|
|
||||||||
Original Report |
1
Department of Radiology, Hospital of the University of Pennsylvania, 3400
Spruce St., Philadelphia, PA 19104.
2
Department of Medicine, Hospital of the University of Pennsylvania,
Philadelphia, PA 19104.
Received April 7, 2000;
accepted after revision May 30, 2000.
Address correspondence to M. S. Levine.
Abstract
|
|
|---|
CONCLUSION. Our experience suggests that portal hypertensive gastropathy may manifest on barium studies as thickened nodular folds in the gastric fundus. Although varices or various forms of gastritis can also produce thickened gastric folds, portal hypertensive gastropathy should be suspected when this finding is detected in patients with known portal hypertension.
|
|
|---|
Although the endoscopic findings of portal hypertensive gastropathy have been well documented [4], little is known about the radiographic findings of this condition. To our knowledge, the medical literature contains only one anecdotal case report of portal hypertensive gastropathy in which a double-contrast upper gastrointestinal examination revealed multiple small polypoid protrusions in the body of the stomach [5]. The purpose of this investigation was to determine the findings of portal hypertensive gastropathy on barium studies and whether there are useful radiographic criteria for diagnosing this condition.
|
|
|---|
Four of the eight patients underwent double-contrast examinations and the other four had single-contrast examinations. Excluding one patient in whom the barium study was performed 6.5 years before endoscopy, the average interval between the barium study and endoscopy was 5.4 months (range, 0-16 months). Four of the eight patients in our study group had undergone endoscopy before undergoing barium studies; hence, the diagnosis of portal hypertensive gastropathy had already been established at the time of the radiographic examinations. Although the other four patients underwent endoscopy after having undergone barium studies, all of these patients had signs or symptoms of portal hypertension (e.g., jaundice, ascites, hepatosplenomegaly, hematemesis) or known chronic liver disease at the time of the radiographic examinations. Thus, portal hypertensive gastropathy was likely present when the barium studies were performed.
The radiographs from these eight upper gastrointestinal examinations were reviewed retrospectively for the presence or absence of varices, abnormally thickened folds (width of folds, >5 mm), ulcers, nodules, erosions, enlarged areae gastricae, or other abnormalities in the stomach. When present, the size, location, and morphologic features of these abnormalities were noted.
Endoscopy reports and photographs were also reviewed to determine the degree of portal hypertensive gastropathy in these eight patients. Portal hypertensive gastropathy was classified as mild or moderate if endoscopy revealed various degrees of a mosaic or "snakeskin" pattern associated with mucosal hyperemia and vascular ectasia and as severe if endoscopy revealed superimposed cherry red spots representing extravasated RBCs [1, 4]. The distribution of portal hypertensive gastropathy within the stomach was not mentioned in the endoscopy reports. These reports were also reviewed for the presence or absence of esophageal or gastric varices, gastritis, or other gastric abnormalities. Endoscopic biopsy specimens were not obtained in any of these eight patients because of concern about the risk of bleeding in patients with underlying coagulopathies associated with chronic liver disease.
The findings from barium studies were then correlated with those from endoscopy. Finally, medical records were reviewed to determine the clinical presentation and treatment.
|
|
|---|
Five of the eight patients with portal hypertensive gastropathy underwent treatment: one received oral propranolol, one underwent an esophagogastric variceal devascularization procedure, one underwent surgical placement of a portacaval shunt and subsequent placement of a transjugular intrahepatic portosystemic shunt, and two received orthotopic liver transplants. The patient who had the variceal devascularization procedure underwent follow-up endoscopy, which revealed improvement of his portal hypertensive gastropathy 2 months after surgery. The other four patients did not undergo follow-up endoscopic examinations after treatment.
Radiographic Findings
Five (63%) of the eight patients with portal hypertensive gastropathy had
thickened gastric folds, which had a mean thickness of 10 mm (range, 8-12 mm).
The enlarged folds involved only the fundus in four patients and the fundus
and body in one. In all five patients, the thickened folds had a nodular
appearance with undulating contours and indistinct borders (Figs.
1,2,3).
The radiographic findings therefore were somewhat different from those of
gastric varices, which classically appear as multiple rounded submucosal
nodules or as serpentine folds in the gastric fundus
[6]
(Fig. 4). When the
double-contrast and single-contrast studies were considered separately,
thickened folds were detected in three (75%) of four patients on
double-contrast examinations and in two (50%) of four on single-contrast
examinations.
|
|
|
|
Thickened nodular folds were seen at endoscopy in the gastric fundus in one of two patients with mild portal hypertensive gastropathy, in two of four with moderate portal hypertensive gastropathy, and in both patients with severe portal hypertensive gastropathy. One patient with portal hypertensive gastropathy also had a benign-appearing antral ulcer, and one had two small antral polyps. However, none of the patients with portal hypertensive gastropathy had gastric erosions or enlarged areae gastricae. Finally, five (71%) of seven patients with esophageal varices at endoscopy had radiographic evidence of esophageal varices.
Endoscopic Findings
Portal hypertensive gastropathy was classified at endoscopy as mild in two
patients, moderate in four, and severe in two. Two patients also had
endoscopic findings of gastric antral vascular ectasia, a distinct endoscopic
and histopathologic entity seen with increased frequency in patients with
portal hypertensive gastropathy
[7]. Seven (88%) of the eight
patients had associated esophageal varices but only two (25%) had gastric
varices at endoscopy; neither of these two patients had positive findings on
barium studies for portal hypertensive gastropathy with thickened fundal
folds. Finally, none of the patients had any endoscopic evidence of gastritis,
but endoscopic biopsy specimens were not obtained.
|
|
|---|
The development of portal hypertensive gastropathy has important clinical implications; it is the second most common cause of upper gastrointestinal bleeding (after esophageal varices) in patients with portal hypertension [10]. Once the initial episode of bleeding from portal hypertensive gastropathy has occurred, the risk of recurrent bleeding during the next 2 years is as high as 75% [11]. Although some patients may have acute, even life-threatening, hematemesis, others may have chronic or intermittent low-grade bleeding that results in severe iron-deficiency anemia [4]. In our study, seven of eight patients with portal hypertensive gastropathy had a history of upper gastrointestinal bleeding, including hematemesis, melena, or both in five patients and signs of low-grade bleeding in two.
Because of the high risk of bleeding in portal hypertensive gastropathy even in the absence of varices, a variety of therapeutic options have been recommended for these patients [4, 12]. The risk of bleeding can be decreased markedly by oral administration of propranolol, a ß-adrenergic blocker. In patients with active bleeding, endoscopic hemostasis can be achieved with photo- or electrocoagulation if the bleeding is acute and relatively localized. If the bleeding is more diffuse, IV infusion of somatostatin or Glypressin (terlipressin; Ferring Pharmaceuticals, Tarrytown, NY) can be used to control bleeding by inducing vasoconstriction and decreased gastric blood flow. In patients with severe portal hypertensive gastropathy, placement of surgical shunts or transjugular intrahepatic portosystemic shunts or even liver transplantation may be performed.
The diagnosis of portal hypertensive gastropathy is usually based on the appearance of the stomach at endoscopy. In all cases, the gastric mucosa has a mosaic or snakeskin appearance with small, slightly protruding polygonal lesions that have clearly demarcated whitish yellow borders [1]. In patients with more severe disease, slightly raised cherry red spots are superimposed on this mosaic background as a result of tiny collections of extravasated blood in the mucosa [4]. The appearance at endoscopy is so characteristic that portal hypertensive gastropathy can be differentiated from gastric varices or gastritis on the basis of endoscopic findings.
In one anecdotal case report, portal hypertensive gastropathy was manifested on a barium study as small nodular filling defects in the gastric body [5]. In our series, however, five (63%) of eight patients with endoscopically diagnosed portal hypertensive gastropathy had thickened nodular gastric folds with undulating contours and indistinct borders that produced similar radiographic findings (Figs. 1,2,3). Although the pathophysiologic basis for this fold thickening is uncertain, it could result from a combination of mucosal hyperemia and dilated submucosal vessels in these patients [1, 10]. Furthermore, in all five patients, the thickened folds were confined to the gastric fundus (n = 4) or the fundus and body (n = 1). Other endoscopic studies have also shown that portal hypertensive gastropathy has a predilection for the gastric fundus, presumably because of the microcirculation and venous drainage pattern of the stomach [1, 13].
Various forms of gastritis, particularly Helicobacter pylori gastritis, may also be manifested as thickened gastric folds, but the thickened folds are not confined to the fundus and are often associated with other abnormalities such as ulcers, erosions, nodules, or enlarged areae gastricae [14]. Lymphoma and Ménétrier's disease are other causes of thickened gastric folds, but the folds are usually much thicker and more lobulated in these patients. Gastric varices may appear as thickened folds on upper gastrointestinal examinations, but the folds tend to have a more serpentine configuration and are often associated with discrete submucosal nodules, which have been likened to the appearance of a bunch of grapes [6] (Fig. 4). It should also be noted that none of our five patients with portal hypertensive gastropathy and thickened folds on barium studies had gross endoscopic findings of varices or gastritis, but endoscopic biopsy specimens were not obtained. The presence of thickened nodular folds in the gastric fundus therefore should suggest the possibility of portal hypertensive gastropathy in patients with portal hypertension.
It is important to be aware of the limitations of our retrospective study, including observer bias and a small sample size, which precluded statistical analysis of the data. Further investigation is therefore needed to determine the frequency of thickened nodular folds in the gastric fundus in a larger population of patients with portal hypertensive gastropathy and also to determine the positive and negative predictive values of this finding for portal hypertensive gastropathy in patients with portal hypertension.
In conclusion, it is important to be aware of portal hypertensive gastropathy because of the increased risk of upper gastrointestinal bleeding in these patients even in the absence of esophageal or gastric varices. Our experience suggests that portal hypertensive gastropathy may manifest on barium studies as thickened nodular folds in the gastric fundus. Although varices or various forms of gastritis may also produce thickened gastric folds, portal hypertensive gastropathy should be suspected when this finding is detected in patients with known portal hypertension.
|
|
|---|
This article has been cited by other articles:
![]() |
S. E. Rubesin, M. S. Levine, and I. Laufer Double-Contrast Upper Gastrointestinal Radiography: A Pattern Approach for Diseases of the Stomach Radiology, January 1, 2008; 246(1): 33 - 48. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |