Professional Documents
Culture Documents
Abstract
Full Text
Figures
References
PDF
PDF Plus
Add to Favorites
Permissions
Download Citation
ABSTRACT
oBJECTIVE. The purpose of this study was to assess whether prone or supine imaging provides
superior small-bowel loop distention during MRI small-bowel follow-through examinations and
whether either position is better with regard to lesion detection and evaluation.
SUBJECTS AND METHODS. Forty consecutively enrolled clinically referred patients with
known or suspected small-bowel abnormalities prospectively underwent 62 MRI small-bowel
follow-through examinations in both the prone and the supine positions. Images were blindly and
independently reviewed by two observers. Each small-bowel segment was assessed with a 3-
point scoring system, and differences in bowel distention in the prone and supine positions were
evaluated with a paired Wilcoxon's test. Differences between rates of lesion detection and
characterization (e.g., ulceration, stricturing) were analyzed with a paired Student's t test.
Interobserver agreement was estimated with the kappa coefficient.
RESULTS. In both normal and diseased small bowel, the prone position had statistically
significantly higher distention scores than did the supine position (p < 0.05) with a high level of
interobserver agreement. This finding, however, did not translate into improved lesion detection
or characterization (p > 0.05).
CONCLUSION. Although use of the prone position results in superior small-bowel distention
during MRI small-bowel follow-through, both the prone and supine positions are equal in terms
of lesion detection and feature visualization.
MRI small-bowel follow-through is an evolving promising diagnostic tool for evaluation of the
small bowel. Providing exquisite soft-tissue contrast resolution, superb mural and extramural
enteric depiction, and multiplanar imaging capabilities, this approach has generated considerable
interest as a potential successor to conventional small-bowel follow-through and CT
enterography in the presence of suspected small-bowel abnormalities. The attraction of this
approach is further augmented by the absence of associated ionizing radiation exposure,
facilitating sequential imaging and derivation of vital information regarding small-bowel
function and the effect of disease thereon [1, 2].
It remains unknown whether one position surpasses the other in terms of bowel distention and
lesion detection. The prone position has been preferred [5], however, because it exerts mild
pressure on the anterior abdominal wall, facilitating separation of small-bowel loops. The
purpose of this study was to assess which patient position, prone or supine, is associated with
better small-bowel distention and lesion detection. To the best of our knowledge, no such study
has been reported in the medical literature.
From November 2004 to July 2005, 40 consecutively enrolled patients (22 women, 18 men;
mean age, 32 years; range, 15–73 years) prospectively underwent 62 MRI examinations with
small-bowel follow-through. Each patient was referred for clinical reasons and had a known or
suspected diagnosis of Crohn's disease (n = 38), small-bowel tumor (n = 8), small-bowel
lymphoma (n = 9), or celiac disease (n = 7). The exclusion criterion was contraindications to
MRI (claustrophobia, certain implanted metallic devices). Institutional review board approval
and patient informed consent were obtained.
Patient Preparation
All patients fasted from midnight of the night before MRI. No bowel preparation, medications to
promote gastric emptying, or bowel relaxation or paramagnetic contrast agents were
administered before imaging. Bowel distention was achieved with a single packet of
polyethylene glycol solution (Klean-Prep, Norgine) diluted in 1,000 mL of water and a small
amount of orange cordial, added for optimization of patient tolerance. Polyethylene glycol
(polyethylene glycol 3350, 59.0 g; anhydrous sodium sulfate, 5.685 g; sodium bicarbonate, 1.685
g; sodium chloride, 1.465 g; potassium chloride, 0.7425 g) is a high-osmolarity non-absorbed
contrast medium that provides excellent intraluminal contrast and distention. Patients were
instructed to continuously ingest the solution over a 10- to 15-minute period at a constant rate of
approximately 75–100 mL/min monitored by an MRI technologist. The patients ingested this
solution while seated in the MRI preparation room and were brought to the MRI unit only when
images were being acquired. Additional oral contrast medium was not administered between
prone and supine imaging acquisitions.
Initial images were acquired 10 minutes after the patient completed ingestion of 1,000 mL of the
oral contrast agent. A second series of images was obtained 30 minutes after completion of
ingestion and every 20 minutes thereafter until the contrast bolus was deemed to have adequately
passed to the colon. For the first set of images, the MRI technologist randomly used the prone or
the supine position. Patients were then repositioned and immediately imaged in the other
orientation.
Svi pacijenti su postili od ponoći pre MR. Pre snimanja nisu davani nikakva
priprema creva, lekovi za podsticanje pražnjenja želuca, opuštanje creva ili
paramagnetni kontrastni agensi. Distenzija creva je postignuta jednim paketom
rastvora polietilen glikola (Klean-Prep, Norgine) razblaženog u 1000 mL vode i
malom količinom narandžastog kordijala, dodatog za optimizaciju tolerancije
pacijenata. Polietilen glikol (polietilen glikol 3350, 59,0 g; anhidrovani
natrijum sulfat, 5,685 g; natrijum bikarbonat, 1,685 g; natrijum hlorid, 1,465
g; kalijum hlorid, 0,7425 g) je odličan kontrastni medijum bez kontrasta sa
visokim kontrastom aluminijuma. i distenzija. Pacijentima je naloženo da
kontinuirano gutaju rastvor tokom perioda od 10 do 15 minuta konstantnom
brzinom od približno 75-100 mL/min koju prati tehnolog za MRI. Pacijenti su
unosili ovaj rastvor dok su sedeli u prostoriji za pripremu MRI i dovođeni su
u jedinicu za magnetnu rezonancu tek kada su se snimale slike. Dodatni oralni
kontrastni medijum nije davan između snimanja na ležećem i ležećem položaju.
Imaging Technique
MRI was performed with a 1.5-T system (Magnetom Symphony, Siemens Medical Solutions)
equipped with high-performance gradient coils characterized by a maximum gradient amplitude
of 52 mT/m and a slew rate of 125 mT/m/ms. A phased-array body coil also was used for
optimization of signal reception. Multiplanar rapid localization was followed by heavily T2-
weighted 2D true fast imaging with steady-state precession sequences in the coronal and axial
planes encompassing the diaphragmatic apex to the groin in craniocaudal extent. These true fast
imaging with steady-state precession acquisitions incorporated the following parameters: TR/TE,
4.6/2.3; flip angle, 70°; matrix size, 205 × 256; field of view, 400 × 400 mm. A slice thickness of
5 mm enabled acquisition of 19 sections within a comfortable single breath-hold of 21 seconds.
Image Analysis
Images were independently and qualitatively assessed by two consultant radiologists with more
than 6 and 10 years of experience (more than 250 and 350 examinations). Observers were
blinded to symptoms and clinical data. A minimum interval of 12 weeks passed between
evaluation of the prone and supine data sets from each patient. Studies were displayed and
reviewed on a PACS workstation without data compression. For image assessment, the small
bowel was considered to comprise five distinct segments: duodenum, jejunum, proximal ileum
(left of midline), distal ileum (right of midline), and terminal ileum. Prone and supine images
were qualitatively evaluated with regard to degree of small-bowel distention with a visual
assessment grading scale that ranged from 1 to 3. A score of 1 represented poor small-bowel
distention; 2, moderate distention; and 3, excellent distention (Figs. 1, 2, and 3).
In assigning scores to each anatomic segment, observers were requested to determine adequacy
of small-bowel distention according to visibility of the bowel lumen, endoluminal folds, and
bowel wall. The grade that described and represented each segment most was the assigned score.
This approach was similar to the grading system used by Laghi et al. [8] and Minowa et al. [11].
Jejunal loops were differentiated from ileal loops both by their respective locations and by the
configuration of their folds, more closely packed folds indicating jejunum, sparse folds
indicating ileum. The number and location of intraluminal, mural, and extraluminal lesions
depicted at each segment were recorded and compared for each observer. Observers were
requested to describe lesion characteristics (e.g., ulceration, effect on adjacent bowel lumen such
as stricture, and aneurysmal dilatation) in the case of each lesion.
Statistical Analysis
Statistical analysis was performed with the SPSS program (version 14, SPSS). Scores assigned to
bowel distention in the diseased segments, all small-bowel segments, and for each MRI small-
bowel follow-through study as a whole (by combining the scores for each segment of bowel for a
given patient) at supine and prone imaging were analyzed with a paired Wilcoxon's test for
ordinal data. Differences between rates of lesion detection in the prone and supine positions were
evaluated with a paired Student's t test. A value of p < 0.05 was considered to represent the
presence of a statistically significant difference. Interobserver agreement with regard to small-
bowel distention was assessed with the kappa coefficient. A kappa value of 0 indicated poor
agreement; 0.01–0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate
agreement; 0.61–0.80, good agreement; and 0.81–1.00, excellent agreement.
Statistička analiza
Twenty-one of the 40 patients included in this study underwent a single examination (prone and
supine), 16 patients underwent two examinations (prone and supine), and three patients
underwent three examinations (prone and supine), for a total of 62 paired (prone and supine)
examinations for qualitative analysis. All examinations were successfully completed, and orally
ingested contrast medium was completely consumed and tolerated by all patients. The duration
of each MRI session, including positioning on the table, coil placement, and image acquisition,
varied from 5 to 8 minutes. The total time from entering the department to completion of the
examination varied from 40 to 80 minutes (average, 51 minutes).
All images were considered of diagnostic quality by both observers. Findings at 34 of the 62
examinations were deemed by both observers to be within normal limits, and 28 examinations
showed pathologic changes, subsequently confirmed as Crohn's disease in 16 patients, bowel
carcinoid in one patient, small-bowel lymphoma in four patients, polyps in two patients, and
extraluminal lesions including a splenic mass in one patient, mesenteric nodes in two patients,
and ovarian cysts in two patients.
For both observers, use of the prone position resulted in significantly higher small-bowel
distention scores in the diseased bowel segments than did use of the supine position (n = 20;
mean supine score observer 1, 2.2 ± 0.8 [SD]); mean prone score observer 1, 2.6 ± 0.4; mean
supine score observer 2, 2.1 ± 0.6; mean prone score observer 2, 2.6 ± 0.4) (p < 0.05). Use of the
prone position also resulted in higher scores for all small-bowel segments (n = 310; mean supine
score observer 1, 2.2 ± 0.4; mean prone score observer 1, 2.5 ± 0.4; mean supine score observer
2, 2.1 ± 0.5; mean prone score observer 2, 2.5 ± 0.5) (p < 0.05). Use of the prone position also
resulted in higher scores for each small-bowel study as a whole, comprising the total scores
assigned to each segment within a patient (n = 62; mean supine score observer 1, 11.4 ± 1.9;
mean prone score observer 1, 12.1 ± 2; mean supine score observer 2, 11.2 ± 1.9; mean prone
score observer 2, 12.2 ± 2.1) (p < 0.05). The distribution of these scores for both observers is
outlined in Table 1. The degree of interobserver agreement for scores assigned to bowel
distention was considered good (κ = 0.79).
Rezultati
Dvadeset i jedan od 40 pacijenata uključenih u ovu studiju podvrgnut je jednom
pregledu (ležeći i ležeći), 16 pacijenata je podvrgnuto dva pregleda (ležeći i
ležeći), a tri pacijenta su podvrgnuta tri pregleda (ležeći i ležeći), ukupno
62 uparena (ležeće i ležeće) preglede za kvalitativnu analizu. Svi pregledi su
uspešno obavljeni, a oralno unet kontrastni medij su u potpunosti konzumirali
i tolerisali svi pacijenti. Trajanje svake MRI sesije, uključujući
pozicioniranje na stolu, postavljanje kalema i dobijanje slike, variralo je od
5 do 8 minuta. Ukupno vreme od ulaska na odeljenje do završetka ispita
variralo je od 40 do 80 minuta (prosečno 51 minut).
All lesions were identified on both prone and supine images by each observer. No significant
difference was detected between rates of identification of various mural characteristics for either
position (Table 2). These characteristics included wall and fold thickening and fold attenuation,
mural ulceration, pseudopolyps, bowel stenosis, strictured segments, bowel dilation, and
separation of bowel loops. A number of extracolonic lesions were found, including two lymph
node masses (6 cm and 4 cm in short-axis diameter), two polyps with a diameter less than 1 cm
(9 mm and 8 mm), a splenic mass (2 cm), and two ovarian cysts (3 × 2 cm and 2.5 × 1.4 cm).
There was no significant difference in visualization, detection, or lesion size between the two
positions (p > 0.05).
Discussion
We found statistically significantly better distention of diseased segments of small bowel alone
and all small-bowel segments and better distention for each MRI small-bowel follow-through
examination with imaging in the prone position than with imaging in the supine position (p <
0.05). Despite this clear qualitative advantage of prone imaging, lesion detection and
characterization did not benefit as a result; similar rates of lesion detection were found for each
position.
Optimization of bowel distention undoubtedly is key to prevention of false-positive [5, 12] and
false-negative [4, 13] results during MRI small-bowel follow-through. However, the results of
this study suggest that although use of the prone position results in significantly better small-
bowel distention than does use of the supine position, this improved distention does not translate
into improved lesion detection or morphologic evaluation. This finding may have positive
implications in the imaging of patients unable to adopt one of the positions, be it as a result of
pain or surgical incision, or in pediatric imaging, in which use of the prone position can cause
increased levels of anxiety.
The importance of optimal small-bowel distention at MRI small-bowel follow-through has been
repeatedly emphasized in the literature. Poor distention can lead to difficulties of interpretation,
false-positive results occurring as the result of apparent thickening of the bowel wall and folds
[5, 12], and false-negative results emanating from lesion obscuration due to suboptimal mural
delineation [4, 13]. A number of approaches to ensuring reliable and reproducible small-bowel
distention have been described [8, 13–23], including optimization of contrast agent, addition of
osmotic and nonosmotic agents, refinement of volumes of contrast material administered [3], and
the timing of image acquisition [3, 24]. The technique described is routinely used at our
institution and has been validated in the literature [14, 25–27].
We investigated the effect of patient position on small-bowel distention, lesion detection, and
lesion visualization. The hypothesis that distention seen in one position may surpass that in a
different position stemmed from studies [28, 29] relating to colonic distention, during which
improved observer confidence and polyp detection were found during CT colonography in both
the supine and prone positions compared with supine acquisition alone. Fletcher et al. [29]
determined that supplementary prone scanning increased sensitivity both by overcoming
segmental collapse and by reducing perceptual errors due to poor polyp conspicuity in
underdistended segments. The same may hold true for MRI small-bowel follow-through with
regard to distribution of oral contrast material and may thus positively affect bowel distention.
Other potential advantages of the prone position include compensation for inability to compress
and separate bowel loops, as in conventional enterography, because the prone position exerts
mild uniform pressure on the anterior abdominal wall, facilitating separation of small-bowel
loops and decreasing the volume of the peritoneal cavity to be imaged.
MRI small-bowel follow-through has a number of advantages over other currently available
techniques. In contrast to conventional enterography, MRI small-bowel follow-through provides
detailed information about the small-bowel wall and has the potential to depict extraluminal
pathologic manifestations of disease, such as abscesses and mesenteric phlegmon, fibrofatty
proliferation, lymphadenopathy, and colonic skip lesions in Crohn's disease [2, 30, 31]. The
cross-sectional nature of this approach renders evaluation of overlapping bowel loops
straightforward, obviating the requirement for ionizing radiation exposure. The latter feature is a
particular advantage because many patients undergoing this examination are young (mean age in
this study, 35 years). In addition to these advantages, MRI small-bowel follow-through does not
involve concessions in small-bowel functional information derived, as is the case with CT
enterography [32]. Although the spatial resolution of conventional enterography may be better
than that of MRI small-bowel follow-through, especially for early superficial mucosal lesions,
the multiplanar imaging capabilities of this approach compensate for and perhaps negate this
limitation [6]. Studies [10, 30, 33] have shown MRI small-bowel follow-through to be as
effective as conventional enterography in the primary diagnosis of Crohn's disease, the main
indication for this examination.
MRI praćenje tankog creva ima brojne prednosti u odnosu na druge trenutno
dostupne tehnike. Za razliku od konvencionalne enterografije, MRI praćenje
tankog creva pruža detaljne informacije o zidu tankog creva i ima potencijal
da prikaže ekstraluminalne patološke manifestacije bolesti, kao što su apscesi
i mezenterični flegmon, fibromasna proliferacija, limfadenopatija i
preskakanje debelog creva. kod Kronove bolesti [2, 30, 31]. Priroda poprečnog
preseka ovog pristupa čini procenu preklapajućih crevnih petlji jednostavnom,
eliminišući zahtev za izlaganjem jonizujućem zračenju. Ova poslednja
karakteristika je posebna prednost jer su mnogi pacijenti koji se podvrgavaju
ovom pregledu mladi (srednja starost u ovoj studiji, 35 godina). Pored ovih
prednosti, MRI praćenje tankog creva ne uključuje ustupke u dobijenim
funkcionalnim informacijama tankog creva, kao što je slučaj sa CT
enterografijom [32]. Iako prostorna rezolucija konvencionalne enterografije
može biti bolja od one kod MRI praćenja tankog creva, posebno za rane
površinske lezije sluzokože, mogućnosti multiplanarnog snimanja ovog pristupa
kompenzuju i možda negiraju ovo ograničenje [6]. Studije [10, 30, 33] su
pokazale da je MRI praćenje tankog creva jednako efikasno kao i konvencionalna
enterografija u primarnoj dijagnozi Kronove bolesti, glavne indikacije za ovo
ispitivanje.
Limitations of this study included our inability to blind the observers to patient position. As the
result of gravity, contrast solution conformed to the dependent position in the stomach and
cecum. We believed, however, that the long interval between evaluation of prone and supine
images would prevent recollection of the small-bowel appearances of particular importance in
the evaluation of diseased bowel segments.
In conclusion, the prone imaging position offers statistically superior small-bowel distention of
both normal and diseased loops during MRI small-bowel follow-through. The positions,
however, are equal in terms of lesion detection and morphologic evaluation. Therefore, patients
may be imaged in either position according to preference or necessity without compromise of
diagnostic accuracy.