| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University Hospital Dr. Peset, 90 Gaspar Aguilar
Ave., 46017, Valencia, Spain.
2
Department of Radiology, Sagunto Hospital, 46017, Valencia, Spain.
Received November 16, 2000;
accepted after revision January 29, 2001.
Address correspondence to T. Ripollés.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
METHODS AND METHODS. A review was made of the radiologic findings of 17 consecutive patients with surgically verified gastrointestinal bezoars over a period of 51 months.
RESULTS. Twelve patients had a history of previous gastric surgery. In no patient was a bezoar clinically suspected. Phytobezoars were recorded in 16 patients and a trichobezoar in only one. A total of 33 bezoars were identified at surgery. Two patients had isolated gastric bezoars, whereas 15 patients had bezoars located in the small bowel. Among the latter group, associated gastric bezoars were found in eight patients, and five patients had multiple intestinal bezoars. Abdominal radiographs revealed bezoars in three patients, sonography revealed bezoars in 15, and CT revealed bezoars in all 17. Seven patients had associated gastric bezoars revealed at CT versus only two patients with gastric bezoars revealed at sonography. CT revealed multiple intestinal bezoars in five patients whereas sonography revealed them in only two patients.
CONCLUSION. Both sonography and CT are reliable methods for diagnosing gastrointestinal bezoars. CT is more accurate, however, and exhibits a quite characteristic bezoar image; in addition, this imaging technique is able to reveal the presence of additional gastrointestinal bezoars.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Findings on radiographs can indicate the presence of bezoars, and barium studies can confirm it [1]. According to some recent publications, sonography and CT can be useful in diagnosing bezoars before surgery [2,3,4,5,6]. Most of these studies have been case reports or studies with only a few patients and have found relatively specific sonographic characteristics and specific CT features for bezoars. The most extensive series published in the literature to date studied seven patients, of whom only one had findings that corresponded to those of a phytobezoar [7]. To our knowledge, no study has yet contrasted the usefulness of the different imaging techniques in application to bezoars.
The present study describes the radiographic, sonographic, and CT characteristics of 17 patients with surgically confirmed gastric or intestinal bezoars. We also assess the comparative value of the different imaging studies in the diagnosis of gastrointestinal bezoars.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Diagnosis was based on previously published data [1,2,3,4,5,6,7]. On the conventional radiographs, bezoars were considered as a diagnosis when mottled radiotransparencies in the interstices of solid matter were identified (Fig. 1). Sonographic diagnosis of a bezoar was established by detection of an intraluminal mass with a hyperechoic arclike surface and a marked acoustic shadow (Fig. 2). Finally, CT diagnosis of bezoar was based on identifying a low-density intraluminal mass containing air bubbles and exhibiting a characteristic mottled appearance (Fig. 3). The imaging study data were obtained from the imaging reports made before surgery—with the exception of bezoar length, which was determined retrospectively.
|
|
|
The sonographic examinations were performed using a scanner (Sonolayer SSH-140A or SSA-250A; Toshiba, Tokyo, Japan) with a convex 3.75-MHz probe. Patients underwent CT performed with one of two models of CT scanners (PQ 2000S, Picker International, Cleveland, Ohio; or Somaton AR, Siemens, Erlangen, Germany). All studies were carried out with 10-mm-thick slices obtained at 10-mm intervals from the diaphragm through the pubic symphysis. None of the patients received oral or IV contrast material.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Surgery was performed in all 17 patients. A gastrotomy with extraction of the foreign body was performed in two patients with gastric bezoars. In 12 patients with intestinal bezoars, the bezoar was manually fragmented, and the remants then were pushed toward the large bowel. In these 12 patients, the stomach was surgically explored, with the extraction of an additional bezoar via gastrotomy in five patients. An enterotomy with bezoar extraction proved necessary in the other three patients with intestinal bezoars because fragmenting the foreign body was not possible. None of the patients had bezoar-induced recurrent obstruction observed at follow-up.
Twelve (75%) of the 16 patients with phytobezoars, all of whom were men, had undergone gastric surgery for peptic disease between 10 and 20 years before admission for treatment of bezoars. Five patients had been subjected to truncal vagotomy and pyloroplasty, whereas seven had undergone Bilroth II partial gastroenterostomy. Another two patients had a definite history of excessive persimmon consumption shortly before the onset of symptoms. A history of hair-eating was established in the girl with a trichobezoar after the radiologic diagnosis had been made. The remaining two patients had no history of gastric surgery or unusual dietary habits.
At surgery, a total of 33 bezoars were identified in the 17 patients—22 small-bowel lesions and 11 gastric lesions. Two individuals presented with isolated stomach bezoars with gastric dilatation but no signs of small-bowel obstruction. The remaining 15 patients had intestinal bezoars. Bezoar-induced small-bowel obstruction was confirmed in 13 patients. In the other two patients, the bezoar was identified in the colon (transverse and sigmoid, respectively). Eight patients (53%) had a gastric bezoar associated with an intestinal bezoar; one patient had two gastric bezoars. Five patients (33%) had multiple intestinal bezoars: three patients had two bezoars, and two patients had three.
The abdominal radiographs showed intestinal obstruction signs in 12 patients. The typical bezoar image, involving a mottled air pattern, was visible in only three patients (18%), two with gastric bezoars and one with a jejunal bezoar.
Sonography, in turn, showed small-bowel obstruction in 14 of the 15 intestinal bezoars. The characteristic image suggesting a bezoar was observed in 15 (88%) of the 17 patients. Sonography revealed only 19 (58%) of the 33 bezoars that were surgically confirmed—specifically, the two isolated gastric bezoars, 13 of the 15 intestinal bezoars, two (25%) of the eight gastric bezoars associated with intestinal bezoars, and two of the five cases involving multiple intestinal bezoars.
Finally, CT revealed small-bowel obstruction with the image of an intestinal bezoar close to the transition point in 15 patients and gastric dilatation with an isolated gastric bezoar in the other two patients (100%) (Fig. 4). Seven of the 15 patients with intestinal bezoars were seen to have associated gastric bezoars; one confirmed at surgery had not been seen on CT scans. Five patients presented multiple intestinal bezoars, coinciding with the number identified at surgery. The percentage of associated gastric bezoars detected was 87%.
|
A total of 32 bezoars (97%) were identified on CT scans. The locations of the intestinal bezoars were as follows: one in the duodenum, six in the jejunum, and 15 in the ileum. Two bezoars located in the middle ileum according to CT findings were found at surgery in the transverse colon and sigmoid. One bezoar located in the jejunum—ileum transition zone was found close to the ileocecal valve, whereas another gastric bezoar was identified at surgery in the duodenum. In the rest of bezoars, the location as defined by CT findings coincided with that described in the surgical report.
Because of the use of axial slices, the approximate length of the bezoars was difficult to determine using CT. Lengths of the intestinal bezoars ranged from 15 to 70 mm (mean, 35 mm). Sixteen of the 22 intraluminal bowel masses were smaller than 40 mm. The gastric bezoars were larger (range, 20-130 mm; mean, 69 mm). Nine of the 11 gastric bezoars measured longer than 40 mm.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patients with phytobezoars often have a history of prior surgery [1]. In our series, 75% of patients with phytobezoars had undergone surgery previously. Most small-bowel obstruction occurring in patients with antecedents of surgery are attributable to adhesions and are treated conservatively, with physicians resorting to surgery only if such treatment fails. However, bezoar-induced obstruction rarely improves with conservative treatment, and early surgery may be required to secure a definitive solution [9, 11]. Therefore, the diagnosis of a bezoar as the cause of obstruction is important because it modifies the approach to treatment, accelerating the use of surgery to correct the condition.
As many as 9% of the patients undergoing surgery for an intestinal bezoar may require a second operation because of recurrent bowel obstruction, presumably caused by the presence of residual bezoars [1, 9]. In our series, a concurrent gastric bezoar was found in 53% of patients with a small-bowel bezoar. Consequently, when an intestinal bezoar is diagnosed, the possible presence of concomitant gastric or intestinal bezoars should be investigated carefully.
Until only a few years ago, bezoars were rarely diagnosed before surgery because the clinical symptoms and radiographic findings are similar to those of intestinal obstruction attributable to other causes [1, 9]. However, findings from recent studies suggest that sonography or CT can help radiologists diagnose bezoars before surgery [2,3,4,5,6,7].
Conventional abdominal radiographs may be enough for a radiologist to detect the signs of bowel obstruction—conventional radiographs were used to identify the bowel obstructions of 12 (80%) of the 15 patients in our series who had such an obstruction. However, radiologists are rarely able to identify a bezoar as the cause of bowel obstruction from radiographs alone. The bezoars could be identified in only 18% (3/17) of the patients from radiographs alone, a figure similar to the 10% reported by Verstanding et al. [1]. The radiographic image of a bezoar is not uniquely characteristic and can be mistaken for an abscess or the presence of feces in the colon.
In the two hospitals in which our study took place, sonography is routinely used to examine both patients with nonspecific acute abdominal pain and those with signs of small-bowel obstruction on radiographs. Many authors have reported that sonography affords good sensitivity and specificity in diagnosing intestinal obstruction [12]. In our series, sonography revealed signs of bowel obstruction in 93% of the patients with intestinal bezoars. It has been reported that sonographic visualization of an intraluminal mass with a hyperechoic arclike surface and a marked acoustic shadow is suggestive of a bezoar [2, 3]. On the basis of this characteristic image, we suspected the presence of a bezoar before surgery in 15 (88%) of the 17 patients of our series. Sonography was unable to reveal a bezoar as the cause of obstruction in two patients because the bezoar was located far from the abdominal surface. Sonography also had difficulties in revealing multiple bezoars—only two of the five cases of multiple bezoars in our series—perhaps because of the impossibility of exploring the entire trajectory of the dilated bowel loops.
Bezoar-induced obstruction should be distinguished from gallstone ileus because bezoars produce the same sonographic image as ectopic lithiasis [13]. Fecal material in the colon can also simulate the image of a bezoar, particularly in the presence of barium concretions. In most patients, sonography can differentiate the small-bowel loops from the colon because of their distinct morphology and location.
Gastric bezoars must be differentiated from the images produced by retained food particles in the stomach. The marked acoustic shadowing behind the echogenic band produced by a bezoar is different from "dirty" shadowing generated by ingested gas and food within the stomach [8]. However, in our experience, sonography is of limited sensitivity in revealing gastric bezoars. Although it afforded images for the correct diagnoses of the two cases of isolated gastric bezoar, it only revealed two (25%) of the eight gastric bezoars associated with intestinal bezoars.
Many studies have described the diagnostic accuracy gained through the use of CT in evaluating small-bowel obstruction [14, 15]. This modality allows the physician to determine the cause, level, and degree of intestinal obstruction. The liquid and gas filling the gut lumen in the presence of high-degree obstruction makes oral contrast medium unnecessary. Moreover, according to some authors, the appearance of obstructing small-bowel bezoars on CT scans is sufficient for diagnosis [4,5,6,7]. On the basis of previously published characteristic CT images, we gave all our patients a preoperative diagnosis of bezoar. CT revealed a well-defined oval intraluminal mass with air bubbles retained within the interstices, a dilated small bowel proximal to the mass, and normal or collapsed distal loops. In our study, the CT results were almost fully coincident with the surgical findings; in fact, CT revealed 32 of the 33 surgically verified bezoars. CT also revealed the five cases of multiple intestinal bezoars and seven of the eight concomitant gastric masses.
Some bezoars—one in the stomach and one in the distal jejunum—were not found at surgery in the location previously indicated at CT. A discrepancy between the CT and surgical localizations may be caused by migration of the bezoar during the interval between the imaging examination and surgery. In two patients, the bezoars located on CT scans in the small bowel were subsequently found at surgery in the large intestine, thus proving that spontaneous expulsion of the bezoar is possible in some patients [16].
According to our experience, a radiologist can use CT to distinguish between retained food particles in the stomach and small gastric bezoars visualized on an image. Small bezoars are rounded or ovoid, tend to float on the water—air surface surrounded by the gastric contents, and show lower density than food particles (Fig. 5). Large bezoars tend to fill all the lumen, exhibiting air bubbles diffusely distributed throughout the mass. However, occasionally it can be difficult to differentiate a bezoar from a large amount of retained food.
|
We have observed that bezoars, especially those located in the stomach, may go undetected if the CT scans are printed at standard abdominal soft-tissue window setting (level, 40 H; width, 350 H). Modifying the window setting by reducing the level to—100 H makes it possible to better identify the characteristic appearance of bezoars (Fig. 6A,6B). Although it is possible to diagnose bezoars using routine soft-tissue window settings, in our opinion, special window settings increase the radiologist's diagnostic confidence and the number of additional lesions detected. All but two of the CT studies in our series were performed to assess a sonographically suggested gastrointestinal bezoar; consequently, in almost all patients, the images were obtained with the modified window setting. This fact may explain the large number of bezoars (97%) revealed on CT before surgery in our study.
|
|
Lastly, CT image of a bezoar must be differentiated from the image of small-bowel feces reported in patients with high-degree intestinal obstruction or cystic fibrosis [17]. Such images correspond to gas bubbles in particulate matter, yielding an elongated form of greater length and a less compact nature than bezoars. The small-bowel feces are not located in the obstruction transition zone, a fact that facilitates differentiation from bezoars, particularly in bezoars of considerable length. (Two of the bezoars in our study measured 70 mm).
The results of our study show CT to be more useful to the radiologist in diagnosing bezoars than either conventional abdominal radiographs or sonography. Although sonography revealed bezoars in 88% of our patients, the corresponding images were less characteristic than those obtained with CT. With sonographic images, identification of additional intestinal or gastric bezoars was difficult. In our opinion, CT may be considered the imaging technique of choice for confirming a diagnosis of gastrointestinal bezoars. Using CT enables radiologists to determine the point of obstruction, reveal the bezoar as the underlying cause of obstruction, and detect the existence of additional intestinal or gastric bezoars.
Because the present study consisted of a retrospective analysis in which all patients had surgically confirmed bezoars, no statistical evaluation is possible. Further prospective surveys involving large groups of patients with small-bowel obstruction are required to define more clearly the sensitivity and specificity of sonography and CT.
In conclusion, the existence of a bezoar as the underlying cause of intestinal obstruction should be considered in patients who have had previous gastric surgery. In contrast to clinical examination, sonography and CT can help physicians rapidly diagnose bezoar-induced intestinal obstruction. However, sonography can produce an image in which a bezoar can be confused with other conditions; the CT image of a gastrointestinal bezoar—an intraluminal mass containing a mottled air pattern—is quite characteristic, and CT is more effective in revealing concomitant gastric and intestinal bezoars.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A N HEWITT, M S LEVINE, S E RUBESIN, and I LAUFER Gastric bezoars: reassessment of clinical and radiographic findings in 19 patients Br. J. Radiol., November 1, 2009; 82(983): 901 - 907. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S-J Lee, S-J Chu, and S-H Tsai Colonic phytobezoar BMJ Case Reports, January 8, 2009; 2009(jan08_1): bcr2006039412 - bcr2006039412. [Full Text]
|
|
|
|
 |
|
 S-J Lee, S-J Chu, and S-H Tsai Colonic phytobezoar Emerg. Med. J., July 1, 2007; 24(7): 518 - 518. [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Frey, M. McKee, R. A. King, and A. Martin Hair Apparent: Rapunzel Syndrome Am J Psychiatry, February 1, 2005; 162(2): 242 - 248. [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Lee, S. E. Jung, and K.-Y. Lee Small-Bowel Obstruction Caused by Phytobezoar: MR Imaging Findings Am. J. Roentgenol., August 1, 2002; 179(2): 538 - 539. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
