| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Departments of Radiology (S.L.Z., M.S.L., S.E.R., I.L.), Medicine (M.C.M., D.A.K.), and Pathology (E.E.F.), Hospital of the University of Pennsylvania, 3400 Spruce St, Philadelphia, PA 19104. Received June 22, 2004; revision requested August 19; revision received September 7; accepted October 4. Address correspondence to M.S.L. (e-mail: marc.levine{at}uphs.upenn.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: The 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. The study was compliant with the Health Insurance Portability and Accountability Act. A review of the pathology and radiology databases at the authors' institution revealed 14 patients (11 men and three women; mean age, 41.3 years) with IEE (defined as more than 20 eosinophils per high-power field in biopsy specimens) who had undergone barium studies and endoscopy. The radiographs were reviewed for strictures, esophagitis, or other abnormalities. The endoscopic reports, clinical records, and laboratory data were also reviewed and compared with the radiographic findings.
RESULTS: Seven of the 14 patients (50%) had a history of allergies, and two of nine patients with complete blood cell counts (22%) had peripheral eosinophilia. Thirteen patients (93%) had dysphagia, six (43%) had food impactions, and six (43%) had reflux symptoms. Ten patients (71%) had a total of 11 strictures at barium studies (two in the upper part of the esophagus, two in the middle part, three in the distal part, one in the middle and distal parts, and three at the gastroesophageal junction). The strictures had a mean length of 5.1 cm. In seven patients (50%), the strictures contained multiple fixed ringlike indentations that produced a ringed esophagus. The ringlike indentations appeared as multiple, fixed, closely spaced, concentric rings traversing the stricture. Four patients (28%) had esophagitis. Of 13 patients who underwent recumbent imaging, 10 (77%) had hiatal hernias and nine (69%) had reflux. Eight of the 10 patients (80%) with strictures underwent endoscopic dilation procedures, which resulted in only temporary relief of dysphagia.
CONCLUSION: The findings suggest that most patients with IEE have esophageal strictures, often with distinctive ringlike indentations that produce a ringed esophagus.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In previous studies, IEE most commonly manifested at esophagography as segmental esophageal strictures (13–17); however, patients with these strictures occasionally may have diffuse esophageal narrowing, which produces a "small-caliber" esophagus (4,6,7,18,19). We recently encountered several patients with IEE in whom barium studies revealed esophageal strictures with distinctive ringlike indentations. Some authors in the gastroenterology literature have reported similar findings in IEE, with thin, fixed ringlike indentations encircling the esophagus that produce a so-called "ringed" esophagus at endoscopy, barium studies, or both (4–6,20,21). To our knowledge, however, such an appearance has not been described as a sign of IEE in the radiology literature. The purpose of our investigation, therefore, was to retrospectively assess the findings of IEE on barium study images and determine the frequency of the ringed esophagus in patients with this condition.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Patient Population
A review of the computerized pathology database at our hospital by one author (S.L.Z.) revealed 141 patients with increased intraepithelial eosinophils in endoscopic biopsy specimens obtained from the esophagus from January 1998 to September 2003. A subsequent review of our computerized radiology database by the same author revealed that 35 of the 141 patients (25%) underwent barium studies of the esophagus during this period. The tissue specimens from these 35 patients were reviewed by another author (E.E.F., a gastrointestinal pathologist with 14 years of experience) to determine the degree of eosinophilic infiltration of the esophageal wall, and 14 patients (40%) were found to have more than 20 eosinophils per high-power field on the histologic sections with the densest eosinophilic infiltrates, which is a widely accepted pathologic criterion for the diagnosis of eosinophilic esophagitis (1,4,5,7).
The indications for the barium studies (including dysphagia, food impaction, and reflux symptoms) were determined from medical records. The mean interval between the barium study and endoscopy was 145 days (range, 1–501 days). Five of the 14 patients (36%) were treated with proton pump inhibitors during this interval, but none of the 14 patients received oral steroids or other antiallergy therapy. Endoscopy was performed within 1 month of the barium studies in seven patients and within 1–17 months in seven patients. The barium studies were performed before endoscopy in 13 patients and after endoscopy in one patient. (This patient did not undergo an endoscopic dilation procedure in the interim.) Our study group comprised these 14 patients. The mean age of the 14 patients was 41.3 years (range, 23–89 years). Eleven patients (79%) were men, and three (21%) were women.
Examination Technique
Twelve of the 14 patients (86%) underwent biphasic esophagography that included upright, left posterior oblique (LPO) and, less commonly, right posterior oblique double-contrast views with an effervescent agent (Baros; Lafayette Pharmaceuticals, Lafayette, Ind) and a 250% wt/vol barium suspension (E-Z-HD; E-Z-EM, Westbury, NY) and prone, right anterior oblique (RAO) single-contrast views with a 50% wt/vol barium suspension (Entrobar; Lafayette Pharmaceuticals). The remaining two patients underwent single-contrast esophagography that included upright LPO and/or prone RAO single-contrast views with a 50% wt/vol barium suspension (Entrobar). These patients underwent a single-contrast study because of suspected esophageal food impaction (n = 1) and debility (n = 1). The quality of the barium study images was classified in consensus by two authors (M.S.L. and S.E.R., gastrointestinal radiologists with 22 and 20 years of experience, respectively) as excellent (optimal distention and mucosal coating) or good (near-optimal distention or mucosal coating) in 13 patients and as fair (suboptimal distention or mucosal coating) in the patient with esophageal food impaction. The latter patient's images nevertheless were believed to be of diagnostic quality. All barium studies were either performed by residents or fellows under the supervision of one of three attending gastrointestinal radiologists (M.S.L., S.E.R., and I.L.) or directly performed by one of the attending gastrointestinal radiologists. All images were interpreted by the attending radiologists.
Three patients underwent follow-up barium studies (two underwent one follow-up study, and one underwent two follow-up studies) a mean interval of 290 days (range, 5–739 days) after the initial examination. In all three patients, the follow-up studies were performed because of persistent or recurrent dysphagia. None of these patients were treated with steroids or other antiallergy medications between the initial examinations and the follow-up studies.
Image Analysis
The radiographs from these 14 barium studies were reviewed in consensus by two authors (M.S.L. and S.E.R.), who were blinded to the clinical and endoscopic findings (but not to the final pathologic diagnosis of IEE), to determine whether strictures, esophagitis (criteria for esophagitis included a granular or nodular mucosa, thickened folds, and ulceration), hiatal hernias, or other abnormalities were present in the esophagus. When strictures were identified, the reviewers assessed the morphologic features of the strictures, including location (ie, cervical esophagus, upper part of the thoracic esophagus [from the thoracic inlet to the aortic arch], middle part of the thoracic esophagus [from the aortic arch to the carina], distal part of the thoracic esophagus [from the carina to the gastro-esophageal junction], or gastroesophageal junction); symmetry (symmetric or concentric vs asymmetric or eccentric); margins (tapered vs abrupt); contours (smooth vs irregular or ulcerated); width (ie, smallest luminal diameter) and length of the narrowed segment; and the presence or absence of ringlike indentations (a ringed esophagus).
When ringlike indentations were present, the images were reviewed to determine whether the rings were better visualized in the collapsed or partially collapsed esophagus or in the distended esophagus or equally well seen regardless of the degree of esophageal distention. The images were also reviewed to determine whether the rings were better seen with double- or single-contrast views or equally well seen with both views.
The original radiology reports were reviewed for the presence or absence of spontaneous gastroesophageal reflux or abnormal esophageal motility. (The presence or absence of reflux was not mentioned in two reports, so we assumed that reflux was not present. Similarly, esophageal motility was not mentioned in four reports, so we assumed motility was normal.) The patient with an esophageal food impaction was not placed in a recumbent position, so esophageal motility and the presence or absence of a hiatal hernia or spontaneous gastroesophageal reflux could not be evaluated.
Comparison of Clinical and Endoscopic Findings
Clinical records, laboratory data, endoscopic reports, and pathology reports were reviewed by one author (S.L.Z.) without knowledge of the radiographic findings (this author did not review the barium study images) to determine the clinical findings (including history of allergies, presence or absence of peripheral eosinophilia, signs and symptoms at presentation, treatment, and subsequent course) and the findings at endoscopy, pathologic examination, and 24-hour esophageal pH monitoring studies or manometry. The radiographic findings then were compared with the clinical and endoscopic findings.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Thirteen of the 14 patients (93%) had a history of dysphagia (mean duration, 7.3 years; range, 1–20 years). Six patients (43%) had a history of at least one food impaction, and six (43%) had reflux symptoms, including heartburn and regurgitation. Four patients (28%) underwent esophageal manometry, which revealed normal peristalsis in two patients, abnormal peristalsis in one patient (this patient was 35 years old, so the abnormal motility could not be attributed to age), and a hypertensive lower esophageal sphincter in one patient. Two patients underwent 24-hour ambulatory esophageal pH monitoring, and results were normal in both.
Radiographic Findings
Ten of the 14 patients (71%) had a total of 11 strictures at barium studies, including two in the upper part of the thoracic esophagus (Fig 1), two in the middle part of the thoracic esophagus (Fig 2), four in the distal part of the thoracic esophagus (Figs 3, 4), and three at the gastroesophageal junction. (One patient had two strictures—one in the distal part of the esophagus and one at the gastroesophageal junction.) The mean length of the strictures was 5.1 cm (range, 0.5–13.0 cm), and the mean smallest luminal diameter was 1 cm (range, 0.2–1.3 cm). The mean length of the strictures at the gastroesophageal junction was considerably shorter (1.4 cm) than that of strictures found elsewhere in the esophagus (6.5 cm). The strictures had tapered margins in all 10 patients (100%), a smooth contour in nine patients (90%) versus an irregular contour in one patient (10%), and a symmetric configuration in seven patients (70%) versus an asymmetric configuration in three patients (30%). The strictures were responsible for dysphagia in 10 of the 13 patients (77%) with this symptom.
|
|
|
|
|
|
|
|
Four of the 14 patients (28%) had esophagitis with a granular mucosa in the middle and distal parts of the esophagus as a result of mucosal edema and inflammation (Fig 5). The remaining 10 patients (71%) had no evidence of esophagitis. Of the 13 patients in whom recumbent imaging was performed, 10 (77%) had hiatal hernias and nine (69%) had spontaneous gastroesophageal reflux at fluoroscopy. Eleven of the 13 patients who underwent recumbent imaging (85%) had normal esophageal motility, one (8%) had weakened primary peristalsis, and one (8%) had incomplete opening of the lower esophageal sphincter. (The latter two patients had similar findings at manometry.)
|
Pathologic Findings
The 14 patients with IEE had a mean count of 49 eosinophils per high-power field (range, 21–86 eosinophils per high-power field) in endoscopic biopsy specimens from the esophagus. Biopsy specimens revealed a mean count of 45 eosinophils per high-power field (range, 28–82 eosinophils per high-power field) in the 10 patients with esophageal strictures at barium studies compared with a mean count of 61 eosinophils per high-power field (range, 21–86 eosinophils per high-power field) in the four patients without strictures. None of the biopsy specimens revealed pathologic findings of Barrett mucosa or tumor in the esophagus. Seven of the 14 patients (50%) also had endoscopic biopsy specimens from the gastric antrum, duodenum, or both, and none of these specimens revealed increased numbers of eosinophils suggestive of eosinophilic gastroenteritis.
Treatment and Follow-Up
After a diagnosis of IEE had been confirmed in endoscopic biopsy specimens, 10 of the 14 patients (71%) received proton pump inhibitors, with symptomatic improvement in one patient, and six patients (43%) were treated with oral steroid inhalation (fluticasone propionate [Flovent; GlaxoSmithKline, Research Triangle Park, NC]), with varying degrees of symptomatic improvement in all six patients (including five with a ringed esophagus). Five of the 14 patients (36%) were also treated with proton pump inhibitors (but not with other antiallergy therapy) before endoscopy, with symptomatic improvement in none. Eight of the 10 patients (80%) with strictures at barium studies underwent endoscopic dilation procedures, and all eight had only temporary relief of dysphagia. Four patients with recurrent dysphagia underwent repeated dilations (mean number of procedures, 2.5; range, 2–4) during a mean period of 4.5 years (range, 1–10 years).
In three patients, esophageal strictures were identified at follow-up barium studies without change from the initial examination. However, one of these patients with a stricture in the middle part of the thoracic esophagus developed a second stricture at the gastroesophageal junction, as seen at a follow-up barium study performed 2 years after the initial examination. Another patient underwent an endoscopic dilation procedure between the initial study and the follow-up examination, and all three received proton pump inhibitors in the interim between these procedures.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In the original clinical description of IEE by Attwood et al (1), only one of 12 patients (8%) with this condition was found to have gastroesophageal reflux at barium studies or 24-hour ambulatory esophageal pH monitoring. Lee (22) and Kelly et al (8), however, reported gastroesophageal reflux in 10 of 11 patients (91%) and seven of 10 patients (70%), respectively, with pathologic findings of IEE in endoscopic biopsy specimens. Similarly, in our study, nine of 13 patients (69%) with pathologic findings of IEE had gastroesophageal reflux and 10 (77%) had hiatal hernias at esophagography that included recumbent imaging. Six of the 14 patients (43%) with IEE also had reflux symptoms. Our findings and those of earlier studies indicate that reflux disease, by virtue of its high prevalence, may be present in a substantial number of patients with IEE. IEE, however, can usually be differentiated from reflux esophagitis on the basis of the degree of eosinophilic infiltration of the esophagus in endoscopic biopsy specimens; most patients with IEE have more than 20 eosinophils per high-power field (1,4,5,7), whereas most patients with reflux esophagitis have fewer than five to 10 eosinophils per high-power field (1,2,5,23). Endoscopic biopsy specimens, therefore, are necessary for differentiating these conditions.
In our series and others, most patients with IEE have been found to have esophageal strictures at barium studies (13–17). In the study by Vitellas et al (16), the largest reported series in the radiology literature, seven of 10 strictures (70%) were located in the upper or middle part of the thoracic esophagus (none of these strictures contained ringlike indentations). In our study, however, seven of 11 strictures (64%) were located in the distal part of the esophagus or at the gastroesophageal junction. Although the reason for this discrepancy is unclear, these strictures apparently may have a variable location in the esophagus. In our study, the mean length of strictures at the gastroesophageal junction (1.4 cm) was considerably shorter than that of strictures elsewhere in the esophagus (6.5 cm). It is possible that these shorter strictures were caused by coexisting reflux disease, whereas longer strictures above the gastroesophageal junction were caused by IEE. In fact, two patients had concomitant strictures at and above the gastroesophageal junction. Our findings suggest that some patients with simultaneous reflux disease and IEE may develop separate strictures from independent pathophysiologic mechanisms.
In our study, patients with IEE who had esophageal strictures were found to have a lower mean number of eosinophils per high-power field than did those with IEE who did not have strictures. Although the reason for this discrepancy is unclear, it is possible that eosinophilic infiltration of the esophageal wall is more intense during the acute phase of IEE and that eosinophilic infiltration subsides in patients with more chronic disease who have developed esophageal strictures.
In seven (50%) of our patients with IEE, the strictures contained distinctive ringlike indentations on barium study images, which resulted in a so-called "ringed" esophagus. These ringlike indentations were characterized by multiple, closely spaced, concentric rings that traversed the stricture. To our knowledge, such rings have not been described previously in the radiology literature in patients with IEE. In retrospect, however, concentric rings appear to have been present in one patient with a ringlike stricture reported by Picus and Frank (13) and in another with a corrugated stricture reported by Feczko et al (15). In our series, the rings were more easily recognized at barium studies than at endoscopy (only four of seven patients with rings at barium studies had rings at endoscopy), so the barium study may be more sensitive than endoscopy in the detection of this finding. Nevertheless, a ringed esophagus has been well documented at endoscopy in patients with IEE (4–6,20,21). Although the pathogenesis of the rings is uncertain, we believe that this finding is strongly suggestive of IEE in patients with esophageal strictures at barium studies or endoscopy.
A ringed esophagus has also been described in patients with congenital esophageal stenosis in whom barium studies or endoscopy revealed corrugated esophageal strictures that contained multiple concentric rings (24,25). These rings have been attributed to aberrant embryologic development with tracheobronchial remnants or actual cartilaginous rings in the esophageal wall (24,25). Like IEE, congenital esophageal stenosis typically occurs in young men with long-standing dysphagia and recurrent food impactions (24–26). Although this condition is not usually associated with an allergic history or peripheral eosinophilia, some patients with congenital esophageal stenosis may have increased numbers of eosinophils in endoscopic biopsy specimens from the esophagus (24). In view of the similarities in the clinical, radiographic, and pathologic findings of these conditions, we suspect that some of the patients reported to have congenital esophageal stenosis had unrecognized IEE as the cause of their disease. Other authors have postulated that the ringed esophagus represents an atypical manifestation of reflux disease (20,27,28); however, intraepithelial eosinophil counts are lacking in these studies. It is therefore difficult to corroborate such a hypothesis on the basis of existing data from the gastroenterology literature.
The differential diagnosis of the ringed esophagus includes fixed transverse folds in patients with peptic strictures (29). Although the concentric rings in IEE are also fixed, they are closely spaced and traverse the stricture, whereas, in general, fixed transverse folds are incomplete and further apart, producing a characteristic stepladder appearance as a result of the trapping of barium between the folds (29). The feline esophagus could also conceivably be mistaken for the ringed esophagus of IEE, but these transverse striations occur as a transient phenomenon and are not associated with stricture formation (30). Finally, nonperistaltic contractions in the esophagus may produce a corrugated esophageal contour, but this form of esophageal dysmotility is also observed as a transient finding without associated stricture formation.
Other authors in the gastroenterology literature have described patients with IEE in whom barium studies revealed diffuse esophageal narrowing that produced a small-caliber esophagus (4,7,18,19). Paradoxically, these extensive areas of narrowing can be more difficult to recognize on barium study images than shorter segments of narrowing because of their long length, uniform luminal diameter, and smooth contour without obvious demarcations from a normal-caliber esophagus abutting the stricture. Despite the frequent subtlety of this finding, IEE should be considered when a small-caliber esophagus is detected at barium studies in the proper clinical setting.
It is important to establish a diagnosis of IEE so that appropriate treatment can be instituted. Antireflux therapy has been found to have little or no value in these patients (4,5). On the basis of the hypothesis that food allergens act as antigenic stimuli for eosinophilic inflammation of the esophagus (2–7), many investigators instead advocate various forms of antiallergy therapy, including oral steroids, topical steroids (eg, swallowing metered doses of aerosolized steroid preparations), and elemental diets, with varying degrees of success (8–12,16). In our study, six of the 14 patients (43%) with IEE received oral steroid inhalation (with symptomatic improvement in all patients), but only after the diagnosis had been confirmed in endoscopic biopsy specimens, as the referring gastroenterologists required a pathologic diagnosis before treatment with steroids. Although patients with strictures that cause intractable dysphagia can also undergo endoscopic dilation procedures, as in our study, these individuals often have only transient relief of dysphagia, so multiple dilation procedures may be required (1,6,23). Some authors believe that the presence of a ringed esophagus increases the risk of esophageal laceration or perforation during such dilation procedures (4), so extreme caution should be exercised when dilating these strictures.
Our investigation has the inherent limitations of a retrospective study, including selection bias and possible inaccurate documentation in medical records. Another concern is the relatively long mean interval between the barium studies and endoscopy. As a result, it can be argued that disease activity might have changed in the interval between these procedures and that the endoscopic findings might not necessarily reflect the radiographic findings. IEE, however, is thought to be a chronic disease, and, without some form of antiallergy therapy, substantial healing would not be likely. The chronic nature of IEE was underscored in a prospective study involving 30 adult patients with this condition (31); dysphagia persisted in all but one patient during a mean follow-up period of 7.2 years.
In summary, our experience suggests that most patients with IEE have esophageal strictures, often with distinctive ringlike indentations that produce a ringed esophagus. Fifty percent or fewer of these patients have classic findings of IEE, such as a history of allergies or peripheral eosinophilia, whereas most have hiatal hernias, gastroesophageal reflux, or both. The presence of reflux symptoms or other findings of reflux disease, therefore, should not preclude a diagnosis of IEE when dense eosinophilic infiltrates are present in endoscopic biopsy specimens from the esophagus. It is important to be aware of the clinical and radiographic features of IEE in adults so that appropriate treatment can be instituted in these patients.
|
FOOTNOTES |
|---|
Abbreviations: IEE = idiopathic eosinophilic esophagitis • LPO = left posterior oblique • RAO = right anterior oblique
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, M.S.L.; study concepts, S.L.Z., M.S.L., S.E.R., D.A.K.; study design, S.L.Z., M.S.L.; literature research, S.L.Z.; clinical studies, M.S.L., S.E.R., I.L., D.A.K.; data acquisition, S.L.Z., M.S.L., S.E.R., M.C.M., E.E.F.; data analysis/interpretation, S.L.Z., M.S.L.; manuscript preparation, S.L.Z., M.S.L.; manuscript definition of intellectual content, M.S.L.; manuscript editing, revision/review, and final version approval, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R Fass and D Sifrim Management of heartburn not responding to proton pump inhibitors Gut, February 1, 2009; 58(2): 295 - 309. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Levine, S. E. Rubesin, and I. Laufer Barium Studies in Modern Radiology: Do They Have a Role? Radiology, January 1, 2009; 250(1): 18 - 22. [Full Text] [PDF]
|
|
|
|
 |
|
 I. T. NONEVSKI, E. DOWNS-KELLY, and G. W. FALK Eosinophilic esophagitis: An increasingly recognized cause of dysphagia, food impaction, and refractory heartburn Cleveland Clinic Journal of Medicine, September 1, 2008; 75(9): 623 - 633. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
