| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 All authors: Department of Radiology, Hospital of the University of Pennsylvania, 3400 Spruce St., Philadelphia, PA 19104.
Received August 31, 2007;
accepted after revision November 3, 2007.
Address correspondence to M. S. Levine
(marc.levine{at}uphs.upenn.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Jejunal diverticula have characteristic findings on CT, appearing as discrete round or ovoid, contrast-, fluid-, or air-containing structures outside the expected lumen of the small bowel, with a smooth, barely discernible wall and no recognizable small-bowel folds. Not infrequently, these structures are seen to communicate directly with an adjoining small-bowel loop, a feature best recognized by scrolling the images. Our experience suggests that jejunal diverticulosis can often be recognized on the basis of the characteristic CT features of this condition.
Keywords: abdominal CT • barium studies • jejunal diverticulosis • small bowel
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Jejunal diverticulosis is readily diagnosed on barium studies of the small bowel by the presence of multiple rounded, variably sized jejunal outpouchings, often seen in profile to have discrete, relatively narrow necks [1, 2]. In patients with massive jejunal diverticulosis, paradoxically, differentiating the diverticula from overlapping loops of small bowel can be more difficult. Although the radiographic features of jejunal diverticulosis are well recognized on barium studies, there have only been anecdotal descriptions of jejunal diverticulosis on CT [6–8]. The purpose of our study therefore was to better characterize the CT findings of jejunal diverticulosis by retrospectively reviewing abdominal CT scans in a series of patients with this condition on barium studies.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Barium Technique
All of the small-bowel follow-through examinations were performed using a
50% weight/volume (w/v) barium suspension (Entrobar, Mallinckrodt). The
fluoroscopists obtained periodic spot images of the entire small bowel
(including the terminal ileum) with graded compression under fluoroscopic
guidance. Small-bowel enemas (enter oclysis) were performed by peroral
intubation of the proximal jejunum, followed by administration of an 80% w/v
barium suspension (Entero-H, E-Z-EM) and methylcellulose. All barium studies
were performed using digital fluoroscopic equipment (Diagnost 76, Philips
Medical Systems; Sireskop SD, Siemens Medical Solutions).
CT Technique
CT was performed using a variety of helical CT scanners (including
single-detector and multidetector scanners) according to an established
protocol. All but one patient (96%) received positive oral contrast material
(1,000 mL of barium sulfate [ReadiCat, E-Z-EM]) 30–60 minutes before the
study. Twenty-five patients (89%) also had IV administration of 150 mL of 60%
iodinated contrast material (iohexol [Omnipaque 300, GE Healthcare]). CT
images were routinely obtained with the patient in a supine position during
full inspiration. Axial images were obtained at 5- or 7-mm slice collimations
(pitch, 1.3:1; 200–220 mAs) and reconstructed with a softtissue algo
rithm, but coronal and sagittal reformatting of the images was not performed.
The CT images were reviewed retrospectively using the same format that was
used for the original review; thus, the images were not reformatted in coronal
or sagittal planes.
Case Review
The abdominal CT scans from all 28 patients were reviewed retrospectively
at a computer workstation by a consensus of two gastrointestinal radiologists
(with 25 and 23 years of experience, respectively). The workstation enabled
interpretation and scrolling of images in the axial plane. The studies were
reviewed to determine the number, size, shape, and content (including air,
fluid, debris, and contrast material) of the diverticula and the thickness of
the diverticular wall and the presence or absence of a discrete neck or
visible small-bowel folds within the diverticulum. Even when a diverticular
neck was not identified, the images were scrolled to determine whether one or
more diverticula could be shown to communicate with adjoining small-bowel
loops. The scans were also reviewed for the presence or absence of associated
findings, including mesenteric fluid or air; free intra peritoneal air;
dilated small bowel; and duodenal, ileal, or colonic diverticula.
The images from the barium studies in these 28 patients were reviewed separately to determine the number, size, and shape of the diverticula and to determine how often discrete diverticular necks were visualized in profile from one or more diverticula. The original radiologic reports were also reviewed to determine whether small-bowel transit time was described as normal or delayed (i.e., longer than 120 minutes for barium to reach the colon).
Jejunal diverticulosis was defined as the presence of two or more diverticula in the jejunum on barium examinations. For the purposes of this investigation, the degree of diverticulosis on barium studies and CT was classified on the basis of the number of jejunal diverticula as mild (2–4 diverticula), moderate (5–10 diverticula), marked (11–30 diverticula), or massive (more than 30 diverticula). The original abdominal CT reports were also reviewed to determine whether the diagnosis of jejunal diverticulosis had been suggested on these reports.
Medical records (including clinical, laboratory, and pathologic findings) were reviewed to determine the clinical presentation of these 28 patients at the time of CT and the treatment and clinical course. Our institutional review board approved all aspects of this retrospective study and did not require informed consent from patients whose records were included in the study.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The patient with melena underwent a laparotomy, which confirmed the presence of jejunal diverticulosis with a bleeding diverticulum. A localized small-bowel resection was performed, and the bleeding resolved postoperatively.
Radiographic Findings
Barium studies—Barium studies revealed jejunal
diverticulosis that was classified as mild in 12 patients (43%), moderate in
eight (29%), marked in seven (25%), and massive in one (4%). The diverticula
appeared as rounded, barium-filled outpouchings that ranged from 0.2 to 5.0 cm
in diameter (Figs. 1B and
2C). Discrete diverticular
necks were visualized in profile for one or more diverticula in 25 patients
(89%). Small-bowel transit time was described as normal on the original
radiologic reports in 24 patients (86%) and as delayed in four (14%) (i.e.,
long er than 120 minutes).
|
|
In contrast, our retrospective review of the abdominal CT scans revealed radiographic findings of jejunal diverticulosis in 21 (75%) of 28 patients with this condition on barium studies. The degree of jejunal diverticulosis was classified on CT as mild in 12 patients (43%) (Fig. 3), moderate in six (21%) (Fig. 1A, 1B), marked in two (7%), and massive in one (4%) (Fig. 2A, 2B, 2C). Using the barium study as the reference standard, CT correctly staged the degree of jejunal diverticulosis in seven patients (25%), underestimated the degree of diverticulosis in 17 patients (61%) (by one level in 11 patients, two levels in five, and three levels in one), and overestimated the degree of diverticulosis in four patients (15%) (by one level in three patients and two levels in one). Thus, CT was far more likely to underestimate than to overestimate the degree of jejunal diverticulosis in these patients. When the CT findings were stratified based on the degree of jejunal diverticulosis on barium studies, CT revealed diverticula in eight (67%) of 12 patients with mild diverticulosis, six (75%) of eight with moderate diverticulosis, six (86%) of seven with marked diverticulosis, and one (100%) of one with massive diverticulosis. Thus, four (58%) of seven patients with jejunal diverticulosis not seen on CT, even in retrospect, had mild disease with fewer than five diverticula on barium studies.
|
|
|
|
In all 21 patients in whom jejunal diverticula were visualized in retrospect on CT, the diverticula appeared as discrete round or ovoid structures outside the expected lumen of the small bowel, with a smooth, barely discernible wall and no recognizable small-bowel folds (Figs. 1A, 1B, 2A, 2B, 2C, 3, 4, 5). One or more diverticula were seen to communicate with an adjacent small-bowel loop in eight (38%) of the 21 patients by scrolling images in the axial plane, and a discrete diverticular neck was visualized in three patients (14%) (Fig. 4). The jejunal diverticula were found to contain air in all 21 patients (100%); air and contrast material (with air–contrast levels) were seen in one or more diverticula in 16 patients (76%) (Figs. 1A and 2A); air and nonenhancing fluid (with air–fluid levels) were seen in one or more diverticula in two patients (10%) (Figs. 4 and 5); and air alone was seen in one or more diverticula in three patients (14%) (Fig. 3). The diverticula ranged from 0.2 to 4.5 cm in diameter; 15 (54%) of the 28 patients had one or more diverticula less than 1 cm in diameter (Fig. 3).
|
|
Six (21%) of the 28 patients with jejunal diverticulosis were found to have associated complications on CT, including jejunal diverticulitis in three, diverticular fecaliths in two (without small-bowel obstruction or diverticulitis), and diverticular pneumatosis with a tiny amount of mesenteric gas in one (Fig. 2A). In all three patients with diverticulitis, CT revealed peridiverticular stranding in the adjacent small-bowel mesentery, and in two, CT revealed associated mesenteric gas (n = 1) (Fig. 5) or gas and fluid with an air–fluid level (n = 1) [9]. One of the three patients with jejunal diverticulitis received antibiotics with resolution of abdominal pain after treatment. In the remaining two cases, follow-up data were not available.
Barium studies or CT also revealed a pulsion diverticulum in the esophagus in one (4%) of the 28 patients and additional diverticula in the duodenum in 19 patients (68%), in the ileum in seven (25%), and in the colon in 23 (82%).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although the radiographic features of jejunal diverticulosis are well recognized on barium studies [1, 2], there have only been anecdotal descriptions of this condition on abdominal CT [6–8]. Jejunal diverticula have been described on CT as rounded outpouchings from the small bowel containing some combination of contrast material, air, and debris [6, 8]. Other investigators have suggested that large (greater than 3 cm in diameter) jejunal diverticula can be differentiated from adjacent small-bowel loops on CT by their differing contents and by the absence of discernible valvulae conniventes [7].
Despite these earlier reports, jejunal diverticulosis is an infrequent diagnosis on abdominal CT. In our study, jejunal diverticulosis was described on the original CT reports in only two (7%) of 28 patients with this condition on barium examinations. In both cases, the barium studies were performed before CT, so the diagnosis of jejunal diverticulosis could have been known at the time of CT. If so, there were no patients with jejunal diverticulosis in whom the diagnosis initially was suggested on the basis of the CT findings. The difficulty detecting jejunal diverticulosis on CT is probably related to a combination of factors including the subtle criteria used to differentiate diverticula from gas- or fluid-filled loops of small bowel as well as that jejunal diverticulosis is an uncommon cause of abdominal symptoms, so radiologists may not search for jejunal diverticula on CT even when the diverticula are causing symptoms.
In contrast, our retrospective review of the CT scans revealed radiographic findings of jejunal diverticulosis in 21 (75%) of 28 patients with this condition on barium studies; four (58%) of the seven patients with diverticula missed on CT had mild disease with fewer than five diverticula on barium studies. CT also was far more likely to underestimate than to overestimate the degree of jejunal diverticulosis in terms of the number of visualized diverticula in comparison with barium studies, again because of the frequent difficulty identifying these structures on CT.
In all 21 patients, the jejunal diverticula appeared on CT as discrete round or ovoid, contrast-, fluid-, or air-containing structures outside the expected lumen of the small bowel, with a smooth, barely discernible wall and no recognizable small-bowel folds (Figs. 1A, 1B, 2A, 2B, 2C, 3, 4, 5). The majority of patients had diverticula less than 1 cm in diameter, so it was possible to detect even small jejunal diverticula on CT. Not infrequently, the site of communication with an adjoining small-bowel loop was delineated by scrolling images in the axial plane (Fig. 4), and a discrete diverticular neck was occasionally visualized on CT (Fig. 4). The latter findings may be easier to recognize on CT images reformatted in coronal or sagittal planes. Because CT enterography enables better luminal distention of small-bowel loops in axial, coronal, and sagittal planes, we suspect this technique will increase the sensitivity of CT for detecting jejunal diverticulosis.
Jejunal diverticulosis was diagnosed on CT in 19 (68%) of 24 patients with nondilated small bowel versus only two (50%) of four with dilated small bowel, so it was more difficult to visualize these structures on CT in patients with dilated small bowel. None of these four patients had signs of small-bowel obstruction, so the dilated small bowel was most likely secondary to the hypomotility known to occur in patients with jejunal diverticulosis [2]. In any case, this limitation of CT is most likely related to effacement of folds and thinning of the wall in dilated small-bowel loops, so it is more difficult to detect jejunal diverticula on axial CT images in the presence of dilated small bowel. CT images reformatted in coronal or sagittal planes may therefore be especially helpful for differentiating dilated small-bowel loops from jejunal diverticula in this setting.
Our investigation has the inherent limitations of a retrospective study. In particular, we recognize that our retrospective review of the CT images in patients with known jejunal diverticulosis on barium studies resulted in interpretation bias that precludes obtaining meaningful data about the sensitivity of CT in diagnosing this condition. However, the purpose of this study was not to determine the sensitivity of CT for detecting jejunal diverticulosis, but rather to better characterize the CT findings in patients with this condition.
In conclusion, jejunal diverticula have characteristic findings on CT, appearing as discrete round or ovoid, contrast-, fluid-, or air-containing structures outside the expected lumen of the small bowel, with a smooth, barely discernible wall and no recognizable small-bowel folds. Not infrequently, these structures are seen to communicate directly with an adjoining small-bowel loop, a feature best recognized by scrolling the images. Our experience suggests that jejunal diverticulosis can often be recognized on the basis of the characteristic CT features of this condition.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
