- Email: [email protected]
Endoscopic ultrasonography in esophageal diseases
0016-5107/89/3503-0220$02.00 GASTROINTESTINAL ENDOSCOPY Copyright © 1989 by the American Society for Gastrointestinal Endoscopy
Endoscopic ultrasonography in esophageal diseases H. Dancygier, MD, M. Classen, MD Offebach and Munich, Federal Republic of Germany
Endoscopic ultrasonography (EUS) was performed prospectively in 38 patients with various esophageal disorders. Twenty-four had a histologically proven carcinoma and EUS was done to assess its ability in preoperative staging. In 9 of 24 patients (37.5%), tumor stenosis could be passed with the endoscope and EUS preoperative findings regarding tumor extension and the presence of enlarged periesophageal lymph nodes were confirmed in those operated on (n = 4). In five patients with achalasia, a periesophageal tumor was reliably excluded by EUS. In one of four patients with Barrett's esophagus, EUS demonstrated a small «1 cm) carcinoma that could not be visualized with any conventional technique. For differentiation of cancer recurrence after operation from periesophageal scar tissue EUS-guided transmural biopsies are needed. Our experience shows that at the present time EUS is the most reliable method to demonstrate small (::s1 cm) intra- and extramural esophageal lesions and that it should therefore be applied early in the work-up of patients with dysphagia. (Gastrointest Endosc
1989;35:220-225)
Endoscopic ultrasonography (EUS) with an ultrasonic transducer attached to the tip of an endoscope provides intracorporeal sonographic images of high resolution. It permits visualization of the entire thickness of the esophageal and upper gastrointestinal tract wall, as well as differentiation of various layers and delineation of closely adjacent structures.!' 2 With current conventional imaging modalities like endoscopy, barium x-ray, and computerized tomography (CT), the diagnosis and localization of esophageal tumors is possible in many cases. But the depth oftumor growth, especially of small esophageal cancers cannot be assessed accurately. We were therefore interested in the ability of EUS to visualize or exclude esophageal tumors and wondered if this tool might provide useful preoperative clinical information in patients with esophageal cancer. Received August 26, 1987. For revision October 20, 1987. Accepted May 25, 1988. From II. Medizinische Klinik und Poliklinik der Technischen Universitiit Munchen, Klinikum rechts der Isar, Munich, Federal Republic of Germany. Reprint requests: H. Dancygier, MD, II, Medizinische Klinik und Poliklinik der Technischen Universitiit Munchen, Klinikum rechts der Isar, Ismaninger Str. 22, 8 Munich 80, Federal Republic of Germany. 220
MATERIALS AND METHODS
Thirty-eight patients (age range, 47 to 82 years) were prospectively evaluated with the ultrasonic endoscope GF UM2/EU M2 (Olympus Opt./Aloca). Twenty-four patients had a histologically proven carcinoma of the esophagus. This diagnosis was known to the investigator and EUS was performed to assess its ability in preoperative cancer staging. Four patients had Barrett's esophagus (BE). In them, as well as in two patients who had undergone esophageal resection because of carcinoma, EUS was done as a follow-up procedure. In five patients with achalasia, EUS was performed after pneumatic dilation of the stenosis in order to exclude an extramural tumor. Prior to pneumodilation, brush cytology, forceps biopsies, and CT were done. Two further patients with endoscopically suspected submucosal tumors and one patient with dysphagia and diffuse thickening of the esophageal wall on CT investigation were examined by EUS. These data are summarized in Table 1. EUS examination was performed in the left lateral decubitus position after premedication with 5 to 12.5 mg of midazolam iv. The instrument used was a radial scanning upper gastrointestinal side-viewing endoscope that employs a frequency of 7.5 to 10 MHz and provides a 360-degree scan perpendicular to the long axis of the endoscope. The ultrasonic transducer measures 2 em in length, the rigid end is 4.2 em long, and the diameter is 1.3 em. In all patients the esophageal wall was evaluated with a water-filled balloon surrounding the sonographic probe (Fig. 1). GASTROINTESTINAL ENDOSCOPY
RESULTS
There were no complications due to premedication or insertion of the instrument. In all 24 cases with previous histologically verified carcinoma, the tumor was also visualized on EUS examination. CT was performed in all patients and failed to disclose tumors in six of them, although the radiologist was aware that he was dealing with esophageal cancer. In these patients the tumors were :s1.5-cm thick (Fig. 2). Esophageal carcinomas presented on EUS as circumferential (n = 15) or localized (n = 9) thickenings of the esophageal wall with disruption of its normal echo pattern. Most tumors were echo poor or had an irregular mixed echo pattern with prevailing echo poor areas (Fig. 3). Periesophageal lymph nodes as small as 0.5 cm in diameter could be clearly demonstrated by EUS (Fig. 4). In 15 of 24 cases the tumor growth had already led to a marked stenosis that proved Table 1. Endoscopic sonography in esophageal diseases Patients Carcinoma Resected carcinoma
BE Achalasia Submucous tumors Malignant lymphoma
24 2 4
5 2 1
impassable with the endoscope, so that only the oral aspect of the carcinoma could be assessed. Four of the nine patients in whom the entire length of the tumor could be passed with the instrument underwent an operation that confirmed the EUS findings regarding tumor extension and the presence of enlarged periesophageal lymph nodes. The remaining carcinoma patients were inoperable because of vessel invasion and/or distant organ metastases and had palliative therapy, i.e., external radiation, afterloading, laser treatment, and/or insertion of a prosthesis, so that in these cases anatomical verification of EUS findings was not possible. The delineation of the two benign submucosal tumors posed no difficulty for EUS, as they were well demarcated within the esophageal wall, had smooth outer margins and a regular echo pattern. On histological examination they were shown to be leiomyomas (Fig. 5). Of the two patients that were studied repeatedly after resection of a carcinoma, newly formed perianastomotic tissue was seen on EUS in one of them (Fig. 6). The EUS characteristics of spur like growth with irregular contours led to an incorrect interpretation of cancer recurrence. The excellent clinical course for nearly 2 years of follow-up-the patient objected to a second operation-strongly suggested that it was chronic granulation scar tissue. In the five patients with achalasia EUS reliably
Figure 1. No~mal esophageal wall (arrows). Balloon inflated around the tip of the instrument. a, descending aorta· rv right
pulmonary vein.
VOLUME 35, NO.3, 1989
'
,
221
Figure 2. Small (1 cm) esophageal cancer (arrows). Note irregular echo pattern with echo poor areas on the periphery of the tumor. a, descending aorta.
Figure 3. Esophageal carcinoma. Note ill-defined outer margin. la, left atrium; a, descending aorta.
excluded a periesophageal tumor, although the investigation in these cases could be performed only after prior pneumatic dilation of the stenosis. Of the four patients with BE, one had dysplastic 222
epithelial alterations and underwent resection after a localized eccentric thickening suggesting tumor growth was demonstrated by EUS. The diagnosis was confirmed on the resected specimen. In another 82GASTROINTESTINAL ENDOSCOPY
Figure 4. Paraesophageallymph node (1 x 0.5 em; black arrow) and paraesophageal vessel (white arrow). a, descending aorta.
Figure 5. Benign submucosal esophageal tumor (arrows). In contrast to "fixed" malignant tumors, these tumors "jump" in front of the ultrasonic probe while they are passed with the balloon. This is an EUS sign of their submucosal benign nature ("jumping sign").
year-old woman with BE, cytologic brushings showed evidence of malignant cells. Because of her advanced age and normal EUS layering of the eosphagus, no VOLUME 35, NO.3, 1989
operation was performed. After 16 months of followup, there is no evidence of esophageal cancer. In one patient with dysphagia, a diffuse thickening of the 223
Figure 6. Newly formed perianastomotic granulation tissue (arrows) after resection of esophageal cancer.
Figure 7. Malignant lymphoma (diagnosis made by histology). Note that the tumor lies inside the central echo poor band. The endosonographic features do not allow differentiation from carcinoma. a, descending aorta.
esophageal wall was seen on CT, a finding also confirmed by EUS. In addition to the demonstration of mere thickening, EUS also permitted localization of 224
the infiltrating tissue between mucosa and muscularis propria (Fig. 7). Histologically, a malignant lymphoma was found that was treated by chemotherapy. GASTROINTESTINAL ENDOSCOPY
DISCUSSION 3
The data presented here, our previous experience, and the results of other authors 4 confirm that EUS is a valuable method to define small ($1 cm) intramural lesions of the esophagus. Alterations missed by hitherto applied morphologic methods are easily depicted by EUS. The excellent resolution is due to the high frequencies (7.5 to 10 MHz) used and the ability to place the ultrasonic transducer in the immediate vicinity of the region of interest. The development of probes that generate a 360-degree radial scan has further facilitated the investigation, since the entire circumference of the esophagus can now be visualized without the need for manipulation of the instrument tip or patient position in order to demonstrate the desired sector. One aim of this study was to assess the efficacy of EUS in demonstrating or excluding an intra- or extramural esophageal tumor. Particular weight was given to the capability of EUS in the preoperative staging of esophageal cancer. In all 24 cases the carcinoma could be demonstrated by EUS. These results compare favorably with CT where the diagnosis was missed in six cases, although the radiologist had prior knowledge of the underlying disease. Tumors that are smaller than 1 cm in diameter are likely to be missed by CT. At the present time, an accurate classification of esophageal carcinoma based on the TNM system and in particular regarding the depth of extension of the primary tumor can only be made in those patients who have undergone radical tumor resection. In our patients complete passage of tumor stenosis-an absolute prerequisite for preoperative staging-was possible only in 9 of 24 cases (37.5%). In the remaining patients only the oral aspect of the tumor could be assessed. All of these tumors were already inoperable and were treated palliatively. In the four patients that were operated on, EUS diagnosis regarding tumor extension and periesophageal lymph node enlargement was confirmed. Care should, however, be exercised in ascribing enlargement of lymph nodes to metastatic involvement. We do not think that irregularities of the outer margin or of the internal echo pattern of a lymph node are reliable indicators of metastatic spread. Most lymph nodes larger than 1 cm in diameter in the draining area of a carcinoma will be malignant 5 • However, in the individual case it is still impossible with EUS to decide precisely on the nature of this enlargement. Therefore, we recommend only describing the size and location of the lymph node(s) and resisting the temptation of making histologic diagnoses by EUS. The esophageal wall is about 4-mm thick. The main
VOLUME 35, NO.3, 1989
central echo poor layer is thought to represent the muscularis propria. 6 Tumors in stages pT1 and pT2 have a significantly better prognosis than pT3 tumors. The emphasis during EUS investigation of esophageal cancer should therefore be placed on the integrity of this layer. Although preoperative EUS staging was possible only in a minority (37.5%) of our patients, this cannot be ascribed to methodological difficulties only. Our results rather mirror the natural history of esophageal cancer since most of our patients presented for EUS examination in a far advanced tumor stage. Due to the expense of the method, it is obvious that EUS cannot be applied as a screening procedure. One consequence of these data should therefore be to apply EUS early in the work-up of every patient with dysphagia, a strategy that we have already adopted in our hospital. Another fact that emerges from our study is the need for further technical improvement, especially the implementation of an instrumentation channel to enable guided transmural fine needle biopsies. This is extremely important in the follow-up of patients after resection for esophageal cancer, when it may be impossible to judge from the echo pattern alone if newly formed perianastomotic tissue is due to a chronic inflammatory granulation reaction or due to cancer recurrence. The construction of forward-viewing endoscopes combined with radial scanning devices may further increase the percentage of stenoses passed with the EUS probe. In regions with a high incidence of esophageal cancer and in high risk populations, our experience indicates that EUS in combination with already practiced cytologic screening methods might lead to an improvement of localization and preoperative staging. It remains to be seen however if this will also be accompanied by an improvement of treatment results and if it finally will have an impact on life expectancy.
REFERENCES 1. Dancygier H. Endoskopische Sonographie-ein diagnostischer Fortschritt im oberen Verdauungstrakt? Leber Magen Darm 1986;16:114-20. 2. Dancygier H, Classen M. Endoskopische Sonographie des oberen Verdauungstraktes. Med Klin 1986;81:92-6. 3. Dancygier H, Classen M. Endoscopic sonography in Barrett's esophagus. N Engl J Med 1987;316:277-8. 4. Takemoto T, Ito T, Aibe T, Okita K. Endoscopic ultrasonography in the diagnosis of esophageal carcinoma with particular regard to staging it for operability. Endoscopy 1986;18(suppl 3):22-5. 5. Feuerbach S, Lukas P, Gmeinwieser J. Fehlinterpretationen im Computertomogramm bei malignen Lympyknotenerkrankungen im Becken und Abdomen. Digit Bilddiagn 1984;4:176-80. 6. Dancygier H, Classen M. How can we diagnose the depth of cancer invasion in the esophagus? Endoscopy 1986;18(suppl 3):19-21.
225
Endoscopic ultrasonography in esophageal diseases H. Dancygier, MD, M. Classen, MD Offebach and Munich, Federal Republic of Germany
Endoscopic ultrasonography (EUS) was performed prospectively in 38 patients with various esophageal disorders. Twenty-four had a histologically proven carcinoma and EUS was done to assess its ability in preoperative staging. In 9 of 24 patients (37.5%), tumor stenosis could be passed with the endoscope and EUS preoperative findings regarding tumor extension and the presence of enlarged periesophageal lymph nodes were confirmed in those operated on (n = 4). In five patients with achalasia, a periesophageal tumor was reliably excluded by EUS. In one of four patients with Barrett's esophagus, EUS demonstrated a small «1 cm) carcinoma that could not be visualized with any conventional technique. For differentiation of cancer recurrence after operation from periesophageal scar tissue EUS-guided transmural biopsies are needed. Our experience shows that at the present time EUS is the most reliable method to demonstrate small (::s1 cm) intra- and extramural esophageal lesions and that it should therefore be applied early in the work-up of patients with dysphagia. (Gastrointest Endosc
1989;35:220-225)
Endoscopic ultrasonography (EUS) with an ultrasonic transducer attached to the tip of an endoscope provides intracorporeal sonographic images of high resolution. It permits visualization of the entire thickness of the esophageal and upper gastrointestinal tract wall, as well as differentiation of various layers and delineation of closely adjacent structures.!' 2 With current conventional imaging modalities like endoscopy, barium x-ray, and computerized tomography (CT), the diagnosis and localization of esophageal tumors is possible in many cases. But the depth oftumor growth, especially of small esophageal cancers cannot be assessed accurately. We were therefore interested in the ability of EUS to visualize or exclude esophageal tumors and wondered if this tool might provide useful preoperative clinical information in patients with esophageal cancer. Received August 26, 1987. For revision October 20, 1987. Accepted May 25, 1988. From II. Medizinische Klinik und Poliklinik der Technischen Universitiit Munchen, Klinikum rechts der Isar, Munich, Federal Republic of Germany. Reprint requests: H. Dancygier, MD, II, Medizinische Klinik und Poliklinik der Technischen Universitiit Munchen, Klinikum rechts der Isar, Ismaninger Str. 22, 8 Munich 80, Federal Republic of Germany. 220
MATERIALS AND METHODS
Thirty-eight patients (age range, 47 to 82 years) were prospectively evaluated with the ultrasonic endoscope GF UM2/EU M2 (Olympus Opt./Aloca). Twenty-four patients had a histologically proven carcinoma of the esophagus. This diagnosis was known to the investigator and EUS was performed to assess its ability in preoperative cancer staging. Four patients had Barrett's esophagus (BE). In them, as well as in two patients who had undergone esophageal resection because of carcinoma, EUS was done as a follow-up procedure. In five patients with achalasia, EUS was performed after pneumatic dilation of the stenosis in order to exclude an extramural tumor. Prior to pneumodilation, brush cytology, forceps biopsies, and CT were done. Two further patients with endoscopically suspected submucosal tumors and one patient with dysphagia and diffuse thickening of the esophageal wall on CT investigation were examined by EUS. These data are summarized in Table 1. EUS examination was performed in the left lateral decubitus position after premedication with 5 to 12.5 mg of midazolam iv. The instrument used was a radial scanning upper gastrointestinal side-viewing endoscope that employs a frequency of 7.5 to 10 MHz and provides a 360-degree scan perpendicular to the long axis of the endoscope. The ultrasonic transducer measures 2 em in length, the rigid end is 4.2 em long, and the diameter is 1.3 em. In all patients the esophageal wall was evaluated with a water-filled balloon surrounding the sonographic probe (Fig. 1). GASTROINTESTINAL ENDOSCOPY
RESULTS
There were no complications due to premedication or insertion of the instrument. In all 24 cases with previous histologically verified carcinoma, the tumor was also visualized on EUS examination. CT was performed in all patients and failed to disclose tumors in six of them, although the radiologist was aware that he was dealing with esophageal cancer. In these patients the tumors were :s1.5-cm thick (Fig. 2). Esophageal carcinomas presented on EUS as circumferential (n = 15) or localized (n = 9) thickenings of the esophageal wall with disruption of its normal echo pattern. Most tumors were echo poor or had an irregular mixed echo pattern with prevailing echo poor areas (Fig. 3). Periesophageal lymph nodes as small as 0.5 cm in diameter could be clearly demonstrated by EUS (Fig. 4). In 15 of 24 cases the tumor growth had already led to a marked stenosis that proved Table 1. Endoscopic sonography in esophageal diseases Patients Carcinoma Resected carcinoma
BE Achalasia Submucous tumors Malignant lymphoma
24 2 4
5 2 1
impassable with the endoscope, so that only the oral aspect of the carcinoma could be assessed. Four of the nine patients in whom the entire length of the tumor could be passed with the instrument underwent an operation that confirmed the EUS findings regarding tumor extension and the presence of enlarged periesophageal lymph nodes. The remaining carcinoma patients were inoperable because of vessel invasion and/or distant organ metastases and had palliative therapy, i.e., external radiation, afterloading, laser treatment, and/or insertion of a prosthesis, so that in these cases anatomical verification of EUS findings was not possible. The delineation of the two benign submucosal tumors posed no difficulty for EUS, as they were well demarcated within the esophageal wall, had smooth outer margins and a regular echo pattern. On histological examination they were shown to be leiomyomas (Fig. 5). Of the two patients that were studied repeatedly after resection of a carcinoma, newly formed perianastomotic tissue was seen on EUS in one of them (Fig. 6). The EUS characteristics of spur like growth with irregular contours led to an incorrect interpretation of cancer recurrence. The excellent clinical course for nearly 2 years of follow-up-the patient objected to a second operation-strongly suggested that it was chronic granulation scar tissue. In the five patients with achalasia EUS reliably
Figure 1. No~mal esophageal wall (arrows). Balloon inflated around the tip of the instrument. a, descending aorta· rv right
pulmonary vein.
VOLUME 35, NO.3, 1989
'
,
221
Figure 2. Small (1 cm) esophageal cancer (arrows). Note irregular echo pattern with echo poor areas on the periphery of the tumor. a, descending aorta.
Figure 3. Esophageal carcinoma. Note ill-defined outer margin. la, left atrium; a, descending aorta.
excluded a periesophageal tumor, although the investigation in these cases could be performed only after prior pneumatic dilation of the stenosis. Of the four patients with BE, one had dysplastic 222
epithelial alterations and underwent resection after a localized eccentric thickening suggesting tumor growth was demonstrated by EUS. The diagnosis was confirmed on the resected specimen. In another 82GASTROINTESTINAL ENDOSCOPY
Figure 4. Paraesophageallymph node (1 x 0.5 em; black arrow) and paraesophageal vessel (white arrow). a, descending aorta.
Figure 5. Benign submucosal esophageal tumor (arrows). In contrast to "fixed" malignant tumors, these tumors "jump" in front of the ultrasonic probe while they are passed with the balloon. This is an EUS sign of their submucosal benign nature ("jumping sign").
year-old woman with BE, cytologic brushings showed evidence of malignant cells. Because of her advanced age and normal EUS layering of the eosphagus, no VOLUME 35, NO.3, 1989
operation was performed. After 16 months of followup, there is no evidence of esophageal cancer. In one patient with dysphagia, a diffuse thickening of the 223
Figure 6. Newly formed perianastomotic granulation tissue (arrows) after resection of esophageal cancer.
Figure 7. Malignant lymphoma (diagnosis made by histology). Note that the tumor lies inside the central echo poor band. The endosonographic features do not allow differentiation from carcinoma. a, descending aorta.
esophageal wall was seen on CT, a finding also confirmed by EUS. In addition to the demonstration of mere thickening, EUS also permitted localization of 224
the infiltrating tissue between mucosa and muscularis propria (Fig. 7). Histologically, a malignant lymphoma was found that was treated by chemotherapy. GASTROINTESTINAL ENDOSCOPY
DISCUSSION 3
The data presented here, our previous experience, and the results of other authors 4 confirm that EUS is a valuable method to define small ($1 cm) intramural lesions of the esophagus. Alterations missed by hitherto applied morphologic methods are easily depicted by EUS. The excellent resolution is due to the high frequencies (7.5 to 10 MHz) used and the ability to place the ultrasonic transducer in the immediate vicinity of the region of interest. The development of probes that generate a 360-degree radial scan has further facilitated the investigation, since the entire circumference of the esophagus can now be visualized without the need for manipulation of the instrument tip or patient position in order to demonstrate the desired sector. One aim of this study was to assess the efficacy of EUS in demonstrating or excluding an intra- or extramural esophageal tumor. Particular weight was given to the capability of EUS in the preoperative staging of esophageal cancer. In all 24 cases the carcinoma could be demonstrated by EUS. These results compare favorably with CT where the diagnosis was missed in six cases, although the radiologist had prior knowledge of the underlying disease. Tumors that are smaller than 1 cm in diameter are likely to be missed by CT. At the present time, an accurate classification of esophageal carcinoma based on the TNM system and in particular regarding the depth of extension of the primary tumor can only be made in those patients who have undergone radical tumor resection. In our patients complete passage of tumor stenosis-an absolute prerequisite for preoperative staging-was possible only in 9 of 24 cases (37.5%). In the remaining patients only the oral aspect of the tumor could be assessed. All of these tumors were already inoperable and were treated palliatively. In the four patients that were operated on, EUS diagnosis regarding tumor extension and periesophageal lymph node enlargement was confirmed. Care should, however, be exercised in ascribing enlargement of lymph nodes to metastatic involvement. We do not think that irregularities of the outer margin or of the internal echo pattern of a lymph node are reliable indicators of metastatic spread. Most lymph nodes larger than 1 cm in diameter in the draining area of a carcinoma will be malignant 5 • However, in the individual case it is still impossible with EUS to decide precisely on the nature of this enlargement. Therefore, we recommend only describing the size and location of the lymph node(s) and resisting the temptation of making histologic diagnoses by EUS. The esophageal wall is about 4-mm thick. The main
VOLUME 35, NO.3, 1989
central echo poor layer is thought to represent the muscularis propria. 6 Tumors in stages pT1 and pT2 have a significantly better prognosis than pT3 tumors. The emphasis during EUS investigation of esophageal cancer should therefore be placed on the integrity of this layer. Although preoperative EUS staging was possible only in a minority (37.5%) of our patients, this cannot be ascribed to methodological difficulties only. Our results rather mirror the natural history of esophageal cancer since most of our patients presented for EUS examination in a far advanced tumor stage. Due to the expense of the method, it is obvious that EUS cannot be applied as a screening procedure. One consequence of these data should therefore be to apply EUS early in the work-up of every patient with dysphagia, a strategy that we have already adopted in our hospital. Another fact that emerges from our study is the need for further technical improvement, especially the implementation of an instrumentation channel to enable guided transmural fine needle biopsies. This is extremely important in the follow-up of patients after resection for esophageal cancer, when it may be impossible to judge from the echo pattern alone if newly formed perianastomotic tissue is due to a chronic inflammatory granulation reaction or due to cancer recurrence. The construction of forward-viewing endoscopes combined with radial scanning devices may further increase the percentage of stenoses passed with the EUS probe. In regions with a high incidence of esophageal cancer and in high risk populations, our experience indicates that EUS in combination with already practiced cytologic screening methods might lead to an improvement of localization and preoperative staging. It remains to be seen however if this will also be accompanied by an improvement of treatment results and if it finally will have an impact on life expectancy.
REFERENCES 1. Dancygier H. Endoskopische Sonographie-ein diagnostischer Fortschritt im oberen Verdauungstrakt? Leber Magen Darm 1986;16:114-20. 2. Dancygier H, Classen M. Endoskopische Sonographie des oberen Verdauungstraktes. Med Klin 1986;81:92-6. 3. Dancygier H, Classen M. Endoscopic sonography in Barrett's esophagus. N Engl J Med 1987;316:277-8. 4. Takemoto T, Ito T, Aibe T, Okita K. Endoscopic ultrasonography in the diagnosis of esophageal carcinoma with particular regard to staging it for operability. Endoscopy 1986;18(suppl 3):22-5. 5. Feuerbach S, Lukas P, Gmeinwieser J. Fehlinterpretationen im Computertomogramm bei malignen Lympyknotenerkrankungen im Becken und Abdomen. Digit Bilddiagn 1984;4:176-80. 6. Dancygier H, Classen M. How can we diagnose the depth of cancer invasion in the esophagus? Endoscopy 1986;18(suppl 3):19-21.
225