Diagnostic Imaging Europe
July 2002
REPORT: GASTROINTESTINAL
MRI rivals traditional small bowel studies
MR enteroclysis can evaluate Crohn's disease, tumors, and intestinal obstruction
By: Panos Prassopoulos, M.D., Nickolas Papanikolaou, M.Sc., And Nicholas Gourtsoyiannis, M.D.
MRI of the small bowel originated in the early 1990s.1,2 Prolonged examination times and increased motion artifacts, due to either respiration or peristalsis, initially limited its diagnostic capability. Use of ultrafast sequences and high-performance gradients, however, has overcome most skepticism about the technique.
Reports now suggest that MRI could become the modality of choice for small bowel imaging in the near future.3,4 Excellent soft-tissue contrast, in conjunction with 3D imaging capabilities and breath-hold acquisition times, produces high-resolution images that are almost insensitive to motion artifacts, allowing detailed evaluation of superficial, transmural, and extraintestinal abnormalities in diseases involving the small bowel.5 Anatomic demonstration of the small bowel, mesenteries, and peritoneal cavity with MRI can be combined with functional information on small bowel motility and activity of Crohn's disease.
MRI's clear advantages over other modalities have encouraged research into its clinical role in the small bowel. Although conventional enteroclysis is considered the gold standard, it cannot demonstrate exoenteric pathology associated with small intestinal disease. CT suffers from relatively poor soft-tissue contrast and acquisition of axial images only. Both techniques also involve significant radiation dose to the patients, who are frequently young.
Contrast agents are fundamental to lumen opacification of the small bowel by MRI. Prerequisites for a suitable intraluminal agent include homogeneous opacification, clear differentiation between the lumen and bowel wall, minimal mucosal absorption, absence of artifacts and severe adverse effects, and availability at low cost. A number of suitable contrast agents have been proposed that exhibit either high signal intensity (positive contrast agents), low signal intensity (negative contrast agents), or behavior that depends on the pulse sequence (biphasic contrast agents); i.e., they are negative on T1-weighted sequences and positive on T2-weighted sequences. Extensive comparative studies are lacking, however, and no consensus exists on the optimal intraluminal contrast agent. It is likely that several of the available products meet the prerequisites for a suitable small bowel contrast agent.
The route of contrast administration may be more important than the selection of the substance itself. Two major approaches are used: by mouth6-8 and through a nasojejunal catheter after intubation.3,4,9 The former technique is more comfortable for the patient but may produce suboptimal depiction of jejunal loops or the terminal ileum.8 Oral administration does not show distention of intestinal loops to the same extent as enteroclysis, raising doubts about identification of subtle abnormalities requiring distention.6
Duodenal intubation in experienced hands is well tolerated by patients and guarantees adequate intestinal distention. MRI of the small bowel after intubation is referred to as MR enteroclysis.4 A maximum of 1.5 liters of an iso-osmotic water solution with polyethylene glycol and electrolytes is administered by a manual pump to patients lying inside the magnet. Prone positioning is suggested so that mild pressure on the anterior abdominal wall facilitates separation of the small bowel loops and the volume of peritoneal cavity to be imaged decreases.5
The solution is administered in two phases. A flow rate of 80 to 150 mL/min is used until the contrast reaches the terminal ileum. This is then increased to 200 mL/min to achieve reflex atony, allowing image acquisition with minimal or no motion artifacts. Another way to avoid motion artifacts is to administer antiperistaltic drugs intravenously, especially prior to motion-sensitive MR sequences. An additional advantage is the ability of MR enteroclysis to evaluate small bowel motility in a fluoroscopic-like mode by using ultrafast pulse sequences such as single-shot turbo spin-echo.3
MRI of the small bowel can also be combined with conventional enteroclysis.10,11 An MR examination is performed immediately after enteroclysis, using 800 mL of barium sulfate solution and 1200 mL of methylcellulose solution mixed with an MRI contrast agent.
Intravenous administration of a paramagnetic contrast agent is also considered integral to MR examinations of the small bowel. Contrast helps with the characterization of space-occupying lesions and evaluation of Crohn's disease.12 Additionally, paramagnetic contrast can improve views of the normal bowel wall on T1-weighted images with fat saturation and negative intraluminal contrast agents. Maximum enhancement of the normal bowel wall is achieved 70 to 80 seconds after the start of injection.13 This scanning delay is suggested to optimize contrast resolution between the enhanced bowel wall, hypointense intraluminal intestinal content, and saturated extraluminal peritoneal fat.
CHOOSING SEQUENCES
Most authors consider T1- and T2-weighted images necessary for a comprehensive MR examination of the small bowel.3-7 Sequences should be fast enough to permit breath-hold imaging. Axial and coronal planes are both useful.
Many researchers suggest gradient-echo T1-weighted sequences in either 2D or 3D acquisition modes. Fat saturation has been also recommended, especially when a negative contrast agent is used for intestinal opacification.14-16 Fat-suppressed fast gradient-echo T1-weighted images, negative intraluminal contrast, and IV administration of gadolinium chelates have proved useful in the detection and characterization of intestinal wall abnormalities.14,15 Fat and intraluminal fluid exhibit low signal intensity, normal bowel wall shows moderate signal intensity, and inflammatory lesions or neoplasms are revealed as high signal intensity, due to Gd uptake.
Spatial resolution can be improved substantially if a 3D FLASH sequence with 2-mm slice thickness and a 512 matrix is used (Figure 1A).16 Antiperistaltic drugs are necessary when using 3D FLASH T1-weighted sequences because peristalsis-induced motion artifacts may blur the intestinal wall.
HASTE (Figure 1B), turbo spin-echo, and TrueFISP (Figure 1C) sequences have been proposed for T2-weighted imaging.3,4,7,16 High-resolution HASTE images can be acquired in less than one second per slice and are relatively insensitive to susceptibility and chemical shift artifacts. Inflammatory and neoplastic lesions exhibit high signal intensity on HASTE images, in comparison with the low signal intensity of the normal bowel wall. HASTE images, however, do not provide information from the mesenteries or mesenteric vessels. Antiperistaltic drugs should be used prior to the sequence to reduce intraluminal flow voids. Single-shot turbo spin-echo sequences, routinely used to demonstrate the small bowel lumen, provide information related to intestinal distention and small bowel motility.3
The TrueFISP sequence can provide high-quality images of the small bowel that are useful for anatomic visualization and identification of abnormalities, especially in Crohn's disease.5 High-resolution hybrid T2-weighted TrueFISP images are almost insensitive to motion, due to the short acquisition time (1.5 seconds or less per slice), eliminating the need for antiperistaltic drugs or breath-holding. So-called black boundary artifacts at the interface of the bowel wall and mesenteric fat are invariably present on TrueFISP images as a thin black line. These can be easily differentiated, however, from the moderate signal intensity of the bowel wall.5 Air trapping may produce susceptibility artifacts that are usually minimal and easily recognized.5
The major advantage of TrueFISP images is their ability to demonstrate mesenteries, mesenteric vessels, and small lymph nodes. This makes the sequence extremely important for depicting small bowel diseases involving adjacent exoenteric tissues.
We suggest the following steps for a complete, detailed, and comprehensive MR examination of the small bowel:17
- fluoroscopically guided duodenal intubation, for optimal small bowel distention;
- administration of iso-osmotic biphasic contrast agent, providing adequate contrast resolution between small bowel lumen and intestinal wall on both T1- and T2-weighted images;
- acquisition of repeated single-shot turbo spin-echo images to monitor the small bowel filling process and for MR fluoroscopy;
- acquisition of axial and coronal TrueFISP images for the anatomic demonstration of the small bowel and mesenteries, identification of wall thickening and superficial abnormalities, and depiction of space-occupying lesions;
- intravenous administration of antiperistaltic substances, such as 1 mg of glucagon, to reduce inherent small bowel motion and related artifacts;
- acquisition of axial and coronal HASTE images for lesion identification and characterization; and
- acquisition of fat-suppressed 3D FLASH images 70 to 80 seconds post-IV Gd administration for depiction of normal and abnormal bowel wall and identification and characterization of inflammatory and tumorous small bowel abnormalities.
The total examination time in the MR suite is approximately 15 minutes, while acquisition time for all MR enteroclysis sequences is less than eight minutes. High-field-strength (1.5T), high-performance gradient systems, and abdominal phased-array coils are all required for a high-quality examination.
CLINICAL APPLICATIONS
The most important clinical applications of small bowel MRI include evaluation of Crohn's disease, small bowel tumors, and intestinal obstruction. Characteristic ulcerations of Crohn's disease can be demonstrated on MR enteroclysis, ideally using the TrueFISP sequence, and also on HASTE images. Longitudinal or transverse deep linear ulcers are manifested as thin high-signal-intensity discrete lines within the thickened bowel wall on TrueFISP images. Cobblestoning on TrueFISP images appears as patchy, incompletely demarcated areas of moderate signal intensity, separated by the high signal intensity of fluid within ulcers. Early, but not specific, superficial alterations of the disease, including mucosal nodularity and aphthous ulcers, may be depicted to a lesser extent than with conventional enteroclysis, due to the lower spatial resolution of MR enteroclysis.
Wall thickening is invariably seen with all MR enteroclysis sequences exhibiting moderate to low signal intensity on TrueFISP (Figure 2) and HASTE images in the absence of significant edema. Lumen narrowing and prestenotic intestinal dilatation can also be noted. MR enteroclysis is as accurate as conventional enteroclysis in detecting and localizing the affected small bowel segments and in estimating wall thickening and length of involvement in patients with Crohn's disease.5
MR enteroclysis can demonstrate exoenteric manifestations or complications of Crohn's disease. Fibrofatty proliferation is revealed by its space-occupying behavior in all sequences, although it is better demonstrated on TrueFISP sequences, where the fatty high-signal-intensity components and fibrotic low-signal-intensity components can be seen. Increased mesenteric vascularity at the level of vasa recta-the "comb" sign-is ideally delineated on TrueFISP images but can also be seen on Gd-enhanced 3D FLASH images with fat saturation. Small lymph nodes, frequently less than 8 mm in diameter, are depicted as low-signal-intensity ovoid structures within bright fat on TrueFISP images. They are not seen clearly on HASTE images because of k-space filtering effects and are detected on post-Gd fat-saturated 3D FLASH images only when manifesting contrast uptake (Figure 3). Sinus tracts and fistulas are diagnosed by the high signal intensity of their fluid content on TrueFISP and HASTE images. Abscesses are best demonstrated when Gd-enhanced 3D FLASH images cause the abscess wall to enhance.
Contrast enhancement of the thickened bowel wall on T1-weighted 3D FLASH images (Figure 3) is considered an important indicator of Crohn's disease activity.12,18 Severity of the disease process can be ranked when the degree of Gd uptake is combined with measurements of wall thickening and length of segment involved.12 High signal intensity of the thickened intestinal wall on T2-weighted images also indicates disease activity.
Cross-sectional MRI is an effective means of demonstrating neoplasms of the small bowel.15 MR enteroclysis may combine the advantages of cross-sectional MRI with those of conventional enteroclysis in tumor detection and characterization. Most intestinal neoplasms, including leiomyoma, leio-myosarcoma, adenocarcinoma, and carcinoid tumors, enhance on T1-weighted images following Gd administration. Three-D FLASH images with fat saturation are preferable, as they provide high contrast resolution between the enhancing tumor, the low signal intensity of the lumen content, and the saturated mesenteric fat. TrueFISP images can depict small bowel tumors as moderate signal, in comparison with the high signal intensity of the distended lumen and peritoneal fat (Figure 4). Peritoneal extension of the neoplasms, associated lymphadenopathy, and the degree of intestinal obstruction can also be appreciated on MR enteroclysis.
MR enteroclysis correlates well with conventional enteroclysis in demonstrating the presence and level of small bowel obstruction (Figure 5).3 Sequential TrueFISP images in a functional cine MRI mode have proved highly accurate in showing obstruction due to postsurgical adhesions.19 Dynamic single-shot turbo spin-echo images are also helpful in the prompt disclosure of obstruction, while contrast-enhanced 3D FLASH images may reveal other causes of obstruction.
MRI appears promising for evaluation of small bowel anatomy and pathology. Initial experience indicates that the modality performs well in Crohn's disease, tumorous lesions, and small bowel obstruction. Adequate bowel distention, homogeneous opacification of the lumen, short acquisition times permitting breath-holding, T1- and T2-weighted imaging, and contrast enhancement are fundamental for achieving state-of-the-art small bowel MRI. MRI can also provide functional information. Future applications may include evaluation of small bowel ischemia and bleeding.
DR. PRASSOPOULOS is an associate professor, DR. PAPANIKOLAOU is a biomedical engineer, and PROF. GOURTSOYIANNIS is chief of radiology, all at University Hospital of Heraklion, Medical School of Crete, Greece. Assisting in the preparation of this article were DR. JOHN GRAMMATIKAKIS, DR. THOMAS MARIS, and DR. MARIA ROUSSOMOUSTAKAKI, all at the same institution.
References
1. Mirowitz SA, Susmann N. Use of a nutritional support formula as a gastrointestinal contrast agent for MRI. JCAT 1992;16:908.
2. Rubin DL, Muller HH, Young SW. Formulation of radiographically detectable gastrointestinal contrast agents for magnetic resonance imaging: effects of barium sulfate additive on MR contrast agent effectiveness. Magn Reson Med 1992;23:154.
3. Umschaden HW, Szolar D, Gasser J, et al. Small-bowel disease: comparison of MR enteroclysis images with conventional enteroclysis and surgical findings. Radiology 2000;215:717-7125.
4. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, et al. Magnetic resonance imaging of the small bowel using a True-FISP sequence after enteroclysis with water solution. Invest Radiol 2000;35(12):707-711.
5. Prassopoulos P, Papanikolaou N, Grammatikakis J, et al. MR enteroclysis imaging findings in Crohn's disease. Radiographics 20011;21 spec no:S161-172.
6. Low RN, Francis IR. MR imaging of gastrointestinal tract with IV gadolinium and diluted barium oral contrast media compared with unenhanced MR imaging and CT. AJR 1997;169:1051-1059.
7. Maccioni F, Viscido A, Broglia L, et al. Evaluation of Crohn's disease activity with magnetic resonance imaging. Abdom Imaging 2000;25:219-228.
8. Minowa O, Ozaki Y, Kyogoku S, et al. MR imaging of the small bowel using water as a contrast agent in a preliminary study with healthy volunteers. AJR 1999;173:581-582.
9. Lomas DJ, Graves MJ. Small bowel MRI using water as a contrast medium. BJR 1999;72:994-997.
10. Aschoff A, Zeitler H, Merkle EM, et al. MR enteroclysis for MR diagnosis of inflammatory small bowel disease with contrast enhancement. Fortschr Rontgenstr 1997;167:387-391.
11. Rieber A, Wruk D, Nussle K, et al. MRI of the abdomen in combination with enteroclysis in Crohn's disease with oral and intravenous Gd-DTPA. Radiologe 1998;38:23-28.
12. Marcos HB, Semelka RC. Evaluation of Crohn's disease using half-Fourier RARE and gadolinium-enhanced SGE sequences: initial results. Magn Reson Imag 2000;18:263-268.
13. Papanikolaou N, Prassopoulos P, Grammatikakis J, et al. Optimization of the scanning delay on post-contrast MR imaging of the small bowel. ECR 2002.
14. Faber SC, Stehling MK, Holzknecht N, et al. Pathologic conditions in the small bowel: findings at fat-suppressed gadolinium-enhanced MR imaging with an optimized suspension of oral magnetic particles. Radiology 1997;205:278-282.
15. Semelka RC, John G, Kelekis N, et al. Small bowel neoplastic disease: demonstration by MRI. JMRI 1996;6:855-860.
16. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, et al. MR enteroclysis protocol optimization: comparison between 3D FLASH with fat saturation after intravenous gadolinium injection and TrueFISP sequences. Eur Radiol 2001;11(6):908-913.
17. Papanikolaou N, Grammatikakis J, Maris T, et al. A comprehensive MR imaging protocol for small bowel studies. ESGAR 2001.
18. Schunk K, Kern A, Oberholzer K, et al. Hydro-MRI in Crohn's disease. Appraisal of disease activity. Invest Radiol 2000;35:431-437.
19. Lienemann A, Sprenger D, Steitz HO, et al. Detection and mapping of intraabdominal adhesions by using functional cine MR imaging: preliminary results. Radiology 2000;217:421-425.
