ORIGINAL ARTICLE
Diffusion-weighted MR Enterography for Evaluating Crohn’s Disease: How Does It Add Diagnostically to Conventional MR Enterography? Kyung-Jo Kim, MD,* Yedaun Lee, MD,† Seong Ho Park, MD,‡ Bo-Kyeong Kang, MD,‡ Nieun Seo, MD,‡ Suk-Kyun Yang, MD,* Byong Duk Ye, MD,* Sang Hyoung Park, MD,* So Yeon Kim, MD,‡ Seunghee Baek, PhD,§ and Hyun Kwon Ha, MD‡
Background: Diffusion-weighted imaging (DWI) is a novel technique to evaluate bowel inflammation in Crohn’s disease (CD). It is unknown whether and how DWI adds to the accuracy of conventional magnetic resonance enterography (MRE).
Methods: Fifty consecutive adults suspected of CD prospectively underwent clinical assessment, conventional MRE and DWI at b ¼ 900 sec/mm2
without water enema, and ileocolonoscopy within 1 week. MRE images were interpreted with proper blinding. Forty-four patients finally diagnosed with CD (male:female, 34:10; 26.9 6 6.1 yr) were analyzed. The per-segment accuracy of MRE for diagnosing active CD was assessed in the terminal ileum, right colon, and rectum using location-by-location matching with endoscopy as the reference standard.
Results: The study evaluated 58 bowel segments with deep or superficial ulcers, 34 with aphthae, erythema, or edema only, and 35 without inflammation. Conventional MRE + DWI was more sensitive for bowel inflammation than conventional MRE alone (83% [76/92] versus 62% [57/92]; P ¼ 0.001) largely because of additional detection of aphthae, erythema, or edema. The sensitivities for deep and overt ulcers were similar regardless of DWI, ranging from 88% to 97%. Conventional MRE + DWI was less specific than conventional MRE alone (60% [21/35] versus 94% [33/35]; P , 0.001), mostly because of many false positives in the colorectum. Positive DWI findings in the bowel showing active inflammation on conventional MRE were associated with higher Crohn’s disease endoscopic index of severity score (P ¼ 0.021) and deep ulcers (P ¼ 0.01; diagnostic odds ratio, 12).
Conclusions: DWI performed without water enema is not useful for incremental detection of bowel inflammation. DWI may help identify more severe inflammation among bowel segments showing active inflammation on conventional MRE. (Inflamm Bowel Dis 2015;21:101–109) Key Words: Crohn’s disease, inflammatory bowel disease, diffusion, magnetic resonance, enterography, diagnostic accuracy
C
rohn’s disease (CD) is a chronic inflammatory bowel disease that can involve the entire gastrointestinal tract, especially the small and large bowels. As chronic inflammation in CD leads
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.ibdjournal.org). Received for publication July 19, 2014; Accepted August 14, 2014. From the *Department of Gastroenterology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea; †Department of Radiology, Inje University College of Medicine, Haeundae Paik Hospital, Busan, Korea; ‡Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea; and §Department of Clinical Epidemiology and Biostatistics, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea. Supported by a grant from Dongkook Pharmaceutical, Seoul, South Korea. S. Ho Park reports grants from DongKook Pharm for the submitted study and from GE Healthcare outside the submitted work. The authors have no conflicts of interest to disclose. Kyung-Jo Kim and Yedaun Lee equally contributed to this work. Reprints: Seong Ho Park, MD, Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpagu, Seoul 138-736, Korea (e-mail: [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000222 Published online 28 October 2014.
Inflamm Bowel Dis Volume 21, Number 1, January 2015
to irreversible bowel damage, monitoring of disease activity and bowel damage is important, especially because mucosal healing using biologics or immunosuppressive therapy is considered a therapeutic goal.1 Magnetic resonance enterography (MRE) plays an important role in the diagnosis and follow-up of patients with CD.2– 4 Recently, increased need for repeat follow-up imaging to monitor responses to therapeutic interventions aimed at mucosal healing has further accentuated the role of MRE, because MRE has an advantage of the lack of radiation exposure over computed tomographic enterography.2–4 Diffusion-weighted imaging (DWI) is a specialized MR imaging technique that maps the diffusion of water molecules in biological tissues and has been used largely for the diagnosis of ischemia in the nervous system and for oncologic imaging. Recently, studies have shown that restricted diffusion in the bowel wall on DWI in patients with CD is associated with the presence of active bowel inflammation and has a good correlation with bowel inflammatory activity estimated using conventional MRE (i.e., T2and T1-weighted images with intravenous contrast enhancement).5– 13 However, to our knowledge, few studies to date have compared DWI and conventional MRE, which is currently the standard method of MRE practice, regarding their relative strengths and www.ibdjournal.org |
101
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Kim et al
weaknesses using independent reference standards. Therefore, although the promising potential of DWI as a new imaging biomarker of bowel inflammation in CD has been revealed by the previous studies, it is still unknown whether DWI has an incremental diagnostic value compared with conventional MRE and whether and how DWI should be added to the current standard practice of MRE. Moreover, many previous studies had methodological weaknesses, including their retrospective study design5–8,10,11; small sample sizes5,6,8,9; and/or lack of proper blinding for the review of DWI, in that knowledge of conventional MRE findings before DWI may have affected/biased the interpretation of DWI results.6–11 For these reasons, it would now be important to evaluate whether DWI has an additive diagnostic value compared with conventional MRE method and to determine how the new technique should be incorporated into the clinical practice of MRE to evaluate CD. This study was to prospectively determine whether and how DWI adds to the diagnostic accuracy of conventional MRE to evaluate bowel inflammation in patients with CD.
MATERIALS AND METHODS This prospective study was approved by the Institutional Review Board of Asan Medical Center. Informed consent was obtained from the participants.
Subjects Between October 2012 and December 2013, adults ($18 yr) referred to the Inflammatory Bowel Disease Center of Asan Medical Center, an academic referral institution, with suspected
CD or after an initial diagnosis of CD were recruited for the study. Patients with a history of bowel resection, those who required emergency care related to bowel disease (such as severe bowel obstruction), and those with contraindications for MR examinations or spasmolytics (i.e., Buscopan) used for MRE were excluded. Fifty-four patients were consecutively enrolled, but 4 withdrew before the study. The remaining 50 patients, consisting of 38 men and 12 women (mean 6 SD age, 27.7 6 6.4 yr), underwent all study procedures. This study was designed as an exploratory study; therefore, the sample size was chosen empirically. The 44 patients finally confirmed as having CD according to conventional clinical, radiologic, endoscopic, and histopathologic criteria14 were analyzed (Fig. 1).
Study Procedures Patients underwent clinical assessments including measurements of Crohn’s disease activity index, laboratory tests, MRE, and ileocolonoscopy within 1 week. Laboratory tests included serum concentrations of erythrocyte sedimentation rate, C-reactive protein, and albumin, platelet, and white blood cell counts in peripheral blood. Patients were not allowed to take any medications for CD over the course of these examinations.
Magnetic Resonance Enterography MRE examinations were performed after 6 hours of fasting. Patients were instructed to drink 1500 mL of 2.5% sorbitol solution to distend the bowel before MR scanning. Water enema was not used. MR scanning was performed using a 3-T system (Ingenia; Philips Healthcare, Best, the Netherlands). MRE imaging consisted of coronal T2-weighted half-Fourier sequences
FIGURE 1. Study flow diagram.
102
| www.ibdjournal.org
Inflamm Bowel Dis Volume 21, Number 1, January 2015
with and without fat suppression, coronal, and axial T2-like steady-state gradient echo sequences with fat suppression, coronal DWI (with b factors of 0 and 900 sec/mm2) and an apparent diffusion coefficient (ADC) map, coronal dynamic fatsuppressed spoiled gradient-echo T1-weighted unenhanced and contrast-enhanced (enteric and portal-venous phases) imaging, and axial fat-suppressed spoiled gradient-echo T1-weighted delayed contrast-enhanced scan. For contrast enhancement, 0.2 mL/kg body weight of gadoterate meglumine (Dotarem; Guerbet, Villepinte, France) was injected intravenously at a rate of 2 mL/sec followed by a saline flush. To prevent bowel peristalsis, 10 mg of scopolamine-N-butyl bromide (Buscopan; Boehringer Ingelheim, Ingelheim, Germany) was administered intravenously 3 times at intervals during the scanning. Field of view was adjusted to include the entire small and large bowels. Further details of scan parameters are provided in the Appendix, Supplemental Digital Content 1, http://links.lww.com/IBD/A623. MRE was interpreted in consensus by 2 board-certified faculty radiologists experienced in MRE of inflammatory bowel disease to achieve a more factual correlation between MRE and endoscopy, given that some degree of interobserver variability exists in interpreting MRE.15,16 Interobserver agreement between the 2 study readers was then separately analyzed (see Appendix, Supplemental Digital Content 1, http://links.lww.com/IBD/ A623). The readers were informed about the length of the terminal ileum evaluated by endoscopy (see Reference Standard: Ileocolonoscopy) so that the same terminal ileal lengths could be evaluated by MRE and endoscopy, resulting in an accurate location-by-location match between MRE and ileocolonoscopy. Otherwise, the readers were blinded to the results of ileocolonoscopy, clinical assessment, and laboratory tests. The technical adequacy of each MRE examination was evaluated, and examinations deemed inadequate for diagnostic evaluation were excluded from analysis. Bowel inflammation was assessed during 2 separate reading sessions. In the first session, conventional MRE (i.e., all sets of images except for DWI and ADC maps) was read. The colorectum and terminal ileum were divided into 5 segments: the terminal ileum, right colon (cecum and ascending), transverse colon, left colon (descending and sigmoid), and rectum according to the Crohn’s disease endoscopic index of severity (CDEIS).17 Signs of active bowel inflammation in each segment were recorded, including mural thickening, mural hyperenhancement, increased mural signals on T2-weighted images (i.e., bowel wall edema) and/or perienteric edema, increased vasa recta (i.e., the comb sign), and perienteric abnormalities related to penetrating disease (i.e., sinus tract, fistula, phlegmon, or abscess). The presence of individual ulcers was not recorded separately because of the difficulty in consistently distinguishing among individual ulcers and distinguishing them from deformed luminal contours on cross-sectional images, although recognition of individual ulcers on MRE is possible in some cases. After a washout period of at least 1 month and random reshuffling of the case review order to prevent recall bias,18 the second reading session interpreted DWI and ADC maps in the absence of conventional MRE images
Diffusion-weighted MRE in CD
to prevent any influence/bias due to prior knowledge of conventional MRE findings. Because of the lack of anatomical details on DWI, the areas to be evaluated on DWI were premarked on the images by a separate coordinating investigator. DWI was interpreted in a dichotomous manner and bowel segments with high signal intensity on DWI, at least similar to that of the spleen and mesenteric lymph nodes on b ¼ 900-sec/mm2 images, and with low signals on the corresponding ADC map were regarded as active inflammation (Figs. 2–4). Semiquantitative MRE scores of bowel inflammatory activity for each segment were determined by assigning one point to each MRE finding of active bowel inflammation and summing the points, in a manner similar to the methods used in related studies.7,19 A per-patient MRE score was then derived by taking the average of the segmental scores in a manner similar to global CDEIS scores are calculated.
Reference Standard: Ileocolonoscopy Ileocolonoscopy findings were used as the reference standard. Ileocolonoscopy was performed by either one of the 3 boardcertified faculty gastroenterologists who were experienced in endoscopic examination of patients with inflammatory bowel disease and blinded to the results of MRE, clinical assessment, and laboratory tests, using a video colonoscope (CF H260AL or CF H260AI; Olympus Optical Co., Tokyo, Japan). Colons were prepared using 4 L of polyethylene glycol. The length of the terminal ileum endoscopically evaluated was recorded. Endoscopic lesions were evaluated according to CDEIS.17 Global and segmental CDEIS scores were determined. The types of inflammatory mucosal lesions were specified, including overt (deep and superficial) ulcers, aphthoid lesions, and erythema or edema. Absence of any of these lesions was considered the state of absence of disease (to constitute the denominator for calculating specificity).
Data and Statistical Analysis Per-patient MRE scores of bowel inflammatory activity obtained either with conventional MRE alone (ranging from 0 to 5) or with combined conventional MRE plus DWI (ranging from 0 to 6) were correlated with global CDEIS scores using the Spearman’s correlation analysis. The results were compared between themselves and with the degree of correlation between Crohn’s disease activity index or serum C-reactive protein and global CDEIS scores. Per-bowel segment analysis was also performed and only included the terminal ileum, right colon, and rectum. The selection of these 3 segments was to avoid any uncertainty related to location mismatches between MRE and endoscopy. It is well known that segmental designations of the same colonic location on radiologic imaging and colonoscopy often do not coincide,20–22 e.g., transverse colon on one method but descending or ascending colon on the other method, particularly in areas where endoscopic landmarks are lacking. Endoscopy can locate the tip more straightforwardly in the 3 segments due to the presence of or the proximity to anatomical landmarks and boundaries, allowing for a more accurate location-by-location correlation between MRE and endoscopy. We analyzed the sensitivity and specificity of conventional MRE www.ibdjournal.org |
103
Kim et al
Inflamm Bowel Dis Volume 21, Number 1, January 2015
FIGURE 3. False-negative conventional MRE but positive DWI findings in the rectum of a 37-year-old man with CD. A, The rectum (arrowheads) shows no signs of active inflammation on a contrast-enhanced T1-weighted image (left), but a diffusely high signal on DWI (middle) and a low signal on an ADC map (right). B, Colonoscopic image of the rectum showing numerous scattered tiny aphthoid lesions.
FIGURE 2. True-positive conventional MRE and DWI findings in the terminal ileum of a 38-year-old woman with CD. A, The terminal ileum (arrowheads) shows diffuse thickening, increased mural signal on a T2weighted image (left), and mural hyperenhancement (right) and prominently increased vasa recta (right). B, The terminal ileum (arrowheads) shows a high signal on DWI (left) and a dark signal on an ADC map (right). C, Endoscopic image of the terminal ileum showing multiple deep longitudinal ulcers.
104
| www.ibdjournal.org
alone and combined conventional MRE plus DWI in diagnosing active bowel inflammation. The combined conventional MRE plus DWI was regarded as positive for active bowel inflammation if either was positive. Subgroup analyses were performed in terms of severity of bowel inflammation, defined as segments with deep ulcers, overt (deep or superficial) ulcers, or less severe lesions only (aphthae, erythema, or edema without overt ulcers)23; and in terms of bowel location, defined as the terminal ileum or colorectum. Statistical comparisons between the 2 MRE methods were made using the generalized linear mixed model to address the correlation between repeated observations in the same bowel segment using different MRE methods and the nested data structure (i.e., multiple segments per patient). Any notable differences in the diagnostic performance were further elaborated on with a cross-tabulation of conventional MRE and DWI interpretations. Second, the incremental role of DWI in bowel areas showing active inflammation on conventional MRE was evaluated. The relationships of DWI findings with segmental CDEIS scores and the presence/absence of deep ulcers in those bowel segments positive for active inflammation as seen on conventional MRE were analyzed using the Student’s t test and the Fisher’s exact test, respectively. The ability of
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Diffusion-weighted MRE in CD
TABLE 1. Characteristics of Study Subjects Patients included (n ¼ 44) Age, mean 6 SD, yr Gender (male:female) Body mass index, mean 6 SD, kg/m2 CDAI, mean 6 SD Laboratory findings ESR, mean 6 SD, mm/h CRP, range (median), mg/dL Albumin, mean 6 SD, g/dL Platelet count, mean 6 SD, ·103/mL White blood cell count, mean 6 SD, ·103/mL Global CDEIS score, mean 6 SD Bowel segments included in the persegment analysis (n ¼ 127) Anatomical location Terminal ileum Right colon Rectum Endoscopic findings Types of lesions Overt (deep or superficial) ulcers FIGURE 4. Negative conventional MRE and false-positive DWI findings in the ascending colon of a 24-year-old woman with CD. A, The lateral wall of the proximal ascending colon (arrowheads) shows no signs of active inflammation on a contrast-enhanced T1-weighted image (left) but a high signal on DWI (middle) and a low signal on an ADC map (right). B, Colonoscopic image of the proximal ascending colon showing no signs of inflammation or ulcer. IC, ileocecal valve.
semiquantitative MRE scores of bowel inflammation to distinguish between segments with and without deep ulcers among those segments positive for active inflammation as seen on conventional MRE was compared between conventional MRE alone and combined conventional MRE plus DWI using the receiver operating characteristic analysis. All statistical analyses were performed using IBM SPSS Statistics for Windows Version 21.0 (IBM Corp., Armonk, NY) and MedCalc (version 13.1; MedCalc Software, Mariakerke, Belgium). P , 0.05 was considered statistically significant.
RESULTS
Aphthae, erythema, or edema only No inflammation Segmental CDEIS score Terminal ileum (range and median) Colorectum (range and median)
26.9 6 6.1 34:10 20.0 6 2.9 201.53 6 111.14 42 6 26 0.1–17.66 (1.86) 3.6 6 0.6 325 6 89 7.3 6 2.5 10.8 6 6.9
41 42 44
58 (terminal ileum, 31; colorectum, 27) 34 (terminal ileum, 3; colorectum, 31) 35 (terminal ileum, 7; colorectum, 28) 0–34 (median, 15) 0–38 (median, 4)
CDAI, Crohn’s disease activity index; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
summarized in Table 1. The time interval between MRE and endoscopy ranged from 0 to 7 days (median, 1 d). The entire colon and terminal ileum were examined endoscopically in 41 patients; the length of the terminal ileum evaluated ranged from 1 to 20 cm (median, 10 cm) and was $10 cm in 36 patients. Of the remaining 3 patients, one each was endoscopically evaluated up to the cecum, the sigmoid colon, and the rectum. As a result, a total of 127 segments, 92 with and 35 without active inflammation, were included in the per-segment analysis (Table 1).
Subjects
Per-patient Analysis
All 50 patients underwent all the study procedures without any adverse events. Of those, 3 were finally diagnosed with a condition other than CD, and 3 others (6% of 47) had MR images deemed below diagnostic quality due to various artifacts. These 6 patients were therefore excluded from analysis. The remaining 44 patients constituted the ultimate cohort for analysis (Fig. 1); their demographic and clinical characteristics are
The correlation coefficient between MRE scores of bowel inflammation and global CDEIS scores was 0.819 (95% confidence interval, 0.690–0.898; P , 0.001) for conventional MRE alone and was similarly 0.823 (0.696–0.900; P , 0.001) for combined conventional MRE plus DWI. These were higher than the correlation coefficients between Crohn’s disease activity index and global CDEIS scores, 0.443 (0.168–0.654; P ¼ 0.003), and www.ibdjournal.org |
105
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Kim et al
TABLE 2. Rates of Positive MRE Results According to the Severity of Bowel Inflammation Endoscopic Findings
Aphthae, Erythema, or Edema Only (n ¼ 34)
Superficial Ulcers (n ¼ 26)
Deep Ulcers (n ¼ 32)
All (n ¼ 92)
18 (6/34) 53 (18/34)
77 (20/26) 73 (19/26)
97 (31/32) 94 (30/32)
62 (57/92) 73 (67/92)
Active inflammation on conventional MRE Active inflammation on DWI
Data are percentages with the number of bowel segments provided in the parentheses.
between serum C-reactive protein and global CDEIS scores, 0.525 (0.271–0.711; P , 0.001).
Per-segment Analysis: MRE Accuracy for Diagnosing Active Bowel Inflammation Both conventional MRE and DWI showed increasing rates of positive test results with increasing severity of bowel inflammation, and DWI revealed an apparently higher test-positive rate compared with conventional MRE in segments with mild inflammatory lesions without overt ulcers (aphthae, erythema, or edema only) (Table 2). Therefore, overall sensitivity for all types of inflammatory bowel lesions became significantly higher when DWI was added to conventional MRE (83% [76/92]) compared with conventional MRE alone (62% [57/92]) (P ¼ 0.001) (Table 3). By contrast, the specificity of combined conventional MRE plus DWI (60% [21/35]) was significantly lower than that of conventional MRE alone (94% [33/ 35]) (P , 0.001) (Table 3). Subgroup analyses showed that the sensitivity of conventional MRE plus DWI was significantly greater than that of conventional MRE alone for mild inflammatory lesions without overt ulcers (62% [21/34] versus 18% [6/34]; P , 0.001), whereas there were no significant differences in sensitivity for deep
and overt ulcers (Table 3). When analyzed according to bowel location, conventional MRE plus DWI showed significantly higher sensitivity but lower specificity than conventional MRE alone in the colorectum (Table 3). Representative findings are shown in Figures 2–4. Conventional MRE and DWI interpretations concurred in 68.5% (87 of 127) of the segments (Table 4). Of 92 bowel segments with active inflammation at endoscopy, 19 were diagnosed as such solely by DWI (Table 4), and 14 of these were in the colorectal segments with only aphthae, erythema, or edema without overt ulcers. DWI falsely diagnosed active bowel inflammation in 12 segments where conventional MRE correctly diagnosed the absence of inflammation (Table 4), with 11 of these segments in the colorectum.
Per-segment Analysis: DWI in the Bowel Segments Showing Active Inflammation on Conventional MRE A total of 59 bowel segments (32 terminal ileal and 27 colorectal segments) were positive for active bowel inflammation as seen on conventional MRE. Of these, the 50 DWI-positive segments showed significantly higher segmental CDEIS scores than the 9 DWInegative segments (21 6 10.1 versus 12.6 6 8.4; P ¼ 0.021). Positive
TABLE 3. Accuracy of MRE for Diagnosing Bowel Inflammation Sensitivity
All types of bowel inflammationa Subgroups according to inflammatory severity Deep ulcers Overt (deep or superficial) ulcers Aphthae, erythema, or edema only Subgroups according to bowel location Terminal ileum Colorectum
Specificity
Conventional MRE
Combined Conventional MRE + DWI
62 (57/92)
83 (76/92)
0.001
97 (31/32) 88 (51/58) 18 (6/34)
97 (31/32) 95 (55/58) 62 (21/34)
1 0.171 ,0.001
NA NA NA
91 (31/34) 45 (26/58)
94 (32/34) 76 (44/58)
0.650 ,0.001
86 (6/7) 96 (27/28)
P
Data are percentages with the number of bowel segments provided in the parentheses. a Includes overt (deep or superficial) ulcers and less severe lesions such as aphthae, erythema, or edema. NA, not applicable.
106
| www.ibdjournal.org
Conventional MRE
Combined Conventional MRE + DWI
P
94 (33/35)
60 (21/35)
,0.001
71 (5/7) 57 (16/28)
0.560 0.003
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Diffusion-weighted MRE in CD
TABLE 4. Cross-tabulation of the Results of MRE Interpretation Segments with Active Inflammationa at Reference Segments without Active Inflammation at Reference Standard (n ¼ 92) Standard (n ¼ 35) DWI Interpretation
Conventional MRE interpretation Positive Negative Total
DWI Interpretation
Positive
Negative
Total
Positive
Negative
Total
48 19 67
9 16 25
57 35 92
2 12 14
0 21 21
2 33 35
Data are numbers of bowel segments. a All types of inflammatory lesions including overt (deep or superficial) ulcers and less severe lesions such as aphthae, erythema, or edema.
DWI findings were significantly associated with the presence of deep ulcers (60% [30/50] versus 11% [1/9]; P ¼ 0.01; diagnostic odds ratio, 12). The area under the receiver operating characteristic curve of the semiquantitative MRE scores for distinguishing bowel segments with and without deep ulcers among bowel segments positive for active inflammation as seen on conventional MRE was significantly higher for combined conventional MRE plus DWI than for conventional MRE alone (0.720 versus 0.661; P ¼ 0.029) (Fig. 5).
DISCUSSION The results presented here reconfirm the association between restricted diffusion in the bowel wall as seen on DWI and active bowel inflammation in patients with CD. The study
FIGURE 5. Receiver operating characteristic curves showing the ability of semiquantitative MRE scores to distinguish bowel segments with and without deep ulcers among segments showing active inflammation on conventional MRE.
also demonstrates that enterographic DWI performed without water enema generally has limited incremental benefit compared with conventional MRE in diagnosing bowel inflammation in CD. The increased detection of bowel inflammation by addition of DWI was mostly because of the increased identification of mild inflammation, consisting of aphthae, erythema, or edema without overt ulcers in the colorectum. By contrast, the sensitivity of conventional MRE was already quite high for overt ulcers, particularly for deep ulcers, with the addition of DWI having little effect on the sensitivity of MRE for such lesions. Although deep ulcers correlate with a more aggressive clinical course24 and should be considered targets for disease-modifying therapy,25 the natural history and clinical significance of mild inflammatory lesions without overt ulcers remain doubtful.26,27 Therefore, it is uncertain whether the ability of DWI to enhance the detection of mild bowel inflammation represents a true clinical benefit. By contrast, it is noteworthy that positive DWI findings may suggest more severe inflammation and deep ulcers in bowel segments showing active inflammation as seen on conventional MRE. Consistent with our result, previous studies suggested that ADC values measured in the bowel wall on DWI may be used as a parameter to assess bowel inflammatory severity in CD as they demonstrated a good correlation with bowel inflammatory activity estimated using conventional MRE.9,12 Taken together, these results indicate that DWI may have a role in further stratifying patients with CD with known bowel inflammation, regarding their inflammatory severity or suitability for disease-modifying therapy rather than identifying bowel inflammation undetected by conventional MRE. Another point raised by our study is the low specificity of DWI especially in the colorectum when performed without water enema. MRE is used in patients with CD primarily to examine bowel segments difficult to reach by endoscopy,2,4 not for preendoscopic screening, making high specificity an important feature of MRE. Therefore, the low specificity of DWI seems worrisome. The low specificity of DWI in our study (60%) was similar to the 62.6% specificity reported in a retrospective analysis of a prospectively developed patient registry.7 Although a few other studies reported numerically greater specificity values (81.4%–82.4% for the small and large bowels and 75.7% for the large bowel),5,6 those www.ibdjournal.org |
107
Kim et al
studies were retrospective in design, included relatively few patients (with many being surgical patients), and used suboptimal reference standards. Moreover, as DWI was interpreted without blinding to conventional MRE results and the definitions of positive DWI findings were somewhat vague (i.e., “higher signal intensity than adjacent normal-appearing bowel loops,” with DWI itself used to distinguish between normal and abnormal bowel segments) in the previous studies,5–7 the previous studies may have overestimated the specificity of DWI. Most of the false-positive DWI results in our study occurred in the colorectum, similar to previous findings.5,6 The high rate of DWI false positives in the colorectum and the finding that most of the additional areas of bowel inflammation detected by DWI were in the colorectum in our study may also suggest the possible fortuitous coincidence of DWI false positives in colorectal segments with subtle inflammation below the resolution of MRE in some cases. The predilection of DWI false positives in the colorectum may be related to the generally suboptimal preparation/distention state of the colorectum when MRE is performed without water enema.28 It may be worthwhile to investigate whether these DWI false positives could be reduced by water enema in addition to oral administration of fluid.23,29,30 Our study had several methodological strengths compared with previous related studies. However, unlike previous studies, we focused on visual interpretation of DWI and did not measure ADC values. Despite the advantages of quantitative analysis, measuring absolute ADC values has generally shown limited reproducibility in the abdomen.31–36 Moreover, obtaining ADC values from the bowel wall in patients with CD is likely further confounded by the presence of multiple diseased segments and small sizes of individual bowel walls, introducing additional variability in measurement.37 Therefore, the quantitative approach is rarely used in clinical practice, although it is plausible in a controlled research setting.9,12 In our study, we intended to evaluate the clinical value of DWI from a more practical viewpoint. This study had limitations. First, the number of terminal ileal segments devoid of inflammation or with only mild inflammation without overt ulcers was small. The incidence of false-positive DWI results in terminal ileal segments should be further confirmed in a larger patient sample. Second, MRE interpretation in this study did not consider the length of bowel involvement, which may have introduced some imprecision in correlating MRE and endoscopy. Third, this study focused on incremental benefits of DWI when it is added to conventional MRE regarding evaluating bowel inflammation at single time points. Other roles of DWI, such as for longitudinal monitoring after medical therapy, as a substitute for intravenous contrast enhancement,9,12 or in evaluating bowel damage and complications, would require further studies. Finally, small bowel segments other than the terminal ileum were not evaluated. However, this is an inevitable methodological limitation of almost any comparative/correlative research studies between radiological bowel imaging and endoscopy because location-by-location matching is virtually impossible for the small bowel except for the terminal ileum. In conclusion, DWI performed without water enema is likely not useful for additional detection of bowel inflammation in
108
| www.ibdjournal.org
Inflamm Bowel Dis Volume 21, Number 1, January 2015
CD undetected by conventional MRE as abnormalities solely detected by DWI were mostly nonspecific inflammation lacking strong clinical relevance or false positives particularly in the colorectum. DWI may help identify more severe inflammation and deep ulcers among bowel segments showing active inflammation on conventional MRE.
REFERENCES 1. Cheifetz AS. Management of active Crohn disease. JAMA. 2013;309: 2150–2158. 2. Benitez JM, Meuwis MA, Reenaers C, et al. Role of endoscopy, crosssectional imaging and biomarkers in Crohn’s disease monitoring. Gut. 2013;62:1806–1816. 3. Makanyanga JC, Taylor SA. Current and future role of MR enterography in the management of Crohn disease. AJR Am J Roentgenol. 2013;201: 56–64. 4. Pariente B, Peyrin-Biroulet L, Cohen L, et al. Gastroenterology review and perspective: the role of cross-sectional imaging in evaluating bowel damage in Crohn disease. AJR Am J Roentgenol. 2011;197:42–49. 5. Kiryu S, Dodanuki K, Takao H, et al. Free-breathing diffusion-weighted imaging for the assessment of inflammatory activity in Crohn’s disease. J Magn Reson Imaging. 2009;29:880–886. 6. Oto A, Zhu F, Kulkarni K, et al. Evaluation of diffusion-weighted MR imaging for detection of bowel inflammation in patients with Crohn’s disease. Acad Radiol. 2009;16:597–603. 7. Oussalah A, Laurent V, Bruot O, et al. Diffusion-weighted magnetic resonance without bowel preparation for detecting colonic inflammation in inflammatory bowel disease. Gut. 2010;59:1056–1065. 8. Oto A, Kayhan A, Williams JT, et al. Active Crohn’s disease in the small bowel: evaluation by diffusion weighted imaging and quantitative dynamic contrast enhanced MR imaging. J Magn Reson Imaging. 2011; 33:615–624. 9. Buisson A, Joubert A, Montoriol PF, et al. Diffusion-weighted magnetic resonance imaging for detecting and assessing ileal inflammation in Crohn’s disease. Aliment Pharmacol Ther. 2013;37:537–545. 10. Neubauer H, Pabst T, Dick A, et al. Small-bowel MRI in children and young adults with Crohn disease: retrospective head-to-head comparison of contrast-enhanced and diffusion-weighted MRI. Pediatr Radiol. 2013; 43:103–114. 11. Ream JM, Dillman JR, Adler J, et al. MRI diffusion-weighted imaging (DWI) in pediatric small bowel Crohn disease: correlation with MRI findings of active bowel wall inflammation. Pediatr Radiol. 2013;43: 1077–1085. 12. Hordonneau C, Buisson A, Scanzi J, et al. Diffusion-weighted magnetic resonance imaging in ileocolonic Crohn’s disease: validation of quantitative index of activity. Am J Gastroenterol. 2014;109:89–98. 13. Tielbeek JA, Ziech ML, Li Z, et al. Evaluation of conventional, dynamic contrast enhanced and diffusion weighted MRI for quantitative Crohn’s disease assessment with histopathology of surgical specimens. Eur Radiol. 2014;24:619–629. 14. Lennard-Jones JE. Classification of inflammatory bowel disease. Scand J Gastroenterol Suppl. 1989;170:2–6; discussion 16–19. 15. Jensen MD, Ormstrup T, Vagn-Hansen C, et al. Interobserver and intermodality agreement for detection of small bowel Crohn’s disease with MR enterography and CT enterography. Inflamm Bowel Dis. 2011;17:1081–1088. 16. Lee SS, Kim AY, Yang SK, et al. Crohn disease of the small bowel: comparison of CT enterography, MR enterography, and small-bowel follow-through as diagnostic techniques. Radiology. 2009;251:751–761. 17. Mary JY, Modigliani R. Development and validation of an endoscopic index of the severity for Crohn’s disease: a prospective multicentre study. Groupe d’Etudes Therapeutiques des Affections Inflammatoires du Tube Digestif (GETAID). Gut. 1989;30:983–989. 18. Sica GT. Bias in research studies. Radiology. 2006;238:780–789. 19. Rimola J, Ordas I, Rodriguez S, et al. Imaging indexes of activity and severity for Crohn’s disease: current status and future trends. Abdom Imaging. 2012;37:958–966. 20. Ellul P, Fogden E, Simpson C, et al. Colonic tumour localization using an endoscope positioning device. Eur J Gastroenterol Hepatol. 2011;23: 488–491.
Inflamm Bowel Dis Volume 21, Number 1, January 2015
21. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med. 2003;349:2191–2200. 22. Shah SG, Saunders BP, Brooker JC, et al. Magnetic imaging of colonoscopy: an audit of looping, accuracy and ancillary maneuvers. Gastrointest Endosc. 2000;52:1–8. 23. Rimola J, Rodriguez S, Garcia-Bosch O, et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn’s disease. Gut. 2009;58:1113–1120. 24. Allez M, Lemann M, Bonnet J, et al. Long term outcome of patients with active Crohn’s disease exhibiting extensive and deep ulcerations at colonoscopy. Am J Gastroenterol. 2002;97:947–953. 25. Peyrin-Biroulet L, Loftus EV Jr, Colombel JF, et al. Early Crohn disease: a proposed definition for use in disease-modification trials. Gut. 2010;59: 141–147. 26. Allez M, Lemann M. Role of endoscopy in predicting the disease course in inflammatory bowel disease. World J Gastroenterol. 2010;16:2626–2632. 27. Buisson A, Petitcolin V. Commentary: diffusion-weighted magnetic resonance imaging—a novel way to assess disease activity in Crohn’s disease? Authors’ reply. Aliment Pharmacol Ther. 2013;37:834. 28. Panes J, Ricart E, Rimola J. New MRI modalities for assessment of inflammatory bowel disease. Gut. 2010;59:1308–1309. 29. Ordas I, Rimola J, Rodriguez S, et al. Accuracy of magnetic resonance enterography in assessing response to therapy and mucosal healing in patients with Crohn’s disease. Gastroenterology. 2014;146:374–382. e371.
Diffusion-weighted MRE in CD
30. Rimola J, Ordas I. MR colonography in inflammatory bowel disease. Magn Reson Imaging Clin N Am. 2014;22:23–33. 31. Bilgili MY. Reproductibility of apparent diffusion coefficients measurements in diffusion-weighted MRI of the abdomen with different b values. Eur J Radiol. 2012;81:2066–2068. 32. Braithwaite AC, Dale BM, Boll DT, et al. Short- and midterm reproducibility of apparent diffusion coefficient measurements at 3.0-T diffusionweighted imaging of the abdomen. Radiology. 2009;250:459–465. 33. Kim SY, Lee SS, Byun JH, et al. Malignant hepatic tumors: short-term reproducibility of apparent diffusion coefficients with breath-hold and respiratory-triggered diffusion-weighted MR imaging. Radiology. 2010; 255:815–823. 34. Kim SY, Lee SS, Park B, et al. Reproducibility of measurement of apparent diffusion coefficients of malignant hepatic tumors: effect of DWI techniques and calculation methods. J Magn Reson Imaging. 2012;36: 1131–1138. 35. Miquel ME, Scott AD, Macdougall ND, et al. In vitro and in vivo repeatability of abdominal diffusion-weighted MRI. Br J Radiol. 2012;85:1507–1512. 36. Lambie H, Tolan D. Commentary: diffusion-weighted magnetic resonance imaging—a novel way to assess disease activity in Crohn’s disease? Aliment Pharmacol Ther. 2013;37:833–834. 37. Maccioni F, Patak MA, Signore A, et al. New frontiers of MRI in Crohn’s disease: motility imaging, diffusion-weighted imaging, perfusion MRI, MR spectroscopy, molecular imaging, and hybrid imaging (PET/MRI). Abdom Imaging. 2012;37:974–982.
www.ibdjournal.org |
109
Diffusion-weighted MR Enterography for Evaluating Crohn’s Disease: How Does It Add Diagnostically to Conventional MR Enterography? Kyung-Jo Kim, MD,* Yedaun Lee, MD,† Seong Ho Park, MD,‡ Bo-Kyeong Kang, MD,‡ Nieun Seo, MD,‡ Suk-Kyun Yang, MD,* Byong Duk Ye, MD,* Sang Hyoung Park, MD,* So Yeon Kim, MD,‡ Seunghee Baek, PhD,§ and Hyun Kwon Ha, MD‡
Background: Diffusion-weighted imaging (DWI) is a novel technique to evaluate bowel inflammation in Crohn’s disease (CD). It is unknown whether and how DWI adds to the accuracy of conventional magnetic resonance enterography (MRE).
Methods: Fifty consecutive adults suspected of CD prospectively underwent clinical assessment, conventional MRE and DWI at b ¼ 900 sec/mm2
without water enema, and ileocolonoscopy within 1 week. MRE images were interpreted with proper blinding. Forty-four patients finally diagnosed with CD (male:female, 34:10; 26.9 6 6.1 yr) were analyzed. The per-segment accuracy of MRE for diagnosing active CD was assessed in the terminal ileum, right colon, and rectum using location-by-location matching with endoscopy as the reference standard.
Results: The study evaluated 58 bowel segments with deep or superficial ulcers, 34 with aphthae, erythema, or edema only, and 35 without inflammation. Conventional MRE + DWI was more sensitive for bowel inflammation than conventional MRE alone (83% [76/92] versus 62% [57/92]; P ¼ 0.001) largely because of additional detection of aphthae, erythema, or edema. The sensitivities for deep and overt ulcers were similar regardless of DWI, ranging from 88% to 97%. Conventional MRE + DWI was less specific than conventional MRE alone (60% [21/35] versus 94% [33/35]; P , 0.001), mostly because of many false positives in the colorectum. Positive DWI findings in the bowel showing active inflammation on conventional MRE were associated with higher Crohn’s disease endoscopic index of severity score (P ¼ 0.021) and deep ulcers (P ¼ 0.01; diagnostic odds ratio, 12).
Conclusions: DWI performed without water enema is not useful for incremental detection of bowel inflammation. DWI may help identify more severe inflammation among bowel segments showing active inflammation on conventional MRE. (Inflamm Bowel Dis 2015;21:101–109) Key Words: Crohn’s disease, inflammatory bowel disease, diffusion, magnetic resonance, enterography, diagnostic accuracy
C
rohn’s disease (CD) is a chronic inflammatory bowel disease that can involve the entire gastrointestinal tract, especially the small and large bowels. As chronic inflammation in CD leads
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.ibdjournal.org). Received for publication July 19, 2014; Accepted August 14, 2014. From the *Department of Gastroenterology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea; †Department of Radiology, Inje University College of Medicine, Haeundae Paik Hospital, Busan, Korea; ‡Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea; and §Department of Clinical Epidemiology and Biostatistics, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea. Supported by a grant from Dongkook Pharmaceutical, Seoul, South Korea. S. Ho Park reports grants from DongKook Pharm for the submitted study and from GE Healthcare outside the submitted work. The authors have no conflicts of interest to disclose. Kyung-Jo Kim and Yedaun Lee equally contributed to this work. Reprints: Seong Ho Park, MD, Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpagu, Seoul 138-736, Korea (e-mail: [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000222 Published online 28 October 2014.
Inflamm Bowel Dis Volume 21, Number 1, January 2015
to irreversible bowel damage, monitoring of disease activity and bowel damage is important, especially because mucosal healing using biologics or immunosuppressive therapy is considered a therapeutic goal.1 Magnetic resonance enterography (MRE) plays an important role in the diagnosis and follow-up of patients with CD.2– 4 Recently, increased need for repeat follow-up imaging to monitor responses to therapeutic interventions aimed at mucosal healing has further accentuated the role of MRE, because MRE has an advantage of the lack of radiation exposure over computed tomographic enterography.2–4 Diffusion-weighted imaging (DWI) is a specialized MR imaging technique that maps the diffusion of water molecules in biological tissues and has been used largely for the diagnosis of ischemia in the nervous system and for oncologic imaging. Recently, studies have shown that restricted diffusion in the bowel wall on DWI in patients with CD is associated with the presence of active bowel inflammation and has a good correlation with bowel inflammatory activity estimated using conventional MRE (i.e., T2and T1-weighted images with intravenous contrast enhancement).5– 13 However, to our knowledge, few studies to date have compared DWI and conventional MRE, which is currently the standard method of MRE practice, regarding their relative strengths and www.ibdjournal.org |
101
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Kim et al
weaknesses using independent reference standards. Therefore, although the promising potential of DWI as a new imaging biomarker of bowel inflammation in CD has been revealed by the previous studies, it is still unknown whether DWI has an incremental diagnostic value compared with conventional MRE and whether and how DWI should be added to the current standard practice of MRE. Moreover, many previous studies had methodological weaknesses, including their retrospective study design5–8,10,11; small sample sizes5,6,8,9; and/or lack of proper blinding for the review of DWI, in that knowledge of conventional MRE findings before DWI may have affected/biased the interpretation of DWI results.6–11 For these reasons, it would now be important to evaluate whether DWI has an additive diagnostic value compared with conventional MRE method and to determine how the new technique should be incorporated into the clinical practice of MRE to evaluate CD. This study was to prospectively determine whether and how DWI adds to the diagnostic accuracy of conventional MRE to evaluate bowel inflammation in patients with CD.
MATERIALS AND METHODS This prospective study was approved by the Institutional Review Board of Asan Medical Center. Informed consent was obtained from the participants.
Subjects Between October 2012 and December 2013, adults ($18 yr) referred to the Inflammatory Bowel Disease Center of Asan Medical Center, an academic referral institution, with suspected
CD or after an initial diagnosis of CD were recruited for the study. Patients with a history of bowel resection, those who required emergency care related to bowel disease (such as severe bowel obstruction), and those with contraindications for MR examinations or spasmolytics (i.e., Buscopan) used for MRE were excluded. Fifty-four patients were consecutively enrolled, but 4 withdrew before the study. The remaining 50 patients, consisting of 38 men and 12 women (mean 6 SD age, 27.7 6 6.4 yr), underwent all study procedures. This study was designed as an exploratory study; therefore, the sample size was chosen empirically. The 44 patients finally confirmed as having CD according to conventional clinical, radiologic, endoscopic, and histopathologic criteria14 were analyzed (Fig. 1).
Study Procedures Patients underwent clinical assessments including measurements of Crohn’s disease activity index, laboratory tests, MRE, and ileocolonoscopy within 1 week. Laboratory tests included serum concentrations of erythrocyte sedimentation rate, C-reactive protein, and albumin, platelet, and white blood cell counts in peripheral blood. Patients were not allowed to take any medications for CD over the course of these examinations.
Magnetic Resonance Enterography MRE examinations were performed after 6 hours of fasting. Patients were instructed to drink 1500 mL of 2.5% sorbitol solution to distend the bowel before MR scanning. Water enema was not used. MR scanning was performed using a 3-T system (Ingenia; Philips Healthcare, Best, the Netherlands). MRE imaging consisted of coronal T2-weighted half-Fourier sequences
FIGURE 1. Study flow diagram.
102
| www.ibdjournal.org
Inflamm Bowel Dis Volume 21, Number 1, January 2015
with and without fat suppression, coronal, and axial T2-like steady-state gradient echo sequences with fat suppression, coronal DWI (with b factors of 0 and 900 sec/mm2) and an apparent diffusion coefficient (ADC) map, coronal dynamic fatsuppressed spoiled gradient-echo T1-weighted unenhanced and contrast-enhanced (enteric and portal-venous phases) imaging, and axial fat-suppressed spoiled gradient-echo T1-weighted delayed contrast-enhanced scan. For contrast enhancement, 0.2 mL/kg body weight of gadoterate meglumine (Dotarem; Guerbet, Villepinte, France) was injected intravenously at a rate of 2 mL/sec followed by a saline flush. To prevent bowel peristalsis, 10 mg of scopolamine-N-butyl bromide (Buscopan; Boehringer Ingelheim, Ingelheim, Germany) was administered intravenously 3 times at intervals during the scanning. Field of view was adjusted to include the entire small and large bowels. Further details of scan parameters are provided in the Appendix, Supplemental Digital Content 1, http://links.lww.com/IBD/A623. MRE was interpreted in consensus by 2 board-certified faculty radiologists experienced in MRE of inflammatory bowel disease to achieve a more factual correlation between MRE and endoscopy, given that some degree of interobserver variability exists in interpreting MRE.15,16 Interobserver agreement between the 2 study readers was then separately analyzed (see Appendix, Supplemental Digital Content 1, http://links.lww.com/IBD/ A623). The readers were informed about the length of the terminal ileum evaluated by endoscopy (see Reference Standard: Ileocolonoscopy) so that the same terminal ileal lengths could be evaluated by MRE and endoscopy, resulting in an accurate location-by-location match between MRE and ileocolonoscopy. Otherwise, the readers were blinded to the results of ileocolonoscopy, clinical assessment, and laboratory tests. The technical adequacy of each MRE examination was evaluated, and examinations deemed inadequate for diagnostic evaluation were excluded from analysis. Bowel inflammation was assessed during 2 separate reading sessions. In the first session, conventional MRE (i.e., all sets of images except for DWI and ADC maps) was read. The colorectum and terminal ileum were divided into 5 segments: the terminal ileum, right colon (cecum and ascending), transverse colon, left colon (descending and sigmoid), and rectum according to the Crohn’s disease endoscopic index of severity (CDEIS).17 Signs of active bowel inflammation in each segment were recorded, including mural thickening, mural hyperenhancement, increased mural signals on T2-weighted images (i.e., bowel wall edema) and/or perienteric edema, increased vasa recta (i.e., the comb sign), and perienteric abnormalities related to penetrating disease (i.e., sinus tract, fistula, phlegmon, or abscess). The presence of individual ulcers was not recorded separately because of the difficulty in consistently distinguishing among individual ulcers and distinguishing them from deformed luminal contours on cross-sectional images, although recognition of individual ulcers on MRE is possible in some cases. After a washout period of at least 1 month and random reshuffling of the case review order to prevent recall bias,18 the second reading session interpreted DWI and ADC maps in the absence of conventional MRE images
Diffusion-weighted MRE in CD
to prevent any influence/bias due to prior knowledge of conventional MRE findings. Because of the lack of anatomical details on DWI, the areas to be evaluated on DWI were premarked on the images by a separate coordinating investigator. DWI was interpreted in a dichotomous manner and bowel segments with high signal intensity on DWI, at least similar to that of the spleen and mesenteric lymph nodes on b ¼ 900-sec/mm2 images, and with low signals on the corresponding ADC map were regarded as active inflammation (Figs. 2–4). Semiquantitative MRE scores of bowel inflammatory activity for each segment were determined by assigning one point to each MRE finding of active bowel inflammation and summing the points, in a manner similar to the methods used in related studies.7,19 A per-patient MRE score was then derived by taking the average of the segmental scores in a manner similar to global CDEIS scores are calculated.
Reference Standard: Ileocolonoscopy Ileocolonoscopy findings were used as the reference standard. Ileocolonoscopy was performed by either one of the 3 boardcertified faculty gastroenterologists who were experienced in endoscopic examination of patients with inflammatory bowel disease and blinded to the results of MRE, clinical assessment, and laboratory tests, using a video colonoscope (CF H260AL or CF H260AI; Olympus Optical Co., Tokyo, Japan). Colons were prepared using 4 L of polyethylene glycol. The length of the terminal ileum endoscopically evaluated was recorded. Endoscopic lesions were evaluated according to CDEIS.17 Global and segmental CDEIS scores were determined. The types of inflammatory mucosal lesions were specified, including overt (deep and superficial) ulcers, aphthoid lesions, and erythema or edema. Absence of any of these lesions was considered the state of absence of disease (to constitute the denominator for calculating specificity).
Data and Statistical Analysis Per-patient MRE scores of bowel inflammatory activity obtained either with conventional MRE alone (ranging from 0 to 5) or with combined conventional MRE plus DWI (ranging from 0 to 6) were correlated with global CDEIS scores using the Spearman’s correlation analysis. The results were compared between themselves and with the degree of correlation between Crohn’s disease activity index or serum C-reactive protein and global CDEIS scores. Per-bowel segment analysis was also performed and only included the terminal ileum, right colon, and rectum. The selection of these 3 segments was to avoid any uncertainty related to location mismatches between MRE and endoscopy. It is well known that segmental designations of the same colonic location on radiologic imaging and colonoscopy often do not coincide,20–22 e.g., transverse colon on one method but descending or ascending colon on the other method, particularly in areas where endoscopic landmarks are lacking. Endoscopy can locate the tip more straightforwardly in the 3 segments due to the presence of or the proximity to anatomical landmarks and boundaries, allowing for a more accurate location-by-location correlation between MRE and endoscopy. We analyzed the sensitivity and specificity of conventional MRE www.ibdjournal.org |
103
Kim et al
Inflamm Bowel Dis Volume 21, Number 1, January 2015
FIGURE 3. False-negative conventional MRE but positive DWI findings in the rectum of a 37-year-old man with CD. A, The rectum (arrowheads) shows no signs of active inflammation on a contrast-enhanced T1-weighted image (left), but a diffusely high signal on DWI (middle) and a low signal on an ADC map (right). B, Colonoscopic image of the rectum showing numerous scattered tiny aphthoid lesions.
FIGURE 2. True-positive conventional MRE and DWI findings in the terminal ileum of a 38-year-old woman with CD. A, The terminal ileum (arrowheads) shows diffuse thickening, increased mural signal on a T2weighted image (left), and mural hyperenhancement (right) and prominently increased vasa recta (right). B, The terminal ileum (arrowheads) shows a high signal on DWI (left) and a dark signal on an ADC map (right). C, Endoscopic image of the terminal ileum showing multiple deep longitudinal ulcers.
104
| www.ibdjournal.org
alone and combined conventional MRE plus DWI in diagnosing active bowel inflammation. The combined conventional MRE plus DWI was regarded as positive for active bowel inflammation if either was positive. Subgroup analyses were performed in terms of severity of bowel inflammation, defined as segments with deep ulcers, overt (deep or superficial) ulcers, or less severe lesions only (aphthae, erythema, or edema without overt ulcers)23; and in terms of bowel location, defined as the terminal ileum or colorectum. Statistical comparisons between the 2 MRE methods were made using the generalized linear mixed model to address the correlation between repeated observations in the same bowel segment using different MRE methods and the nested data structure (i.e., multiple segments per patient). Any notable differences in the diagnostic performance were further elaborated on with a cross-tabulation of conventional MRE and DWI interpretations. Second, the incremental role of DWI in bowel areas showing active inflammation on conventional MRE was evaluated. The relationships of DWI findings with segmental CDEIS scores and the presence/absence of deep ulcers in those bowel segments positive for active inflammation as seen on conventional MRE were analyzed using the Student’s t test and the Fisher’s exact test, respectively. The ability of
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Diffusion-weighted MRE in CD
TABLE 1. Characteristics of Study Subjects Patients included (n ¼ 44) Age, mean 6 SD, yr Gender (male:female) Body mass index, mean 6 SD, kg/m2 CDAI, mean 6 SD Laboratory findings ESR, mean 6 SD, mm/h CRP, range (median), mg/dL Albumin, mean 6 SD, g/dL Platelet count, mean 6 SD, ·103/mL White blood cell count, mean 6 SD, ·103/mL Global CDEIS score, mean 6 SD Bowel segments included in the persegment analysis (n ¼ 127) Anatomical location Terminal ileum Right colon Rectum Endoscopic findings Types of lesions Overt (deep or superficial) ulcers FIGURE 4. Negative conventional MRE and false-positive DWI findings in the ascending colon of a 24-year-old woman with CD. A, The lateral wall of the proximal ascending colon (arrowheads) shows no signs of active inflammation on a contrast-enhanced T1-weighted image (left) but a high signal on DWI (middle) and a low signal on an ADC map (right). B, Colonoscopic image of the proximal ascending colon showing no signs of inflammation or ulcer. IC, ileocecal valve.
semiquantitative MRE scores of bowel inflammation to distinguish between segments with and without deep ulcers among those segments positive for active inflammation as seen on conventional MRE was compared between conventional MRE alone and combined conventional MRE plus DWI using the receiver operating characteristic analysis. All statistical analyses were performed using IBM SPSS Statistics for Windows Version 21.0 (IBM Corp., Armonk, NY) and MedCalc (version 13.1; MedCalc Software, Mariakerke, Belgium). P , 0.05 was considered statistically significant.
RESULTS
Aphthae, erythema, or edema only No inflammation Segmental CDEIS score Terminal ileum (range and median) Colorectum (range and median)
26.9 6 6.1 34:10 20.0 6 2.9 201.53 6 111.14 42 6 26 0.1–17.66 (1.86) 3.6 6 0.6 325 6 89 7.3 6 2.5 10.8 6 6.9
41 42 44
58 (terminal ileum, 31; colorectum, 27) 34 (terminal ileum, 3; colorectum, 31) 35 (terminal ileum, 7; colorectum, 28) 0–34 (median, 15) 0–38 (median, 4)
CDAI, Crohn’s disease activity index; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
summarized in Table 1. The time interval between MRE and endoscopy ranged from 0 to 7 days (median, 1 d). The entire colon and terminal ileum were examined endoscopically in 41 patients; the length of the terminal ileum evaluated ranged from 1 to 20 cm (median, 10 cm) and was $10 cm in 36 patients. Of the remaining 3 patients, one each was endoscopically evaluated up to the cecum, the sigmoid colon, and the rectum. As a result, a total of 127 segments, 92 with and 35 without active inflammation, were included in the per-segment analysis (Table 1).
Subjects
Per-patient Analysis
All 50 patients underwent all the study procedures without any adverse events. Of those, 3 were finally diagnosed with a condition other than CD, and 3 others (6% of 47) had MR images deemed below diagnostic quality due to various artifacts. These 6 patients were therefore excluded from analysis. The remaining 44 patients constituted the ultimate cohort for analysis (Fig. 1); their demographic and clinical characteristics are
The correlation coefficient between MRE scores of bowel inflammation and global CDEIS scores was 0.819 (95% confidence interval, 0.690–0.898; P , 0.001) for conventional MRE alone and was similarly 0.823 (0.696–0.900; P , 0.001) for combined conventional MRE plus DWI. These were higher than the correlation coefficients between Crohn’s disease activity index and global CDEIS scores, 0.443 (0.168–0.654; P ¼ 0.003), and www.ibdjournal.org |
105
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Kim et al
TABLE 2. Rates of Positive MRE Results According to the Severity of Bowel Inflammation Endoscopic Findings
Aphthae, Erythema, or Edema Only (n ¼ 34)
Superficial Ulcers (n ¼ 26)
Deep Ulcers (n ¼ 32)
All (n ¼ 92)
18 (6/34) 53 (18/34)
77 (20/26) 73 (19/26)
97 (31/32) 94 (30/32)
62 (57/92) 73 (67/92)
Active inflammation on conventional MRE Active inflammation on DWI
Data are percentages with the number of bowel segments provided in the parentheses.
between serum C-reactive protein and global CDEIS scores, 0.525 (0.271–0.711; P , 0.001).
Per-segment Analysis: MRE Accuracy for Diagnosing Active Bowel Inflammation Both conventional MRE and DWI showed increasing rates of positive test results with increasing severity of bowel inflammation, and DWI revealed an apparently higher test-positive rate compared with conventional MRE in segments with mild inflammatory lesions without overt ulcers (aphthae, erythema, or edema only) (Table 2). Therefore, overall sensitivity for all types of inflammatory bowel lesions became significantly higher when DWI was added to conventional MRE (83% [76/92]) compared with conventional MRE alone (62% [57/92]) (P ¼ 0.001) (Table 3). By contrast, the specificity of combined conventional MRE plus DWI (60% [21/35]) was significantly lower than that of conventional MRE alone (94% [33/ 35]) (P , 0.001) (Table 3). Subgroup analyses showed that the sensitivity of conventional MRE plus DWI was significantly greater than that of conventional MRE alone for mild inflammatory lesions without overt ulcers (62% [21/34] versus 18% [6/34]; P , 0.001), whereas there were no significant differences in sensitivity for deep
and overt ulcers (Table 3). When analyzed according to bowel location, conventional MRE plus DWI showed significantly higher sensitivity but lower specificity than conventional MRE alone in the colorectum (Table 3). Representative findings are shown in Figures 2–4. Conventional MRE and DWI interpretations concurred in 68.5% (87 of 127) of the segments (Table 4). Of 92 bowel segments with active inflammation at endoscopy, 19 were diagnosed as such solely by DWI (Table 4), and 14 of these were in the colorectal segments with only aphthae, erythema, or edema without overt ulcers. DWI falsely diagnosed active bowel inflammation in 12 segments where conventional MRE correctly diagnosed the absence of inflammation (Table 4), with 11 of these segments in the colorectum.
Per-segment Analysis: DWI in the Bowel Segments Showing Active Inflammation on Conventional MRE A total of 59 bowel segments (32 terminal ileal and 27 colorectal segments) were positive for active bowel inflammation as seen on conventional MRE. Of these, the 50 DWI-positive segments showed significantly higher segmental CDEIS scores than the 9 DWInegative segments (21 6 10.1 versus 12.6 6 8.4; P ¼ 0.021). Positive
TABLE 3. Accuracy of MRE for Diagnosing Bowel Inflammation Sensitivity
All types of bowel inflammationa Subgroups according to inflammatory severity Deep ulcers Overt (deep or superficial) ulcers Aphthae, erythema, or edema only Subgroups according to bowel location Terminal ileum Colorectum
Specificity
Conventional MRE
Combined Conventional MRE + DWI
62 (57/92)
83 (76/92)
0.001
97 (31/32) 88 (51/58) 18 (6/34)
97 (31/32) 95 (55/58) 62 (21/34)
1 0.171 ,0.001
NA NA NA
91 (31/34) 45 (26/58)
94 (32/34) 76 (44/58)
0.650 ,0.001
86 (6/7) 96 (27/28)
P
Data are percentages with the number of bowel segments provided in the parentheses. a Includes overt (deep or superficial) ulcers and less severe lesions such as aphthae, erythema, or edema. NA, not applicable.
106
| www.ibdjournal.org
Conventional MRE
Combined Conventional MRE + DWI
P
94 (33/35)
60 (21/35)
,0.001
71 (5/7) 57 (16/28)
0.560 0.003
Inflamm Bowel Dis Volume 21, Number 1, January 2015
Diffusion-weighted MRE in CD
TABLE 4. Cross-tabulation of the Results of MRE Interpretation Segments with Active Inflammationa at Reference Segments without Active Inflammation at Reference Standard (n ¼ 92) Standard (n ¼ 35) DWI Interpretation
Conventional MRE interpretation Positive Negative Total
DWI Interpretation
Positive
Negative
Total
Positive
Negative
Total
48 19 67
9 16 25
57 35 92
2 12 14
0 21 21
2 33 35
Data are numbers of bowel segments. a All types of inflammatory lesions including overt (deep or superficial) ulcers and less severe lesions such as aphthae, erythema, or edema.
DWI findings were significantly associated with the presence of deep ulcers (60% [30/50] versus 11% [1/9]; P ¼ 0.01; diagnostic odds ratio, 12). The area under the receiver operating characteristic curve of the semiquantitative MRE scores for distinguishing bowel segments with and without deep ulcers among bowel segments positive for active inflammation as seen on conventional MRE was significantly higher for combined conventional MRE plus DWI than for conventional MRE alone (0.720 versus 0.661; P ¼ 0.029) (Fig. 5).
DISCUSSION The results presented here reconfirm the association between restricted diffusion in the bowel wall as seen on DWI and active bowel inflammation in patients with CD. The study
FIGURE 5. Receiver operating characteristic curves showing the ability of semiquantitative MRE scores to distinguish bowel segments with and without deep ulcers among segments showing active inflammation on conventional MRE.
also demonstrates that enterographic DWI performed without water enema generally has limited incremental benefit compared with conventional MRE in diagnosing bowel inflammation in CD. The increased detection of bowel inflammation by addition of DWI was mostly because of the increased identification of mild inflammation, consisting of aphthae, erythema, or edema without overt ulcers in the colorectum. By contrast, the sensitivity of conventional MRE was already quite high for overt ulcers, particularly for deep ulcers, with the addition of DWI having little effect on the sensitivity of MRE for such lesions. Although deep ulcers correlate with a more aggressive clinical course24 and should be considered targets for disease-modifying therapy,25 the natural history and clinical significance of mild inflammatory lesions without overt ulcers remain doubtful.26,27 Therefore, it is uncertain whether the ability of DWI to enhance the detection of mild bowel inflammation represents a true clinical benefit. By contrast, it is noteworthy that positive DWI findings may suggest more severe inflammation and deep ulcers in bowel segments showing active inflammation as seen on conventional MRE. Consistent with our result, previous studies suggested that ADC values measured in the bowel wall on DWI may be used as a parameter to assess bowel inflammatory severity in CD as they demonstrated a good correlation with bowel inflammatory activity estimated using conventional MRE.9,12 Taken together, these results indicate that DWI may have a role in further stratifying patients with CD with known bowel inflammation, regarding their inflammatory severity or suitability for disease-modifying therapy rather than identifying bowel inflammation undetected by conventional MRE. Another point raised by our study is the low specificity of DWI especially in the colorectum when performed without water enema. MRE is used in patients with CD primarily to examine bowel segments difficult to reach by endoscopy,2,4 not for preendoscopic screening, making high specificity an important feature of MRE. Therefore, the low specificity of DWI seems worrisome. The low specificity of DWI in our study (60%) was similar to the 62.6% specificity reported in a retrospective analysis of a prospectively developed patient registry.7 Although a few other studies reported numerically greater specificity values (81.4%–82.4% for the small and large bowels and 75.7% for the large bowel),5,6 those www.ibdjournal.org |
107
Kim et al
studies were retrospective in design, included relatively few patients (with many being surgical patients), and used suboptimal reference standards. Moreover, as DWI was interpreted without blinding to conventional MRE results and the definitions of positive DWI findings were somewhat vague (i.e., “higher signal intensity than adjacent normal-appearing bowel loops,” with DWI itself used to distinguish between normal and abnormal bowel segments) in the previous studies,5–7 the previous studies may have overestimated the specificity of DWI. Most of the false-positive DWI results in our study occurred in the colorectum, similar to previous findings.5,6 The high rate of DWI false positives in the colorectum and the finding that most of the additional areas of bowel inflammation detected by DWI were in the colorectum in our study may also suggest the possible fortuitous coincidence of DWI false positives in colorectal segments with subtle inflammation below the resolution of MRE in some cases. The predilection of DWI false positives in the colorectum may be related to the generally suboptimal preparation/distention state of the colorectum when MRE is performed without water enema.28 It may be worthwhile to investigate whether these DWI false positives could be reduced by water enema in addition to oral administration of fluid.23,29,30 Our study had several methodological strengths compared with previous related studies. However, unlike previous studies, we focused on visual interpretation of DWI and did not measure ADC values. Despite the advantages of quantitative analysis, measuring absolute ADC values has generally shown limited reproducibility in the abdomen.31–36 Moreover, obtaining ADC values from the bowel wall in patients with CD is likely further confounded by the presence of multiple diseased segments and small sizes of individual bowel walls, introducing additional variability in measurement.37 Therefore, the quantitative approach is rarely used in clinical practice, although it is plausible in a controlled research setting.9,12 In our study, we intended to evaluate the clinical value of DWI from a more practical viewpoint. This study had limitations. First, the number of terminal ileal segments devoid of inflammation or with only mild inflammation without overt ulcers was small. The incidence of false-positive DWI results in terminal ileal segments should be further confirmed in a larger patient sample. Second, MRE interpretation in this study did not consider the length of bowel involvement, which may have introduced some imprecision in correlating MRE and endoscopy. Third, this study focused on incremental benefits of DWI when it is added to conventional MRE regarding evaluating bowel inflammation at single time points. Other roles of DWI, such as for longitudinal monitoring after medical therapy, as a substitute for intravenous contrast enhancement,9,12 or in evaluating bowel damage and complications, would require further studies. Finally, small bowel segments other than the terminal ileum were not evaluated. However, this is an inevitable methodological limitation of almost any comparative/correlative research studies between radiological bowel imaging and endoscopy because location-by-location matching is virtually impossible for the small bowel except for the terminal ileum. In conclusion, DWI performed without water enema is likely not useful for additional detection of bowel inflammation in
108
| www.ibdjournal.org
Inflamm Bowel Dis Volume 21, Number 1, January 2015
CD undetected by conventional MRE as abnormalities solely detected by DWI were mostly nonspecific inflammation lacking strong clinical relevance or false positives particularly in the colorectum. DWI may help identify more severe inflammation and deep ulcers among bowel segments showing active inflammation on conventional MRE.
REFERENCES 1. Cheifetz AS. Management of active Crohn disease. JAMA. 2013;309: 2150–2158. 2. Benitez JM, Meuwis MA, Reenaers C, et al. Role of endoscopy, crosssectional imaging and biomarkers in Crohn’s disease monitoring. Gut. 2013;62:1806–1816. 3. Makanyanga JC, Taylor SA. Current and future role of MR enterography in the management of Crohn disease. AJR Am J Roentgenol. 2013;201: 56–64. 4. Pariente B, Peyrin-Biroulet L, Cohen L, et al. Gastroenterology review and perspective: the role of cross-sectional imaging in evaluating bowel damage in Crohn disease. AJR Am J Roentgenol. 2011;197:42–49. 5. Kiryu S, Dodanuki K, Takao H, et al. Free-breathing diffusion-weighted imaging for the assessment of inflammatory activity in Crohn’s disease. J Magn Reson Imaging. 2009;29:880–886. 6. Oto A, Zhu F, Kulkarni K, et al. Evaluation of diffusion-weighted MR imaging for detection of bowel inflammation in patients with Crohn’s disease. Acad Radiol. 2009;16:597–603. 7. Oussalah A, Laurent V, Bruot O, et al. Diffusion-weighted magnetic resonance without bowel preparation for detecting colonic inflammation in inflammatory bowel disease. Gut. 2010;59:1056–1065. 8. Oto A, Kayhan A, Williams JT, et al. Active Crohn’s disease in the small bowel: evaluation by diffusion weighted imaging and quantitative dynamic contrast enhanced MR imaging. J Magn Reson Imaging. 2011; 33:615–624. 9. Buisson A, Joubert A, Montoriol PF, et al. Diffusion-weighted magnetic resonance imaging for detecting and assessing ileal inflammation in Crohn’s disease. Aliment Pharmacol Ther. 2013;37:537–545. 10. Neubauer H, Pabst T, Dick A, et al. Small-bowel MRI in children and young adults with Crohn disease: retrospective head-to-head comparison of contrast-enhanced and diffusion-weighted MRI. Pediatr Radiol. 2013; 43:103–114. 11. Ream JM, Dillman JR, Adler J, et al. MRI diffusion-weighted imaging (DWI) in pediatric small bowel Crohn disease: correlation with MRI findings of active bowel wall inflammation. Pediatr Radiol. 2013;43: 1077–1085. 12. Hordonneau C, Buisson A, Scanzi J, et al. Diffusion-weighted magnetic resonance imaging in ileocolonic Crohn’s disease: validation of quantitative index of activity. Am J Gastroenterol. 2014;109:89–98. 13. Tielbeek JA, Ziech ML, Li Z, et al. Evaluation of conventional, dynamic contrast enhanced and diffusion weighted MRI for quantitative Crohn’s disease assessment with histopathology of surgical specimens. Eur Radiol. 2014;24:619–629. 14. Lennard-Jones JE. Classification of inflammatory bowel disease. Scand J Gastroenterol Suppl. 1989;170:2–6; discussion 16–19. 15. Jensen MD, Ormstrup T, Vagn-Hansen C, et al. Interobserver and intermodality agreement for detection of small bowel Crohn’s disease with MR enterography and CT enterography. Inflamm Bowel Dis. 2011;17:1081–1088. 16. Lee SS, Kim AY, Yang SK, et al. Crohn disease of the small bowel: comparison of CT enterography, MR enterography, and small-bowel follow-through as diagnostic techniques. Radiology. 2009;251:751–761. 17. Mary JY, Modigliani R. Development and validation of an endoscopic index of the severity for Crohn’s disease: a prospective multicentre study. Groupe d’Etudes Therapeutiques des Affections Inflammatoires du Tube Digestif (GETAID). Gut. 1989;30:983–989. 18. Sica GT. Bias in research studies. Radiology. 2006;238:780–789. 19. Rimola J, Ordas I, Rodriguez S, et al. Imaging indexes of activity and severity for Crohn’s disease: current status and future trends. Abdom Imaging. 2012;37:958–966. 20. Ellul P, Fogden E, Simpson C, et al. Colonic tumour localization using an endoscope positioning device. Eur J Gastroenterol Hepatol. 2011;23: 488–491.
Inflamm Bowel Dis Volume 21, Number 1, January 2015
21. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med. 2003;349:2191–2200. 22. Shah SG, Saunders BP, Brooker JC, et al. Magnetic imaging of colonoscopy: an audit of looping, accuracy and ancillary maneuvers. Gastrointest Endosc. 2000;52:1–8. 23. Rimola J, Rodriguez S, Garcia-Bosch O, et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn’s disease. Gut. 2009;58:1113–1120. 24. Allez M, Lemann M, Bonnet J, et al. Long term outcome of patients with active Crohn’s disease exhibiting extensive and deep ulcerations at colonoscopy. Am J Gastroenterol. 2002;97:947–953. 25. Peyrin-Biroulet L, Loftus EV Jr, Colombel JF, et al. Early Crohn disease: a proposed definition for use in disease-modification trials. Gut. 2010;59: 141–147. 26. Allez M, Lemann M. Role of endoscopy in predicting the disease course in inflammatory bowel disease. World J Gastroenterol. 2010;16:2626–2632. 27. Buisson A, Petitcolin V. Commentary: diffusion-weighted magnetic resonance imaging—a novel way to assess disease activity in Crohn’s disease? Authors’ reply. Aliment Pharmacol Ther. 2013;37:834. 28. Panes J, Ricart E, Rimola J. New MRI modalities for assessment of inflammatory bowel disease. Gut. 2010;59:1308–1309. 29. Ordas I, Rimola J, Rodriguez S, et al. Accuracy of magnetic resonance enterography in assessing response to therapy and mucosal healing in patients with Crohn’s disease. Gastroenterology. 2014;146:374–382. e371.
Diffusion-weighted MRE in CD
30. Rimola J, Ordas I. MR colonography in inflammatory bowel disease. Magn Reson Imaging Clin N Am. 2014;22:23–33. 31. Bilgili MY. Reproductibility of apparent diffusion coefficients measurements in diffusion-weighted MRI of the abdomen with different b values. Eur J Radiol. 2012;81:2066–2068. 32. Braithwaite AC, Dale BM, Boll DT, et al. Short- and midterm reproducibility of apparent diffusion coefficient measurements at 3.0-T diffusionweighted imaging of the abdomen. Radiology. 2009;250:459–465. 33. Kim SY, Lee SS, Byun JH, et al. Malignant hepatic tumors: short-term reproducibility of apparent diffusion coefficients with breath-hold and respiratory-triggered diffusion-weighted MR imaging. Radiology. 2010; 255:815–823. 34. Kim SY, Lee SS, Park B, et al. Reproducibility of measurement of apparent diffusion coefficients of malignant hepatic tumors: effect of DWI techniques and calculation methods. J Magn Reson Imaging. 2012;36: 1131–1138. 35. Miquel ME, Scott AD, Macdougall ND, et al. In vitro and in vivo repeatability of abdominal diffusion-weighted MRI. Br J Radiol. 2012;85:1507–1512. 36. Lambie H, Tolan D. Commentary: diffusion-weighted magnetic resonance imaging—a novel way to assess disease activity in Crohn’s disease? Aliment Pharmacol Ther. 2013;37:833–834. 37. Maccioni F, Patak MA, Signore A, et al. New frontiers of MRI in Crohn’s disease: motility imaging, diffusion-weighted imaging, perfusion MRI, MR spectroscopy, molecular imaging, and hybrid imaging (PET/MRI). Abdom Imaging. 2012;37:974–982.
www.ibdjournal.org |
109
