Fluoroscopic studies typically require use of enteric contrast material for diagnosis.3,4 Barium, an inert substance that is not absorbed, remains the primary contrast medium used in fluoroscopic procedures, whether it is administered orally to evaluate the esophagus and upper GI tract or rectally in a contrast enema. Several barium preparations are available; barium sulfate powder (96% wt/wt) can be diluted with sterile water for infant upper GI examinations to the desired concentration of 40% to 60% wt/vol. Premixed suspensions (60% wt/vol) can be used in older children, adolescents, and adults. Enema kits containing 97% barium wt/wt can be mixed with water to a final concentration of 15% to 33% barium wt/vol for infants, older children, and adolescents. Although adverse reactions to barium products are rare—reportedly 2 per million or less⁵—they do occur and may present as a rash, loss of consciousness, and anaphylaxis, typically related to any one of several additives, such as methylparaben and carboxymethylcellulose.^5–7 Aspiration of barium in small quantities is tolerated, but aspiration of a large volume of barium can be fatal.^8,9

Barium is contraindicated in cases in which viscus perforation is suspected. In such cases, a low-osmolality, nonionic, water-soluble iodinated medium such as iohexol is used (Table 85-1). It is important that hypertonic media such as ionic or high-osmolality media (e.g., diatrizoate or iothalamate) not be used orally because of the risk of aspiration and consequent pulmonary edema.^10–12

Table 85-1

Commonly Used Types of Contrast Media in Fluoroscopy of the Gastrointestinal Tract

Data from package inserts.

Gastrografin (diatrizoate meglumine and diatrizoate sodium) is an ionic, markedly hypertonic iodine solution with an osmolality (mOsm/kg) of approximately 1600; a 1 : 5 dilution approximates serum osmolality (285) but also dilutes the iodine concentration. Ionic hyperosmolar media can be absorbed from the GI tract and thus pose a risk in patients with a history of hypersensitivity, particularly to iodine, and potentially in patients with thyroid disease. Hyperosmolar media cause severe pulmonary complications of edema and pneumonitis if aspirated and can cause major fluid shifts into the bowel lumen, leading to a decrease in intravascular volume, an increase in serum osmolarity, and a decrease in cardiac output. In patients with underlying bowel disease, additional injury is possible.13,14

Omnipaque (iohexol) is a nonionic water-soluble iodinated contrast medium that is available in concentrations of 140, 180, 240, 300, and 350 mg of iodine. It is poorly absorbed from the intact GI tract, with renal excretion of 0.1% to 0.5% of the administered dose. Isovue (iopamidol) also has been used in the evaluation of the pediatric GI tract, but currently only Omnipaque is officially approved for this purpose. It must be emphasized that the osmolality of both of these media is greater than that of blood and that no agent is safe in the tracheobronchial tree, and thus great care and close fluoroscopic monitoring is necessary in all patients in whom aspiration is a potential complication.¹⁵

Barium is the standard agent used in the evaluation of the colon. However, in cases of potential perforation, water-soluble agents are used and can be diluted to approximate the tonicity of serum. Higher osmolality contrast media are used rectally for therapeutic purposes in cases of uncomplicated meconium ileus after diagnosis with a low-osmolarity agent. Gastrografin (diatrizoate meglumine and diatrizoate sodium) was the original agent described for this purpose.¹⁶ However, this agent can be associated with large fluid shifts and systemic complications in severely ill infants.¹³ Full-strength iothalamate meglumine 30% also can be used successfully for this purpose. Close attention to water and electrolyte balance, along with surgical standby, are mandatory.

Air also can be used during fluoroscopic procedures. For example, air provides an excellent way to distend a viscus during fluoroscopic transpyloric tube placement without obscuring the tube or adjacent bowel loops, and it is the preferred agent in the reduction of intussusception.

Specific procedures are outlined in the following sections, and indications and imaging protocols are discussed.

Indications and Protocols

Esophagram and Upper gastrointestinal Series

An esophagram and an upper GI series usually are performed in conjunction and include evaluation of swallowing, along with evaluation of the esophagus, stomach, and duodenum to the duodenojejunal junction. Common indications include evaluation of esophageal problems, such as complications of esophageal atresia repair, postoperative strictures or acute postoperative leaks, radiolucent foreign bodies such as impacted food, and the degree and efficacy of peristalsis. This examination is not indicated to diagnose esophageal atresia in most cases, because a chest radiograph with use of a coiled enteric tube typically is diagnostic, and the study may lead to unintended aspiration. Evaluation of the stomach in young infants includes assessment of gastric emptying as well as evaluation of the mucosa and focal lesions such as gastric duplication cysts. Evaluation of the duodenum is crucial in pediatric patients to document normal intestinal rotation. Small bowel follow-through has been largely supplanted by cross-sectional imaging in the diagnosis and monitoring of inflammatory bowel disease and therefore is undertaken much less frequently than in years past.

The examination is begun in the lateral projection, with the child lying on his or her left side to maintain the ingested contrast agent within the fundus of the stomach. Images of the esophagus are obtained from the nasopharynx to the esophagogastric junction, with special attention paid to nasopharyngeal aspiration, tracheal aspiration, masses, fistulas, and esophageal peristalsis and distensibility. The child is then laid supine, and the esophagus is examined in the anteroposterior projection. When the evaluation of the esophagus is completed, the barium in the fundus will be directed into the duodenum by turning the child into the prone right anterior oblique position. Gastric emptying is assessed, along with distensibility of the antrum, pylorus, duodenal bulb, and descending duodenum. Once the contrast material has reached the junction of second and third portions of the duodenum, the child is quickly placed in the supine position for assessment of the duodenojejunal junction, which is visible through the air-filled antrum. The duodenojejunal junction should lie to the left of the spine, at approximately the same level as the duodenal bulb. Once this assessment is accomplished, the child is quickly turned again, this time for a lateral projection to document the posterior course of the ascending and descending limbs of the normally rotated retroperitoneal duodenum. Evaluation for reflux can be performed after this portion of the study, if desired, or this can be done through other means such as scintigraphy or esophageal probe. A final image documents gastric emptying (Fig. 85-1).

Figure 85-1 Typical upper gastrointestinal series showing reflux in otherwise healthy infant.
A and B, Lateral and anteroposterior views of the esophagus during drinking show full distensibility without intrinsic or extrinsic mass lesions. C, Oblique imaging (right anterior oblique) directs ingested contrast material to the gastric outlet and shows prompt emptying and a normal pylorus and first and second portions of the duodenum. D, An anteroposterior image immediately following C shows progress of contrast material to the normally located gastroduodenal junction at the ligament of Treitz. E, A subsequent lateral view of the duodenum shows the posterior retroperitoneal location of parallel ascending and descending limbs. F, After further drinking and gastric filling, an episode of reflux to the cervical esophagus is documented. Note the wide open gastroesophageal junction, a typical appearance during reflux. G, An image recorded at the completion of the study shows good progress of contrast material through the small bowel.

A small bowel follow-through procedure usually requires ingestion of a larger amount of a contrast agent, typically barium, although in premature infants one can use a nonionic water-soluble contrast agent. Radiographs are obtained at regular intervals based on the course of the contrast medium through the bowel loops, with fluoroscopic evaluation as needed. Images of the terminal ileum with and without compression are obtained once the contrast has reached the cecum. In ill infants in the neonatal intensive care unit who do not require visualization of the ligament of Treitz, a “portable” small bowel follow-through procedure can be done, with the contrast material administered at the bedside and portable radiographs obtained at the appropriate intervals.

Contrast Enema

Although contrast enema has been superseded by other procedures (such as endoscopy) for previous indications (such as polyp identification), this examination remains extremely useful in many pediatric clinical settings, such as evaluation of distal bowel obstruction in the neonate, evaluation of complications of surgery or of disease such as necrotizing enterocolitis, and reduction of ileocolic intussusception.

The choice of contrast medium and the technique used for the contrast enema vary with the indication for the procedure, as previously discussed. Barium typically is used unless a perforation is suspected, in which case an iso-osmolal water-soluble medium is used. In newborns suspected of having distal bowel obstruction, an iso-osmolal water-soluble medium is used, and changed to a hyperosmolal medium is used as a therapeutic option if meconium ileus is encountered. Air is the contrast of choice in patients with intussusception during fluoroscopic reduction.

A catheter with a small tip is placed in the rectum and secured with tape to both buttocks, which are then taped together using manual pressure. Use of a balloon-tipped catheter is usually unnecessary and, we believe, inadvisable in young infants because of the potential for rectal injury. Fluoro-grab images can be recorded liberally to document the progression of contrast material and to document findings; spot filming is needed in areas in which increased detail is important, such as mucosal abnormalities or if a subtle leak or perforation is encountered.

Sonography

Overview

Sonography has become an extraordinarily useful modality in the evaluation of patients with GI symptoms. In addition to the availability of spectral as well as frequency shift and amplitude color Doppler imaging, advances with newer equipment include recording of motion cine images, harmonic imaging, extended field-of-view imaging, and three-dimensional (3D) capability.

Indications and Protocols

The utility and application of sonography in pediatric patients is extensive. The major advantage of sonography over CT is the obvious lack of ionizing radiation, but other very important advantages include the lack of need for sedation, and the multiplanar imaging capabilities. In general, scanning should be performed with the transducer that has the highest frequency necessary to penetrate the anatomy to be imaged and that will allow optimal spatial resolution; linear transducers are preferred if the access window will allow it. Curved array transducers allow a broader field of view; sector transducers may be necessary when the access window is limited or to image deeper structures in larger patients.

The primary role of sonography in the diagnosis of pyloric stenosis has become firmly established. Sonography is also extremely useful in the assessment of patients with clinically equivocal symptoms of appendicitis, although this setting punctuates its well-known operator dependence (e-Fig. 85-2), with published sensitivities ranging between 40% and 100%.^17–19 Sonography also is extremely useful in the evaluation of mesenteric adenopathy (e-Fig. 85-3), in highly detailed assessment of the bowel wall (Fig. 85-4),²⁰ in evaluation of small and large bowel intussusception (e-Fig. 85-5),²¹ and coupled with Doppler, in the effective estimation of disease activity in patients with Crohn disease.²²

Figure 85-4 Ultrasound appearance of normal and abnormal loops of bowel.
A, A transverse image through the right lower quadrant demonstrates normal, collapsed loops of small bowel. B, A transverse image through the mid abdomen in a 14-month-old boy with vomiting and diarrhea consistent with gastroenteritis reveals multiple distended, fluid-filled loops of bowel.

e-Figure 85-2 A, Appendicitis in a 15-year-old girl. The appendix is abnormal, measuring 15 mm in diameter, and is surrounded by marked periappendiceal edema. B, A 4-year-old girl. The appendix contains a shadowing fecalith; proximal to the fecalith it measured 5 mm and distally it measured 13 mm with purulent material.

e-Figure 85-3 A, Mesenteric adenopathy in an 8-year-old boy with abdominal pain. An ultrasound through the right lower quadrant reveals a chain of enlarged mesenteric nodes. The distance between cursors is 14 mm. B, An 11-year-old girl with abdominal pain and enlarged mesenteric nodes. Doppler interrogation reveals adjacent flow.

e-Figure 85-5 Small bowel intussusception.
Transverse (A) and longitudinal (B) images through incidental small bowel intussusception in a 7-month-old infant with gastroenteritis. The length between calipers in B was 2 cm. A follow-up study several hours later showed resolution of the intussusception, and the child remained asymptomatic.

The role of sonography in the diagnosis of solid organ pathology is likewise extensive and can be the final diagnostic tool for many abnormalities. Sonographic detail is particularly well visualized in young children, in whom high frequency and linear transducers can be used to access even the deeper abdominal structures. Indications include evaluation of biliary tract abnormalities, such as choledochal cysts, although additional imaging may be performed with MR depending on the specific clinical circumstances. Sonography may be performed as the initial modality to investigate suspected liver masses, and any subsequent CT or MRI protocols can be tailored on the basis of sonographic findings.

Although vascular structures are seen easily with contrast-enhanced CT, the direction and velocity of flow can be evaluated with Doppler sonography. Analysis of waveform pattern can identify hepatofugal flow in collateral vessels in patients with portal hypertension, along with vascular stenosis or thrombus. In patients with heterotaxy, abdominal sonography is helpful in assessing a splenic mass (located along the greater curvature of the stomach) and the associated vascular anomalies, such as interruption of the inferior vena cava, a preduodenal portal vein, and infradiaphragmatic total anomalous pulmonary venous connection 23,24 (e-Fig. 85-6).

e-Figure 85-6 Ultrasound findings in patients with heterotaxy.
A, A preduodenal portal vein. A longitudinal image through the upper abdomen in a child with polysplenia shows the course of the superior mesenteric vein into the liver anterior to the gas-filled duodenal bulb. B, Infradiaphragmatic total anomalous pulmonary venous connection. A longitudinal image through the upper abdomen in an infant with asplenia shows the anomalous vessel entering the abdomen from the chest at the esophageal hiatus. C, Multiple splenules in 1-day-old girl with abnormal results of a prenatal ultrasound. Transverse ultrasound of the right upper quadrant reveals multiple splenules between the stomach (S) and the ipsilateral adrenal gland (arrows). Azygous continuation of the inferior vena cava was documented elsewhere. She did not have congenital heart disease. L, Lateral extension of liver, the bulk of which was in the left upper quadrant.

Computed Tomography

Overview

CT is a particularly useful modality in pediatric abdominal imaging. The introduction of scanners with multichannel technology and volumetric acquisition permits very rapid examinations with isotropic reconstructions in multiple planes, with decreasing need for sedation.25,26 These new capabilities require development of newer protocols to accommodate more complex and sophisticated diagnostic demands. The timing and rate of contrast administration, with the ability to scan during a specific phase of intravascular contrast distribution, demand particular attention to technical details and new approaches to image interpretation.^27,28 The pediatric radiologist is further challenged by the need to balance image detail with radiation dose and implementation of the ALARA, or “as low as reasonably achievable” concept, with the increasing recognition of the potential risks of radiation exposure for pediatric patients.^29,30 Improvements in equipment aimed at reducing radiation exposure include innovations such as improved collimators and iterative reconstructive algorithms. Although significant challenges persist, much progress has been made through educational and awareness-raising social marketing campaigns such as the Image Gently Campaign of The Alliance for Radiation Safety in Pediatric Imaging (www.imagegently.org).³¹

Indications and Protocols

Unlike sonography, CT images are sequential, standardized, and much less operator dependent, and therefore CT is particularly useful in patients with complex disease affecting multiple organ systems, because it provides reliable monitoring of change in the extent of the disease during therapy and follow-up. CT also helps solve problems in patients with unusual multi-organ abnormalities. Evaluation of both intraabdominal and extraabdominal multiorgan pathology can be accomplished with great anatomic detail (e-Figs. 85-7 and 85-8) because evaluation of solid organ, hollow viscera, and peritoneal cavity pathology is rapidly accomplished with great anatomic detail and physiologic information. The relative lack of operator dependence and high sensitivity and specificity of CT in the imaging diagnosis of appendicitis has led to its increasing use when this diagnosis is clinically equivocal, with a documented reduction in negative appendectomies.³² However, this success has led to overuse in patients with abdominal pain; therefore physical examination, followed by ultrasound when the diagnosis is clinically uncertain, is recommended by most pediatric radiologists, with CT reserved for more difficult cases.^17,18 MRI is receiving increasing attention as a viable substitute for CT scanning in many indications, such as inflammatory bowel disease.³³

e-Figure 85-7 Pericardial hernia.
A, An abdominopelvic radiograph of a newborn infant with a prenatal diagnosis of hydrothorax. The radiograph suggests absence of most of the liver shadow from the right upper quadrant. The posterior costophrenic sulci are well outlined, indicating that chest density is not related to bilateral hydrothorax. Coronal (B) and sagittal (C) reformats of contrasted computed tomography scans outline herniation of the liver into the pericardial cavity with displacement of the heart and associated pericardial effusion.

e-Figure 85-8 A choledochal cyst with intrahepatic ductal dilatation and absence of the portal vein.
A and B, Axial contrasted computed tomography images through the liver and porta hepatis outline intrahepatic cystic dilatation of the biliary tree (Caroli disease) and cystic dilatation of the common bile duct at the porta hepatis (arrow). C and D, Coronal reconstructions better define the cystic intrahepatic ductal dilatation, absence of the portal vein, and splenomegaly.

CT protocols vary and undergo change with the ongoing introduction of new applications and advances in equipment capability; generalizable protocols applicable to pediatric patients can be downloaded at http://www.imagegently.org. However, some underlying principles underscore most successful pediatric examinations. Administration of intravenous contrast material is extremely important, particularly in pediatric patients in whom a paucity of intraabdominal fat decreases intrinsic intraabdominal contrast.^27,34 CT angiography requires a rapid contrast bolus injection, which in pediatrics can be challenging because of the caliber of IV access. Lowering kVp is important in patients in whom high-contrast structures are of interest, such as those undergoing angiography or bone examinations; in neonates, the kVp can be decreased to as low as 80, with some adjustment of the milliamperes-second (mAs) to produce acceptable image quality.²⁹ Precontrast images are seldom necessary and serve to increase the radiation exposure without adding diagnostic information. If necessary (e.g., to identify the presence of calcifications in an abdominal mass), the mAs of the precontrast scans can be decreased significantly and the scan should be limited to the appropriate specific area (e.g., scan only the mass, not the entire abdomen). The use of oral contrast material is usually important when outlining some types of intraperitoneal pathology, such as abscess or masses, but in other cases, its use is more controversial.³⁵ Positive oral contrast material will mask mucosal enhancement; use of water-density contrast material may be more appropriate in such cases.

Despite radiation concerns, CT remains an important life-saving modality in pediatric diagnosis. As with any other tool, it needs to be used judiciously, according to the principles of appropriateness, justification, optimization, and training.³⁶

Magnetic Resonance Imaging

Overview

MRI increasingly is being used for pediatric GI applications. The introduction and increasing use of imaging at 3 T and advances in parallel imaging and coil design expand the range of pediatric applications. In addition to congenital and acquired hepatobiliary abnormalities, indications for MRI increasingly include bowel abnormalities, such as inflammatory bowel disease and appendicitis.

Patient Preparation and Equipment Requirements

Adequate patient preparation is essential, because MRI often is time-consuming and requires extensive hospital resources. Considerations include the need for sedation, fasting requirements, and the need for oral contrast media.

Preparation begins with assessment of the need for sedation or anesthesia. Generally, children younger than 6 years who cannot hold their breath for 20 seconds will need sedation or anesthesia. Sedation is avoided in patients undergoing MR enterography examinations because of the need to use oral contrast media.

Patients should refrain from oral intake for 4 hours before the examination to ensure gallbladder distension in examinations of the biliary tree, as well as to minimize artifacts from bowel peristalsis. For all cases requiring vascular assessment, power injection of contrast material in appropriate-aged children is optimal and requires adequate peripheral venous access. Finally, once the patient is on the MR table, a respiratory monitoring pillow or belt should be placed with care.

Use of an oral contrast agent is essential for enterography examinations. A number of choices are acceptable, although most regimens consist of a biphasic agent, that is, one that gives the bowel lumen a long T2 and T1 relaxation time. Agents include VoLumen (E-Z-Em, New York, NY), mannitol, polyethylene glycol l, and locust bean gum (a type of galactomannan) solutions. Little difference is seen in efficacy of these choices, although patient tolerance may vary.³⁷ Most important is rapid consumption of a large volume of the contrast agent; 25 mL per kilogram of body weight over an hour is adequate. Placing the patient in the right decubitus position for the final 15 minutes before the start of imaging aids in emptying of the stomach.

Antiperistaltic agents also are required for enterography examinations. Glucagon, the more commonly used intravenous agent, can be used with either of two strategies: (1) half of the dose at the beginning of the examination and half just before administration of intravenous contrast material, or (2) the entire dose just before intravenous contrast material is administered. The dose, from 0.5 mg to 1 mg, varies between facilities. Although glucagon causes nausea and vomiting, it is well tolerated by most patients if it is given as a slow intravenous push over 1 minute and if, before it is administered, the patient is instructed to expect a brief period of nausea.

Equipment specifications are important because of children’s smaller sizes, with the consequent need for improved signal to noise and faster acquisition times to decrease the need for and length of sedation. Although the literature to date is still sparse on pediatric abdominal imaging at 3 T,38,39 increasing experience suggests that most children will benefit from the higher signal. Phased array surface coils are now standard, typically with eight to 32 channels. In the following situations, 1.5 T often provides improved image quality compared with 3 T: when the patient is very large, when ascites is present, when enterography examinations are being performed (1.5 T results in fewer banding artifacts in steady-state imaging), and during hepatic iron quantification.

Indications and Protocols

Hepatic tumors are well evaluated by MRI relative to CT,⁴⁰ with the goal of imaging being tumor characterization, staging, and assessment of resectability. For lesion characterization, determination of the T2-weighted signal and enhancement characteristics is essential.⁴¹ Staging and resectability of tumors (such as hepatoblastomas) require delineation of anatomic boundaries, lymph node involvement, vascular invasion, and delineation of the biliary tree, according to accepted staging systems such as PRETEXT (PRETreatment tumor EXTension) outlined by the International Childhood Liver Tumor Strategy Group.^42,43 Biliary and pancreatic diseases also are well assessed by MRI.^44,45 Common indications include cholelithiasis, pancreatitis, sclerosing cholangitis,^46,47 ductal plate malformations and choledochal cysts,^48–50 and biliary complications of liver transplantation. In the case of liver transplantation, assessment of vascular complications often is essential. Diffuse liver disease, such as fibrosis, steatosis,^51–53 and iron deposition, can be quantified by MRI. Fibrosis has been quantitatively assessed by elastography,⁵⁴ as well as qualitatively by T2-weighted imaging and delayed contrast enhancement.⁵⁵ Although steatosis can be assessed by spectroscopic methods,⁵⁶ more commonly steatosis, as well as iron deposition, are assessed by multi-echo gradient echo imaging.^53,57 MR enterography is most commonly performed for evaluation of inflammatory bowel disease.^58–60 The goals of MR include detection of bowel inflammation, distinction of active inflammation from chronic fibrosing disease, and fistula/abscess detection and characterization. Fistulography, particularly for fistula in ano, is performed well by MRI.⁶¹ The goals of the examination include detection of fistulae, classification (i.e., intersphincteric, transsphincteric, suprasphincteric, or extrasphincteric), and abscess detection.

Suggested protocols are provided in Table 85-2, with examples in Figures 85-9 and 85-10 and e-Figures 85-11 and 85-12 and details provided in the following sections. In general, matrix, field of view, and slice thickness should be adjusted to the patient size and thus are not emphasized in the following sections.

Table 85-2

Common Pulse Sequences for Hepatobiliary and Bowel Magnetic Resonance Imaging

CUBE, three-dimensional turbo spin echo (TSE) with variable flip angle (GE Healthcare); FFE, fast field echo; FIESTA, fast imaging with steady-state acquisition; FSE, fast spin echo; HASTE, half-Fourier acquisition single shot TSE; LAVA, volume interpolated gradient refocused echo (GRE) (GE Healthcare); MRA, magnetic resonance angiography; MRCP, magnetic resonance cholangiopancreatography; SPACE, three-dimensional TSE with variable flip angle (Siemens); SPGR, spoiled gradient; SSFSE, single shot fast spin echo; SSH-TSE, single shot TSE; TrueFISP, true fast imaging with steady-state precession (Siemens); TSE, turbo spin echo; VIBE, volume interpolated GRE (Siemens).

Figure 85-9 Volumetric, diffusion, and angiographic imaging of a 1-year-old girl with a left lobe lateral segment liver transplant.
A, A localizer is performed quickly. B, Axial fat-suppressed T2-weighted imaging shows lymph nodes (solid arrow) and surrounding mesenteric vessels (dashed arrow), suggesting posttransplant lymphoproliferative disorder. C, Maximum intensity projection of a volumetric T2-weighted image showing patent bile ducts (arrow). D, Axial diffusion again showing lymph nodes (arrow). E, A contrast-enhanced angiogram (spoiled gradient with fat suppression) in the arterial phase showing a patent hepatic artery (black arrow). F, Venous phase imaging shows the proximal portal vein (small arrow), poststenotic dilation of the intrahepatic portal vein (dashed arrow), hemorrhoidal varices (medium arrows), and splenomegaly (large arrow).

Figure 85-10 Magnetic resonance enterography. A 20-year-old girl with typical findings of Crohn disease.
A, Single shot imaging shows a thickened terminal ileum with submucosal edema (arrows). B, Balanced steady state imaging shows wall thickening (arrow). C, Hyperemia and early transmural enhancement is seen after contrast administration, as well as a “comb sign” (dashed arrow). D, Delayed imaging shows persistent enhancement of the terminal ileum (arrow).

e-Figure 85-11 Diffusion and contrast enhancement in a 16-year-old girl with ulcerative colitis.
A, Diffusion highlights colonic wall (arrow) as well as a small pericolonic lymph node (dashed arrow). B, Contrast-enhanced spoiled gradient shows cecal mucosal enhancement (arrow) and prominent adjacent vessels (dashed arrow). C, Enhancement is pancolonic; mucosal enhancement is denoted by the signal intensity difference between the luminal (arrow) and serosal (dashed arrow) portions of the bowel.

e-Figure 85-12 Balanced steady state and small bowel imaging in a 14-year-old boy with jejunitis as an atypical manifestation of Crohn disease.
A, Jejunal wall thickening (arrow) and prominent mesenteric lymph nodes (dashed arrow) on balanced steady-state imaging. B, Submucosal edema (arrows) on single shot imaging. C, Contrast-enhanced spoiled gradient echo with fat suppression immediately after intravenous administration of glucagon shows hyperemic bowel and again demonstrates thickened folds (arrows).

Localizers

At 1.5 T, three-plane single shot fast spin echo with low bandwidth (20 kHz), matrix, and repetition time are performed as localizers. If diagnostic single shot images are required, adjustments can be made to the matrix (320 × 256), slice thickness (4 to 5 mm), and echo time (TE) (200 ms) and images acquired with respiratory triggering.⁶² At 3 T, localizers may be more efficiently achieved with balanced steady-state imaging to decrease specific absorption rate, especially for small patients. All localizer sequences should employ parallel imaging.

Conventional T2-Weighted Imaging

Conventional T2-weighted imaging can be performed with fast spin echo. Here parallel imaging is not optimal, because the primary emphasis is on tissue characterization with high signal/noise ratio (SNR). Sequences should be performed at least in the axial plane, and respiratory triggering or navigation is significantly helpful in improving image quality.63,64 Typical TEs are 80 to 90 ms at 1.5 T and 70 to 80 ms at 3 T. Fast recovery may be used to improve SNR. Current literature is mixed on the ability of T2-weighted single shot images to provide an equivalent alternative to longer conventional sequences,⁶⁵ and thus protocols almost always include conventional T2-weighted imaging.

Volumetric T2-Weighted Imaging

Volumetric T2-weighted sequences have been described mostly for musculoskeletal imaging and neuroimaging.66–68 These pulse sequences are similar to fast spin echo in that a 90° excitation is required, which excites a slab rather than a slice. These sequences permit thin slices (1 to 2 mm) and reformatting in arbitrary planes. Parallel imaging is essential to maintain a reasonable scan time of 4 to 5 minutes. Navigation or respiratory triggering is important in optimizing image quality, along with a high bandwidth (e.g., 62 kHz). For dedicated MR with cholangiopancreatography examinations, a higher TE (>500 ms) can be used to permit excellent maximum intensity reformation of the biliary tree and pancreatic duct. For other examinations, a TE of 70 to 90 ms permits delineation of relevant anatomy.

Single Shot Imaging

Although volumetric imaging usually displays the bile ducts well,⁶⁹ image quality likely will be suboptimal in patients with voluntary or involuntary motion, such as irregular breathing and peristalsis. Thus single shot imaging is complementary⁷⁰ for biliary imaging and essential for bowel imaging. In patients with irregular breathing, the user can hand trigger acquisition of each slice by observing the respiratory belt tracing, enabling reasonable assessment of bile ducts in even the most challenging patients.

T1-Weighted Imaging: Dual echo imaging (spoiled gradient with both in-phase and opposed-phase echoes) offers volumetric images at high spatial resolution, as well as with fat-water separation. These sequences can be performed while the patient hold his or her breath, and data can be processed to yield four image sets from one acquisition: water, in-phase, opposed-phase, and fat. Typical parameters are high bandwidth (100 kHz), 4-mm slice thickness, and 12° to 15° flip angle.

Dynamic Gadolinium

Several contrast agents are now available. Consideration should be given to macrocyclic agents, such as gadobutrol (Gadavist), because they may provide enhanced safety in the setting of renal insufficiency. If the primary clinical concern is evaluation of the vascular structures, gadobenate dimeglumine (MultiHance) is a reasonable option, given its higher relaxivity and longer intravascular residence time. The agent has approximately 5% hepatobiliary elimination. Gadofoveset trisodium (Ablavar) recently has been approved by the Food and Drug Administration for assessment of aortoiliac disease in adults, and it may be an option if vascular evaluation is the sole clinical question. Finally, gadoxetate disodium (Eovist) may be a good option if the primary clinical concern is biliary.71–77 This agent has 50% hepatobiliary excretion in the setting of normal hepatic and renal function and thus provides a functional as well as an anatomic MR with cholangiopancreatography examination. This agent also is preferred in the setting of evaluation of suspected focal nodular hyperplasia. Although the literature is still limited, gadoxetate may prove useful for characterization of other liver tumors. Even though the gadoxetate dose of 0.025 mmol/kg contains only a quarter of the gadolinium of other agents, its higher T1 relaxivity still allows adequate first-pass hepatic imaging. For all agents, the rate of single dose administration at 1 mL/sec followed by a saline solution flush is acceptable.

For dynamic gadolinium studies, a 3D spoiled gradient sequence with intermittent fat suppression can be performed. Generally, if a patient can hold his or her breath, the matrix and number of slices are adjusted to make the scan time match the patient’s ability. If the patient cannot hold his or her breath, quiet breathing with a scan time of approximately 30 seconds can be used. Three phases are acquired in rapid succession followed by delayed imaging at 3 minutes after administration of contrast material.

Scan delay and choice of echo time are significantly different depending on whether the goal of the examination is tumor characterization or vascular assessment. If tumor characterization or detection of bowel wall enhancement is the primary goal, a scan delay should be set to ensure that the center of k-space is acquired at 30 seconds after half the contrast agent is administered:

For tumor characterization, the minimum full echo time should be chosen to maximize SNR.

If vascular assessment is the primary goal, a timing run can be performed with fluoroscopic triggering and centric k-space acquisition or sequential k-space acquisition and calculation of the scan delay, substituting the time to abdominal aortic peak enhancement for 30 seconds in the aforementioned equation. For vascular or bowel enhancement, the minimum echo time (i.e., a fractional echo) should be used to minimize spin dephasing artifacts from flowing blood or field inhomogeneity near the bowel wall from enteric gas, as well as to minimize scan time.

Noncontrast Magnetic Resonance Angiography

Although contrast-enhanced MR angiography (MRA) has become the dominant MR method of vascular assessment, the administration of gadolinium contrast material may be contraindicated in some patients with renal insufficiency. Additionally, the technique has little room for error because intravenous contrast material can be given only once; an injection-scan timing mismatch or patient motion during the scan cannot be rectified, which is particularly problematic in patients who require sedation. Thus noncontrast-enhanced techniques improve the reliability of MRI as a modality for assessment of abdominal vasculature.

Although time-of-flight–based approaches have been used, these sequences produce limited image quality in the abdomen. Noncontrast-enhanced techniques based on balanced steady-state approaches have been gaining favor.⁷⁸ A variation of this method is based on a respiratory-triggered inversion pulse covering the imaged volume, as well as a region inferior to it,⁷⁹ followed by a balanced steady-state echo train. Thus blood flowing from superior to the inverted region is bright, producing MRA. These sequences have pitfalls similar to time-of-flight–based techniques, including flow-related dephasing, slow-flow–related signal dropout, and intrinsic high T1-weighted signal, but on the whole they provide a nice complement to contrast-enhanced MRA.

Iron/Fat Quantification

Although MRI methods of hepatic iron quantification based on a T2-weighted signal and a composite of signals (proton density, T2-weighted, T2*-weighted, and T1-weighted) have been used, the most common approach is based solely on T2*-weighted imaging. T2*-weighted measurement can be performed on any scanner, although a dedicated pulse sequence and image reconstruction facilitates completion of the examination. These examinations generally are performed on 1.5 T scanners, because the T2*-weighted value depends on field strength, and the vast majority of the calibration literature is based on 1.5 T data.

T2* may be calculated by running a gradient echo sequence multiple times at different TE values but with a fixed repetition time of approximately 150 ms. Typically values should range from approximately 1 to 20 ms, and the user should avoid prescanning between sequences (or the transmit and receive gains may change). A region of interest then can be drawn in the same area of the liver on each of the resulting series of images. Thus mean signal at each TE is known and will demonstrate an exponential decay relationship (signal = Ae^–R2*•TE + B, where R2* = 1/T2*). R2* and hence T2* can be determined by various software packages using logarithms and linear regression.

This process for determining T2* is time-consuming for data acquisition, because the patient must hold his or her breath for each TE. An alternative is multi-echo gradient echo sequences, in which a series of gradient echoes are obtained at various TEs after each excitation. The multi-echo approach can be performed with two-dimensional (2D) acquisition of a single slice of the liver or a 3D volume acquisition. Whether 2D or 3D, the acquisition requires that the patient hold his or her breath just once, which has the advantage of avoiding slice misregistration issues; it also is easier on the patient and facilitates patient throughput. When coupled with an image reconstruction algorithm that performs a fit to determine R2*, image maps are obtained. Although the pulse sequences and reconstruction programs are not widely available, all major MRI vendors are actively addressing this issue.

One of the challenges of pediatric hepatic iron determination is the wide range of T2*-weighted values that are acquired in practice. Obtaining accurate T2*-weighted values over such a wide range can be challenging. A long series of echoes with long TEs in a patient with a short T2* will have mean signal over a region of interest dominated by noise, giving an overestimation of T2*. Conversely, a series of echoes with short TEs to address this situation will yield minimal signal decay in a patient with a long T2*, again resulting in poor estimation of T2*. Thus the longest TE used to determine T2* should be based in part on the T2*. One approach is to acquire two datasets (one with a long TE and one with a short TE) and use the appropriate one.

For assessment of steatosis, qualitative evaluation may be performed by dual echo imaging (in/opposed phase gradient echo). For quantitative assessment, low flip angle multi-echo imaging may be used, or alternatively spectroscopy can be performed.80–82 In general, the gradient echo methods are faster and easier, although considerable care must be taken for quantitative accuracy to be maintained.

Nuclear Medicine

Overview

Nuclear scintigraphy plays an important role in the evaluation of hepatobiliary dysfunction and disorders of the gastrointestinal tract in infants and children. Although some of the studies offer unique diagnostic information, others provide functional information complementary to that obtained with sonography, CT, MRI, and fluoroscopy. Radionuclide imaging of the GI system can be divided into two major categories: imaging of the hepatobiliary system and spleen (Table 85-3) and imaging of the GI tract (Table 85-4).

Table 85-3

Radionuclide Imaging of the Liver, Biliary System, and Spleen

FNH, Focal nodular hyperplasia; IDA, iminodiacetic acid derivative; LFT, liver function test; RBC, red blood cell; Tc, technetium.

Table 85-4

Radionuclide Imaging of the Gastrointestinal Tract

GE, Gastroesophageal; GI, gastrointestinal; HMPAO, hexamethylpropyleneamine oxime; RBC, red blood cell; Tc, technetium; WBC, white blood cell.

Hepatobiliary scintigraphy provides an anatomic and dynamic physiologic evaluation of biliary function. Splenic scintigraphy is useful in the evaluation of splenic sequestration, accessory spleens, posttraumatic splenosis, postsurgical residual splenic tissue, and a wandering spleen with torsion. GI scintigraphy allows the dynamic assessment of swallowing, gastric emptying, gastroesophageal reflux, esophageal transit, colonic transit, and tracheobronchial aspiration. Red blood cell labeling allows the anatomic localization of intestinal bleeding, and white blood cell labeling can be used for the localization of sites of abdominal infection inflammation. Finally, ectopic mucosa scintigraphy is used for evaluation of suspected Meckel diverticulum.

Patient Preparation and Equipment Requirements

Sedation usually is not required for GI radionuclide examinations. However, correct anatomic positioning beneath the camera is essential, and some type of restraint usually is required even with cooperative children. Rarely, patient sedation is necessary for lengthy examinations.

Some of the studies discussed in this section require a period of fasting before imaging. Hepatobiliary imaging typically is performed after premedication with phenobarbital, and Meckel diverticulum imaging can be enhanced with pentagastrin (not currently available in the United States), histamine H2 blockers, or glucagon.⁸³

Standard imaging parameters vary, but imaging usually is performed with children in the supine position with either a single-head or dual-head gamma camera, using a low-energy, all-purpose collimator or, preferably, a low-energy, high-resolution collimator. The photopeak and window settings should be predetermined for technetium-99m (^99mTc) (140 keV, 15% to 20%).

Radiopharmaceutical Agents

The radionuclide used in all but positron emission tomography (PET) and molecular imaging studies is ^99mTc, which is administered either intravenously or orally, usually combined with a nonradioactive compound (pharmaceutical agent). The resulting radiopharmaceutical agent (e.g., ^99mTc sulfur colloid) is directed to the target tissue (e.g., the reticuloendothelial system of the liver, spleen, and bone marrow). Ionizing radiation from the internalized radiopharmaceutical agent emanates in the form of gamma rays that deposit energy into the imaging detector. The information sought by the clinician dictates the specific radiopharmaceutical agent used.

The dose of a radiopharmaceutical agent used in a child should be determined by the minimal dose required to yield a high-quality diagnostic examination. Administered activity typically is calculated on the basis of either body surface area or weight. Recent work by Gelfand and colleagues ⁸⁴ determined that weight-based formulas result in lower radiation exposures, especially in young infants. The group has worked with The Alliance for Radiation Safety in Pediatric Imaging (www.imagegently.org) and published weight-based consensus guidelines for recommended doses of radiopharmaceutical agents in children⁸⁴ and is supported by the Society for Pediatric Radiology, the Society of Nuclear Medicine, and the American College of Radiology.

Indications and Protocols

Hepatobiliary Scintigraphy

Hepatobiliary scintigraphy is a useful adjunct to sonography in the evaluation of infants with jaundice and cholestasis, as well as in older children with hyperbilirubinemia. Its value in jaundiced neonates is the timely differentiation of neonatal hepatitis and biliary atresia, because surgical intervention in patients with biliary atresia is most successful when it is performed early in life. Postoperatively, hepatobiliary imaging may be requested to confirm patency of the bilioenteric anastomosis. Imaging with sonography, CT, and MRI frequently is diagnostic for choledochal cysts, but hepatobiliary imaging is a useful adjunct. Similarly, the evaluation of Caroli disease often is facilitated with hepatobiliary imaging. Cystic duct patency can be assessed in older children with suspected acute cholecystitis. Hepatobiliary imaging is highly sensitive and specific for the detection of spontaneous, postoperative, or posttraumatic biliary leaks.⁸⁵ Other clinical indications are listed in Table 85-3.

Currently, ^99mTc mebrofenin is used widely for hepatobiliary imaging because it has a high hepatic extraction; it is transported into the hepatocytes and then excreted with bile into the bile ducts. The radiopharmaceutical agent is administered intravenously, and static anterior images of the abdomen are acquired in the supine position every 5 minutes for 30 minutes. A simultaneous dynamic acquisition is suggested during the first 30 minutes. Delayed images may be obtained at 45 minutes, 60 minutes, and up to 24 hours, as necessary, to visualize the gallbladder and biliary tree and excretion into the duodenum. A normal scan typically demonstrates radiopharmaceutical uptake in the liver by 5 minutes, in the biliary tree by 15 minutes, and in the small bowel by 15 to 45 minutes (Fig. 85-13). In patients with neonatal hepatitis, delayed and diminished uptake by the liver occurs, but the radiopharmaceutical agent eventually reaches the small bowel. In patients with biliary atresia, radiopharmaceutical uptake by the liver is usually adequate, but the radiotracer never reaches the bowel, even on delayed 24-hour images (e-Fig. 85-14).

Figure 85-13 Normal results of a hepatobiliary study in a 5-year-old with right upper abdominal pain.
Anterior static images of the abdomen were obtained every 5 minutes for 45 minutes after the intravenous (IV) administration of technetium (Tc)-99m mebrofenin. Homogeneous radiotracer accumulation is seen throughout the liver, with prompt visualization of the intrahepatic ducts by 5 minutes, the common bile duct and gallbladder by 10 minutes, and the small bowel by 25 minutes.

e-Figure 85-14 Biliary atresia in an 8-week-old infant with jaundice and conjugated hyperbilirubinemia.
Technetium (Tc)-99m mebrofenin biliary scintigraphy demonstrates relatively good hepatic extraction of the radiotracer, nonvisualization of the biliary system, and no evidence of radiotracer passage into the bowel. 0 to 60-minute (A), 4-hour (B), and 24-hour (C) anterior projection images are shown. All images reveal normal, faint radiopharmaceutical agent accumulation in the urinary bladder.

When differentiation between biliary atresia and neonatal hepatis is required, premedication with phenobarbital (5 mg/kg per day for 5 days) improves hepatic extraction of tracer,⁸⁶ and 3 to 5 days of premedication is recommended in the setting of elevated conjugated hyperbilirubinemia. A child should fast for 2 to 4 hours; 2 hours is sufficient for infants. Prolonged fasting is not recommended.

When the possibility of cholecystitis is a concern, the same protocol is used without phenobarbital premedication but with continuous imaging for 1 hour. When the gallbladder is visualized, a right anterior oblique or right lateral view is acquired. Delayed images at 2, 4, and 24 hours may be necessary. If the gallbladder is not visualized at 45 to 60 minutes, in lieu of further delayed imaging, intravenous morphine sulfate (0.04 mg/kg) may be administered, followed by dynamic 1-minute imaging for 30 minutes. If cholecystokinin cholescintigraphy is requested, 0.02 µg/kg of cholecystokinin (Sincalide or Kinevac) is administered as a slow infusion over 30 minutes after the gallbladder is visualized. A normal gallbladder ejection fraction is >35%. When leakage of bile is a concern, the same protocol is used, without administration of phenobarbital.

A study with normal findings reveals radiotracer in the stomach on the initial images, followed by progressive emptying of the stomach (see Fig. 85-15). Computer processing is performed by drawing a region of interest around the stomach at selected time points to calculate fractional emptying. A time-activity curve is then generated and plotted on a linear scale using the geometric mean of the anterior and posterior counts. A half-time for emptying (T_½) is calculated, that is, the length of time required for the initial number of counts to decrease by 50%. In normal studies the time-activity curve should exhibit a continuous decline in activity over time.

Reported normal values are variable. Seibert and colleagues ⁸⁷ showed 60-minute gastric emptying values of 48% ± 16% in infants and 51% ± 7% in children who were fed a radiolabeled milk formula (roughly, a gastric emptying T^½ of 60 minutes in each age group). A different study performed with healthy infants who were fed radiolabeled milk demonstrated a T_½ of 87 ± 29 minutes.⁸⁸ Singh et al.⁸⁹ developed a standard solid meal (^99mTc-labeled “Technecrispy cake”), determined a standard meal volume (30 g), and prospectively established normal gastric emptying values in healthy children, which typically are normalized by each individual nuclear medicine laboratory or imaging department.

Esophageal Transit Scintigraphy

Esophageal transit scintigraphy may be used to evaluate esophageal motility disorders such as achalasia, diffuse esophageal spasm, and impaired motility from esophagitis and esophageal atresia/tracheoesophageal fistula.

Patient preparation requires fasting for 2 to 4 hours. The child is positioned supine with the mouth at the very top of the field of view and the stomach at the lower field of view, A small radioactive marker may be placed over the cricoid cartilage for an anatomic reference. The child is instructed to swallow a ^99mTc sulfur colloid radiolabeled bolus of liquid (water or milk) as computer acquisition begins. A practice swallow of 5 mL of unlabeled water is recommended. Posterior images are acquired over the mouth and stomach at 5 seconds/frame for 1 to 3 minutes. Regions of interest are drawn around the upper, middle, and lower thirds of the esophagus and stomach, and a time-activity curve is generated. The normal transit time through the esophagus is generally less than 10 seconds.

Colonic Transit Scintigraphy

Colonic transit scintigraphy is useful in patients with chronic constipation to establish whether the cause is slow colonic transit or functional fecal retention. Differentiation between these two patterns is important because the treatment strategy is different for each.

Patient preparation requires discontinuance of laxatives for 5 days before the transit study, and 4 hours of patient fasting is recommended. A ^99mTc sulfur colloid radiolabeled liquid or solid meal is administered orally. Anterior and posterior images are obtained in the supine position at 0 to 2, 6, 24, 30, and 48 hours. Radioactivity is measured in six regions (precolonic, ascending, transverse, descending, rectosigmoid colon, and evacuated feces). An examination with normal results shows radioactivity in the cecum by 6 hours and evacuation by 30 to 58 hours. Retention in the proximal colon at 48 hours indicates “slow colonic transit,” whereas retention in the rectum at 48 hours indicates “functional fecal retention.”90,91 Quantitative assessment of transit also can be performed with geometric center analysis.⁹²

Salivagram

The salivagram allows the dynamic assessment of swallowing in patients suspected of having primary aspiration with feeding.

Fasting is not required. The child is positioned supine, and a small dose of ^99mTc sulfur colloid mixed in 0.1 to 0.5 mL of water or saline solution is placed on the anterior tongue and allowed to mix with oral secretions. Rapid dynamic images of the neck, chest, and upper abdomen are acquired from the posterior projection every 60 seconds for 1 hour. Static images are then obtained at 1 hour and 3 hours. Detection of any radiopharmaceutical agent in the tracheobronchial tree is abnormal (Fig. 85-16).

Figure 85-16 A salivagram.
A, A radionuclide salivagram (posterior images) obtained in a child with recurrent pneumonia and neuromuscular incoordination demonstrates technetium-99m sulfur colloid (TcSC) outlining the tracheobronchial tree. Some of the swallowed radiopharmaceutical agent is seen in the stomach. B, A chest radiograph obtained the same day reveals bibasilar lung opacity, consistent with aspiration pneumonitis.

Meckel Scintigraphy

Meckel diverticulum is a congenital anomaly resulting from incomplete closure of the omphalomesenteric duct; it is present in approximately 2% of the population.⁹³ Most of these diverticula are asymptomatic and are lined by ileal mucosa. However, those that contain ectopic gastric mucosa are capable of producing hydrochloric acid and pepsin, thereby inducing mucosal ulceration. Peptic ulceration causes acute GI hemorrhage in children, usually in the first 2 years of life. A Meckel diverticulum can be quite small and difficult to differentiate from bowel loops by anatomic imaging. The Meckel scan is the study of choice for unexplained GI bleeding in children. Those diverticula with ectopic gastric mucosa can be detected with scintigraphy because the radiopharmaceutical agent accumulates within the ectopic gastric mucosa. Bleeding enteric duplication cysts containing ectopic gastric mucosa also can be detected.

Pharmacologic pretreatment is not considered necessary to produce a high-quality Meckel scan. However, if a false-negative study is suspected, a follow-up examination with pentagastrin or an H2 blocker can increase diagnostic sensitivity. Pentagastrin stimulates gastric secretions and increases gastric mucosa uptake of the pertechnetate. It also stimulates secretion of pertechnetate and GI motility. Histamine H2 blockers (e.g., cimetidine, ranitidine, and famotidine) block secretions from the cells and increase gastric mucosa uptake. Glucagon may be given to decrease intestinal peristalsis.⁹⁴

The child is placed in the supine position under the gamma camera and receives an intravenous injection of ^99mTc pertechnetate. Multiple static anterior images of the abdomen are acquired for 30 minutes. Dynamic images also are obtained at 60 seconds per image. The images are processed to allow cinematic display of the dynamic images and summed 5-minute static images.

A study with normal findings reveals radiotracer accumulation in the stomach and the urinary bladder, sometimes with faint uptake by the kidneys. An abnormal study demonstrates accumulation of radiotracer in ectopic tissue simultaneously with the appearance of gastric mucosa, usually between 6 and 10 minutes after injection. The abnormality is usually a small, rounded focus of radiopharmaceutical agent uptake in the right lower abdomen (e-Fig. 85-17). Single photon emission tomography (SPECT) imaging co-registered with simultaneously acquired low-dose CT (SPECT/CT) on a hybrid scanner has proved useful in discriminating between a Meckel diverticulum and possible artifact from urinary or other visualized activity.^95,96

Gastrointestinal Bleeding Scintigraphy

Lower GI bleeding in infants and children has many causes that are largely age-specific. When entities such as volvulus and intussusception are excluded, radionuclide imaging is useful for localizing a bleeding source. Radiolabeled red blood cells are useful in patients who are bleeding slowly or intermittently.

No patient preparation is necessary. A sample of the patient’s blood is obtained (1 to 3 mL), anticoagulated, and labeled with ^99mTc. The tagged autologous cells subsequently are reinjected into the patient. The child is placed supine, with the imaging camera positioned anteriorly over the abdomen. After a bolus intravenous injection of ^99mTc-labeled red blood cells, a dynamic flow study is performed for 1 minute, followed by a static image obtained every 5 minutes for 30 minutes. If the images do not immediately reveal bleeding, delayed images are obtained at 45 minutes and 60 minutes; they can be obtained up to 24 hours after injection. A cinematic display also is processed when bleeding is detected. A positive study reveals activity in a location and configuration consistent with bowel, displays progressively increased intensity over time, and exhibits movement through the GI tract over time.

Gastrointestinal Inflammation and Infection Scintigraphy

The radionuclide labeling of white blood cells can be used for the localization of sites of abdominal infection, including appendicitis,⁹⁷ and for the localization and evaluation of the intensity and extent of inflammatory bowel disease (Crohn disease and ulcerative colitis).

Data suggest that ^99mTc hexamethylpropyleneamine oxime–labeled leukocyte imaging may be superior to CT ⁹⁸ for assessing the extent and activity of inflammatory bowel disease, and a review of these techniques has been published more recently.⁹⁹ Labeled leukocyte imaging may be used as a screening study to determine if a child should undergo more invasive testing. The study also is useful for patients who refuse endoscopy or contrast radiography or for those who have luminal narrowing that precludes endoscopic evaluation.

The radiopharmaceutical agent used is ^99mTc hexamethylpropyleneamine oxime. In children, the minimal amount of blood needed for labeling is 10 to 15 mL, depending on the child’s size and circulating leukocyte count. Images are acquired at 30 minutes and 2 to 3 hours after the injection of the labeled white blood cells. The patient should void before each image acquisition. SPECT imaging may be performed after planar imaging for improved localization of the disease. A study with normal findings reveals uptake of the radiolabeled leukocytes by the liver, spleen, bone marrow, kidneys, and urinary bladder. Normal bowel activity due to hepatobiliary excretion is seen in 20% to 30% of children by 1 hour. Abnormal bowel activity may be seen in 15 to 30 minutes and usually increases in intensity during the next 2 to 3 hours.

Positron Emission Tomography

PET using fluorine-18 (¹⁸F) fluorodeoxyglucose has been reported as a noninvasive, sensitive alternative to conventional studies in the identification and localization of intestinal inflammatory and infectious processes in children. Specific inflammatory conditions investigated using ¹⁸F-fluorodeoxyglucose PET imaging include inflammatory bowel disease,^100–102 chronic granulomatous disease,³¹ appendicitis,¹⁰³ and fever of unknown origin.¹⁰⁴

PET imaging using ¹⁸F-dihydroxyphenylalanine has been used to distinguish focal from diffuse pancreatic disease in infants with hyperinsulism. Patients with focal disease may undergo resection of the offending adenoma, whereas those with diffuse disease may be treated with octreotide or subtotal pancreatectomy.^105,106 The imaging techniques for these PET studies are beyond the scope of this chapter.