- Email: [email protected]
Magnesium sulfate as an oral contrast medium in magnetic resonance imaging of the small intestine
European Journal of Radiology 81 (2012) e370–e375
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Magnesium sulfate as an oral contrast medium in magnetic resonance imaging of the small intestine Hao Shi ∗ , Cun Liu, Hong Yu Ding, Chun Wei Li Department of Diagnostic Imaging, Shandong Qianfoshan Hospital, Jingshi Street 16766, Jinan, Shandong Province 250014, PR China
a r t i c l e
i n f o
Article history: Received 15 May 2011 Received in revised form 23 November 2011 Accepted 25 November 2011 Keywords: Administration Oral Contrast media Small intestine
a b s t r a c t Background: To explore the use of magnesium sulfate (MgSO4 ) as an oral contrast medium (CM) in MRI of the small intestine. Methods: By comparing MgSO4 SNRs at different concentrations, we determined that 2.5% MgSO4 is the ideal concentration for small bowel MRI. Twenty volunteers underwent MRI after drinking 2.5% MgSO4 . Thirty-one patients with clinical suspicion of small intestinal pathology underwent both MRI and the air-barium contrast examination. The patient’s tolerance, side effects and complications were noted. Results: 2.5% MgSO4 can decrease the absorption of water and fully fill the enteric cavity, thereby increasing the contrast between the intestinal wall and lumen and facilitating radiographic examination of the small bowel. The mean diameter of the small intestine was 19.8 ± 1.21 mm in the 20 volunteers consuming 2.5% MgSO4 and 12.7 ± 0.84 mm in the 20 volunteers given water. There was a significant difference (P < 0.05) between the diameters of the small intestine of the two groups. But there were no significant differences (P > 0.05) in side effects between MgSO4 and water groups. Small intestinal MRI was successfully performed in all 31 patients, who were also examined by the double contrast barium, which gave almost identical diagnoses to MRI in all cases except for 1 patient with small intestinal hemorrhage. Conclusions: MRI with 2.5% MgSO4 can demonstrate intestinal abnormalities. Therefore, 2.5% MgSO4 solution is an ideal oral CM for small bowel MRI. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Magnetic resonance imaging (MRI) has been used extensively in clinical practice as it offers high resolution of soft tissues, allows for multiplanar reconstructions and does not expose the patient to radiation. However, MRI has not been extensively applied in the clinical evaluation of diseases of the gastrointestinal tract, although its potential utility in this regard is currently under investigation [1–3]. The use of oral contrast media (OCM) is critical to the MRI examination of the small intestine. Most OCM are expensive, poorly tolerated by patients, have considerable side effects and imaging artifacts, all of which limit their use in intestinal MRI [1,4]. Water was frequently used as an OCM for MRI examination of the small intestine. However, water cannot completely fill the small intestine as it is rapidly absorbed [1–4]. Magnesium sulfate (MgSO4 ) can be easily dissolved in the water. Mg2+ and SO4 2+ cannot be absorbed in the small intestine and can produce a high osmotic pressure (OP)
∗ Corresponding author at: Department of Diagnostic Imaging, Shandong Qianfoshan Hospital, Jingshi Street 16766, Jinan, Shandong Province 250021, PR China. Tel.: +86 531 89268750; mobile: +86 18653113953. E-mail addresses: [email protected], [email protected] (H. Shi). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.11.049
[5]. In turn, this high OP may slow the absorption of water and allow for adequate filling of the small intestinal lumen. We developed a new MRI OCM and had performed MRI examination of the small intestine with oral administration of MgSO4 to assess its utility in the diagnosis of small bowel pathology.
2. Materials and methods 2.1. MRI technique All patients were examined by 1.5 Tesla magnetic resonance tomography (Sonata, Siemens Medical Solutions, Germany) with an abdominal phased-array coil with 6 segments. Routine scanning sequences for abdominal examination were used, which included GRE (gradient echo) T1 WI (Flash2d) with parallel imaging and without gating technique (TR/TE 110/4.8 ms, FOV 298 × 400 mm, matrix 256 × 192, thickness 6 mm, turbo factor 2, scan time 19 s), GRE (gradient echo) FS-T1 WI (Flash2d) without parallel imaging and gating technique (TR/TE 226/7.2 ms; FOV 261 × 350 mm, matrix 256 × 192, thickness 6 mm, scan time 21 s), GRE (gradient echo) T2 WI (trueFISP, true Fast Imaging with Steady-state Procession) without parallel imaging and gating technique (TR/TE 3.7/1.8 ms, FOV 360 × 360 mm, matrix 512 × 512, thickness 6 mm, scan time
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
e371
19 s) and TSE (turbo Spin echo) T2 WI (Haste, Half-Fourier Acquisition Single-shot) without parallel imaging and gating technique (TR/TE 1040/72 ms, FOV 269 × 360 mm, matrix 512 × 384, thickness 6 mm, scan time 17 s). Since MgSO4 presented high signal intensity like water, T2 WI sequences were more used for a clearer contrast between the media and surrounding structures. Scanning directions included axial, coronal and sagittal planes. All study protocols were approved by our institutional review board, and informed consent was obtained prior to each examination. 2.2. Determination of the optimal concentration of contrast The different concentrations of MgSO4 sample (50%, 40%, 30%, 20%, 10%, 5%, 2.5%, 1.25%, 1% and 0.5%) and water (0% as control group) were placed into the different test tubes, and then scanned using the different MRI sequences described above. The OPs of the different concentrations were calculated by the Van’t Hoff OP formula (Table 1) and SNR of the different concentrations of MgSO4 samples were compared. The concentration of MgSO4 with the highest osmotic pressure and most distinct imaging signal and SNR was selected for subsequent use. 2.3. Clinical study in healthy subjects To study the utility of the selected contrast medium, twenty volunteers (11 males, 9 females, aged between 18 and 50 years, mean 30 years) were tested with the selected OCM. None of the volunteers had any history of abdominal surgery, GI pathology or any contraindications to MRI. Subjects had not consumed a barium meal or CT oral contrast agents within 3 days, nor had received intravenous contrast within 24 h of study participation. The volunteers were not permitted to drink or eat within 12 h of the examination. Twenty volunteers were randomly divided into an experimental group (10 volunteers) and a control group (10 volunteers). 1000 ml of 2.5% MgSO4 was given orally 20 min before the examination. The volunteers in the experimental group drank 2000 ml of water just before the examination to completely fill the stomach and the duodenum. In the control group, the volunteers drank 1000 ml of the water 20 min before and 2000 ml just prior to the examination. 10 mg of anisodamine was injected IM 10 min before the examination in all volunteers to reduce artifacts that may result from peristalsis. For all volunteers included in both the experimental and control groups, examinations were performed at least 2 weeks apart.
Fig. 1. Schematic diagram of intestinal grouping. Group 1: duodenum; group 2: superior segment of jejunum; group 3: inferior segment of jejunum; group 4: superior segment of ileum; group 5: middle segment of ileum; group 6: inferior segment of ileum.
assigned. The small intestine was divided into six groups according to X-ray anatomy (Fig. 1). The dilation of the small intestinal tract and the image quality were scored and graded according to the criteria outlined in Tables 2 and 3. Adverse reactions occurring within 24 h of the examination were also recorded and graded by the criteria defined in Table 4. MRI and gas-barium double contrast fluoroscopy of the small intestine were performed in 31 patients with suspected small bowel pathology. All studies were interpreted independently by two radiologists who were not informed to which arm of the study the patients had been assigned. The radiologists evaluated the appearance of the mucosa, the intestinal wall and surrounding tissues and described the location, characteristics and involved structure of all lesions. Finally, the results were analyzed and compared with the pathological findings of biopsies and surgical specimens.
2.4. Clinical study in patients with small bowel pathology 2.6. Statistical analysis Between November 2004 and April 2006, 31 patients with clinical suspicion of small intestinal pathology, underwent intestinal MR examination after oral administration of MgSO4 . The methods of MR examination were the same as previously described above. The patients included 20 men and 11 women, aged between 25 and 80 years (mean age 34 years). Signs and symptoms suggestive of small bowel pathology included abdominal pain, diarrhea, weight loss, anemia, fever, abdominal masses and fecal occult blood. The course of the disease was from 2 months to 10 years. All 31 patients underwent both MRI and gas-barium double contrast fluoroscopy, which was conducted with either the SHIMADZU XED150L-20 500 mA X-ray machine, the Philips D-93 digit X-ray machine or the SHIMADZU PRO-series digit X-ray machine. 2.5. Evaluation of results Images of 20 volunteers (both the experimental and control groups) were analyzed by two radiologists with experience in MRI who was blinded as to which arm of the study volunteers had been
The dilation of the small intestine, image quality and adverse reactions between the experiment group and the control group were calculated with the rank-sum test (Wilcoxon rank test, Mann–Whitney test and Kruskal–Wallis test respectively). The means of the dilation were presented in mean ± SD. All data were statistically analyzed with SSPS 12.0 software. A p-value of less than 0.05 was considered significant. 3. Results 3.1. Optimization of MgSO4 concentration The use of MgSO4 resulted in a shortening of T1 and T2 relaxation times. However, this shortening of T1 and T2 relaxation times decreased with decreasing concentrations of MgSO4 (Fig. 2). At a concentration of 2.5%, the OP was 417 mmol/l, there was low signal intensity and SNR on T1 WI and high signal intensity and SNR on T2 WI (Fig. 2). The SNR of MgSO4 was similar to water, which
e372
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
Table 1 The osmotic pressure of MgSO4 at different concentrations. Concentration (%) Osmotic pressure (mmol/l)
50 8333
40 6666
30 5000
20 3333
10 1667
5 833
2.5 417
1.25 208
1 166
0.5 83
0 0
Table 2 Grading criteria for the dilation of the small intestine. Grade
Degree of dilation of the small intestine
Score
I II III IV
The diameter of the intestine was ≤1 cm, unable to distinguish the intestinal lumen and wall, poor dilation The diameter of the intestine was 1–1.5 cm, able to distinguish the intestinal lumen and wall, moderate dilation The diameter of the intestine was 1.5–2.0 cm, liquid contained in the intestinal lumen, good dilation The diameter of the intestine was >2.0 cm, lumen completely filled by liquid, excellent dilation
1 2 3 4
Table 3 Grading criteria for image quality. Grade
Image quality of the small intestine
Score
I
There is no signal distinction between the intestinal lumen and wall; poor resolution between the wall and surrounding tissues. There is signal distinction between the intestinal lumen and the wall; inhomogeneous signal in the lumen and moderate resolution between the wall and surrounding tissues. There is homogeneous signal in the intestinal lumen and good resolution between the lumen, wall and surrounding tissues. There is a homogeneous signal in the intestinal lumen and excellent resolution between the lumen wall and surrounding tissues.
1
II III IV
2 3 4
Table 4 Grading criteria for adverse reactions. Grade
Description of the adverse reaction
Score
I II III IV
There were no adverse reactions and no effects on the patient’s work or daily activities There was a slight adverse reaction which had minimal impact on the patient’s work and daily activities. There was a tolerable adverse reaction which affected the patient’s work and daily activities. There was a significant adverse reaction which could not be tolerated and prohibited the patient from working or engaging in other activities.
1 2 3 4
has low SNR on T1 WI and high SNR on T2 WI (Fig. 2). T2 WI had more markedly contrast between the lumen, the wall and the surrounding structures of the intestine and had a more useful value than T1 WI. Therefore, we selected 2.5% MgSO4 as an oral contrast medium for the following reasons. First, it is hyperosmolar and could therefore retain more water in the intestinal lumen. Second, a concentration of 2.5% is relatively low and would be less likely to result in the adverse reactions seen with the higher concentrations used for bowel preparation prior to colonoscopy.
3.2. Results of the clinical study Since T2 WI presented a hyperintensity on the images, it had a more obvious contrast between the lumen and the wall of intestine than T1 WI. Therefore, the clinical comparison study was performed mainly on T2 WI (trueFISP and Haste) sequences. The small intestine of all volunteers in the experimental group was fully dilated. On T2 WI, the liquid in the lumen of the small intestine had high signal intensity; the intestinal wall had moderate signal intensity.
Fig. 2. SNR of different concentrations of MgSO4 and water in different sequences.
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
e373
Table 5 Comparison of dilation scoring between the experimental and control groups. Location
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
Experiment group I
II
III
3 0 0 0 4 8
31 7 8 1 15 15
5 21 22 21 16 13
Control group IV 1 12 10 8 5 4
I
II
7 0 7 15 23 33
Rank-sum test
III
28 9 13 17 17 7
5 23 16 8 0 0
IV 0 8 4 0 0 0
p > 0.05 p > 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05
Table 6 Comparison of image quality scoring between the two groups.
Fig. 3. On T2 WI (TrueFISP), the small intestine is well-filled and completely dilated after the administration of 2.5% MgSO4 .
Hence, it was possible to distinguish between the lumen, wall and surrounding tissues (Fig. 3). The mean dilation of the small intestine was 19.8 ± 1.21 mm. The small intestine of volunteers in the control group did not adequately dilate and it was difficult to distinguish the lumen from the intestinal wall in some parts of the small intestine (Fig. 4). The average dilation of the small intestine was 12.7 ± 0.84 mm. There was a significant difference between the two groups in the dilation of the small intestine (P < 0.05). Grading of the dilation of the small intestine demonstrated that the small intestine in the experimental group was dilated better; the dilation of groups 3–6 of the small intestine in the control group was inferior to the experimental group; these differences were statistically significant (P < 0.05). The difference of the dilation between groups 1 and 2 was not statistically significant (P > 0.05, Table 5). Grading of image quality revealed significantly better image quality in the experimental group compared to the control group, in particular in groups 2–6 (P < 0.05, Table 6).
Location
Experiment group
Control group
I
II
III
IV
I
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
4 0 0 0 4 4
33 6 11 11 16 18
3 26 21 24 16 15
0 8 8 5 4 3
8 0 8 21 27 33
II
Rank-sum test
III
27 15 17 14 13 7
5 23 15 5 0 0
IV 0 2 1 0 0 0
p > 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05
Diarrhea occurred more frequently in the experimental group than the control group (P < 0.01). Urinary frequency occurred more frequently in the control group (P < 0.01). Differences in the frequencies of other adverse reactions, such as nausea, vomiting and enterospasm, between the two groups were not statistically significant (P > 0.05, Table 7). After administration of 2.5% MgSO4 , the normal bowel wall, which was sufficiently dilated, displayed moderate signal with 2.54 ± 4.35 mm of thickness both on T1 WI and T2 WI sequences and looked smooth and regular. With contrast, the bowel wall enhanced homogeneously and could be clearly distinguished from the intestinal lumen and extramural structures. The lumen of the small intestine had low signal intensity on T1 WI and remarkably high signal intensity on T2 WI. There were more mucosal folds in the jejunum, which were perpendicular to the long axis of the small intestine and looked like a spring. Relatively fewer mucosal folds were observed in the ileum, which were perpendicular or oblique to long axis of the small intestine. The number of the folds gradually decreased from the proximal to distal segments of the ileum. Almost no mucosal folds were seen in the terminal ileum. 3.3. Results of the clinical study in patients with small bowel pathology MRI was successfully performed in 31 patients with lesions of the small intestine. Twenty-eight patients exhibited good tolerance to 2.5% MgSO4 and the small intestine was adequately dilated in same extent as the experimental group in the clinical study. Slight nausea and abdominal distention occurred in three patients. Although these three patients took less than 1000 ml of 2.5% MgSO4 , the small bowel was still adequately dilated and a diagnosis could be made in each case without difficulty. In no case was the MR exam terminated due to diarrhea. MR examination with oral Table 7 Comparison of adverse reaction scoring in the two groups. Adverse reaction
Experiment group I
Fig. 4. On T2 WI (TrueFISP) the small intestine is not completely dilated and the wall of the small intestine is not clearly displayed.
Diarrhea 0 Nausea and vomiting 16 17 Enterospasm 0 Urinary frequency
II 6 4 3 12
III 10 0 0 8
Control group
Rank-sum test
IV
I
II
III
IV
4 0 0 0
18 15 16 0
2 5 4 4
0 0 0 10
0 0 0 6
p < 0.01 p > 0.05 p > 0.05 p < 0.01
e374
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
administration of 2.5% MgSO4 accurately demonstrated bowel-wall thickening, strictures, ulcers and masses of the small intestine, as well as extra-mural extension of lesions and their relationships to surrounding structures (Fig. 5). MRI examination with 2.5% MgSO4 , which permits visualization of the intestinal wall, lumen and also surrounding structures, is a better than gas-barium double contrast examination. The tissue pathology specimens from surgical procedures and endoscopy showed that there were 20 cases of Crohn’s disease, 8 cases of interstitialoma, 1 case of small intestinal hemorrhage, 1 case of intussusception and 1 case of tuberculosis. MRI diagnosed Crohn’s disease in 20 cases, intestinal tumor in 8 cases, intussusception in 1 case and tuberculosis in 1 case. MRI gave the correct diagnosis in all cases except for the single patient with small intestinal hemorrhage. The sensitivity of MRI was 96.7%. MRI and double contrast fluoroscopy gave identical diagnoses in 100% of patients.
4. Discussion Historically, gas-barium double contrast examination was the principal modality for imaging the small intestine. However, it is limited in its ability to image structures outside of the intestinal lumen. Interpretation may also be challenging when loops of bowel overlap, making it difficult to pinpoint the location of certain lesions. It also involves exposure to radiation and may not be well-tolerated by all patients. For these reasons, CT examination of gastrointestinal tract is now beginning to replace gas-barium fluoroscopy. However, CT also involves radiation exposure and may not provide adequate contrast of soft tissue structures. Until recently, the clinical utility of MRI in evaluation the GI tract was limited. MRI required long examination times and had low signal-to-noise ratios (SNR) [6]. Moreover, there was no suitable oral contrast medium [3,4]. In addition, image interpretation was limited by numerous artifacts from respiratory, cardiac and gastrointestinal motion. Presently, improvements (such as parallel imaging, gating technique) in MR technology have lead to shorter examination times and high SNR, both of which improve the applicability of this modality to imaging of the GI tract [7]. Nonetheless, it is necessary to identify an ideal OCM. The ideal OCM would be able to help the radiologist accurately visualize normal structures and characterize lesions of the small bowel [8]. The ideal OCM would have the following characteristics. (1) It provides distinct contrast between the contrast medium in the intestinal lumen and the intestinal wall and surrounding tissues without artifacts in any sequences. (2) It can completely fill and distribute throughout the lumen without being rapidly absorbed. (3) It is stable nature and has uniform contrast effects throughout the GI tract. (4) It is safe, non-toxic, does not stimulate the GI tract or result in adverse side effects, and has a taste tolerable to patients. (5) It is relatively inexpensive and can be produced and preserved easily [8–10]. Like mannitol, MgSO4 can stimulate peristalsis and catharsis and is used to prepare the GI tract for colonoscopy [11]. In this application, it is fast, convenient and has few side effects [11]. It is easily dissolved in water, but Mg2+ and SO4 2+ cannot be absorbed in the small intestine. The presence of these two ions in the lumen creates a hyperosmolar state, which helps retain water, thereby dilating the lumen. Our study indicated that MgSO4 at high concentrations shows high signal intensity on T1 WI and low signal intensity on T2 WI. However, at lower concentrations, we observed a lower signal intensity and SNR on T1 WI and high signal intensity and SNR on T2 WI; thus, at lower concentrations, this profile of signal intensities and SNR is similar to water. Therefore, 2.5% MgSO4 could be used as a hypointense OCM in T1 WI and a hyperintense OCM in
Fig. 5. Crohn’s disease. On coronal (a), saggital (b) and transverse (c) T2W (TrueFISP) images, the lumen of the terminal ileum is irregularly narrowed (arrow). The membrane mucosa is derangement and thickness, and the wall is thickened and exhibits cobble stoning on T2 WI.
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
T2 WI. Since 2.5% MgSO4 has a relative high OP (417 mmol/l), it can reduce absorption of water in the GI tract and dilate the lumen sufficiently. It was superior to water in terms of its ability to dilate the small intestine; compared to the control group, it offered better image quality. All patients took 1000 ml of 2.5% MgSO4 20 min before the examination, and 2000 ml of water immediately prior to the examination. This allowed for dilation of the small intestine as 2.5% MgSO4 is hyperosmolar and not easily absorbed. The stomach and duodenum was filled with water in order to decrease the dose of MgSO4 and reduce artifacts from intestinal peristalsis, which could result from stimulation of the GI tract by MgSO4 . After administration of MgSO4 , there were no obvious adverse reactions, such as nausea, vomiting or constipation. However, patients reported a bitter in flavor and occasionally slight diarrhea, both of which were tolerable. Therefore, 2.5% MgSO4 is safe and is well tolerated as an OCM for MRI examination of the small intestine. This study demonstrated that MgSO4 as an OCM in MRI examination of the small intestine can facilitate the diagnosis of intestinal pathology, manifested by abnormalities of the intestinal mucosa, strictures and/or dilation of the lumen and thickening and masses of the intestinal wall. It is useful in estimating the infiltrative depth of the tumors in the intestinal wall and evaluating extension to structures outside of the GI tract. The results of MRI examination with administration of 2.5% MgSO4 in 31 patients with the diseases of the small intestine had a high concordance with gas-barium double contrast examination and tissue pathology. In comparison with gas-barium double contrast examination, the superiority of MRI with 2.5% MgSO4 as an OCM was: (1) the patient was not exposed to radiation; (2) image interpretation was not complicated by overlapping bowel loops, as typically occurs with gas-barium double contrast examinations; (3) soft tissue resolution was good; (4) it demonstrated lesions in the small intestine and surrounding structures, in particular metastases to other organs and the lymph nodes; (5) contrast media could be given, which further helps to characterize lesions; (6) MR hydrography could be performed, which can obtain images similar to gas-barium double contrast examination and observe the small intestine at arbitrary angles, without the limitations of patient positioning and (7) multi-planar and multi-parameter reconstructions, as well as other
e375
post-processing techniques such as virtual colonography, could be employed to assist the radiologist in rendering an accurate diagnosis. In conclusion, 2.5% MgSO4 as an OCM can adequately dilate the small intestine and permit visualization of the intestinal lumen, wall, and surrounding structures. It can facilitate the detection of lesions of the small intestine and their extramural extension. Therefore, MgSO4 is an excellent OCM for MRI examination of the small intestine. Conflicts of interest The authors (including Hao SHI, Cun LIU, Hong Yu DING, Chun Wei LI) have no conflicts of interest to declare. And also there is no financial support. References [1] Geraldes CF, Laurent S. Classification and basic properties of contrast agents for magnetic resonance imaging. Contrast Media Mol Imaging 2009;4(1):1–23. [2] Gilja OH, Hatlebakk JG, Odegaard S, Berstad A, Viola I, Giertsen C, et al. Advanced imaging and visualization in gastrointestinal disorders. World J Gastroenterol 2007;13(9):1408–21. [3] Maccioni F. Current status of gastrointestinal MRI. Abdom Imaging 2002;27(4):358–60. [4] Briggs RW, Wu Z, Mladinich CR, Stoupis C, Gauger J, Liebig T, et al. In vivo animal test of an artifact-free contrast agent for gastrointestinal MRI. Magn Reson Imaging 1997;15(5):559–66. [5] Smilkstein MJ, Steedle D, Kulig KW, Marx JA, Rumack BH. Magnesium levels after magnesium-containing cathartics. J Toxicol Clin Toxicol 1988;26(1–2):51–65. [6] Wang CY, Gastrointestinal MRI. Foreign Med Sci Clin Radiol Fascicle 2004;27(3):185–7. [7] Marin D, Husarik DB, Boll DT, Merkle EM. Abdominal magnetic resonance imaging at 3 T: oncological applications. Top Magn Reson Imaging 2010;21(3):149–56. [8] Asbach P, Breitwieser C, Diederichs G, Eisele S, Kivelitz D, Taupitz M, et al. Cine magnetic resonance imaging of the small bowel: comparison of different oral contrast media. Acta Radiol 2006;47(9):899–906. [9] McKenna DA, Roche CJ, Murphy JM, McCarthy PA. Polyethylene glycol solution as an oral contrast agent for MRI of the small bowel in a patient population. Clin Radiol 2006;61(11):966–70. [10] Frokjaer JB, Larsen E, Steffensen E, Nielsen AH, Drewes AM. Magnetic resonance imaging of the small bowel in Crohn’s disease. Scand J Gastroenterol 2005;40(7):832–42. [11] Cui SQ, Jin DQ. Applying experience of 25% magnesium sulfate for intestinal clearance in 1100 patients. J Chin Endosc 2004;10(3):78–9.
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Magnesium sulfate as an oral contrast medium in magnetic resonance imaging of the small intestine Hao Shi ∗ , Cun Liu, Hong Yu Ding, Chun Wei Li Department of Diagnostic Imaging, Shandong Qianfoshan Hospital, Jingshi Street 16766, Jinan, Shandong Province 250014, PR China
a r t i c l e
i n f o
Article history: Received 15 May 2011 Received in revised form 23 November 2011 Accepted 25 November 2011 Keywords: Administration Oral Contrast media Small intestine
a b s t r a c t Background: To explore the use of magnesium sulfate (MgSO4 ) as an oral contrast medium (CM) in MRI of the small intestine. Methods: By comparing MgSO4 SNRs at different concentrations, we determined that 2.5% MgSO4 is the ideal concentration for small bowel MRI. Twenty volunteers underwent MRI after drinking 2.5% MgSO4 . Thirty-one patients with clinical suspicion of small intestinal pathology underwent both MRI and the air-barium contrast examination. The patient’s tolerance, side effects and complications were noted. Results: 2.5% MgSO4 can decrease the absorption of water and fully fill the enteric cavity, thereby increasing the contrast between the intestinal wall and lumen and facilitating radiographic examination of the small bowel. The mean diameter of the small intestine was 19.8 ± 1.21 mm in the 20 volunteers consuming 2.5% MgSO4 and 12.7 ± 0.84 mm in the 20 volunteers given water. There was a significant difference (P < 0.05) between the diameters of the small intestine of the two groups. But there were no significant differences (P > 0.05) in side effects between MgSO4 and water groups. Small intestinal MRI was successfully performed in all 31 patients, who were also examined by the double contrast barium, which gave almost identical diagnoses to MRI in all cases except for 1 patient with small intestinal hemorrhage. Conclusions: MRI with 2.5% MgSO4 can demonstrate intestinal abnormalities. Therefore, 2.5% MgSO4 solution is an ideal oral CM for small bowel MRI. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Magnetic resonance imaging (MRI) has been used extensively in clinical practice as it offers high resolution of soft tissues, allows for multiplanar reconstructions and does not expose the patient to radiation. However, MRI has not been extensively applied in the clinical evaluation of diseases of the gastrointestinal tract, although its potential utility in this regard is currently under investigation [1–3]. The use of oral contrast media (OCM) is critical to the MRI examination of the small intestine. Most OCM are expensive, poorly tolerated by patients, have considerable side effects and imaging artifacts, all of which limit their use in intestinal MRI [1,4]. Water was frequently used as an OCM for MRI examination of the small intestine. However, water cannot completely fill the small intestine as it is rapidly absorbed [1–4]. Magnesium sulfate (MgSO4 ) can be easily dissolved in the water. Mg2+ and SO4 2+ cannot be absorbed in the small intestine and can produce a high osmotic pressure (OP)
∗ Corresponding author at: Department of Diagnostic Imaging, Shandong Qianfoshan Hospital, Jingshi Street 16766, Jinan, Shandong Province 250021, PR China. Tel.: +86 531 89268750; mobile: +86 18653113953. E-mail addresses: [email protected], [email protected] (H. Shi). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.11.049
[5]. In turn, this high OP may slow the absorption of water and allow for adequate filling of the small intestinal lumen. We developed a new MRI OCM and had performed MRI examination of the small intestine with oral administration of MgSO4 to assess its utility in the diagnosis of small bowel pathology.
2. Materials and methods 2.1. MRI technique All patients were examined by 1.5 Tesla magnetic resonance tomography (Sonata, Siemens Medical Solutions, Germany) with an abdominal phased-array coil with 6 segments. Routine scanning sequences for abdominal examination were used, which included GRE (gradient echo) T1 WI (Flash2d) with parallel imaging and without gating technique (TR/TE 110/4.8 ms, FOV 298 × 400 mm, matrix 256 × 192, thickness 6 mm, turbo factor 2, scan time 19 s), GRE (gradient echo) FS-T1 WI (Flash2d) without parallel imaging and gating technique (TR/TE 226/7.2 ms; FOV 261 × 350 mm, matrix 256 × 192, thickness 6 mm, scan time 21 s), GRE (gradient echo) T2 WI (trueFISP, true Fast Imaging with Steady-state Procession) without parallel imaging and gating technique (TR/TE 3.7/1.8 ms, FOV 360 × 360 mm, matrix 512 × 512, thickness 6 mm, scan time
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
e371
19 s) and TSE (turbo Spin echo) T2 WI (Haste, Half-Fourier Acquisition Single-shot) without parallel imaging and gating technique (TR/TE 1040/72 ms, FOV 269 × 360 mm, matrix 512 × 384, thickness 6 mm, scan time 17 s). Since MgSO4 presented high signal intensity like water, T2 WI sequences were more used for a clearer contrast between the media and surrounding structures. Scanning directions included axial, coronal and sagittal planes. All study protocols were approved by our institutional review board, and informed consent was obtained prior to each examination. 2.2. Determination of the optimal concentration of contrast The different concentrations of MgSO4 sample (50%, 40%, 30%, 20%, 10%, 5%, 2.5%, 1.25%, 1% and 0.5%) and water (0% as control group) were placed into the different test tubes, and then scanned using the different MRI sequences described above. The OPs of the different concentrations were calculated by the Van’t Hoff OP formula (Table 1) and SNR of the different concentrations of MgSO4 samples were compared. The concentration of MgSO4 with the highest osmotic pressure and most distinct imaging signal and SNR was selected for subsequent use. 2.3. Clinical study in healthy subjects To study the utility of the selected contrast medium, twenty volunteers (11 males, 9 females, aged between 18 and 50 years, mean 30 years) were tested with the selected OCM. None of the volunteers had any history of abdominal surgery, GI pathology or any contraindications to MRI. Subjects had not consumed a barium meal or CT oral contrast agents within 3 days, nor had received intravenous contrast within 24 h of study participation. The volunteers were not permitted to drink or eat within 12 h of the examination. Twenty volunteers were randomly divided into an experimental group (10 volunteers) and a control group (10 volunteers). 1000 ml of 2.5% MgSO4 was given orally 20 min before the examination. The volunteers in the experimental group drank 2000 ml of water just before the examination to completely fill the stomach and the duodenum. In the control group, the volunteers drank 1000 ml of the water 20 min before and 2000 ml just prior to the examination. 10 mg of anisodamine was injected IM 10 min before the examination in all volunteers to reduce artifacts that may result from peristalsis. For all volunteers included in both the experimental and control groups, examinations were performed at least 2 weeks apart.
Fig. 1. Schematic diagram of intestinal grouping. Group 1: duodenum; group 2: superior segment of jejunum; group 3: inferior segment of jejunum; group 4: superior segment of ileum; group 5: middle segment of ileum; group 6: inferior segment of ileum.
assigned. The small intestine was divided into six groups according to X-ray anatomy (Fig. 1). The dilation of the small intestinal tract and the image quality were scored and graded according to the criteria outlined in Tables 2 and 3. Adverse reactions occurring within 24 h of the examination were also recorded and graded by the criteria defined in Table 4. MRI and gas-barium double contrast fluoroscopy of the small intestine were performed in 31 patients with suspected small bowel pathology. All studies were interpreted independently by two radiologists who were not informed to which arm of the study the patients had been assigned. The radiologists evaluated the appearance of the mucosa, the intestinal wall and surrounding tissues and described the location, characteristics and involved structure of all lesions. Finally, the results were analyzed and compared with the pathological findings of biopsies and surgical specimens.
2.4. Clinical study in patients with small bowel pathology 2.6. Statistical analysis Between November 2004 and April 2006, 31 patients with clinical suspicion of small intestinal pathology, underwent intestinal MR examination after oral administration of MgSO4 . The methods of MR examination were the same as previously described above. The patients included 20 men and 11 women, aged between 25 and 80 years (mean age 34 years). Signs and symptoms suggestive of small bowel pathology included abdominal pain, diarrhea, weight loss, anemia, fever, abdominal masses and fecal occult blood. The course of the disease was from 2 months to 10 years. All 31 patients underwent both MRI and gas-barium double contrast fluoroscopy, which was conducted with either the SHIMADZU XED150L-20 500 mA X-ray machine, the Philips D-93 digit X-ray machine or the SHIMADZU PRO-series digit X-ray machine. 2.5. Evaluation of results Images of 20 volunteers (both the experimental and control groups) were analyzed by two radiologists with experience in MRI who was blinded as to which arm of the study volunteers had been
The dilation of the small intestine, image quality and adverse reactions between the experiment group and the control group were calculated with the rank-sum test (Wilcoxon rank test, Mann–Whitney test and Kruskal–Wallis test respectively). The means of the dilation were presented in mean ± SD. All data were statistically analyzed with SSPS 12.0 software. A p-value of less than 0.05 was considered significant. 3. Results 3.1. Optimization of MgSO4 concentration The use of MgSO4 resulted in a shortening of T1 and T2 relaxation times. However, this shortening of T1 and T2 relaxation times decreased with decreasing concentrations of MgSO4 (Fig. 2). At a concentration of 2.5%, the OP was 417 mmol/l, there was low signal intensity and SNR on T1 WI and high signal intensity and SNR on T2 WI (Fig. 2). The SNR of MgSO4 was similar to water, which
e372
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
Table 1 The osmotic pressure of MgSO4 at different concentrations. Concentration (%) Osmotic pressure (mmol/l)
50 8333
40 6666
30 5000
20 3333
10 1667
5 833
2.5 417
1.25 208
1 166
0.5 83
0 0
Table 2 Grading criteria for the dilation of the small intestine. Grade
Degree of dilation of the small intestine
Score
I II III IV
The diameter of the intestine was ≤1 cm, unable to distinguish the intestinal lumen and wall, poor dilation The diameter of the intestine was 1–1.5 cm, able to distinguish the intestinal lumen and wall, moderate dilation The diameter of the intestine was 1.5–2.0 cm, liquid contained in the intestinal lumen, good dilation The diameter of the intestine was >2.0 cm, lumen completely filled by liquid, excellent dilation
1 2 3 4
Table 3 Grading criteria for image quality. Grade
Image quality of the small intestine
Score
I
There is no signal distinction between the intestinal lumen and wall; poor resolution between the wall and surrounding tissues. There is signal distinction between the intestinal lumen and the wall; inhomogeneous signal in the lumen and moderate resolution between the wall and surrounding tissues. There is homogeneous signal in the intestinal lumen and good resolution between the lumen, wall and surrounding tissues. There is a homogeneous signal in the intestinal lumen and excellent resolution between the lumen wall and surrounding tissues.
1
II III IV
2 3 4
Table 4 Grading criteria for adverse reactions. Grade
Description of the adverse reaction
Score
I II III IV
There were no adverse reactions and no effects on the patient’s work or daily activities There was a slight adverse reaction which had minimal impact on the patient’s work and daily activities. There was a tolerable adverse reaction which affected the patient’s work and daily activities. There was a significant adverse reaction which could not be tolerated and prohibited the patient from working or engaging in other activities.
1 2 3 4
has low SNR on T1 WI and high SNR on T2 WI (Fig. 2). T2 WI had more markedly contrast between the lumen, the wall and the surrounding structures of the intestine and had a more useful value than T1 WI. Therefore, we selected 2.5% MgSO4 as an oral contrast medium for the following reasons. First, it is hyperosmolar and could therefore retain more water in the intestinal lumen. Second, a concentration of 2.5% is relatively low and would be less likely to result in the adverse reactions seen with the higher concentrations used for bowel preparation prior to colonoscopy.
3.2. Results of the clinical study Since T2 WI presented a hyperintensity on the images, it had a more obvious contrast between the lumen and the wall of intestine than T1 WI. Therefore, the clinical comparison study was performed mainly on T2 WI (trueFISP and Haste) sequences. The small intestine of all volunteers in the experimental group was fully dilated. On T2 WI, the liquid in the lumen of the small intestine had high signal intensity; the intestinal wall had moderate signal intensity.
Fig. 2. SNR of different concentrations of MgSO4 and water in different sequences.
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
e373
Table 5 Comparison of dilation scoring between the experimental and control groups. Location
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
Experiment group I
II
III
3 0 0 0 4 8
31 7 8 1 15 15
5 21 22 21 16 13
Control group IV 1 12 10 8 5 4
I
II
7 0 7 15 23 33
Rank-sum test
III
28 9 13 17 17 7
5 23 16 8 0 0
IV 0 8 4 0 0 0
p > 0.05 p > 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05
Table 6 Comparison of image quality scoring between the two groups.
Fig. 3. On T2 WI (TrueFISP), the small intestine is well-filled and completely dilated after the administration of 2.5% MgSO4 .
Hence, it was possible to distinguish between the lumen, wall and surrounding tissues (Fig. 3). The mean dilation of the small intestine was 19.8 ± 1.21 mm. The small intestine of volunteers in the control group did not adequately dilate and it was difficult to distinguish the lumen from the intestinal wall in some parts of the small intestine (Fig. 4). The average dilation of the small intestine was 12.7 ± 0.84 mm. There was a significant difference between the two groups in the dilation of the small intestine (P < 0.05). Grading of the dilation of the small intestine demonstrated that the small intestine in the experimental group was dilated better; the dilation of groups 3–6 of the small intestine in the control group was inferior to the experimental group; these differences were statistically significant (P < 0.05). The difference of the dilation between groups 1 and 2 was not statistically significant (P > 0.05, Table 5). Grading of image quality revealed significantly better image quality in the experimental group compared to the control group, in particular in groups 2–6 (P < 0.05, Table 6).
Location
Experiment group
Control group
I
II
III
IV
I
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
4 0 0 0 4 4
33 6 11 11 16 18
3 26 21 24 16 15
0 8 8 5 4 3
8 0 8 21 27 33
II
Rank-sum test
III
27 15 17 14 13 7
5 23 15 5 0 0
IV 0 2 1 0 0 0
p > 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05
Diarrhea occurred more frequently in the experimental group than the control group (P < 0.01). Urinary frequency occurred more frequently in the control group (P < 0.01). Differences in the frequencies of other adverse reactions, such as nausea, vomiting and enterospasm, between the two groups were not statistically significant (P > 0.05, Table 7). After administration of 2.5% MgSO4 , the normal bowel wall, which was sufficiently dilated, displayed moderate signal with 2.54 ± 4.35 mm of thickness both on T1 WI and T2 WI sequences and looked smooth and regular. With contrast, the bowel wall enhanced homogeneously and could be clearly distinguished from the intestinal lumen and extramural structures. The lumen of the small intestine had low signal intensity on T1 WI and remarkably high signal intensity on T2 WI. There were more mucosal folds in the jejunum, which were perpendicular to the long axis of the small intestine and looked like a spring. Relatively fewer mucosal folds were observed in the ileum, which were perpendicular or oblique to long axis of the small intestine. The number of the folds gradually decreased from the proximal to distal segments of the ileum. Almost no mucosal folds were seen in the terminal ileum. 3.3. Results of the clinical study in patients with small bowel pathology MRI was successfully performed in 31 patients with lesions of the small intestine. Twenty-eight patients exhibited good tolerance to 2.5% MgSO4 and the small intestine was adequately dilated in same extent as the experimental group in the clinical study. Slight nausea and abdominal distention occurred in three patients. Although these three patients took less than 1000 ml of 2.5% MgSO4 , the small bowel was still adequately dilated and a diagnosis could be made in each case without difficulty. In no case was the MR exam terminated due to diarrhea. MR examination with oral Table 7 Comparison of adverse reaction scoring in the two groups. Adverse reaction
Experiment group I
Fig. 4. On T2 WI (TrueFISP) the small intestine is not completely dilated and the wall of the small intestine is not clearly displayed.
Diarrhea 0 Nausea and vomiting 16 17 Enterospasm 0 Urinary frequency
II 6 4 3 12
III 10 0 0 8
Control group
Rank-sum test
IV
I
II
III
IV
4 0 0 0
18 15 16 0
2 5 4 4
0 0 0 10
0 0 0 6
p < 0.01 p > 0.05 p > 0.05 p < 0.01
e374
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
administration of 2.5% MgSO4 accurately demonstrated bowel-wall thickening, strictures, ulcers and masses of the small intestine, as well as extra-mural extension of lesions and their relationships to surrounding structures (Fig. 5). MRI examination with 2.5% MgSO4 , which permits visualization of the intestinal wall, lumen and also surrounding structures, is a better than gas-barium double contrast examination. The tissue pathology specimens from surgical procedures and endoscopy showed that there were 20 cases of Crohn’s disease, 8 cases of interstitialoma, 1 case of small intestinal hemorrhage, 1 case of intussusception and 1 case of tuberculosis. MRI diagnosed Crohn’s disease in 20 cases, intestinal tumor in 8 cases, intussusception in 1 case and tuberculosis in 1 case. MRI gave the correct diagnosis in all cases except for the single patient with small intestinal hemorrhage. The sensitivity of MRI was 96.7%. MRI and double contrast fluoroscopy gave identical diagnoses in 100% of patients.
4. Discussion Historically, gas-barium double contrast examination was the principal modality for imaging the small intestine. However, it is limited in its ability to image structures outside of the intestinal lumen. Interpretation may also be challenging when loops of bowel overlap, making it difficult to pinpoint the location of certain lesions. It also involves exposure to radiation and may not be well-tolerated by all patients. For these reasons, CT examination of gastrointestinal tract is now beginning to replace gas-barium fluoroscopy. However, CT also involves radiation exposure and may not provide adequate contrast of soft tissue structures. Until recently, the clinical utility of MRI in evaluation the GI tract was limited. MRI required long examination times and had low signal-to-noise ratios (SNR) [6]. Moreover, there was no suitable oral contrast medium [3,4]. In addition, image interpretation was limited by numerous artifacts from respiratory, cardiac and gastrointestinal motion. Presently, improvements (such as parallel imaging, gating technique) in MR technology have lead to shorter examination times and high SNR, both of which improve the applicability of this modality to imaging of the GI tract [7]. Nonetheless, it is necessary to identify an ideal OCM. The ideal OCM would be able to help the radiologist accurately visualize normal structures and characterize lesions of the small bowel [8]. The ideal OCM would have the following characteristics. (1) It provides distinct contrast between the contrast medium in the intestinal lumen and the intestinal wall and surrounding tissues without artifacts in any sequences. (2) It can completely fill and distribute throughout the lumen without being rapidly absorbed. (3) It is stable nature and has uniform contrast effects throughout the GI tract. (4) It is safe, non-toxic, does not stimulate the GI tract or result in adverse side effects, and has a taste tolerable to patients. (5) It is relatively inexpensive and can be produced and preserved easily [8–10]. Like mannitol, MgSO4 can stimulate peristalsis and catharsis and is used to prepare the GI tract for colonoscopy [11]. In this application, it is fast, convenient and has few side effects [11]. It is easily dissolved in water, but Mg2+ and SO4 2+ cannot be absorbed in the small intestine. The presence of these two ions in the lumen creates a hyperosmolar state, which helps retain water, thereby dilating the lumen. Our study indicated that MgSO4 at high concentrations shows high signal intensity on T1 WI and low signal intensity on T2 WI. However, at lower concentrations, we observed a lower signal intensity and SNR on T1 WI and high signal intensity and SNR on T2 WI; thus, at lower concentrations, this profile of signal intensities and SNR is similar to water. Therefore, 2.5% MgSO4 could be used as a hypointense OCM in T1 WI and a hyperintense OCM in
Fig. 5. Crohn’s disease. On coronal (a), saggital (b) and transverse (c) T2W (TrueFISP) images, the lumen of the terminal ileum is irregularly narrowed (arrow). The membrane mucosa is derangement and thickness, and the wall is thickened and exhibits cobble stoning on T2 WI.
H. Shi et al. / European Journal of Radiology 81 (2012) e370–e375
T2 WI. Since 2.5% MgSO4 has a relative high OP (417 mmol/l), it can reduce absorption of water in the GI tract and dilate the lumen sufficiently. It was superior to water in terms of its ability to dilate the small intestine; compared to the control group, it offered better image quality. All patients took 1000 ml of 2.5% MgSO4 20 min before the examination, and 2000 ml of water immediately prior to the examination. This allowed for dilation of the small intestine as 2.5% MgSO4 is hyperosmolar and not easily absorbed. The stomach and duodenum was filled with water in order to decrease the dose of MgSO4 and reduce artifacts from intestinal peristalsis, which could result from stimulation of the GI tract by MgSO4 . After administration of MgSO4 , there were no obvious adverse reactions, such as nausea, vomiting or constipation. However, patients reported a bitter in flavor and occasionally slight diarrhea, both of which were tolerable. Therefore, 2.5% MgSO4 is safe and is well tolerated as an OCM for MRI examination of the small intestine. This study demonstrated that MgSO4 as an OCM in MRI examination of the small intestine can facilitate the diagnosis of intestinal pathology, manifested by abnormalities of the intestinal mucosa, strictures and/or dilation of the lumen and thickening and masses of the intestinal wall. It is useful in estimating the infiltrative depth of the tumors in the intestinal wall and evaluating extension to structures outside of the GI tract. The results of MRI examination with administration of 2.5% MgSO4 in 31 patients with the diseases of the small intestine had a high concordance with gas-barium double contrast examination and tissue pathology. In comparison with gas-barium double contrast examination, the superiority of MRI with 2.5% MgSO4 as an OCM was: (1) the patient was not exposed to radiation; (2) image interpretation was not complicated by overlapping bowel loops, as typically occurs with gas-barium double contrast examinations; (3) soft tissue resolution was good; (4) it demonstrated lesions in the small intestine and surrounding structures, in particular metastases to other organs and the lymph nodes; (5) contrast media could be given, which further helps to characterize lesions; (6) MR hydrography could be performed, which can obtain images similar to gas-barium double contrast examination and observe the small intestine at arbitrary angles, without the limitations of patient positioning and (7) multi-planar and multi-parameter reconstructions, as well as other
e375
post-processing techniques such as virtual colonography, could be employed to assist the radiologist in rendering an accurate diagnosis. In conclusion, 2.5% MgSO4 as an OCM can adequately dilate the small intestine and permit visualization of the intestinal lumen, wall, and surrounding structures. It can facilitate the detection of lesions of the small intestine and their extramural extension. Therefore, MgSO4 is an excellent OCM for MRI examination of the small intestine. Conflicts of interest The authors (including Hao SHI, Cun LIU, Hong Yu DING, Chun Wei LI) have no conflicts of interest to declare. And also there is no financial support. References [1] Geraldes CF, Laurent S. Classification and basic properties of contrast agents for magnetic resonance imaging. Contrast Media Mol Imaging 2009;4(1):1–23. [2] Gilja OH, Hatlebakk JG, Odegaard S, Berstad A, Viola I, Giertsen C, et al. Advanced imaging and visualization in gastrointestinal disorders. World J Gastroenterol 2007;13(9):1408–21. [3] Maccioni F. Current status of gastrointestinal MRI. Abdom Imaging 2002;27(4):358–60. [4] Briggs RW, Wu Z, Mladinich CR, Stoupis C, Gauger J, Liebig T, et al. In vivo animal test of an artifact-free contrast agent for gastrointestinal MRI. Magn Reson Imaging 1997;15(5):559–66. [5] Smilkstein MJ, Steedle D, Kulig KW, Marx JA, Rumack BH. Magnesium levels after magnesium-containing cathartics. J Toxicol Clin Toxicol 1988;26(1–2):51–65. [6] Wang CY, Gastrointestinal MRI. Foreign Med Sci Clin Radiol Fascicle 2004;27(3):185–7. [7] Marin D, Husarik DB, Boll DT, Merkle EM. Abdominal magnetic resonance imaging at 3 T: oncological applications. Top Magn Reson Imaging 2010;21(3):149–56. [8] Asbach P, Breitwieser C, Diederichs G, Eisele S, Kivelitz D, Taupitz M, et al. Cine magnetic resonance imaging of the small bowel: comparison of different oral contrast media. Acta Radiol 2006;47(9):899–906. [9] McKenna DA, Roche CJ, Murphy JM, McCarthy PA. Polyethylene glycol solution as an oral contrast agent for MRI of the small bowel in a patient population. Clin Radiol 2006;61(11):966–70. [10] Frokjaer JB, Larsen E, Steffensen E, Nielsen AH, Drewes AM. Magnetic resonance imaging of the small bowel in Crohn’s disease. Scand J Gastroenterol 2005;40(7):832–42. [11] Cui SQ, Jin DQ. Applying experience of 25% magnesium sulfate for intestinal clearance in 1100 patients. J Chin Endosc 2004;10(3):78–9.