ABDOMINAL IMAGING 1767 MR Imaging in Gastrointestinal

1767ABDOMINAL IMAGING

Jaroslaw N. Tkacz, MD • Stephan A. Anderson, MD • Jorge Soto, MD

Accurate and rapid diagnostic imaging is essential for the appropriate management of acute gastrointestinal conditions. Computed tomog-raphy (CT) is the modality most often used in this setting because of its widespread availability and the relative speed, ease, and uniformity with which evaluations can be performed. CT allows the diagnosis of a wide spectrum of acute gastrointestinal diseases with the adjustment of only a few variables in the acquisition protocol. For example, the contrast material volume, injection rate, and delay before image acqui-sition can be manipulated to enhance vascular or organ-specific con-trast for myriad gastrointestinal diagnoses. Magnetic resonance (MR) imaging has similarly robust potential, although its integration into the acute care setting requires greater technical and logistical effort. Im-proved MR imaging sequences, advances in coil technology, stream-lined imaging protocols, and increased technical and professional familiarity with the modality make it an increasingly attractive option when there is concern about patient radiation exposure or allergy to iodinated contrast material. A variety of acute abdominal conditions, including pancreatic and biliary tract trauma, choledocholithiasis, gallbladder disease, acute pancreatitis, and appendicitis can be rapidly and accurately demonstrated with MR imaging. MR imaging also can play a vital role in the follow-up assessment of treatment response and in the diagnosis of indeterminate findings at CT or ultrasonography. Nevertheless, incompatibility of patient monitoring devices with the MR magnet, lack of MR imaging system availability, and the acuity of illness may limit the use of the modality.©RSNA, 2009 • radiographics.rsna.org

MR Imaging in Gastrointestinal Emergencies1

Abbreviations: ERCP = endoscopic retrograde cholangiopancreatography, GRE = gradient-recalled echo, MIP = maximum intensity projection, SE = spin echo, SPIR = spectral presaturation with inversion recovery, 3D = three-dimensional

RadioGraphics 2009; 29:1767–1780 • Published online 10.1148/rg.296095509 • Content Codes: 1From the Department of Radiology, Boston University Medical Center, 820 Harrison Ave, 3rd Floor, Boston, MA 02118. Received March 9, 2009; revision requested April 13 and received June 8; accepted June 17. J.S. is a research consultant for General Electric, Bracco Group, and Bayer; all other authors have no financial relationships to disclose. Address correspondence to J.N.T. (e-mail: [email protected]).

©RSNA, 2009

See last page

TEACHING POINTS

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

1768 October Special Issue 2009 radiographics.rsna.org

IntroductionThe number of emergency department visits has increased substantially in the past decade (1–3). Increased patient volume has led to an acceler-ated effort to rapidly and accurately diagnose acute gastrointestinal disease to allow appropriate triage and management. The spectrum of acute gastrointestinal disease is broad and is gener-ally subdivided according to whether the cause of disease is traumatic or nontraumatic. The use of computed tomography (CT) in examining hemodynamically stable patients with traumatic injuries is well established (4,5). By contrast, the higher cost, increased acquisition time, and limited availability of magnetic resonance (MR) imaging have been major deterrents to its use for assessing acute gastrointestinal conditions, whether the cause is traumatic or nontraumatic. However, certain advantages exist with MR imag-ing in some patient subgroups (6). For example, MR imaging is particularly useful for evaluating pregnant patients who have symptoms of appen-dicitis (7). MR imaging also can be used to fur-ther delineate or follow up equivocal CT findings in patients with multiple traumatic injuries (eg, pancreatic and biliary duct disruptions) and to monitor previously diagnosed acute gastrointesti-nal disease, especially in young patients, in whom radiation exposure is of concern.

The article surveys the current indications for the use of MR imaging in patients with acute gastrointestinal diseases of both traumatic and nontraumatic causes. MR imaging protocols are discussed in detail, with particular attention to focused and rapid image acquisition techniques. The MR imaging appearances of various acute gastrointestinal disorders are described and, where appropriate, compared with CT findings.

MR Imaging ProtocolsMR imaging protocols in the acute care setting should be tailored to concisely address diagnostic considerations, and the sequences should be cho-sen to minimize acquisition time. Negative oral contrast agents are helpful in improving the spec-

ificity of MR cholangiopancreatography and MR enterographic studies, but their use is optional and depends on the stability of the patient. At our institution, we use a 50/50 vol/vol combination of ferumoxsil (Gastromark; Mallinckrodt Medical, St Louis, Mo) and barium sulfate suspension. The use of intravenous gadolinium-based con-trast agents also is optional because many acute gastrointestinal conditions involve inflammation, edema, and structural derangements that can be demonstrated or inferred by combining fluid-sen-sitive sequences with three-dimensional (3D) vol-umetric techniques for anatomic correlation. We have found utility in intravenously administered gadoxetate disodium (Eovist; Bayer Healthcare Pharmaceuticals, Wayne, NJ), which has a partial hepatobiliary excretion route. The contrast agent is administered in a dose of 0.1 mL per kilogram of body weight and at a rate of 2 mL/sec, with an imaging delay sufficient to allow evaluation of bile duct injuries and bile leaks.

In our experience, a free breathing or re-spiratory triggering technique is preferable to breath-hold protocols in the acute setting because patients often have difficulty with breath hold-ing, and respiratory motion may compromise image quality. Single-excitation half-Fourier T2-weighted turbo spin-echo (SE) sequences and two-dimensional T1-weighted gradient-recalled echo (GRE) sequences may be applied with either free breathing or respiratory triggering. Useful fluid-sensitive sequences include short inversion time inversion-recovery, or STIR, and respiratory-triggered T2-weighted chemically fat-suppressed turbo SE techniques. Coherently balanced steady-state sequences provide clear depiction of the abdominal and pelvic anatomy and may be used as a white-blood imaging technique for evalua-tion of the major vessels without requiring the use of a gadolinium-based contrast material. Three-dimensional isotropic volumetric T1-weighted se-quences (variously named THRIVE, VIBE, LAVA, and FAME by their respective vendors) require a typical breath hold of 15–20 seconds for coverage of a single organ. Although that requirement may limit their usefulness in the acute setting, these sequences produce high-resolution 2–3-mm-thick interpolated sections and allow the acquisition of

TeachingPoint

RG ■ Volume 29 • Number 6 Tkacz et al 1769

dynamic phase images after the administration of an intravenous gadolinium-based contrast agent (rate, 2 mL/sec; volume, 20 mL).

Our MR imaging protocol for emergent evaluation of the abdomen and pelvis includes the following sequences: axial breath-hold in-phase and out-of-phase T1-weighted GRE, axial and coronal breath-hold T2-weighted single-shot turbo SE, axial T2-weighted fat-suppressed turbo SE with respiratory triggering (or short inversion time inversion recovery), and steady-state free precession. Thick-slab MR cholan-giopancreatographic, respiratory-triggered T2-weighted turbo SE, and volumetric T1-weighted sequences are added as necessitated by the clinical indication and with consideration of the total acquisition time and the patient’s ability to tolerate breath-hold acquisitions (Table). The duration of the MR imaging examination is typi-cally 20–30 minutes (6,8).

Clinical Applications

Traumatic Injuries to the Pancreas and Biliary TractVisceral injuries in patients with blunt abdominal trauma most commonly involve the liver, spleen, and kidneys (9). Injuries to these organs usually can be diagnosed without difficulty at CT. By con-trast, injuries to the pancreas and biliary tract may be more subtle and may be overlooked in patients with trauma involving multiple organs. Although pancreatic and biliary injuries are uncommon, delays in their diagnosis are associated with high morbidity and mortality, and their early diagnosis

MR Imaging Protocols for the Evaluation of Gastrointestinal Emergencies

Parameter

T1-weighted In-Phase and Out-of-Phase

T2-weighted Turbo SE

T2-weighted Single-Shot Turbo SE

Balanced Turbo FE

MR Cholangio-pancreatography THRIVE

FOV (mm2) 40 40 40 39 30 40Scanning tech-

niqueFast FE SE SE Fast FE SE Fast FE

Fast imaging mode

None Turbo SE Turbo SE, half Fourier

Turbo FE Turbo SE Turbo FE

Scanning mode

Multisection, dual echo

Multisection Multisection 2D

Multisec-tion 2D

Multisection 2D or 3D

3D

TR (msec) 300 2000 550 3.6 8000 3.6TE (msec) 2.3/4.6 80 80 1.8 800 1.7Section thick-

ness (mm)5 5 5 7 40 for 2D, 0.8

for 3D2

Flip angle (de-grees)

90 90 90 60 90 15

SENSE reduc-tion factor

1.8 2 2 2 2 1.7

Respiration control tech-nique

Respiratory triggering



Free breath-ing

Breath holding or respiratory triggering

Breath holding

Fat suppression technique

None SPIR SPIR or none None None SPIR

Note.—All sequences listed are available on a commercial 1.5-T MR imaging system (Achieva; Philips Health-care, Andover, Mass). FE = field echo, FOV = field of view, SENSE = sensitivity encoding, SPIR = spectral presaturation with inversion recovery, TE = echo time, THRIVE = T1-weighted high-resolution isotropic volume excitation, TR = repetition time, 2D = two-dimensional.


The forces necessary to sustain a pancreatic injury are commonly high enough to impact adjacent structures such as the stomach, liver, duodenum, spleen, and vasculature. Indeed, isolated pancre-atic injuries are rare; three or more organs are usu-ally involved. The pancreatic body or neck is most commonly injured (65%), followed by the head or, with equal frequency, the tail (12). Injuries to the pancreatic head, which is in close proximity to the inferior vena cava, superior mesenteric vein, and portal vein, are more commonly fatal because of acute hemorrhage (13,14).

Figure 1. Pancreatic laceration in a patient with blunt abdominal trauma. (a) Axial contrast-enhanced CT scan demonstrates a peripancreatic region with intermediate attenuation characteristic of fluid and a focus of abnormal hypoattenuation in the pancreas (arrow), findings suggestive of contusion or lacera-tion. (b) Axial unenhanced MR image obtained with a T2-weighted fat-suppressed turbo SE sequence depicts a laceration in the pancreatic neck (arrow). (c) Axial contrast-enhanced MR image obtained with a T1-weighted fat-suppressed GRE sequence shows the extent of the pancreatic injury (arrow).

other injuries, and the symptoms and clinical find-ings are often unreliable and nonspecific. MR im-aging of the pancreas and biliary tract in patients with blunt trauma to multiple organs can provide important diagnostic information (11).

Pancreatic injuries are uncommon; they occur in less than 2% of patients with blunt abdominal trauma. However, their early diagnosis is critical because delayed complications such as fistulas, abscesses, hemorrhage, and sepsis may lead to sig-nificant morbidity and mortality. The pancreas is susceptible to crush injury from impaction against the vertebral column. Common mechanisms of pancreatic trauma include bicycle injuries (in chil-dren) and automobile and motorcycle accidents.

therefore is critical (10). CT findings in pancreatic and biliary injuries may be nonspecific or over-looked because they are associated with multiple Teaching

Point


Figure 3. Pseudocyst in a patient with pancreatic duct injury. Axial MR image obtained with a T2-weighted single-shot turbo SE sequence shows a pseu-docyst (arrow), a delayed complication of previously diagnosed pancreatic duct disruption due to blunt abdominal trauma.

Figure 2. Pancreatic injury due to blunt abdominal trauma in a 42-year-old patient. (a) Axial MR image obtained with a T2-weighted single-shot turbo SE sequence demonstrates a pancreatic fluid collection that communicates with the main pancreatic duct through a ductal disruption (arrow). (b) Maximum intensity projection (MIP) image from 3D MR cholangiopancreatography shows the extent of the peripancreatic fluid collection (arrows) and depicts a normal common bile duct (arrowhead).

this initial imaging examination. The integrity of the main pancreatic duct is of critical importance in the diagnosis of pancreatic trauma. Ductal lac-eration is not typically directly visualized at CT; however, secondary CT findings such as a deep parenchymal laceration may be suggestive of duc-tal injury. In patients in whom the CT findings are not definitive but only suggestive of pancreatic trauma or in whom CT findings of pancreatic injury are present but assessment of the pancre-atic duct is needed, MR imaging may be useful. Contusions and lacerations of the glandular paren-chyma are well demonstrated with MR imaging. The depth of a pancreatic laceration, especially the distinction between a complete transection of the gland and a partial injury, also can be determined with MR imaging (Fig 1) (15). However, evalu-ation of the integrity of the main pancreatic duct remains the critical role of MR imaging (Fig 2). Disruption of the main duct is treated surgically or endoscopically with stent placement. After ductal disruption is diagnosed, monitoring for associ-ated delayed complications (eg, fistula, abscess, or pseudocyst) and treatment follow-up also can be performed with MR imaging (Fig 3). The use of

All blunt abdominal trauma patients who are hemodynamically stable routinely undergo an initial CT evaluation to determine the scope and severity of their internal injuries. Both direct and indirect signs of pancreatic injury may be seen in

TeachingPoint


sional forces. Associated hepatic (91%), splenic (54%), and duodenal (54%) injuries are common and may mask underlying biliary injuries (17). Although CT features may be suggestive of gall-bladder injury, findings are typically nonspecific, with fluid and hemorrhage appearing in and around the gallbladder. Biliary duct injuries may be suspected on the basis of the severity of liver lacerations observed on CT images but are not often directly depicted.

The use of MR cholangiopancreatography in the setting of gallbladder trauma is not well es-tablished. The MR findings are nonspecific and mimic those seen at CT, with collapse of the gall-bladder, pericholecystic fluid, and intraluminal

MR imaging instead of CT for these purposes is especially important in children and young adults, who constitute a large percentage of patients with pancreatic trauma and for whom the lifetime risk from repeated radiation exposure might well out-weigh the benefits (Fig 4).

Blunt abdominal trauma may result also in biliary tract injury, although such occurrences are rare. The gallbladder is the site most often af-fected, followed by the common bile duct and the intrahepatic ducts (16). Like pancreatic injury, biliary tract injury is associated with multiple or-gan injuries due to compressive, shearing, or tor-

Figure 4. Pancreatic duct injury in a 15-year-old pa-tient with blunt trauma from a bicycle accident. Axial T2-weighted single-shot turbo SE MR image (a) and ra-dial thick-slab MR cholangiopancreatographic image (b) demonstrate ductal disruption (straight arrow), a small associated pancreatic fluid collection (curved arrow), and a mildly dilated proximal duct (arrowhead), findings con-firmed at endoscopic retrograde cholagiopancreatography (ERCP) (c).


hemorrhage being the most common (18–20). MR cholangiopancreatography with gadoxetate disodium used as a hepatobiliary contrast agent for delayed phase imaging can depict the origin and extent of bile leaks (Fig 5). Further investi-gation is necessary to determine the optimal use of gadoxetate disodium for the evaluation of bile duct disruptions.

CholedocholithiasisCholedocholithiasis typically results from the passage of gallbladder stones through the cystic duct into the common bile duct; less commonly, it occurs with the primary formation of stones within the common bile duct. The pathophysiol-ogy of the formation of stones is complex, with bile stasis, bactibilia, pH imbalances, increased bilirubin excretion, and sludge formation playing principal roles (21). Regardless of the source of

stones, their presence in the common bile duct may lead to acute biliary obstruction with jaun-dice, fever, cholangitis, pancreatitis, and sepsis.

The possibility of coexistent choledocholithia-sis in patients with acute cholecystitis complicates diagnostic imaging and management. Choledo-cholithiasis is found in 10%–15% of the approxi-mately 600,000 cholecystectomies performed in the United States each year (22,23). The preferred imaging study for detection depends on the clinical manifestations and the level of clinical suspicion. Nearly all patients with right upper quadrant pain and suspected biliary obstruction undergo ultrasonography (US) for evaluation of the gallbladder and bile ducts. The classic findings of acute cholecystitis—including

Figure 5. Biliary duct leak in a patient with a liver laceration in segment II due to blunt abdominal trauma. (a) Axial MR image obtained with a T2-weighted fat-suppressed turbo SE sequence depicts a small fluid collec-tion (arrow). (b, c) Axial 3D MR images obtained with a T1-weighted GRE sequence at the same level as a, before (b) and 20 hours after (c) the administration of gadoxetate disodium, show the accumulation of the par-tially hepatically excreted T1-hyperintense contrast agent in the fluid collection (arrow), a finding indicative of bile leakage.


MR cholangiopancreatography has shown con-sistently high sensitivity and specificity (97%–99% and 95%–99%, respectively) for the detection of bile duct abnormalities (28). It may play a pivotal role in the management of patients with chole-cystitis in whom choledocholithiasis is suspected, because it can help avoid diagnostic procedures that are unnecessarily invasive in the absence of common bile duct stones (eg, ERCP) or help di-rect surgical management in the presence of stones (Fig 6). Complications of acute cholecystitis, such as gangrenous cholecystitis, can be readily de-picted with MR imaging (Fig 7). Impacted cystic duct stones can be demonstrated with MR chol-angiopancreatography, enabling the diagnosis of Mirizzi syndrome (Fig 8).

Figure 6. Cholecystitis and choledocholithiasis. (a, b) Axial image from T2-weighted single-shot turbo SE (a) and axial subtraction image from con-trast-enhanced 3D T1-weighted GRE (b) MR ac-quisitions demonstrate a distended gallbladder with wall thickening (arrow in a) and hyperemic change in liver segment V (arrow in b), findings indicative of acute cholecystitis. (c) MIP image from 3D MR cholangiopancreatography shows a stone in the dis-tal common bile duct (arrow).

gallbladder luminal distention (>5 cm transverse diameter), wall thickening (>3 mm), hyperemia, and stones—can be depicted accurately by US, which has a reasonably high sensitivity and specificity (24). An abnormal caliber (>8 mm) of the common bile duct is suggestive of choledo-cholithiasis. Measurement of the ductal caliber usually can be performed accurately at the level of the porta hepatis but is more difficult at the suprapancreatic or intrapancreatic level, mostly because of overlying duodenal or antral bowel gas or patient body habitus (25). CT is often used to evaluate patients with nonspecific abdominal pain; however, standard abdominal CT tech-niques have inherent limitations for detecting common bile duct stones (26,27). Therefore, MR cholangiopancreatography is typically performed as a follow-up study when clinical suspicion of the presence of choledocholithiasis is high or CT findings are negative or equivocal and clinical suspicion is not allayed.

TeachingPoint


Figure 7. Gangrenous cholecystitis. Axial T1-weighted in-phase (a) and out-of-phase (b) MR images show an area of low signal intensity that produces a blooming artifact in the gallbladder wall (arrow), a finding suggestive of gas. Gangrenous cholecystitis was found in a surgical specimen.

Figure 8. Mirizzi syndrome in a 45-year-old man with right upper quadrant pain and jaundice. (a) Radial thick-slab MR cholangiopancreatographic image demonstrates multiple gallbladder stones (arrowhead) with dilatation of the intrahepatic ducts and common hepatic duct (short arrow) but a normal appearance of the common bile duct (curved arrow). A filling defect is visible at the insertion of the cystic duct (long arrow). (b) Axial T2-weighted single-shot turbo SE MR image obtained at the level of the cystic duct insertion depicts an impacted stone (arrow) obstructing the common hepatic duct. (c) ERCP image helps confirm the diagnosis of obstruction of the common hepatic duct by a stone (arrow).


Figure 9. Acute pancreatitis due to choledocholithiasis in a 37-year-old man with severe pain and elevated serum levels of amylase and lipase. (a) Respiratory-triggered T2-weighted fat-suppressed turbo SE image of the abdomen depicts pancreatic edema and peripancreatic fluid (arrows), findings suggestive of pancreatitis. (b) Coronal MIP im-age from 3D MR cholangiopancreatography clearly depicts a stone (arrow) lodged in the distal common bile duct.

Gallstones are implicated in 50% of patients with acute pancreatitis (29). Acute pancreatitis may occur when stones migrate into the distal common bile duct and obstruct the ampulla of Vater. The management of gallstone-induced pancreatitis involves endoscopic sphincterotomy and retrieval of any impacted stones. Although the technique is highly successful when per-formed by an experienced person, complications may occur in as many as 10% of cases and may include cholangitis (2%), hemorrhage (1%), pan-creatitis (2%), and duodenal injury (1%) (30). Therefore, the judicious use of ERCP should be a foremost consideration, whereas noninvasive imaging is typically performed when the presence of gallstone-induced pancreatitis is suspected.

The clinical manifestations of acute pancreatitis differ from those of acute cholecystitis. Patients with acute pancreatitis more commonly experi-ence severe and persistent epigastric pain with nausea and vomiting and have elevated serum levels of amylase and lipase. CT is helpful for the initial diagnosis in equivocal cases or for early as-sessment of disease severity with use of the clini-cally validated CT severity index (31). The utility

of MR cholangiopancreatography has proved comparable to that of contrast-enhanced CT for early assessment of acute pancreatitis and deter-mination of disease severity (26). Fluid-sensitive sequences with fat suppression can clearly depict early signs of pancreatitis, including glandular enlargement (edema), periglandular inflamma-tory changes, and intra- and peripancreatic fluid collections. The added value of MR cholangio-pancreatography lies in its ability to help identify choledocholithiasis as the cause of pancreatitis and thus help direct interventional care (Fig 9). Other potential complications of acute pancreatitis that may be readily diagnosed with MR cholangiopan-creatography include hemorrhagic pancreatitis, a potentially fatal complication (32). Hemorrhagic pancreatitis occurs when severe inflammation and regional necrosis cause major vessel erosion (with or without pseudoaneurysm formation) with resul-tant massive bleeding into the pancreatic bed, ret-roperitoneum, gastrointestinal tract, or peritoneal cavity; the condition requires immediate surgical intervention. Standard T1-weighted sequences de-pict hemorrhage as a region of high signal intensity within the pancreas or adjacent structures (Fig 10). Regardless of the cause of acute pancreatitis, MR cholangiopancreatography is useful for moni-toring and follow-up.

TeachingPoint


Figure 10. Hemorrhagic pancreatitis in a 53-year-old woman with severe epigastric pain and a low hematocrit level. (a) Axial T1-weighted GRE MR image shows high-signal-intensity methe-moglobin within an enlarged pancreas (arrow), findings suggestive of hemorrhagic pancreatitis. (b) Axial T2-weighted fat-suppressed turbo SE MR image obtained at the same level as a demon-strates peripancreatic edema and inflammatory change (arrow).

Gastrointestinal Tract EmergenciesMR imaging of the gastrointestinal tract poses several challenges: The selection of an appropri-ate sequence for the specific indication is impor-tant to obtain diagnostically useful images; for example, bowel peristalsis causes image blurring if a 3D volumetric technique is used. Moreover, underdistention of the small bowel with oral contrast agents can lead to decreased sensitivity for the detection of small-bowel abnormalities (33). On the other hand, MR imaging also pro-vides the advantages of excellent contrast resolu-tion and lack of ionizing radiation. It has proven value for several indications in the diagnosis of gastrointestinal disease, and it has become the imaging modality of choice for the evaluation of pregnant patients in whom the presence of acute appendicitis is suspected (34).

Acute appendicitis is the most common gastro-intestinal emergency requiring surgery in pregnant patients. The classic manifestations of appendici-tis (leukocytosis, fever, and right lower quadrant pain) are nonspecific during pregnancy, and leu-kocytosis may be a normal physiologic condition in this setting. In addition, the appendix in a preg-nant woman is typically displaced from its usual

location in the right lower quadrant by the gravid uterus, which may make it more difficult to iden-tify the appendix as the source of pain (34).

Current American College of Radiology (ACR) appropriateness criteria for the evaluation of preg-nant patients with leukocytosis and fever recom-mend the use of targeted right-lower-quadrant US with graded compression as the first-line imaging modality. MR imaging is second, followed by ab-dominal CT with an intravenous contrast agent (35). The goal of these recommendations is to en-able the diagnosis of appendicitis without exposing the fetus unnecessarily to ionizing radiation. Any effects of fetal exposure to high magnetic fields, rapid gradient shifts, and radiofrequency energy deposition during clinical MR imaging remain largely unknown; to our knowledge, no short- or long-term effects have yet been documented or proved (36,37). At our institution, all patients are informed of this fact, and written consent for MR imaging is routinely obtained.

In cooperation with the emergency department and obstetric-gynecologic service at our institu-tion, we developed an algorithm based on the


Figure 11. Acute appendicitis in a 27-year-old pregnant woman with acute abdominal pain and leukocytosis. (a) Axial T2-weighted single-shot turbo SE MR image shows an abnormally dilated appendix with a low-signal-intensity filling defect, findings indicative of an appendicolith (arrow). (b) Axial T2-weighted fat-suppressed single-shot turbo SE MR image depicts periappendiceal inflammation (arrow).

intravenous gadolinium- or iodine-based contrast agents for either of these examinations.

Our imaging protocol includes (a) single-shot T2-weighted turbo SE sequences applied in the three orthogonal planes with and without SPIR for fat suppression and (b) axial T1-weighted in-phase and opposed-phase GRE sequences. Respiratory triggering is used if necessary. The total MR image acquisition time is 20–30 minutes (8).

MR imaging has high reported sensitivity (97%–100%) and specificity (92%–93%) for the diagnosis of acute appendicitis (8). The anatomic imaging features of acute appendicitis are similar regardless of the modality used and include a cross-sectional appendiceal diameter of more than 7 mm and wall thickness of more than 2 mm. Single-shot turbo SE sequences are particularly helpful for identifying the location of the appen-dix. Edema and inflammation appear as T2 signal hyperintensity within the wall or surrounding the appendix; fat-suppressed single-shot sequences are useful for depicting these features (Fig 11). A fluid-filled appendix is another suggestive find-ing that is readily identified on MR images. It may be difficult, with the abundance of nearby blood vessels and compression or displacement of the appendix by the gravid uterus, to delineate the blind-ended tubular structure that represents the normal appendix. T1-weighted in-phase and

ACR criteria for evaluation of potential appendici-tis in pregnant patients. All pregnant patients with leukocytosis, fever, and suspected appendicitis are initially evaluated with a directed US study. We follow the method suggested by Pedrosa et al (8) and use a negative oral contrast agent composed of 300 mL of dilute barium sulfate suspension (Redi-Cat; Bracco, Princeton, NJ) and 300 mL of ferumoxsil (Gastromark; Mallinckrodt, St Louis, Mo). The patient begins drinking this mixture in the emergency department, before transport to a US suite in the radiology department, and scanning is performed 2 hours later. It has been our experience, and early results show, that US delineation of a normal appendix in a pregnant woman is poor (38), particularly in the second and third trimesters of pregnancy. It is our assumption that the patient will likely require MR imaging of the appendix, and we try to minimize the delay between initial presentation and MR imaging. If the MR imaging results are nondiagnostic, the patient proceeds immediately to CT. If necessary, the preparation for CT can be streamlined, since the oral contrast agent given in preparation for US contains barium sulfate. We do not administer


Figure 12. Normal appendix in a pregnant patient. Axial T1-weighted in-phase (a) and out-of-phase (b) MR images demonstrate a blooming artifact, a feature that helps distinguish the normal appendix (arrow) from adjacent vessels.

out-of-phase GRE sequences can help locate the appendix when it is air filled; the appendix pro-duces a blooming artifact on in-phase images, which are acquired with a longer echo time (Fig 12). Blooming artifact also may be observed if the appendix fills with an orally administered contrast agent that contains ferumoxsil. In either instance, acute appendicitis is less likely. Some pregnant patients with MR imaging findings suggestive of appendicitis are found to have another acute intraabdominal condition, such as cholecystitis, pyelonephritis, or diverticulitis.

ConclusionsIt is technically feasible to perform rapid MR im-aging in patients with acute gastrointestinal disor-ders. Although CT and US remain the frontline diagnostic imaging methods in this setting, the superior soft-tissue contrast and lack of ionizing radiation make MR imaging a valuable and at-tractive asset for use in selected patient popula-tions. MR cholangiopancreatography can play a pivotal role in the diagnosis of pancreatic duct disruption and biliary tract trauma when CT re-sults are equivocal or nondiagnostic. MR cholan-giopancreatography is also of value for detecting choledocholithiasis in patients with symptoms of acute cholecystitis or acute pancreatitis, a di-agnosis that may lead to a change in therapeutic management. Acute pancreatitis and its poten-tially fatal complications, such as hemorrhagic

pancreatitis, are readily evaluated with MR im-aging. MR imaging is also the most appropriate imaging modality for detecting acute appendicitis in pregnant patients when US findings are non-diagnostic. Given the increasing general concern for avoiding unnecessary radiation exposure, MR imaging provides an ideal way to follow the prog-ress of acute gastrointestinal diseases.

References 1. Cunningham PJ. What accounts for differences in

the use of hospital emergency departments across the U.S. communities? Health Aff (Millwood) 2006; 25(5):w324–w336.

2. Bhargavan M, Sunshine JH. Utilization of radiol-ogy services in the United States: levels and trends in modalities, regions, and populations. Radiology 2005;234(3):824–832.

3. Smith S, Heffler SK, Calfo S, et al. National health projections through 2008. Health Care Financ Rev 1999;21(2):211–237.

4. Wolfman NT, Bechtold RE, Scharling ES, Meredith JW. Blunt upper abdominal trauma: evaluation by CT. AJR Am J Roentgenol 1992;158(3):493–501.

5. Wing VW, Federle MP, Morris JA, Jeffrey RB, Bluth R. The clinical impact of CT for blunt ab-dominal trauma. AJR Am J Roentgenol 1985;145 (6):1191–1194.

6. Pedrosa I, Rofsky NM. MR imaging in abdominal emergencies. Radiol Clin North Am 2003;41(6): 1243–1273.


7. Birchard KR, Brown MA, Hyslop WB, Firat Z, Se-melka R. MRI of acute abdominal and pelvic pain in pregnant patients. AJR Am J Roentgenol 2005;184 (2):452–458.

8. Pedrosa I, Levine D, Eyvazzadeh AD, Siewert B, Ngo L, Rofsky NM. MR imaging evaluation of acute appendicitis in pregnancy. Radiology 2006; 238(3):891–899.

9. Matthes G, Stengel D, Seifert J, Rademacher G, Mutze S, Ekkernkamp A. Blunt liver injuries in polytrauma: results from a cohort study with the regular use of whole-body helical computed to-mography. World J Surg 2003;27(10):1124–1130.

10. Bradley EL 3rd, Young PR Jr, Chang MC, et al. Di-agnosis and initial management of blunt pancreatic trauma: guidelines from a multi-institutional review. Ann Surg 1998;227(6):861–869.

11. Gupta A, Stuhlfaut JW, Fleming KW, Lucey BC, Soto JA. Blunt trauma of the pancreas and biliary tract: a multimodality imaging approach to diagno-sis. RadioGraphics 2004;24(5):1381–1395.

12. Soto JA, Alvarez O, Múnera F, Yepes NL, Sepúlveda ME, Pérez JM. Traumatic disruption of the pancre-atic duct: diagnosis with MR pancreatography. AJR Am J Roentgenol 2001;176(1):175–178.

13. Akhrass R, Yaffe MB, Brandt CP, Reigle M, Fallon WF Jr, Malangoni MA. Pancreatic trauma: a ten-year multi-institutional experience. Am Surg 1997; 63(7):598–604.

14. Lin BC, Chen RJ, Fang JF, Hsu YP, Kao JL. Man-agement of blunt major pancreatic injury. J Trauma 2004;56(4):774–778.

15. Ragozzino A, Manfredi R, Scaglione M, De Ritis R, Romano S, Rotondo A. The use of MRCP in the detection of pancreatic injuries after blunt trauma. Emerg Radiol 2003;10(1):14–18.

16. Burgess P, Fulton RL. Gallbladder and extrahepatic biliary duct injury following abdominal trauma. In-jury 1992;23(6):413–414.

17. Takishima T, Hirata M, Kataoka Y, et al. Pancreato-graphic classification of pancreatic ductal injuries caused by blunt trauma to the pancreas. J Trauma 2000;48(4):745–751; discussion 751–752.

18. Wong YC, Wang LJ, Chen RJ, Chen CJ. Magnetic resonance imaging of extrahepatic bile duct disrup-tion. Eur Radiol 2002;12(10):2488–2490.

19. Jeffrey RB Jr, Federle MP, Laing FC, Wing VW. Com-puted tomography of blunt trauma to the gallbladder. J Comput Assist Tomogr 1986;10(5):756–758.

20. Vitellas KM, El-Dieb A, Vaswani K, et al. Detection of bile duct leaks using MR cholangiography with mangafodipir trisodium (Teslascan). J Comput As-sist Tomogr 2001;25(1):102–105.

21. Wosiewitz U, Schenk J, Sabinski F, Schmack B. In-vestigations on common bile duct stones. Digestion 1983;26(1):43–52.

22. Collins C, Maguire D, Ireland A, Fitzgerald E, O’Sullivan GC. A prospective study of common bile

duct calculi in patients undergoing cholecystectomy: natural history of choledocholithiasis revisited. Ann Surg 2004;239(1):28–33.

23. Phillips EH, Toouli J, Pitt HA, Soper NJ. Treatment of common bile duct stones discovered during cholecystectomy. J Gastrointest Surg 2008;12(4): 624–628.

24. Cronan JJ. US diagnosis of choledocholithiasis: a reappraisal. Radiology 1986;161(1):133–134.

25. Laing FC, Jeffrey RB Jr, Wing VW, Nyberg DA. Bil-iary dilatation: defining the level and cause by real-time US. Radiology 1986;160(1):39–42.

26. Brakel K, Laméris JS, Nijs HG, Terpstra OT, Steen G, Blijenberg BC. Predicting gallstone composition with CT: in vivo analysis. Radiology 1990;174(2): 337–341.

27. Baron RL, Rohrmann CA Jr, Lee SP, Shuman WP, Teefey SA. CT evaluation of gallstones in vitro: correlation with chemical analysis. AJR Am J Roent-genol 1988;151(6):1123–1128.

28. Varghese JC, Farrell MA, Courtney G, Osborne H, Murray FE, Lee MJ. A prospective comparison of magnetic resonance cholangiopancreatography with endoscopic retrograde cholangiopancreatography in evaluation of patients with suspected biliary tract disease. Clin Radiol 1999;54(8):513–520.

29. Stimac D, Miletić D, Radić M, et al. The role of nonenhanced magnetic resonance imaging in the early assessment of acute pancreatitis. Am J Gas-troenterol 2007;102(5):997–1004.

30. Masci E, Toti G, Curioni S, et al. Complications of diagnostic and therapeutic ERCP: a prospective multicenter study. Am J Gastroenterol 2001;96(2): 417–423.

31. Balthazar EJ, Robinson DL, Megibow AJ, Ranson JH. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990;174(2):331–336.

32. Frey CF. Hemorrhagic pancreatitis. Am J Surg 1979;137(5):616–623.

33. Erturk SM, Mortelé KJ, Oliva MR, Barish MA. State-of-the-art computed tomographic and mag-netic resonance imaging of the gastrointestinal sys-tem. Gastrointest Endosc Clin N Am 2005;15(3): 581–614.

34. Mourad J, Elliott JP, Erickson L, Lisboa L. Appen-dicitis in pregnancy: new information that contra-dicts long-held clinical beliefs. Am J Obstet Gynecol 2000;182(5):1027–1029.

35. Bree RL, Blackmore CC, Foley WD, et al. Right lower quadrant pain. In: ACR appropriateness crite-ria. Reston, Va: American College of Radiology, 2005.

36. Hand JW, Li Y, Thomas EL, Rutherford MA, Hajnal JV. Prediction of specific absorption rate in mother and fetus associated with MRI examinations during pregnancy. Magn Reson Med 2006;55(4):883–893.

37. Kanal E, Barkovich AJ, Bell C, et al. ACR guidance document for safe MR practices: 2007. AJR Am J Roentgenol 2007;188(6):1447–1474.

38. Israel GM, Malguria N, McCarthy S, Copel J, Wein-reb J. MRI vs. ultrasound for suspected appendicitis during pregnancy. J Magn Reson Imaging 2008;28 (2):428–433.

RG Volume 29 • October Special Issue 2009 Tkacz et al

MR Imaging in Gastrointestinal Emergencies Jaroslaw N. Tkacz et al

Page 1768 MR imaging protocols in the acute care setting should be tailored to concisely address diagnostic considerations, and the sequences should be chosen to minimize acquisition time. Page 1770 CT findings in pancreatic and biliary injuries may be nonspecific or overlooked because they are associated with multiple other injuries, and the symptoms and clinical findings are often unreliable and nonspecific. MR imaging of the pancreas and biliary tract in patients with blunt trauma to multiple organs can provide important diagnostic information. Page 1771 In patients in whom the CT findings are not definitive but only suggestive of pancreatic trauma or in whom CT findings of pancreatic injury are present but assessment of the pancreatic duct is needed, MR imaging may be useful. Page 1774 MR cholangiopancreatography has shown consistently high sensitivity and specificity (97%--99% and 95%--99%, respectively) for the detection of bile duct abnormalities. It may play a pivotal role in the management of patients with cholecystitis in whom choledocholithiasis is suspected, because it can help avoid diagnostic procedures that are unnecessarily invasive in the absence of common bile duct stones (eg, ERCP) or help direct surgical management in the presence of stones. Page 1776 The added value of MR cholangiopancreatography lies in its ability to help identify choledocholithiasis as the cause of pancreatitis and thus help direct interventional care. Other potential complications of acute pancreatitis that may be readily diagnosed with MR cholangiopancreatography include hemorrhagic pancreatitis, a potentially fatal complication.

RadioGraphics 2009; 29:1767–1780 • Published online 10.1148/rg.296095509 • Content Codes:

ABDOMINAL IMAGING 1767 MR Imaging in Gastrointestinal

Documents

Transcript of ABDOMINAL IMAGING 1767 MR Imaging in Gastrointestinal