DOI: 10.1148/radiol.2362040748
(Radiology 2005;236:535-544.)
© RSNA, 2005
Superficial Esophageal Cancer: Esophagographic Findings Correlated with Histopathologic Findings1
Seung Soo Lee, MD,
Hyun Kwon Ha, MD,
Jae Ho Byun, MD,
Yong Moon Shin, MD,
Hyung-Jin Won, MD,
Ah Young Kim, MD,
Pyo Nyun Kim, MD,
Moon-Gyu Lee, MD,
Soon Jin Lee, MD,
Byung Hee Lee, MD and
Soo Yil Chin, MD2
1 From the Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Poongnap-dong, Songpa-gu, Seoul 138-736, Korea (S.S.L., H.K.H., J.H.B., Y.M.S., H.J.W., A.Y.K., P.N.K., M.G.L.); Department of Radiology and Center for Imaging Science, Sungkyunkwan University School of Medicine, Seoul, Korea (S.J.L.); and Department of Radiology, Korea Cancer Center Hospital, Seoul, Korea (B.H.L., S.Y.C.). Received April 25, 2004; revision requested July 12; revision received September 1; accepted October 15.
Address correspondence to H.K.H. (e-mail: hkha{at}www.amc.seoul.kr).
 |
ABSTRACT
|
|---|
PURPOSE: To retrospectively determine and evaluate the findings of superficial esophageal cancer at esophagography and to correlate the esophagographic findings with the depth of tumor invasion.
MATERIALS AND METHODS: The institutional review boards required neither their approval nor informed patient consent for this retrospective study. One hundred thirteen patients with superficial esophageal cancer who underwent esophagectomy at three institutions were included in this study. Double-contrast esophagograms were reviewed independently by two reviewers. For assessment of histopathologic findings, pathology reports were reviewed. Findings at esophagography, including morphologic type of the lesion, lesion extent, presence or absence of elevated or depressed component, margin and extent of elevated or depressed component, presence or absence of nodularity, extent of nodularity, esophageal luminal narrowing, and esophageal wall rigidity, were compared between mucosal and submucosal cancers by using 
2, Fisher exact, and independent-sample t tests.
RESULTS: Of 122 histopathologically proved superficial esophageal cancers in 113 patients, 100 (82%) were detected at esophagography. The most common morphologic type was the plaquelike form; 50 (50%) such lesions were depicted at esophagography. Morphologic types were significantly different between the mucosal and submucosal cancers (P < .001). Protruded and plaquelike lesions were more frequent among submucosal cancers, whereas most flat lesions were mucosal cancers. An elevated component (P < .001), a rigid esophageal wall (P < .001), and a lobulated or irregular margin of the elevated component (P = .023) were significantly more frequent among submucosal cancers. Also, total extent of the lesion (P < .001), size of the largest nodule (P < .001), and extent of nodularity (P = .036) were significantly larger in the submucosal cancers.
CONCLUSION: In the evaluation of patients with superficial esophageal cancer, esophagography appears to be helpful for diagnosing the tumor and differentiating mucosal from submucosal cancers.
© RSNA, 2005
|
INTRODUCTION
|
|---|
Because of recent advances in diagnostic techniques, the number of detected superficial esophageal cancers, defined as esophageal carcinomas with tumor invasion limited to the submucosa regardless of metastasis to the lymph nodes, has increased (1). Although patients with superficial esophageal cancer have a favorable clinical outcome compared with those who have more advanced cancers, the prognosis is variable, depending on the depth of tumor invasion. Patients with mucosal cancer rarely have metastases to the lymph nodes, and their 5-year survival rates are higher than 80% (2–5). On the other hand, the prevalence of lymph node metastasis in patients with submucosal cancer is 35.3%–53.5%, and the 5-year survival rate for these patients is 54.9%–78.4% (2–5). Given these clinicopathologic data, endoscopic mucosal resection, a minimally invasive surgical procedure, has been proposed as a radical resection treatment for mucosal cancer without evidence of lymphatic spread (4,6–9). Therefore, predicting the depth of tumor invasion during diagnostic work-up is important for determining the appropriate treatment for patients with superficial esophageal cancer.
Although the use of double-contrast esophagography enables visualization of even minimal esophageal abnormalities, as well as fine mucosal changes, investigators in a few studies have described the radiographic findings of superficial esophageal cancer in only a limited number of cases (10–14). Furthermore, there has been a paucity of published reports of radiographic findings of superficial esophageal carcinoma that were correlated with pathologic findings (15,16), and little about the radiographically diagnosed depth of tumor invasion has been published. Thus, the purposes of our study were to retrospectively determine and evaluate the findings of superficial esophageal cancer seen at esophagography and to correlate the esophagographic findings with the depth of tumor invasion.
|
MATERIALS AND METHODS
|
|---|
Patient Population
We retrospectively identified all patients who had histopathologically proved superficial esophageal cancer at two institutions (Asan Medical Center and Korea Cancer Center Hospital) from January 1991 to December 1999 and at one institution (Samsung Medical Center) from January 1998 to December 2000. A total of 2283 patients with esophageal cancers during these periods were identified, and 153 (6.7%) of them had superficial esophageal cancer. Of these 153 patients, 113 patients who underwent preoperative esophagography and surgical resection were included in this study. This patient population comprised 107 men and six women aged 38–82 years (mean age, 62.1 years ± 7.8 [standard deviation]). All 113 of these patients underwent transthoracic esophagectomy with lymphadenectomy. No preoperative radiation therapy and/or chemotherapy was performed. The institutional review boards of all three institutions required neither their approval nor informed patient consent for this type of retrospective study.
Review of Pathology Reports and Medical Records
One author (S.S.L.) reviewed the pathology reports of these patients to assess the locations and numbers of the lesions, the depths of tumor invasion, the histologic and morphologic types of the lesions, and the tumor stages. The depth of tumor invasion was classified as mucosal or submucosal. Mucosal cancer was defined as malignancy in which the tumor invasion was limited to within the muscularis mucosa, and submucosal cancer was defined as malignancy in which the tumor was invading the submucosa (1).
The radiologist (S.S.L.) retrospectively assessed the morphologic types of the tumors on the basis of descriptions in the pathology reports without knowledge of the esophagographic findings. The gross morphologic appearances of the lesions were classified into four types: protruded, plaquelike, flat, and depressed or ulcerative (Fig 1). Polypoid or elevated lesions with heights greater than 5 mm were considered to be protruded, and those with heights of less than 5 mm were considered to be plaquelike. Both protruded and plaquelike lesions may contain central ulceration or a depressed component. Flat lesions were lesions without a definite elevated or depressed component and lesions that manifested as a mucosal color change or a fine mucosal nodularity or irregularity. Ulcerative or depressed lesions were, respectively, depressed or ulcerative lesions without a definite elevated component.
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Morphologic types of superficial esophageal cancer. Protruded type is a polypoid or protuberant lesion with a height of 5 mm or greater. Plaquelike type is a slightly elevated lesion or sessile polyp with a height of less than 5 mm. Both protruded and plaquelike types may have a central ulceration or a depressed component. Flat type is a lesion without a definite elevated or depressed component or a lesion that manifests as a mucosal color change or a mucosal granularity or irregularity. Ulcerative or depressed type is a depressed or ulcerative lesion without a definite elevated component.
|
|
Spreading superficial esophageal carcinomas were classified according to the morphologic type of the main lesion. When there was no definite main lesion and findings of only spreading superficial esophageal cancer, such as nodular and whitish mucosal thickening (17), were present, the lesion was considered to be of the flat type.
To localize the lesion site, we arbitrarily divided the esophagus into three areas: the upper, middle, and lower regions. The upper esophagus was the cervical portion of the esophagus that extends from the upper part of the esophageal sphincter to the thoracic inlet; this part of the esophagus is approximately 5 cm long. The next 10-cm segment of the esophagus, which extends from the thoracic inlet to the carina level, was considered to be the middle esophagus. The remaining lower thoracic and intraabdominal portions of the esophagus combined formed the lower esophagus. The information noted in the pathology report was sufficient for morphologic classification in all cases. The same radiologist (S.S.L.) also reviewed the patients' medical reports to assess their presenting symptoms at the time of diagnosis.
Imaging Techniques
Double-contrast esophagography was performed in all 113 patients. Senior residents or fellows performed all of these examinations with the supervision of radiology staff members (H.K.H., S.J.L., B.H.L., and S.Y.C.) who had 10–30 years of experience. To reduce esophageal motility, each patient received an intramuscular injection of 20 mg of scopolamine-N-butyl bromide (Buscopan; Boehringer Ingelhein, Seoul, Korea) 5 minutes before undergoing esophagography. After the patient ingested an effervescent agent, double-contrast esophagograms were obtained while the patient, in the upright position, rapidly drank a barium suspension (Solotop Powder; Taejoon Pharmacal, Seoul, Korea). Different compositions of barium suspension were used at the three institutions: A 220% wt/vol barium suspension was used at Asan Medical Center, and a 250% wt/vol suspension was used at both Samsung Medical Center and Korea Cancer Center Hospital. Left posterior oblique views, the acquisition of which can prevent overlapping of the esophagus and the spine, were routinely obtained. Anteroposterior, right posterior oblique, and lateral images occasionally were obtained to achieve enface and in-profile views of the lesions.
Image Interpretation
All esophagograms were independently reviewed by one general radiologist with 4 years of experience (S.S.L.) and one abdominal radiologist with 25 years of experience (H.K.H.). Only film hard-copy images were reviewed; videotaped images were not evaluated. Each reviewer retrospectively evaluated the esophagograms and completed a study worksheet. When reviewing the esophagograms, each reviewer knew that all of the patients included in the study had superficial esophageal cancer but was blinded to both the results of the pathology report review and the interpretations made by the other reviewer. On the study worksheet, the reviewers detailed the following: lesion location, morphologic type of the tumor, extent of the lesion in its longest dimension, extent of the elevated component, margin of the elevated component, presence or absence of an ulcerative or depressed component, margin of the depressed or ulcerative component, presence or absence of nodularity, extent of nodularity, size of the largest nodule, presence or absence of esophageal luminal narrowing, degree of esophageal wall rigidity, and image quality.
The depicted extents of the lesion, elevated component, and nodularity; the size of the largest nodule; and image quality were evaluated by one reviewer (S.S.L.). Both reviewers evaluated the other items listed on the study worksheet independently. The same morphologic classifications that were used when reviewing the pathology reports were used to evaluate the morphologic types of the lesions at esophagography.
The margin of the elevated component was classified as smooth or as lobulated or irregular. The margin of a depressed or ulcerative lesion was classified as sharp or unsharp. Nodularity was considered to be present when nodular or granular lesions were seen on the esophageal mucosal surface or on the surfaces of protruded or plaquelike lesions. The size of the largest nodule was measured by selecting the largest nodular lesion among many nodules clustered on the surface of the esophagus. The degree of wall rigidity was assessed on the in-profile view and classified as rigid, equivocal, or nonrigid. We considered the esophageal wall to be rigid when the base of the lesion was always seen as a semilunar or wedged depression on every in-profile view of the lesion. In contrast, when this depressed lesion base was not seen and the esophageal wall near the lesion showed normal peristalsis, we considered the esophageal wall to be nonrigid. When only a mild depression was seen at the lesion base on some of the in-profile views, wall rigidity was considered to be equivocal.
Image quality was classified as excellent, fair, or poor. When the esophagus was adequately distended, adequate double-contrast views were obtained, and both the enface and the in-profile views of the visible lesion were available, we considered the image quality of the esophagogram to be excellent. If one or more of these criteria were not fulfilled, the image quality of the esophagogram was considered to be fair or poor.
Consensus Interpretations and Comparisons
After completing the image interpretations, the two reviewers (S.S.L., H.K.H.) then compared their image finding worksheets. If any discrepancies were present, the reviewers repeated the image interpretation, discrepancies were resolved by consensus, and the results of the consensus reading were used for later comparison with the pathology report findings. To assess whether the lesions detected at esophagography were true cancers, one author (S.S.L.) compared the lesion locations noted in the pathology reports with the lesion locations on the esophagograms. If the lesion location assessed at esophagography was different from that described in the pathology report, the reviewers were considered to have detected a false lesion and the lesion was considered to have been missed at esophagography.
Statistical Analyses
We compared the depth of tumor invasion, the morphologic type of the tumor based on pathology report findings, the lesion location, the lesion extent, and the image quality between the lesions detected at esophagography and those missed at esophagography to identify the important factor in determining the visibility of true lesions. Agreement between the two reviewers regarding the morphologic classification of the lesions, the margin of elevated lesions, the margin of depressed or ulcerative lesions, the presence or absence of nodularity, the presence or absence of esophageal luminal narrowing, and the degree of wall rigidity was calculated by using the Cohen 
test. Values greater than 0.75 were considered to represent excellent agreement beyond chance; values less than 0.40, poor agreement; and values between 0.40 and 0.75, fair to good agreement beyond chance.
To evaluate the relationships between the esophagographic findings of superficial esophageal cancer and the depth of tumor invasion, the esophagographic findings of superficial esophageal cancer, including tumor morphologic type, lesion extent, elevated component extent, elevated component margin, presence or absence of ulcerative or depressed component, depressed or ulcerative component margin, presence or absence of nodularity, nodularity extent, largest nodule size, presence or absence of esophageal luminal narrowing, and degree of esophageal wall rigidity, were compared between the mucosal and submucosal cancers by using 2, independent sample t, and Fisher exact tests. We considered a statistically significant difference to be present when the P value was less than .05.
To assess the accuracy of the esophagographic assessment of tumor morphology, we compared the tumor morphologic types assessed by using the pathology reports with those assessed at esophagography. The results of comparison of the two worksheets and the results of the consensus readings were used to compare the morphologic types assessed at esophagography with the morphologic types assessed by using the pathology reports and to compare the esophagographic findings with the pathologic findings. Data analyses were performed by using a statistical software package (SPSS, version 9.0; SPSS, Chicago, Ill).
|
RESULTS
|
|---|
Clinical Features and Pathologic Findings
The presenting symptoms of the 113 patients included dysphagia in five (4.4%) patients; chest discomfort in one (0.9%) patient; no symptoms in 98 (86.7%) patients; and other gastrointestinal symptoms such as dyspepsia, abdominal discomfort, and indigestion in nine (8.0%) patients. Eight (7.1%) patients had multicentric esophageal cancer: Two separate cancers were detected in each of seven patients, and three cancers were detected in one patient. Therefore, a total of 122 superficial esophageal cancers were detected in the 113 patients. There were variable histologic types of esophageal cancer: squamous cell carcinoma in 111 (90.1%) lesions, small cell carcinoma in four (3.3%) lesions, basaloid squamous cell carcinoma in three (2.5%) lesions, adenosquamous carcinoma in two (1.6%) lesions, mixed small cell and squamous cell carcinoma in one (0.8%) lesion, and sarcomatoid squamous cell carcinoma in one (0.8%) lesion.
Seventeen (13.9%) lesions were located in the upper esophagus; 73 (59.8%) lesions, in the middle esophagus; and 32 (26.2%) lesions, in the lower esophagus. Of the 122 superficial esophageal cancers in the 113 patients, 27 (22.1%), in 19 patients, were mucosal and 95 (77.9%), in 94 patients, were submucosal. Eighteen (15.9%) of the 113 patients—17 (18.1%) of the 94 patients with submucosal cancer and one (5.3%) of the 19 patients with mucosal cancer—had metastases to the lymph nodes.
Esophagographic Findings of Superficial Esophageal Cancer
Ninety-nine (81.1%) of the 122 histopathologically proved superficial esophageal cancers were identified at esophagography by both radiologists. Five (4.1%) of the 122 lesions were detected by only one radiologist, and 18 (14.8%) lesions were missed by both radiologists. In the eight patients with multicentric cancer, only one of the two lesions in each of six patients and one of the three lesions in one patient were detected at esophagography by both radiologists and the two lesions in one patient were missed at esophagography by both radiologists. At comparison of the lesions detected at esophagography with those described in the pathology reports, all 99 lesions identified by both radiologists and one of the five lesions detected by one of the two radiologists were considered to be true cancers. The remaining four lesions detected by one radiologist were considered to be false cancers. Therefore, a total of 100 (82%) true superficial esophageal cancers were detected at esophagography and analyzed for their radiographic findings.
At comparison of the characteristics between the lesions detected and the lesions missed on esophagography, there were significant differences in tumor morphologic type (P = .003): Only 14 (58.3%) of the 24 flat types were detected at esophagography, whereas most of the protruded and plaquelike types were detected (Table 1). The extents of the lesions detected at esophagography (mean extent, 28.0 mm ± 13.9 [standard deviation]) were significantly larger than the extents of the lesions missed (mean extent, 16.5 mm ± 10.3) (P < .001). Image quality was significantly different between the lesions detected and the lesions missed at esophagography (P < .001).
The most common tumor morphologic type was the plaquelike form, which appeared as a slightly elevated lesion or a sessile polyp with a smooth, lobulated, or irregular surface (Tables 2, 3); 50 (50.0%) lesions were depicted as this type at esophagography. Twenty-one (42%) of the 50 plaquelike lesions had surface nodularity and central ulceration (Figs 2, 3). The next most common morphologic type was protruded (Fig 4)—25 (25.0%) lesions were depicted as such at esophagography—followed, in order of decreasing frequency, by the flat type (Fig 5)—13 (13.0%) lesions were depicted as such at esophagography—and the depressed or ulcerative type (Figs 6, 7)—12 (12.0%) lesions were depicted as such at esophagography.
View this table:
[in this window]
[in a new window]
|
TABLE 2. Comparison between Mucosal and Submucosal Superficial Esophageal Cancers according to Morphologic Type Depicted at Esophagography
|
|
View larger version (61K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Plaquelike lesion in 65-year-old man with submucosal cancer. (a) Left posterior oblique double-contrast esophagogram shows a slightly elevated lesion (arrows) with a lobulated surface in the lower esophagus. The lesion has surface nodularity, and the size of the largest nodule is 7 mm. There is no definite esophageal wall rigidity or luminal narrowing. (b) Resected surgical specimen shows a slightly elevated lesion (arrows) with a lobulated margin and surface nodularity.
|
|
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Plaquelike lesion in 65-year-old man with submucosal cancer. (a) Left posterior oblique double-contrast esophagogram shows a slightly elevated lesion (arrows) with a lobulated surface in the lower esophagus. The lesion has surface nodularity, and the size of the largest nodule is 7 mm. There is no definite esophageal wall rigidity or luminal narrowing. (b) Resected surgical specimen shows a slightly elevated lesion (arrows) with a lobulated margin and surface nodularity.
|
|
View larger version (73K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Plaquelike lesion with central ulceration in 69-year-old man with submucosal cancer. (a) Left posterior oblique double-contrast esophagogram shows a slightly elevated lesion (straight arrows) that has a smooth margin, with a central depression in the lower esophagus. A depressed lesion (curved arrow) with a sharp margin is seen in the center of the elevated lesion. Wall rigidity (arrowheads) is considered to be equivocal. (b) Resected surgical specimen confirms the morphologic findings seen in a: an elevated lesion (arrows) with a central depressed lesion (*).
|
|
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Plaquelike lesion with central ulceration in 69-year-old man with submucosal cancer. (a) Left posterior oblique double-contrast esophagogram shows a slightly elevated lesion (straight arrows) that has a smooth margin, with a central depression in the lower esophagus. A depressed lesion (curved arrow) with a sharp margin is seen in the center of the elevated lesion. Wall rigidity (arrowheads) is considered to be equivocal. (b) Resected surgical specimen confirms the morphologic findings seen in a: an elevated lesion (arrows) with a central depressed lesion (*).
|
|
View larger version (58K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Protruded-type lesion in 38-year-old man with submucosal cancer. (a) Anteroposterior double-contrast esophagogram shows a protruded lesion (arrows) with a height of more than 5 mm. The margin of the elevated component is smooth. Esophageal wall rigidity (arrowheads) also is present. (b) Resected esophagus also shows a protruded lesion (arrows) in the distal one-third of the esophagus. (c) Histopathologic section shows esophageal cancer (T) invading the submucosal layer (SM). The muscularis propria layer (M) is not invaded by tumor.
|
|
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Protruded-type lesion in 38-year-old man with submucosal cancer. (a) Anteroposterior double-contrast esophagogram shows a protruded lesion (arrows) with a height of more than 5 mm. The margin of the elevated component is smooth. Esophageal wall rigidity (arrowheads) also is present. (b) Resected esophagus also shows a protruded lesion (arrows) in the distal one-third of the esophagus. (c) Histopathologic section shows esophageal cancer (T) invading the submucosal layer (SM). The muscularis propria layer (M) is not invaded by tumor.
|
|
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Protruded-type lesion in 38-year-old man with submucosal cancer. (a) Anteroposterior double-contrast esophagogram shows a protruded lesion (arrows) with a height of more than 5 mm. The margin of the elevated component is smooth. Esophageal wall rigidity (arrowheads) also is present. (b) Resected esophagus also shows a protruded lesion (arrows) in the distal one-third of the esophagus. (c) Histopathologic section shows esophageal cancer (T) invading the submucosal layer (SM). The muscularis propria layer (M) is not invaded by tumor.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Flat-type lesion in 64-year-old man with mucosal cancer. (a) Left posterior oblique double-contrast esophagogram of a highly distended (left) and a moderately distended (right) esophageal lumen shows a flat lesion in the middle one-third of the esophagus. There is no definite elevated or depressed component, and the only finding is a cluster of fine 1–2-mm surface nodules (arrows) in the esophageal mucosa. This cluster of nodules is a typical finding of flat-type lesions at esophagography. There is no esophageal wall rigidity or esophageal luminal narrowing. (b) Close-up view of a (right image) shows fine granularity and mucosal irregularity (arrows) more clearly. (c) Resected surgical specimen shows a flat lesion (arrows).
|
|
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Flat-type lesion in 64-year-old man with mucosal cancer. (a) Left posterior oblique double-contrast esophagogram of a highly distended (left) and a moderately distended (right) esophageal lumen shows a flat lesion in the middle one-third of the esophagus. There is no definite elevated or depressed component, and the only finding is a cluster of fine 1–2-mm surface nodules (arrows) in the esophageal mucosa. This cluster of nodules is a typical finding of flat-type lesions at esophagography. There is no esophageal wall rigidity or esophageal luminal narrowing. (b) Close-up view of a (right image) shows fine granularity and mucosal irregularity (arrows) more clearly. (c) Resected surgical specimen shows a flat lesion (arrows).
|
|
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Flat-type lesion in 64-year-old man with mucosal cancer. (a) Left posterior oblique double-contrast esophagogram of a highly distended (left) and a moderately distended (right) esophageal lumen shows a flat lesion in the middle one-third of the esophagus. There is no definite elevated or depressed component, and the only finding is a cluster of fine 1–2-mm surface nodules (arrows) in the esophageal mucosa. This cluster of nodules is a typical finding of flat-type lesions at esophagography. There is no esophageal wall rigidity or esophageal luminal narrowing. (b) Close-up view of a (right image) shows fine granularity and mucosal irregularity (arrows) more clearly. (c) Resected surgical specimen shows a flat lesion (arrows).
|
|
View larger version (77K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Left posterior oblique double-contrast esophagograms of a depressed-ulcerative lesion in a 65-year-old man with mucosal cancer. (a) Image of highly distended esophageal lumen shows a localized area of mucosal nodularity and granularity (arrows) in the middle esophagus; this is a finding of spreading superficial esophageal cancer. (b) Image of moderately distended esophageal lumen shows a faint barium pooling with an unsharp margin (arrowheads) due to an ulcerative lesion, which is not depicted in a. However, the mucosal nodularity due to the spread of superficial cancer seen in a is not depicted on this view. Therefore, analysis of multiple spot views of different projections and degrees of luminal distention was necessary for accurate assessment of the morphologic features and the extent of tumor in this patient.
|
|
View larger version (53K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Left posterior oblique double-contrast esophagograms of a depressed-ulcerative lesion in a 65-year-old man with mucosal cancer. (a) Image of highly distended esophageal lumen shows a localized area of mucosal nodularity and granularity (arrows) in the middle esophagus; this is a finding of spreading superficial esophageal cancer. (b) Image of moderately distended esophageal lumen shows a faint barium pooling with an unsharp margin (arrowheads) due to an ulcerative lesion, which is not depicted in a. However, the mucosal nodularity due to the spread of superficial cancer seen in a is not depicted on this view. Therefore, analysis of multiple spot views of different projections and degrees of luminal distention was necessary for accurate assessment of the morphologic features and the extent of tumor in this patient.
|
|
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Findings in 67-year-old man with two submucosal cancers. (a) Left posterior oblique double-contrast esophagogram shows a plaquelike lesion (arrows) in the middle one-third of the esophagus. Ulceration (arrowhead) is seen in the center of the elevated component. Definite esophageal wall rigidity and esophageal luminal narrowing also are present. A lesion in the distal portion of the esophagus cannot be definitively demonstrated. (b) Resected surgical specimen shows a plaquelike lesion with central ulceration (arrows) in the middle one-third of the esophagus, as seen in a. In the distal one-third of the esophagus, a flat-type lesion (arrowheads), which is not depicted in a, is seen. At histopathologic examination, both lesions were found to be submucosal cancers.
|
|
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Findings in 67-year-old man with two submucosal cancers. (a) Left posterior oblique double-contrast esophagogram shows a plaquelike lesion (arrows) in the middle one-third of the esophagus. Ulceration (arrowhead) is seen in the center of the elevated component. Definite esophageal wall rigidity and esophageal luminal narrowing also are present. A lesion in the distal portion of the esophagus cannot be definitively demonstrated. (b) Resected surgical specimen shows a plaquelike lesion with central ulceration (arrows) in the middle one-third of the esophagus, as seen in a. In the distal one-third of the esophagus, a flat-type lesion (arrowheads), which is not depicted in a, is seen. At histopathologic examination, both lesions were found to be submucosal cancers.
|
|
The protruded type was seen as an intraluminal polypoid lesion with a smooth, lobulated, or irregular surface. Five (20%) of the 25 protruded lesions had central ulceration, and 10 (40%) had surface nodularity. The flat type manifested as a fine nodular or granular lesion or a subtle irregularity of the esophageal mucosa. The depressed or ulcerative type manifested as a faint or clear barium collection in the esophagus, with or without surrounding mucosal nodularity or irregularity. The extents of the lesions in their longest dimension ranged from 5 to 60 mm (mean, 25.7 mm ± 14.2).
Interobserver Agreement
Agreement regarding evaluation of the following characteristics was excellent between the two reviewers: esophageal luminal narrowing ( = 0.915, P < .001), classification of morphologic type ( = 0.796, P < .001), presence or absence of nodularity ( = 0.784, P < .001), and margin of depressed or ulcerative lesion ( = 0.776, P < .001). Interobserver agreement was fair in the evaluation of elevated lesion margins ( = 0.712, P < .001) and esophageal wall rigidity ( = 0.523, P < .001).
Comparison of Esophagographic and Pathologic Findings
There were significant differences (Tables 2, 3) in morphologic type between the mucosal and submucosal superficial esophageal cancers (P < .001). Among the 13 flat-type lesions (Fig 5) depicted at esophagography, all except one were mucosal cancers. In contrast, all 25 protruded-type lesions (Fig 4) and 46 (92%) of the 50 plaquelike lesions (Figs 2, 3) depicted at esophagography were submucosal cancers. Among the 50 plaquelike lesions, all 21 lesions with central ulceration were submucosal cancers (P = .076). Lesions with an elevated component were more frequent among the submucosal cancers (71 [90%] lesions) than among the mucosal cancers (four [19%] lesions) (P < .001) (Table 3). The margin of the elevated component was smooth in the mucosal cancers (four [100%] lesions), whereas the margin was frequently lobulated or irregular in the submucosal cancers (48 [68%] lesions) (P = .023).
There were significant differences in esophageal wall rigidity between the mucosal and submucosal cancers (P < .001): All 45 lesions associated with a rigid esophageal wall were submucosal cancers (Figs 4, 7), whereas 15 of 21 lesions associated with a nonrigid esophageal wall were mucosal cancers (Figs 5, 8). The extents of the lesions (P < .001), sizes of the largest nodules (P < .001), and extents of nodularity (P = .036) were significantly larger among the submucosal cancers than among the mucosal cancers. Esophageal luminal narrowing was present with only 11 submucosal cancers (Fig 8), but there was no significant difference in the presence of esophageal luminal narrowing between the mucosal and submucosal cancers (P = .114). Although none of the five mucosal cancers with a depressed component had a sharp margin (Fig 6), there was no significant difference in the margin of the depressed component between the mucosal and submucosal cancers (P = .111).
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Depressed-ulcerative lesion in 61-year-old man with mucosal cancer, which was classified as the flat type at esophagography. (a) Left posterior oblique double-contrast esophagogram of a highly distended (left) and a moderately distended (right) lumen shows an area of fine nodularity and mucosal irregularity (arrows) encircling an approximately 3-cm segment of the distal portion of the esophagus. Possible differential diagnoses include spreading superficial esophageal cancer and focal esophagitis such as drug-induced esophagitis. If this lesion is superficial esophageal cancer, the esophagographic morphologic type is flat type because there is no definite elevated or depressed component. Neither esophageal wall rigidity nor esophageal luminal narrowing is seen. (b) Resected surgical specimen shows whitish mucosal thickening and surface nodularity (arrowheads), which are findings of spreading superficial esophageal cancer. The extent of the lesion is larger than that seen in a: The tumor encompasses nearly a 5-cm segment of the esophagus. A small depressed lesion (*) that is not depicted in a also is seen. The histopathologically determined morphologic type of this tumor is depressed or ulcerative.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Depressed-ulcerative lesion in 61-year-old man with mucosal cancer, which was classified as the flat type at esophagography. (a) Left posterior oblique double-contrast esophagogram of a highly distended (left) and a moderately distended (right) lumen shows an area of fine nodularity and mucosal irregularity (arrows) encircling an approximately 3-cm segment of the distal portion of the esophagus. Possible differential diagnoses include spreading superficial esophageal cancer and focal esophagitis such as drug-induced esophagitis. If this lesion is superficial esophageal cancer, the esophagographic morphologic type is flat type because there is no definite elevated or depressed component. Neither esophageal wall rigidity nor esophageal luminal narrowing is seen. (b) Resected surgical specimen shows whitish mucosal thickening and surface nodularity (arrowheads), which are findings of spreading superficial esophageal cancer. The extent of the lesion is larger than that seen in a: The tumor encompasses nearly a 5-cm segment of the esophagus. A small depressed lesion (*) that is not depicted in a also is seen. The histopathologically determined morphologic type of this tumor is depressed or ulcerative.
|
|
For 81 (81%) of the 100 lesions detected at esophagography, the morphologic types assessed at esophagography were concordant with the morphologic types described in the pathology reports. The morphologic types assessed by using esophagography are compared with those assessed by using the pathology reports in Table 4. All of the histopathologically proved protruded-type lesions except one and 42 (86%) of the 49 proved plaquelike lesions were accurately assessed at esophagography. In contrast, only nine (64%) of the true 14 flat-type lesions and 10 (62.5%) of the 16 true depressed or ulcerative lesions were accurately assessed at esophagography (Fig 8, Table 4).
View this table:
[in this window]
[in a new window]
|
TABLE 4. Comparison between Tumor Morphologic Types Assessed at Esophagography and Those Described in Pathology Report
|
|
|
DISCUSSION
|
|---|
Investigators in several studies have evaluated the roles of endoscopy and endoscopic ultrasonography (US) in the assessment of the depth of invasion by superficial esophageal cancer (3,18–20). To our knowledge, however, before now, there were few published reports comparing the radiographic findings of superficial esophageal carcinoma with the depth of tumor invasion (16). Ueyama et al (16) analyzed the radiographic findings of 46 early esophageal cancers and concluded that the sharpness of depressed lesions and the esophageal wall rigidity associated with both elevated and depressed lesions correlated strongly with the depth of tumor invasion.
In addition to the results of the Ueyama et al study (16), our study results demonstrated that other radiographic findings, such as morphologic type, lesion or nodularity extent, largest nodule size, presence or absence of an elevated component, and elevated component margin, also are helpful in determining the depth of tumor invasion. Among the four morphologic types of superficial esophageal cancer, the protruded type was found only among the submucosal cancers at esophagography, whereas most of the flat types were found among the mucosal cancers. An elevated component was seen more frequently among the submucosal cancers than among the mucosal cancers, and a lobulated or irregular surface was seen only among the submucosal carcinomas that had an elevated component.
As in the other study (16), esophageal wall rigidity, which cannot be easily assessed at endoscopy, was another finding that was valuable for determining the depth of tumor invasion in our study. However, the interpretation of esophageal wall rigidity seemed to be somewhat subjective. Interobserver agreement regarding the interpretation of this finding was fair, but it was worst ( = 0.523, P < .001) with regard to the radiographic findings evaluated. Therefore, we believe that esophageal wall rigidity must be interpreted carefully and that only definite wall rigidity should be considered useful as a radiographic finding suggestive of submucosal invasion.
Although differences in the margin of the depressed component were not significant between the mucosal and submucosal cancers, all lesions that had a depressed component with a sharp margin and esophageal luminal narrowing were submucosal cancers. The lack of statistical significance seemed to be due to the small number of mucosal cancers in our study. We believe that these findings also may be helpful in assessing the depth of tumor invasion when they are considered together with other radiographic findings.
In our study, the rate of superficial esophageal carcinoma detection with esophagography was 82%, which is similar to or higher than the range of detection rates (73%–81%) reported in previous studies (12,13,21). Because we reviewed the esophagograms with knowledge that all of them had been obtained in patients with histopathologically proved superficial esophageal cancer, the detection rate reported in our study was probably overestimated. If the data of healthy control subjects and those of patients with esophageal abnormalities other than cancer had been included for review, the accuracy of esophagography in the detection and diagnosis of superficial esophageal cancer might have been lower than that reported in our study. However, because our purpose in this study was not to determine the accuracy of esophagography in the detection of superficial esophageal cancer but rather to evaluate the findings of this malignancy at esophagography and then compare them with the depth of tumor invasion, it seemed unnecessary to include a control group.
In our study, the main factors that determined the detectability of a malignant lesion were esophagogram image quality, lesion extent, and morphologic type. Among the four morphologic types assessed in our study, the flat type was the most difficult to detect. Only 14 (58%) of the 24 true flat-type lesions were detected at esophagography. Most of the flat-type lesions detected at esophagography were not completely flat. These lesions can have mild mucosal elevations or depressions, which result in fine nodular or irregular surface patterns on esophagograms. Lugol dye endoscopy has facilitated the diagnosis of superficial esophageal carcinoma, which appears very normal or shows only a color change on these endoscopic images (22). These lesions were difficult to detect on even double-contrast esophagograms of excellent image quality. To improve the image quality at double-contrast esophagography, adequate esophageal distention and hypotonia are mandatory. Instead of using glucagon, which is not widely used in our country, we routinely administered 20 mg of scopolamine-N-butyl bromide by means of intramuscular injection to prevent esophageal muscle contraction. Scopolamine-N-butyl bromide, a competitive antagonist of acetylcholine, can effectively induce hypotony of the esophagus, stomach, or duodenum (23–25). Thus, the use of scopolamine-N-butyl bromide in esophagography facilitates double-contrast examinations of the esophagus with adequate distention of the esophageal lumen (26).
In eight patients with multicentric esophageal carcinoma, we found only one of the two lesions in each of seven patients and failed to find any lesions in one patient. Because the reported frequency of multicentric tumors is 14%–23% (2,27,28), other sites in the esophagus should be carefully examined, even when one lesion is detected during both the acquisition and the interpretation of an esophagogram. In our study, one patient with multicentric mucosal cancer had lymph node metastasis. This patient had two separate mucosal cancers: One invaded the superficial mucosal layer, and the other invaded the muscularis mucosa. Histopathologically, both tumors were undifferentiated squamous cell carcinomas. Although the risk of lymphatic spread is negligible with tumors limited to the intraepithelial layer, this risk increases with progressive infiltration depth. Lymph node metastasis is present in about 5%–10% of cases of superficial esophageal cancer when the muscularis mucosa is invaded (3–5). In addition to the depth of invasion, the differentiation grade has been found to affect the prognosis and the risk of lymphatic spread (4). The lymph node metastasis in one patient with mucosal cancer in our study could be attributed to tumor invasion into the muscularis mucosa layer and a high histopathologic grade.
Several esophageal abnormalities may mimic superficial esophageal cancer at esophagography. Squamous papilloma usually appears as small sessile polyps, which are sometimes indistinguishable from superficial esophageal cancer (29,30). Esophageal leiomyoma and other submucosal tumors may mimic protruded-type superficial esophageal cancer; a smooth surface and an obtuse border, characteristics of submucosal tumors, may be helpful in the differential diagnosis (29). Flat-type superficial esophageal cancer, especially the spreading form, should be differentiated from other esophageal abnormalities that manifest as fine, diffuse nodular lesions, including glycogen acanthosis, acanthosis nigricans, leukoplakia, and esophagitis (10,29,30). In the assessment of suspected superficial esophageal cancer showing localized superficial ulceration or mucosal irregularity, drug-induced esophagitis should be included in the differential diagnosis. This is especially true when the lesion is located at the level of the aortic arch or the left main bronchus, which are common locations of drug-induced esophagitis (31).
Although Barrett esophagus is uncommon in Asian countries and no data on patients with adenocarcinoma were evaluated in this study, the prevalence of adenocarcinoma arising from Barrett esophagus has increased in Western countries (32,33). The radiographic findings of early adenocarcinoma secondary to Barrett esophagus may be similar to those of superficial esophageal cancer. Therefore, Barrett esophagus–derived adenocarcinoma should be included in the differential diagnosis, especially in patients with preexisting peptic stricture, reflux esophagitis, or Barrett esophagus (30). Advanced esophageal cancer also should be differentiated from superficial esophageal cancer. Because we did not include data on patients with advanced esophageal cancer in this study, we could not assess the esophagographic differentiation between superficial and advanced cancers.
Endoscopic US has been widely used to diagnose the depth of tumor invasion in patients with superficial esophageal cancer. The main advantage of performing endoscopic US is the direct visualization of esophageal wall structures (19,20). The use of a relatively recently developed high-frequency catheter miniprobe has enabled more accurate staging of superficial esophageal cancers owing to the capability for more detailed visualization of layer structures with this probe (20). The overall accuracy of the determination of tumor invasion depth with this miniprobe is reported to be 75%–92% (19,20). Endoscopic US also facilitates the detection of lymph node metastasis with an overall accuracy of 75%–87% (19,34). Therefore, despite the promising results regarding the role of esophagography in the diagnosis of tumor invasion depth in patients with superficial esophageal cancer reported in our study, endoscopic US still seems to be the method of choice for this diagnosis.
There were several limitations in the present study. First, a relatively small number of mucosal cancers were assessed: 27 (22.1%) of the 122 lesions were mucosal cancers. Because only a small proportion of superficial esophageal cancers are mucosal tumors, it is difficult to collect a large number of cases of mucosal cancer, even in a multi-institutional study such as ours. Second, as discussed earlier, no data from a control group of subjects without esophageal abnormalities, with advanced esophageal cancer, or with esophageal disease other than esophageal cancer were evaluated in our study.
In conclusion, in the evaluation of patients with superficial esophageal cancer, esophagography appears to be helpful for diagnosing the tumor and differentiating mucosal from submucosal cancers.
|
ACKNOWLEDGMENTS
|
|---|
We thank In Suk Lee, PhD, of the Department of Statistics, College of Natural Science, Kyungpook National University, Taegu, Korea, for his advice on the statistical analysis of the data and Bonnie Hami, MA, of the Department of Radiology, University Hospitals of Cleveland, Ohio, for her editorial assistance.
|
FOOTNOTES
|
|---|
2 Department of Radiology, Seoul Medical Center, Seoul, Korea 
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, S.S.L., H.K.H.; study concepts, H.K.H.; study design, S.S.L., H.K.H.; literature research, S.S.L., J.H.B., Y.M.S.; clinical studies, H.J.W., A.Y.K., P.N.K., M.G.L.; data acquisition, J.H.B., S.J.L., B.H.L., S.Y.C.; data analysis/interpretation, Y.M.S., H.J.W., A.Y.K., P.N.K., M.G.L.; statistical analysis, S.S.L.; manuscript preparation and editing, S.S.L., H.K.H.; manuscript definition of intellectual content, H.K.H.; manuscript revision/review and final version approval, all authors
|
References
|
|---|
- Japanese Society for Esophageal Disease. Guidelines for the clinical and pathologic studies on carcinoma of the esophagus. 8th ed. Tokyo, Japan: Kanehara, 1992; 32–39.
- Kato H, Tachimori Y, Watanabe H, Yamaguchi H, Ishikawa T, Itabashi M. Superficial esophageal carcinoma: surgical treatment and the results. Cancer 1990; 66:2319–2323.[CrossRef][Medline]
- Tachibana M, Yoshimura H, Kinugasa S, et al. Clinicopathological features of superficial squamous cell carcinoma of the esophagus. Am J Surg 1997; 174:49–53.[CrossRef][Medline]
- Tajima Y, Nakanishi Y, Ochiai A, et al. Histopathologic findings predicting lymph node metastasis and prognosis of patients with superficial esophageal carcinoma: analysis of 240 surgically resected tumors. Cancer 2000; 88:1285–1293.[CrossRef][Medline]
- Nishimaki T, Tanaka O, Suzuki T, Aizawa K, Watanabe H, Muto T. Tumor spread in superficial esophageal cancer: histopathologic basis for rational surgical treatment. World J Surg 1993; 17:766–772.[CrossRef][Medline]
- Giovannini M, Bernardini D, Moutardier V, et al. Endoscopic mucosal resection (EMR): results and prognostic factors in 21 patients. Endoscopy 1999; 31:698–701.[CrossRef][Medline]
- Nwakakwa V, Fleisher D. Endoscopic mucosal resection of the esophagus: band ligation technique. Gastrointest Endosc Clin N Am 2001; 11:479–488.[Medline]
- Makuuchi H. Endoscopic mucosal resection for mucosal cancer in the esophagus. Gastrointest Endosc Clin N Am 2001; 11:445–458.[Medline]
- Nomura T, Boku N, Ohtsu A, et al. Recurrence after endoscopic mucosal resection for superficial esophageal cancer. Endoscopy 2000; 32:277–280.[CrossRef][Medline]
- Itai Y, Kogure T, Okuyama Y, Akiyama H. Diffuse finely nodular lesions of the esophagus. AJR Am J Roentgenol 1977; 128:563–566.[Abstract]
- Itai Y, Kogure T, Okuyama Y, Akiyama H. Superficial esophageal carcinoma: radiological findings in double contrast studies. Radiology 1978; 126:597–601.[Abstract]
- Moss AA, Koehler RE, Margulis AR. Initial accuracy of esophagograms in detection of small esophageal carcinoma. AJR Am J Roentgenol 1976; 127:909–913.[Abstract]
- Chin SY, Lee BH, Kim KH, Koh JS, Zo JI. Diagnosis of superficial esophageal cancer: comparative study of double contrast esophagography and endoscopy. J Korean Radiol Soc 2000; 43:203–208.
- Levine MS, Dillon EC, Saul SH, Laufer I. Early esophageal cancer. AJR Am J Roentgenol 1986; 146:507–512.[Abstract/Free Full Text]
- Chung J, Kim CB, Kim HK, Kim MJ, Lee JT, Yu HS. Esophagographic findings of early esophageal cancer: comparison with pathologic results. J Korean Radiol Soc 1998; 38:869–874.
- Ueyama T, Kawamoto K, Yamada Y, Masuda K. Early esophageal carcinoma: evaluation of depth of invasion based on double-contrast esophagography. Acta Radiol 1998; 39:133–137.[Medline]
- Soga J, Tanaka O, Sasaki K, Kawaguchi M, Muto T. Superficial spreading carcinoma of the esophagus. Cancer 1982; 50:1641–1645.[CrossRef][Medline]
- Nagawa H, Kaizaki S, Seto Y, Tominaga O, Muto T. The relationship of macroscopic shape of superficial esophageal carcinoma to depth of invasion and regional lymph node metastasis. Cancer 1995; 75:1061–1064.[CrossRef][Medline]
- Murata Y, Suzuki S, Mitsunaga A, et al. Endoscopic ultrasonography in diagnosis and mucosal resection for early esophageal cancer. Endoscopy 1998; 30(suppl):A44–A46.
- Hasegawa N, Niwa Y, Arisawa T, Hase S, Goto H, Hayakawa T. Preoperative staging of superficial esophageal carcinoma: comparison of an ultrasound probe and standard endoscopic ultrasonography. Gastrointest Endosc 1996; 44:388–393.[CrossRef][Medline]
- Haruma K, Tokutomi T, Tsuda T, Yoshihara M, Sumii K, Kajiyama G. Superficial esophageal carcinoma: a report of 27 cases in Japan. Am J Gastroenterol 1991; 86:1723–1728.[Medline]
- Misumi A, Harada K, Murakami A, et al. Role of lugol dye endoscopy in the diagnosis of early esophageal cancer. Endoscopy 1990; 22:12–16.[Medline]
- Guignard JP, Herxheimer A, Greenwood RM. Effect of hyoscine butylbromide on gut motility. Clin Pharmacol Ther 1968; 9:745–748.[Medline]
- Channer KS, Wolinski A, Kaye B, Virjee J. The effect of hyoscine butylbromide on the swallowing of capsules. Br J Clin Pharmacol 1983; 15:560–563.[Medline]
- Niemann H, Jakob G. Esophagomanometric finding in man after intravenous administration of a new cholinolytic (pramiverine), atropine and scopolamine butylbromide. Arzneimittelforschung 1976; 26:728–730.[Medline]
- Levine MS, Laufer I. Esophagus. In: Laufer I, Levine MS, eds. Double contrast gastrointestinal radiology. 2nd ed. Philadelphia, Pa: Saunders, 1992; 109.
- Bogomoletz WV, Molas G, Gayet B, Potet F. Superficial esophageal cell carcinoma of the esophagus: a report of 76 cases and review of the literature. Am J Surg Pathol 1989; 13:535–546.[Medline]
- Yoshinaka H, Shimazu H, Fukumoto T, Baba M. Superficial esophageal carcinoma: a clinicopathological review of 59 cases. Am J Gastroenterol 1991; 86:1413–1418.[Medline]
- Levine MS. Benign tumor. In: Gore RM, Levine MS, Laufer I, eds. Textbook of gastrointestinal radiology. Philadelphia, Pa: Saunders, 1994; 431–443.
- Levine MS, Halvorsen RA. Esophageal carcinoma. In: Gore RM, Levine MS, Laufer I, eds. Textbook of gastrointestinal radiology. Philadelphia, Pa: Saunders, 1994; 446–475.
- Levine MS. Other esophagitides. In: Gore RM, Levine MS, Laufer I, eds. Textbook of gastrointestinal radiology. Philadelphia, Pa: Saunders, 1994; 403–430.
- Lee JI, Park H, Jung H, Rhee PL, Song CW, Choi MG. Prevalence of Barrett's esophagus in an urban Korean population: a multicenter study. J Gastroenterol 2003; 38:23–27.[CrossRef][Medline]
- Tutuian R, Castell DO. Barrett's esophagus prevalence and epidemiology. Gastrointest Endosc Clin N Am 2003; 13:227–232.[CrossRef][Medline]
- Natsugoe S, Yoshinaka H, Morinaga T, et al. Ultrasonographic detection of lymph-node metastases in superficial carcinoma of the esophagus. Endoscopy 1996; 28:674–679.[Medline]
TOP
ABSTRACT
INTRODUCTION
