gastric acid suppressants, aluminum-containing antacids, or somatostatin) is discontinued for an appropriate period (depending on specific pharmacokinetics), unless the scintigraphic test is performed specifically to assess the effect of such drugs on gastric motility. Tobacco and alcohol is also withheld for at least 24 h.

The first radiopharmaceutical described for gastric-emptying studies was chicken liver labeled by injecting sulfur colloid into the chicken. The liver was harvested, cooked, and administered to the patient. An advantage of this approach was the stability of the label. However, this procedure is not practical for routine use (150). A more convenient alternative is to mix 11–18 MBq (300–500 μCi) of ^99mTc-sulfur colloid with eggs or a fat-free egg substitute (EggBeaters; ConAgra Foods, Inc.) to obtain a stable label for the solid phase (Table 4) (151). Besides the radiolabeled component, the test meal typically contains other unlabeled products because a caloric content of at least 200–300 kcal is needed for complete activation of digestive functions (152). In a multicenter study, a standardized, low-fat meal showed excellent reproducibility for gastric-emptying studies. The meal consisted of fat-free egg substitute (EggBeaters) served with 2 slices of bread, 30 g of strawberry jam, and 120 mL of water (150). Standardizing the content of the meal, the proportion of liquid and solid, the amount of fat, the number of calories, and the rate of consumption is important for establishing a reproducible response (153,154).

¹¹¹In-DTPA is usually used for liquid-phase studies, as it is chemically inert and does not bind to the solid phase of the meal, thus allowing simultaneous assessment of both solid and liquid emptying.

The radiolabeled meal is ingested within 10 min, under standardized environmental conditions (ambient noise, ambient light, and patient comfort). If any portion of the meal is not eaten, the uneaten amount is recorded.

Imaging can be performed while the patient is either standing, sitting, or supine, provided the position does not change during the study. Because movement of solids from the posteriorly located fundus to the more anteriorly located antrum results in a nonuniform attenuation of the gastric activity, anterior and posterior imaging with subsequent calculation of the geometric mean is suggested to correct for this attenuation nonuniformity; alternatively, a single left anterior oblique view can be used (155–159). Images are acquired in at least a 64 × 64 pixel matrix. Continuous data recording at a framing rate of 30–60 s per image for at least 90 min is generally recommended (160) (up to 2–3 h for larger-volume meals or meals with a higher calorie, fat, carbohydrate, or protein content). Alternatively, data recorded at discrete 15-min intervals (or even at 0, 60, 120, and 180 min after meal ingestion) have been shown to provide reliable gastric-emptying data. However, with this approach no information is available on the lag phase, and rapid gastric emptying (dumping) may not be fully characterized (32,161). Images recorded at 4 h may be necessary to characterize delayed gastric emptying in cases with borderline 2-h emptying (162).

The evaluation of indigestible solids (fibers or pellets) requires prolonged patient observation and provides information on the interdigestive period rather than postprandial motility (32).

Data Analysis and Quantitative Parameters

After the set of geometric mean images has been created, the full study is initially reviewed in cine mode to detect patient motion. After the cine review, it is helpful to add the images from the serial acquisition to define the borders of the stomach and the adjacent loops of bowel. ROIs are drawn around the gastric area, adjacent esophagus, duodenum, and jejunum. The curves are corrected for radioactive decay.

The percentage of gastric retention at a specific time point (usually 2, 3, or even 4 h) is calculated with respect to the immediate postingestion value at the zero time point. Because this parameter does not have a normal, gaussian-type distribution, the median and 90th or 95th percentile, rather than the mean and SD, are used to define normal values (32).

When continuous imaging is recorded, half-emptying time (T₅₀) is determined as the time it takes to reach half the peak counts, whereas when static, discrete-interval images are acquired, the data points should first be analyzed by nonlinear least-squares fit. Because solid gastric emptying follows a sigmoidal rather than a simple exponential curve, at least 2 parameters are required for an optimal description. At least 2 computational procedures have proven capable of adequately describing the pattern of biphasic solid gastric emptying: the Elashoff power exponential function, y(t) = (e^–kt)^β (163), and a modified power exponential function, y(t) = (1 − [1 − e^kt]β) (25), where y(t) is the gastric retained activity at time (t), k is the gastric-emptying rate, and β is the intercept on the y-axis back-extrapolated from the terminal portion of the curve.

When the Elashoff function is applied, T₅₀ = (ln 2^1/β)/k, whereas when the modified power exponential function is applied, T₅₀ = −ln ([1 − 0.5]^1/β)/k.

The lag phase represents the time required for transfer of solid food from the fundus to the antrum and then for the antrum to grind it into small particles (suspended in the gastric fluid) that can begin to empty into the duodenum (164). Various methods have been proposed for evaluating this parameter. With continuous imaging, it is relatively easy to visually define the transition point in the gastric time–activity curve: the point of first appearance of duodenal activity or the time of maximal antral filling. When data are acquired as serial static images, the lag phase can arbitrarily be calculated as a given percentage of drop (5%–10%) from peak counts; however, this approach is highly dependent on the slope of the second phase (165). In mathematic terms, the lag time corresponds to the inflection point of the gastric time–activity curve, where the second derivative equals zero. When the Elashoff function is applied, lag time = ([β − 1]/β)^1/β/k, whereas when the modified power exponential function is applied, lag time = β/k.

The lag phase calculated by the modified power-exponential function correlates poorly with the data derived by other methods, and the accuracy of such an estimate is greatly limited by the rate of temporal sampling. It has therefore been suggested that the lag phase calculated by this method possibly represents a more complex phenomenon than just emptying of solid food into the duodenum (163).

Dynamic Antral Scintigraphy

Although manometry is the most accurate method for assessing gastric motor function, this invasive procedure is not well accepted by patients and is not widely available (166). Real-time ultrasound evaluation of gastric motor function is often hampered by poor image quality because of echogenic food particles (167). Simultaneous scintigraphic evaluation of both gastric motility and gastric emptying optimizes the characterization of gastric function with minimal additional cost, time, and technical effort.

Movements of the stomach wall induce changes in the distribution of radioactive food within the stomach. Therefore, dynamic scintigraphy as performed for a gastric-emptying study can evaluate gastric motility as well. Because antral hypomotility (reduction of phasic contractions in the distal stomach) has been reported for a variety of dyspeptic syndromes, almost all studies have evaluated only antral motility (168–175). The most common procedure is to add a short, high-definition dynamic acquisition (1–2 s per frame for 4–5 min) at the end of a standard gastric-emptying study, performed either with dynamic imaging or with serial static imaging. Data analysis is performed after attenuation correction, to generate time–activity curves for ROIs defined on the proximal, middle, and distal antral regions. Further data processing includes elimination of artifacts due to patient movement or to translational movement of the stomach (performed by baseline restoration of raw data using a polynomial fit) and elimination of background noise (performed by an autocorrelation function). A refined Fourier transform is then performed on antral time–activity curves to determine the amplitude and frequency of antral contractions.

Consistent with manometric (176), real-time ultrasonographic, and electrogastrographic estimates (177), in several scintigraphic studies the frequency of antral contractions has been found to be about 3 per minute (127,167–173). This normal frequency does not seem to change significantly in patients with diabetes, gastritis, functional dyspepsia, progressive systemic sclerosis, or pyloric stenosis, either after administration of gastric acid suppressants or after gastric surgery. In contrast, the amplitude of antral contractions seems to discriminate various patient groups. In functional dyspepsia, paradoxically increased amplitude has been reported, possibly explained by predominantly inefficient (nonexpulsive) antral contractility or by pyloric dysmotility (170), whereas decreased amplitude has been observed in diabetic patients (168) and in patients who previously underwent Billroth type I or type II gastrectomy or truncal vagotomy (178).

Unfortunately, variability in the definition of the antral ROIs, as well as in the normalization and filtering steps, hampers the reproducibility of measurement of antral function. Two methods avoiding the autocorrelation function have therefore been proposed to improve reproducibility. Both involve a postsynchronization method using either a single distal antral ROI (179) or Fourier analysis of condensed images of the whole stomach (180).

Gastric SPECT

Dyspeptic patients frequently have defective gastric accommodation contributing to postprandial symptoms such as early satiety, distension, and nausea. This condition has been reported in functional dyspepsia, postfundoplication dyspepsia, postvagotomy gastric surgery, and neuropathic diabetic gastroparesis (181–186). Although proximal gastric accommodation can be measured by a barostatic balloon, and real-time ultrasonography can measure antral accommodation, neither of these 2 methods can evaluate whole-stomach accommodation. SPECT has been used to measure gastric accommodation in response to a meal, based on the well-known ability of both the parietal (oxyntic) and the nonparietal (mucous) gastric cells to take up and excrete ^99mTc-pertechnetate (187).

The procedure is as follows: A 370–740 MBq (10–20 mCi) dose of ^99mTc-pertechnetate is injected intravenously after an overnight fast. Dynamic tomographic acquisition is performed starting 10 min after injection, using a multiorbit mode; 6° images are acquired in a 128 × 128 matrix at 3 s per step (total acquisition time, about 180 s). After 2 initial orbits have been recorded as baseline, the patient drinks the test meal while supine, with the aid of a straw. Immediately afterward, 7 additional orbits are acquired over about 21 min to evaluate gastric accommodation to the meal challenge.

A semiautomated segmentation algorithm is used to reconstruct the stomach. Total gastric volume is measured during fasting, during the 3- to 12-min postprandial period, and during the 12- to 21-min postprandial period. The procedure has recently been simplified by acquiring 3 complete 360° orbits every 10 min (6° steps, 64 × 64 matrix, 10 s per step) (188). Intra- and interobserver coefficients of variation in estimating fasting and postprandial gastric volumes are 8%–9% and 12%–13%, respectively. The greatest postprandial volume increase was in the proximal stomach (about 2- to 4-fold), with no significant difference between early (3–12 min) and late (12–21 min) volume changes.

The method has recently been validated by a phantom study using a balloon filled with ^99mTc-DTPA, and changes in gastric volume measured in response to a meal proved to be as accurate as those derived by the barostat balloon technique, with no effects of age (at least in adults) and body mass index on the fasting or postprandial gastric volume ratio (189). Gastric SPECT has also demonstrated impaired accommodation in patients with idiopathic nonulcer dyspepsia and in patients after fundoplication (187,190,191).

Clinical Application

Gastric scintigraphy (Fig. 8) is generally indicated in several clinical conditions when morphologic investigations fail to reveal the cause of dyspepsia, such as postprandial nausea, vomiting, upper abdominal discomfort, early satiety and bloating, suspected gastroparesis, poor diabetic control, and severe reflux esophagitis unresponsive to medical treatment (46,158).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Examples of gastric-emptying curves (corrected for physical radioisotopic decay) obtained by sequential imaging at discrete intervals after ingestion of radiolabeled solid meal. (A) Results for healthy volunteer (in our institution, T_1/2 values of gastric time–activity curves for healthy subjects range from 110 to 120 min). Curve is characterized by distinct lag phase lasting about 20–30 min before emptying begins. (B) Results for patient with gastroesophageal reflux; this curve shows virtually no lag phase and shows markedly delayed gastric emptying, as observed in an average 40% of these patients (236,237). On the basis of this finding, the surgical treatment chosen was fundoplication and pyloroplasty, instead of isolated fundoplication, which would have induced postoperative symptoms (e.g., bloating) with no obvious lag phase. (C) Gastric-emptying curve obtained for same patient after surgical treatment demonstrates slightly accelerated gastric emptying (due to pyloroplasty), with reappearance of a discernible lag phase.

The pathophysiologic basis of functional dyspepsia has variously been ascribed to delayed gastric emptying of solids (192,193), failure of fundic relaxation (184), visceral hypersensitivity (40), and Helicobacter pylori infection (194,195). Although more than 40% of patients with functional dyspepsia have delayed solid gastric emptying, whether such an abnormality explains the clinical symptoms remains controversial (190,196–199). However, some correlation has been found between delayed gastric emptying and relevant postprandial fullness or severe vomiting and with being female (191), the total symptom score being higher in those patients with delayed gastric emptying (200). In diabetic patients, independent predictors of delayed gastric emptying included abdominal bloating, the female sex, and high blood glucose levels and body mass index (201).

Another controversial issue concerns the effect of prokinetic drugs, inducing accelerated gastric emptying, on symptom relief. Surprisingly few studies have evaluated whether baseline gastric emptying is a predictor of the efficacy of these agents or whether accelerated gastric emptying after therapy is a predictor of symptom improvement (202). Helicobacter pylori infection in patients with nonulcer dyspepsia is usually not associated with delayed gastric emptying (203–205), although both gastric emptying and symptom scores have been reported to improve after Helicobacter pylori eradication (206).

Delayed gastric emptying, most likely due to autonomic neuropathy (resulting in a prolonged lag phase (168)), has been identified in 27%–58% of diabetic patients; however, hyperglycemia further delays gastric emptying, and delayed gastric emptying itself may contribute to poor glycemic control (207). Delayed gastric emptying is also associated with various systemic syndromes (Parkinson’s disease, amyloidosis, myotonic dystrophy, polymyositis, HIV infection, or cytomegalovirus infection) and with local conditions such as vagotomy without pyloroplasty, and gastroesophageal reflux (208–215).

As expected, accelerated gastric emptying is found after pyloroplasty or after surgical treatment of reflux (216–219).

Another clinical application of gastric scintigraphy is postsurgical evaluation of the gastroesophageal junction. Achalasia is the esophageal disorder whose treatment has been most substantially affected by the development of minimally invasive surgery—that is, a laparoscopic approach to Heller’s myotomy (220)—widely reported to yield favorable results in up to 90% of these patients (221–224). Because the support structures of the antireflux mechanism, such as the phrenoesophageal ligament and the LES, are sacrificed during laparoscopic dissection of the esophagus and myotomy, an antireflux procedure is also required (225–227). Residual dysphagia can be linked to persistent esophageal narrowing due to incomplete myotomy. Esophageal transit scintigraphy is well suited as an addition to esophageal manometry and 24-h pH monitoring for follow-up and monitoring of this complication.

The efficacy of medical therapy in patients with GERD seems to wane with time, and for many patients the dose of proton pump inhibitors needs to be readjusted to control symptom relapse. However, surgical therapy for GERD has seen dramatic changes over the last decade, making it much more accessible and less invasive than in the past (228). The goal of effective antireflux surgery is to restore the competency of the cardia. Incompetence of the gastroesophageal junction has a multifactorial etiology, including such abnormalities as transient relaxation of the LES (relaxation not associated with a swallow), primary hypotension of the LES, gastric-emptying abnormalities, and absence of a flap valve. The presence of hiatal hernia exacerbates reflux by impairing esophageal emptying and LES function. An ideal operation, thus, is one that addresses each of these aspects. The most effective way (shown to yield favorable results in about 90% of patients) to permanently restore the competency of the cardia is to create some form of plication over the esophagus just proximal to the cardioesophageal junction, according to the following principles. First, the wrap must be constructed over the esophagus, just proximal to the gastroesophageal junction, and must be fixed to the esophagus to remain in that position permanently. Second, the wrap must be constructed without tension using the fundus of the stomach and leaving enough space to accommodate a 42- to 50-French bougie inside the esophagus. Third, a total wrap, such as the total fundoplication described by Nissen, should measure about 2 cm in length in its anterior aspect. Fourth, the wrapped portion of the esophagus must lie below the diaphragm without tension. Finally, the diaphragmatic hiatus must be gently closed around the esophagus, above the wrap.

Esophageal transit scintigraphy and, above all, gastric-emptying scintigraphy are well suited for following up patients who undergo fundoplication because of GERD (Fig. 8) (229).