| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Servicio Patologia Digestiva, Hospital General Universitario de Valencia,
Valencia, Spain.
2 CT and MRI Unit, Hospital General Universitario de Valencia, Avda. Tres Cruces
s/n, Valencia, Valencia, Spain, 46014.
3 Servicio de Gastroenterologia, Hospital Clinico Universitario de Valencia,
Valencia, Spain.
4 Servicio de Diagnostico por Imagen, Hospital General Universitario de
Valencia, Valencia, Spain.
Received March 16, 2004;
accepted after revision September 16, 2004.
Address correspondence to V. Martinez Sanjuan
(martinez_vicsanj{at}gva.es).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Twenty patients with histologically proven CD were prospectively evaluated with MRI and ileocolonoscopy over a 2-year period. The MRI protocol included axial and coronal T2-weighted and contrast-enhanced T1-weighted sequences. MRI examinations were performed twice, once during an acute relapse of CD and the other at clinical remission. The terminal ileum and colon were divided into six segments/patient, and the endoscopy and histology findings were considered the standard of reference. These were compared on a segmental basis with the quantitative MRI findings regarding wall thickness and contrast enhancement. The results obtained in active and remission CD phases were likewise compared with the findings in 10 control subjects who underwent complete ileocolonoscopy for other reasons and had no pathological findings on ileocolonoscopy.
RESULTS. Fifty three of 120 (44.2%) bowel segments showed pathologic changes on endoscopy and histology consistent with CD in active phase. On changing from the active disease phase to clinical remission, a significant decrease was observed in the wall thickness and contrast enhancement of the affected bowel wall. In the active phase of CD, the pathologic bowel segments presented with significantly greater contrast enhancement and wall thickness values compared with the healthy segments of CD and controls. On converting clinically into remission, contrast enhancement tended to normalize, whereas bowel wall thickness remained increased compared with the controls.
CONCLUSION. MRI is able to detect pathologic bowel segments in CD, as it allows the measurement of significant variations in wall thickness and contrast enhancement on changing from the active phase of the disease to remission.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This study evaluates the capability of MRI to accurately assess inflammatory bowel changes when clinical remission occurs.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Study Subjects: Patients
Between August 1999 and September 2001, all patients with histologically
proven and active CD [6] in
whom an abdominal MRI was clinically indicated were referred by their
gastroenterologist for MRI bowel wall assessment and its comparison with
colonoscopy and biopsy samples of the inflammatory active lesions. The
inclusion criteria were (a) CDAI >150, (b) colonoscopic lesions suggestive
of active CD (regardless of lesion grade), and (c) at least one elevated acute
phase reactant (erythrocyte sedimentation rate [ESR] or C-reactive protein
[CRP]). The exclusion criteria were (a) contraindications for MRI
(electrically, magnetically, or mechanically activated devices; central
nervous system hemostasia clips, or the inability to administer a gadolinium
contrast agent because of known allergic problems), (b) pregnancy, and (c) HIV
infection.
|
|
|
|
Subjects: Control Group
The control group consisted of 10 individuals (5 men and 5 women; mean age,
57.7 ± 13.29 years; range, 34-79 years old) who underwent colonoscopy
for the following reasons: rectal bleeding/hemorrhoids in six cases, weight
loss in two, and abdominal pain in two patients. In all cases, after complete
exploration of the colon and terminal ileum, no evidence of organic lesions
was found. In identical fashion as the patient group, these controls underwent
an abdominal MRI within 24 hr after colonoscopy.
Colonoscopy
Colonoscopies were performed by three experienced endoscopists in conscious
and sedated patients (midazolam, 2-4 mgr IV). In every patient, the colon was
divided into six segments: rectosigmoid, descending colon including the
splenic angle, transverse colon, ascending colon and hepatic angle, cecum and
ileocecal valve, and terminal ileum. From each bowel segment, a minimum of two
biopsies were obtained, regardless of whether endoscopically manifest disease
activity was detected. Pathologic segments were defined as those segments
presenting any endoscopic lesion suggestive of CD (e.g., erythema/loss of
vascular pattern, aphthoid ulcerations, polymorphic ulcerations, or a
cobblestone appearance) and/or the existence of at least one of the following
biopsy findings indicative of acute inflammatory activity: (a) presence of
polymorphonuclear cells in the sample, (b) alterations in mucosal epithelial
integrity, or (c) presence of granulomas or microgranulomas. Normal bowel
segments were in turn defined as those segments showing no endoscopic lesions
or histologic findings suggestive of acute inflammation. In the comparative
study with MRI, the distal rectum was not taken into consideration due to
balloon placement used for enema retention during the MRI examination.
MRI Acquisition
MRI of the whole abdomen was performed using a 1.5-T MRI system (Signa
LX-II, GE Healthcare) in 44 examinations (identified as scanner A); the
Magnetom Sonata scanner (Siemens Medical Solutions) was used in 16
examinations (identified as scanner B). Gradient amplitudes of 23 and 40 mT/m,
respectively, were used, with a phased-array multicoil for the body. Patients
fasted for 7 hr before the MRI examination. Thirty min before imaging,
patients were given 1,500 mL of a 2% diluted barium sulfate solution orally.
Water (1,500-2,000 mL) was administered rectally for colonic distention just
before imaging.
After acquiring standard three-plane scout images, the precontrast protocol included the following sequences: (1) axial T2-weighted (scanner A, single-shot fast spin-echo [SSFSE]; TR/TE, infinite/120; scanner B, half-fourier single-shot turbo spin-echo, HASTE; TR/TE, 1200/120, or true fast imaging with steady-state free precession [FISP]; TR/TE 4.3/2.1, 80° flip angle) from the xiphoid process to the pubis (around 40 sections); (2) coronal fat-suppressed T2-weighted (SSFSE/HASTE or true FISP) (around 20 sections); (3) coronal T1-weighted spoiled gradient-echo chemical shift (in/out-phase) imaging (scanner A, fast multiplanar spoiled gradient-recalled echo [FMPSPGR], TR 220, 90° flip angle, field of view, 48 cm; scanner B, turbo fast low-angle shot [turbo FLASH], TR 147, 70° flip angle, field of view, 40 cm); (4) coronal and axial fat-suppressed T1-weighted (scanner A, FMPSPGR, TR/TE 180/3, 90° flip angle, field of view, 40-48 cm; scanner B: turbo FLASH, TR/TE 105/1.9, 60° flip angle, field of view, 38 cm). In these sequences, the number of acquisitions was one, the section thickness was 6-7 mm with a 10-20% intersectional gap, and the imaging matrix was 256 x 151-168 (phase-frequency-encoding). A gadolinium chelate (gadopentetate dimeglumine, Magnevist, Schering) was injected in a dose of 0.1 mmol/kg of body weight as a bolus injection at 3 mL/sec using a power injector (Ulrich). This was followed by the acquisition of a coronal T1-weighted fat-suppressed 2D spoiled gradient-recalled echo sequence (machine A, FMPSPGR; machine B, 2D FLASH and 2D turbo FLASH and 3D volume interpolated breath-hold examination [VIBE]; TR/TE 4.2/1.8, 12° flip angle, field of view, 400 mm, 2-3 mm thickness; number of excitations, 1; matrix, 256 x 144). This sequence was acquired three consecutive times, the first time 20-30 sec after the start of contrast agent injection, and the following two times spaced 30 sec apart. A single late acquisition in the axial plane was obtained 4-5 min after contrast agent injection with the same sequence.
The locations of the different sequences and the number of planes were always the same as in the axial and coronal T2-weighted acquisitions.
Pre- and postgadolinium wall signal intensity and wall-thickness measurements were obtained with slices corresponding to three different T1-weighted sequences: FMPSPGR, 13 examinations in CD patients and three in controls or FLASH (13 examinations in CD patients and 12 in controls); turbo FLASH Fat Sat (11 examinations in CD patients); and 3D VIBE (three examinations in CD patients and five in controls).
Image Analysis
Pre- and postcontrast T1-weighted images were quantitatively analyzed for
colonic wall contrast enhancement (CE) and thickness with a dedicated
postprocessing workstation (Advantage Windows v4.0, GEMS; Leonardo, Siemens).
This image analysis was performed by the same board-certified radiologist
experienced in the interpretation of MRI, who was blinded to the clinical
score, symptoms, and colonoscopy findings. Images were also shown to the
reviewer in a random order for analysis.
Quantitative measurement of CE was assessed as the percentage increase in wall signal intensity (WSI) on administering gadolinium contrast with respect to the baseline signal before contrast agent injection using the following formula [7, 8]: CE = (WSI post-gadopentetate dimeglumine - WSI pre-gadopentetate dimeglumine/WSI pre-gadopentetate dimeglumine x 100 x (SD noise pre-/SD noise post-), where WSI corresponds to the average of five WSI measurements, SD noise pre- is the SD of the signal intensity measured outside of the body in the region anterior to the abdomen (noise region of interest [ROI], 1 cm2) before gadopentetate dimeglumine injection, and SD noise post- corresponds to the SD of the same noise ROI after gadopentetate dimeglumine administration.
The ROI within the colon wall was placed at the same segment location, preferably in the areas with larger thickening, avoiding scantly distended or artifactual segments both after contrast agent injection (WSI pre-gadopentetate dimeglumine) and after contrast administration (WSI post-gadopentetate dimeglumine) (Figs. 1A, 1B, 1C, and 1D). On average, the MRI analysis time per patient was 10-15 min. The ROI was placed in the wall of the affected segment on the precontrast image; after this, the ROI was copied and pasted in the postcontrast image.
Quantitative assessment of bowel thickening was measured in millimeters. Bowel wall thickening secondary to inflammatory edema and transmural fibrosis allowed differentiation between pathologic and normal bowel segments in patients with CD [5, 9]. Bowel wall thickness was measured manually at the active colon segments using the distance-measuring tool available on the workstation. This was done in well-distended, orthogonally imaged loops to both axial or coronal fat suppressed contrast-enhanced images (Figs. 1A, 1B, 1C, and 1D).
Statistical Analysis
All statistical analyses were performed using commercially available
dedicated software (Microsoft Excel 2000). The Student's t test for
paired data was used to evaluate differences between CE and thickness of the
active bowel segments in the different activity phases of CD, while the
Student's t test for independent variables was applied to compare
these same values between the group of patients with CD and the control group.
A statistically significant difference between the groups was considered to be
p < 0.05.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Regarding patient treatment in the active phases of CD, all subjects received combined therapy with two or more drugs (systemic corticoids in 55% of cases, topical corticoid [budesonide] in 10%, 5-aminosalicylic acid in 45%, and immune suppressors in 15%). During the remission periods, 16 patients received maintenance monotherapy in the form of 5-aminosalicylic acid, while four received combined aminosalicylates and immune suppressors. In the CD group, one abscess was diagnosed, along with one enterocutaneous fistula and two enteroenteric fistulas.
Colonoscopy
For comparison purposes, 53 segments were taken and regarded as pathologic;
the remaining 67 were regarded as normal (healthy segments) in the 20 patients
with CD, according to the gold-standard test. Sixty segments belonged to the
10 control patients. A total of 265 biopsies were taken from the 120 bowel
segments initially examined in the CD group. There were coincidental results
with visual assessment by endoscopy regarding disease activity in 83%
(220/265) of the biopsies. In the remaining 17% (45/265) of biopsies, there
was no correlation due to lack of endoscopical findings in the presence of
histologic changes (granulomas in 20/265 biopsies, 7.6%) or lack of acute
activity microscopical changes despite evident endoscopical changes (25/265
biopsies, 9.4%).
|
|
|
|
|
|
Quantitative MRI Analysis
Contrast Enhancement.—Within the CD patient group, a
decrease in the mean values of CE occurred from the active to the remission
phase, which showed a statistical significance compared on a
segment-by-segment basis (paired Student's t test)
(Table 1), (Figs.
2A,
2B,
2C,
2D,
2E,
2F,
3A,
3B,
3C,
3D,
3E,
3F, and
4). The mean CE of the active
bowel segments in patients with CD was greater than in the control group
(p < 0.05), although this difference disappeared upon reaching
remission. Likewise, no statistical significant differences in mean CE were
recorded between the healthy bowel segments in the CD group and the control
series (Table 2).
|
|
|
|
|
|
|
|
|
Bowel Wall Thickness.—In the CD patient group, a statistically significant decrease in wall thickness occurred within the segments between active and remission status (Table 1). In addition, mean bowel wall thickness of the pathologic segments in the CD group was also significantly greater than in the control group, in both the active and remission phases. No differences in wall thickness were observed in the healthy bowel segments of the CD group versus the controls (Table 2).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Thus, assessment of treatment response after recurrence continues to depend on clinical and laboratory testing, especially acute phase reactants [15]. Imaging techniques are reserved for evaluating the disease's complications; MRI has been shown to be useful in CD [2, 3, 5, 9, 16-19]. Some authors have reported a great sensitivity and specificity per bowel segment in detecting inflammatory activity and differentiating it from the inactive segments [2]. Other investigators have in turn observed a good correlation between MRI and endoscopy [1, 10] or between MRI and clinical settings [2] in patients with CD. Madsen et al. [5] reported variations in CE after medical treatment, although their study was limited by the few patients enrolled. Other authors have reported similar results with other techniques, such as computed axial tomography [15]. Our results were similar to these previous studies, as we also measured a significant decrease in CE on evolving from the active to the remission phase of CD.
For CE measurements, we applied a formula similar to that used by Semelka et al. [8], since it is simple and reproducible, and the results (expressed as a percentage) allowed us to include all the MRI examinations regardless of the pulse-sequence type used.
Despite CE decrease in the pathologic segments from active to remission phase, these segments remained pathologically thickened in both phases compared with the control group. This may be explained by the fact that thickening possesses an acute and therefore reversible component (inflammation and edema), and a chronic component (fibrosis) that does not reverse despite treatment response.
We decided to exclude patients with HIV infection, since CD is very rare in immune-suppressed patients. In addition, colitis and ileitis of infectious origin, such as tuberculosis, Mycobacterium avium, or cytomegalovirus cause lesions endoscopically similar to those of CD.
Our study was limited by the reduced number of bowel segments compared in CD patients and controls. This was due to the need to include only the pathologic segments (a mere 53 of the 120 total segments) and the high cost of MRI compared with other imaging techniques, such as sonogram or barium contrast enema. We tried to obtain a control group of similar age to the patient group, but the corresponding young age population is less inclined to undergo a colonoscopy; thus, colonoscopy is usually performed in an older population. We established an interval between both MRI studies of over 1 month (mean, 233 days), since a shorter interval would possibly only reflect clinical changes, with little variation in inflammation at a local level. We also did not consider it advisable to repeat colonoscopy in the remission phase, since it has been shown to be of little benefit in the follow-up of CD patients [20]. To have a well-distended colon and terminal ileum, both barium sulfate orally and water as an enema were given to the patient simultaneously. Since both act as negative contrast agents on T1-weighted sequences, postgadolinium measurements of the bowel wall can be done better.
Another limitation of our study was the manual placement of the ROI over the bowel wall, as this implies bias, because the cursor may not be placed precisely on target. We calculated the mean of five measurements to overcome this problem. Nevertheless, it must be considered that all measurements were made by the same radiologist, who has extensive experience in the field, instead of by two or more reviewers as usual. We decided to sustain the study on numeric values lacking subjectiveness, taken from each of the 120 segments of the CD patients, with total blinding to the colonoscopy and histology findings. Our routine imaging protocol for CD patients comprises T1- and T2-weighted sequences, although the measurements of the present study are limited to the former. Due to practical daily working needs and the fact that we use MRI systems from two different vendors, we adapted the protocol over time with various T1-weighted sequences. Initially we used FMPSGR and FLASH sequences; later, we incorporated the fat-suppressed 2D turbo FLASH and 3D VIBE sequences, which, in our opinion, improved image quality compared with the former sequences. In our institution, we use abdominal MRI in CD patients mainly for assessing internal and external abdominal fistulas, other small bowel disorders than those affecting the terminal ileum, and for completing the study of colon stenoses and their degree of activity. In addition, we use MRI to avoid ionizing radiation exposure in the pediatric CD population and in women during pregnancy. We continue to use CT to evaluate intraabdominal fluid accumulations (phlegmons and abscesses) or mesenteric fat disorders.
Our results show that MRI clearly allows the distinction of pathologic from normal bowel wall in CD, as it detects significant variations in bowel wall thickness with clinical improvement and is able to reflect pathologic inflammatory changes at the bowel wall based on variations in the CE. Consequently, this technique is reliably applicable to the follow-up of patients with CD. More extensive series should be investigated that involve defined disease patterns (fibrosing, inflammatory, fistulizing) to more precisely define the role of MRI in patients with Crohn's disease.
MRI is able to detect significant variations in bowel wall thickness and contrast enhancement, reflecting favorable clinical response to medical treatment of CD's relapse. In addition to its lack of ionizing radiation, this may allow MRI to be the imaging technique of choice for the follow-up of patients with active CD.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. K. Hara, S. Alam, R. I. Heigh, S. R. Gurudu, J. G. Hentz, and J. A. Leighton Using CT Enterography to Monitor Crohn's Disease Activity: A Preliminary Study Am. J. Roentgenol., June 1, 2008; 190(6): 1512 - 1516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Sinha, C. Nwokolo, and P. D Murphy Magnetic resonance imaging in Crohn's disease BMJ, February 2, 2008; 336(7638): 273 - 276. [Full Text] [PDF]
|
|
|
|
 |
|
 F. Booya, J. G. Fletcher, J. E. Huprich, J. M. Barlow, C. D. Johnson, J. L. Fidler, C. A. Solem, W. J. Sandborn, E. V. Loftus Jr, and W. S. Harmsen Active Crohn Disease: CT Findings and Interobserver Agreement for Enteric Phase CT Enterography Radiology, December 1, 2006; 241(3): 787 - 795. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
