World J. Surg. 24, 216 –226, 2000 DOI: 10.1007/s002689910035

WORLD Journal of

SURGERY © 2000 by the Socie´te´ Internationale de Chirurgie

Endoscopic Ultrasonography of the Esophagus Morimichi Fukuda, M.D., Ph.D., Kenichiro Hirata, M.D., Ph.D., Hiroshi Natori, M.D., Ph.D. Department of Diagnostic Ultrasound and Medical Electronics, Sapporo Medical University Hospital, Minami 1-jo, Nishi 16-chome, Chuo-ku, Sapporo, Hokkaido, Japan Abstract. Endoscopic ultrasonography (EUS) is a generally accepted technique for the preoperative staging of malignant tumors in the upper and lower gastrointestinal tracts. In particular, EUS has been considered the method of choice in diagnosing esophageal carcinoma due to the relative ease in performing the examination and the accuracy of staging based on high-resolution ultrasonic imaging from within the lumen of the esophagus. This comprehensive review covers currently available EUS instruments, image characteristics of esophageal carcinoma, and images by the recently introduced miniprobe scanner. The role of the method in diagnosing superficial esophageal carcinoma and the possible treatment by endoscopic mucosal resection of this particular disease entity are discussed.

Noninvasive diagnostic imaging is commonly performed in patients with suspected gastrointestinal diseases and it often includes an upper gastrointestinal (GI) series and endoscopic examinations. Endoscopy is an extremely effective procedure for evaluating mucosal lesions of the GI tract, ulceration, polyps, carcinoma, and inflammatory changes, based on inspection of the mucosa through the endoscope. However, these approaches have limitations in diseases involving submucosal and extramural layers of the GI tract. Likewise, precise staging of carcinoma of the GI tract in terms of the depth of penetration is almost impossible. Furthermore, invasion of carcinoma to an adjacent organ or tissue is extremely difficult to diagnose by these standard procedures. To overcome these limitations, noninvasive imaging methods [X-ray computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound (US)] are used. The limited resolution of X-ray CT and MRI and the difficulty of penetration of sonic waves into air-containing hollow viscera, respectively, have significantly limited their diagnostic application in gastroenterology. The situation is changing rapidly following an introduction of high-resolution real time sonography and of endoscopic ultrasonography (EUS) [1–10]. The malignant neoplasm of the esophagus, esophageal carcinoma, poses another difficulty in management: insidious onset and paucity of subjective symptoms in its very early stage which often result in delay in introducing effective therapeutic measures. Because of its specific anatomy, the almost entirely intrathoCorrespondence to: M. Fukuda, M.D., Ph.D.

racic location, the esophagus is rarely accessible for US imaging except for the cervical or abdominal esophagus, hence the use of endoscopic ultrasound (EUS) is indicated. History of EUS The first attempt to investigate the GI tract by endosonography has been recorded in the literature as early as 1957 [4, 11]. However, EUS, endosonography conducted under optical guidance by endoscopic means, has been in use about 18 years. Preliminary reports on the use of EUS have been made at 2 different sites, Japan and the United States [8 –10]; one was by mechanical scanning and the other with linear array scanning methods. Stimulated by these pioneering works, a number of commercial firms began the manufacture of such instruments. Among these, gradual but steady advances were made by one of the endoscope manufacturers, Olympus Optical Co. Ltd. (Tokyo, Japan), as demonstrated by the production of a series of ultrasonoendoscopes with mechanically rotating transducers attached to the head of the gastrofiberscope. Recently, other manufacturers also have been producing sonoendoscopes equipped with either electronic linear or convex arrays attached to the gastrofiberscope. Although experience is still limited, EUS has already yielded impressive results in the diagnosis of a variety of diseases involving the GI tracts including malignant tumors, myogenic masses, nonmalignant mucosal lesions, achalasia, and varices. The recent introduction of microtip technology for higher frequency US has also opened up a new arena in examination of the GI tract by means of endosonography. Equipment A series of sonoendoscopes manufactured for the respective purposes presented are shown in Table 1 and Figures 1 and 2. These are classified into 2 major subgroups, those with a radially scanning transducer(s) and those with an electronic array transducer. Another model specifically designed for esophageal scanning is shown in Figure 3. The thickness of the scope is reduced to 7.9 mm in diameter and the shape of the scan head is changed to a conus-shaped attachment to facilitate safer insertion through the help of a guidewire previously introduced into an esophageal lumen narrowed due to an infiltrating tumor mass. The shortcom-

Fukuda et al.: EUS of Esophagus

217

Table 1. Commercially available ultrasonoendoscope and miniature probes for upper gastrointestinal examinations.

Instrument Type Endoscopy (degrees) Length (cm) Diameter (mm) Frequency (MHz)

Olympus

Olympus

Pentax

Olympus

Fujinon

Aloka

Toshiba

GF-UM20 Radial scan 45 105 13.2 7.5, 12

MH908 Esophageal probe — 60 8 7.5

FG-32UA Convex array 105 125 12 5, 7.5

UM-2R/3R Mini — 205 2.4 12, 20

SP-501 Mini — 170 2.6 15, 20

SSD550 Mini — 146 2.4 15, 20

PMD1524 Mini — 180 2.4 10, 15, 20

Fig. 1. Series of sonoendoscopes for endoscopic ultrasonography of gastrointestinal (GI) tracts. Left to right: GF-UM3 for upper GI, CF-UM3 for colon, and JF-UM3 for pancreas and biliary systems, manufactured by Olympus Co. Ltd. (Tokyo, Japan).

ings are the removal of the fiberoptics and fixed transducer frequency of 7.5 MHz with a slightly reduced crystal size that caused somewhat degraded image quality in contrast to those obtained by GF-UM3 and GF-UM20, which are equipped with dual frequencies of 7.5 and 12 MHz (standard Olympus EUS series). The electronic scanning sonoendoscope, on the other hand, has the disadvantage of a narrower width compared with the radial scanners described above; the advantage is that the scope has fine-needle aspiration (FNA) capability under direct US guidance [12–14]. Another feature, 2-dimensional color flow Doppler from within the lumen of the GI tract, is useful especially for injection sclerotherapy of esophageal varices. The latest topic in endosonography is the introduction of a thin diameter scanning device, the microtip US probe, to be used through the forceps channel of a therapeutic endoscope (Figs. 4, 5) [15–19]. Three types are available: 2 types of rotating transducers to be used with the therapeutic endoscope for scanning the upper GI tract and 1 type for linear scans examining the narrower lumen of the stenotic GI tract caused by advanced carcinoma or narrower spaces in the common bile duct or pancreatic duct. The microtip scanner generally utilizes US of much higher frequency ranges than for conventional EUS so that higher resolutions are obtained; a drawback is limited penetration of high frequency US. Thus the series of equipment currently available for EUS examinations have indicated clearly that the method has reached an established phase whereby it can cover any area in which the sonoendoscope can reach. Also these developments undoubtedly

Fig. 2. Schematic drawing of Olympus GF-UM3 sonoendoscope. A disk transducer (dual disc) in the scanning chamber is rotated by the motor unit housed in the proximal end of the scope via the flexible shaft in the scope. a. Schematic drawing of the scope. b. Drawing of the scan head.

paved the way for the evolution of intraluminal sonography not only for the detection, diagnosis, and treatment of diseases of the GI tracts but also the possibility for much wider applications in cardiology, urology, gynecology, and other fields of medicine. Scanning Techniques for the Esophagus To obtain clear ultrasonic images of the upper GI tracts including the esophagus and the stomach by conventional EUS, the patient is fasted overnight; an antispasmodic may be injected before the examination to suppress peristalsis. After mild pharyngeal anesthesia, an echoendoscope is inserted with the patient in the left decubitus position. After the insertion of the scope, the tip is gently advanced until it reaches the cardiac portion of the stomach. Guided by visual observation through the fiberoptics of the scope, the tip is gently advanced toward the greater curvature of the stomach with continuous inflation of air as for routine endoscopy. With further insertion, the tip will pass the gastric antrum to reach the pylorus.

218

World J. Surg. Vol. 24, No. 2, February 2000

Fig. 3. Olympus MH908 esophageal endoscanner. Due to the absence of the optical system, the diameter of the scope became smaller (7.9 mm) and 360-degree scanning can be made. The hole for the guidewire facilitates safe insertion into the narrower space. The arrow indicates the conusshaped scan head.

Fig. 4. Olympus UM-2R/3R miniprobe scanner with imaging unit. The motor unit in the proximal end rotates the small disc transducer of either 12 or 20 MHz and the received echo signals are transmitted to the imaging unit. a. Imaging unit with UM-2R. b. Olympus UM-2R miniprobe scanner.

After completion of the endoscopic observations of the esophagus and stomach, degassed water sufficient to distend the gastric lumen is gently infused through the fine water channel in the scope. The water distension method is extremely helpful in examining the gastric as well as the esophageal wall, as an undistended organ tends to be folded, obscuring minute lesions of the wall, and also it is mandatory to maintain an optimal distance between the probe and mucosal surface or lesions exactly at a focal range.

Examination of both the stomach and esophagus is also necessary to survey the esophagogastric junction to exclude the possibility of Barrett’s esophagus with or without carcinoma. Also it is essential to include transgastric scanning of the liver and celiac trunk to exclude any metastatic lesions. The patient is scanned in the right decubitus position and then is asked to take the left decubitus position to continue observation of the upper parts of the stomach and esophagogastric junction. The scope is then gently withdrawn

Fukuda et al.: EUS of Esophagus

219

ing is essential to analyze the image in detail, frame by frame, after each examination. Image Interpretation and Description The sonographic appearance of the GI tract is, in general, almost identical over the entire length with some deviation according to the location; the tubular structure of the wall consists of the following layers: mucosa, submucosa, muscularis propria with serosal layers, outermost. It is well established that the wall of the stomach and esophagus resolves into 5 distinct layers by endosonography. A number of published reports have confirmed that the innermost layer with increased echogenicity and a thin hypoechoic layer immediately deep to it correspond mainly to the mucosa and partly to the muscularis mucosae, and that the next echogenic layer corresponds to the submucosa. The fourth hypoechoic layer is the muscularis propria layer and the outermost echogenic layer is the serosal layer with fat appendage. Normal Structure of the Esophagus Obtained by Miniprobe Scanners Fig. 5. The miniprobe and drawing of the scan head.

to the esophagus and continuous observation and recording are carried out over the entire length of the organ. In routine observation of only the esophagus by conventional EUS using a GF-UM2, GF-UM3, or GF-UM20/EM20 instrument, the scan head is covered by a balloon to ensure good contact between the scope and the esophageal wall. In using high frequency miniprobe systems, on the other hand, the use of the balloon is mostly unnecessary and continuous water infusion can be used to secure an adequate acoustic window for optimal scanning. Fortunately, a balloon sheath for the microtip catheter became available recently so that it may be used in an appropriate indication [16]. To achieve maximum image resolution of the surface of the esophageal mucosa, optimum positioning of the high frequency probe over the suspected lesion of mucosa is essential. Particularly, the distance between the scanning probe and the target is critical, and for this the surface of the esophagus should be placed perpendicular to the direction of the scanning US beam. It is important to note that the superficial change of mucosal cancer of the esophagus is often so subtle that it is difficult to discern even by the eye of the skilled examiner in the presence of definite tissue diagnosis of carcinoma of the biopsied specimen. The assessment of minute changes in the thickness of the mucosa is of utmost importance in diagnosing minute carcinoma of the esophagus, as there will be minimal or no apparent difference in echogenicity in the presence of carcinoma and, in fact, subtle changes in mucosal thickness may be the only sign of mucosal carcinoma. The penetration of high frequency US of 20 MHz was initially limited merely to the layer of the proper muscle; complete imaging of the whole invading cancer generally required the use of lower frequency probes or conventional EUS apparatus. For image recording and archiving, the use of videotape record-

Although the normal structure of the esophagus is the key finding in the detection and differentiation of abnormalities in the esophagus, the appearance of the normal structure itself also may vary considerably according to the level of the esophagus, degree of distension, differences in frequency, and aperture of the transducer used. The miniprobe of 20 MHz frequency gives the image with the highest resolution, as shown in Figure 6, indicating 7 layers instead of 5 layers by the conventional EUS. The layer of the muscularis mucosa (mm layer), which generally appears as a thin hypoechoic layer, is essential to interpret the extent of tumor invasion to the lamina propria. The finding, however, is less consistent and is positive in ⬍15% of cases examined. The layering structure of the normal esophagus in general is less distinct compared with that of the normal gastric mucosa, in which a 5-layer structure is more prominent, especially with a distinct echo difference between the mucosa, submucosa, and proper muscle layers. Furthermore, passive movement of the esophagus to the closely situated cardiac tissues sometimes brings blurring of EUS images, thus resulting in difficulty in discriminating individual layering structures due to the cardiac motion. The wall structure imaged by miniprobe scanning or EUS often reveals characteristic features as shown in Figure 7. The image was obtained from a case that underwent total esophagectomy followed by surgical reconstruction of the jejunal tubes. The image shows striking differences from those of the esophageal or gastic wall. Generally speaking, if no abnormal findings are identified on the esophageal wall by EUS, it is safe to say that the esophagus is intact. It should be considered that reflectivity of any particular layer of the GI tract can be influenced by the law of physics, e.g., the angle of insonation, the original power output, and changes in gain settings of the instrument. The descriptions of the principle findings such as echotexture, contour, and reflectivity strength as isoechoic, hypoechoic, and hyperechoic have to be made in an objective manner, inasmuch as the comparisons with those of the

220

World J. Surg. Vol. 24, No. 2, February 2000 Table 2. Indications of endoscopic ultrasound.

Fig. 6. Ultrasonotomogram of the normal esophagus. M: mucosal layer; MM: thin hypoechoic muscularis mucosa layer; PM: 2 layers of the muscularis propria separated by a thin echogenic layer; SM: thin echogenic submucosa. The extent of tumor invasion is interpreted in reference to these basic structures.

Stomach Staging of gastric cancer invasion Identification of submucosal tumors Differentiation of gastric ulcer, benign and malignant Gastric varices Assessment of therapeutic effect of carcinostatic agent Extragastric compression, elucidation of cause Esophagus and mediastinum Staging of esophageal cancer Differential diagnosis of esophageal tumor Esophageal varices and assessment of sclerosing therapy Imaging of aortic wall structure Transesophageal scanning of the heart Pancreas and biliary tree Diagnosis of pancreatitis and pancreatolithiasis Staging of pancreatic cancer Differential diagnosis of cystic tumors of pancreas Obstructive jaundice, differential diagnosis Peripancreatic lymph nodes Assessment of duodenal papilla and uncus region Bile duct stones Bile duct carcinoma Colon and rectum Staging of colonic carcinoma Differentiation of colonic ulceration Diagnosis of ulcerative colitis and Crohn’s disease Imaging and differentiation of regional lymph nodes Diagnosis of submucosal tumors

Abnormal Findings in EUS Examination The diagnostic indications of EUS are expanding rapidly; the established uses are summarized in Table 2. Diseases of the Esophagus

Fig. 7. Image of jejunal (J) mucosa after reconstruction of esophagectomy. The image was recorded by conventional EUS (GF-UM3, 12 MHz) using the water-filled balloon. Note the difference in architecture with the normal esophageal wall. Arrows indicate the reconstructed esophageal wall made of jejunal tube. N: lymph node.

corresponding normal structures stand at an equidistance from the scanning probe, the sounding source. The most important findings to be observed by EUS will be the presence or absence of localized abnormality, tumor or ulceration, thickening of the wall, lymph node swelling along the esophagus, and presence or absence of variceal vessels. In the presence of any endoscopic abnormalities, area of discoloration, reddening, polyps, and irregular ulceration, these have to be examined thoroughly first by pigment endoscopy, chromoendoscopy, or after staining with Lugol’s solution, then should be followed by either conventional EUS or miniprobe scanning to visualize any possible change due to early esophageal carcinoma.

The diagnostic applications of EUS in esophageal disease are mostly similar to those of the stomach; the main applications are differential diagnosis of mucosal and submucosal tumors and staging of esophageal malignancies. Mucosal cancer confined to the very surface of the mucosa was often difficult to image by conventional EUS because of the limited resolution of the frequency used, the near-field effect, and distortion caused by the rubber balloon, however, interpretation of carcinomatous infiltration in advanced cases was mostly precise, including assessment of tumor invasion to the thoracic aorta and visualization of regional lymph node involvement. These findings were frequently very helpful in interpreting resectability of the mass lesions. Benign lesions, leiomyomas, and cysts are readily identified by their characteristic sonographic findings. Diagnosis of esophageal varices also has been well documented. Assessment of Esophageal Cancer Invasion Using High Frequency Miniature Probes The use of the miniprobe in the diagnosis of esophageal carcinoma has received controversial evaluation by investigators. We also have tested various miniprobe systems for their ability to diagnose esophageal malignancy. Before this trial, we conducted a comparative study on the use of various models of miniprobe scanners for the assessment of gastric cancer invasion. A total of 120 cases with various stages of gastric carcinoma (94 cases with

Fukuda et al.: EUS of Esophagus

221

Table 3. Histologic staging of esophageal carcinoma according to the Japanese Society for Esophageal Disease. Stage 0 1 2 3 4

Depth of invasion

Lymph node metastasis

Organ metastasis

Pleural dissemination

m, sm mp a1 a2 a3

n(⫺) n(⫺) n1(⫹) n2(⫹) n3(⫹), n4(⫹)

M0 M0 M0 M0 M1

Pl0 Pl0 Pl0 Pl0 Pl1

m: invasion confined to the mucosa; sm: invasion confined to the submucosa; mp: invasion into the muscularis propria; a1: invasion reaching to the adventitia; a2: definite invasion into the adventitia; a3: invasion into the adjacent structures; n(⫺): no metastasis; n1(⫹): metastasis to group 1; n2(⫹): metastasis to group 2; n3(⫹): metastasis to group 3; n4(⫹): metastatic node beyond group 3; M0: no organ metastasis; M1: organ metastasis positive; Pl0: no pleural dissemination; Pl1: pleural dissemination positive.

early-stage gastric cancer, mucosal cancer, and submucosal cancer, and 26 cases with advanced cancer) were examined consecutively between August 1990 and September 1997 using 4 types of miniprobe scanners: UM-1W, 7.5 MHz (Olympus); SP-101, 20 MHz (Fujinon, Tokyo, Japan); and UM-2R/3R, 12 and 20 MHz (Olympus). The combined use of SP-101 and UM-1W gave 88.6% accuracy in discriminating early from advanced gastric carcinoma. The combined use of 2 mechanical scanners with different frequencies, Olympus UM-2R (12 MHz) and UM-3R (20 MHz), however, also gave identical results of 90.4% accuracy in discriminating the advanced from the early cancers. When the accuracy rates were assessed on each disease grading of mucosa (m), submucosa (sm), muscularis propria (mp), and subsero-se (ss-se), the results were 72.2%, 66.7%, 66.7%, and 30%, respectively [4]. In both groups, the major causes of erroneous interpretation stem from various reasons: low detectability of the muscularis mucosae, inflammatory changes around the invading tumor tissue, fibrosis, ulcerative changes in the early-stage cancers, and ultrasonic attenuation in cancers of the advanced stage. Assessment of Esophageal Cancer Staging Using High Frequency Miniature Probes The equipment used in the study was similar to those of previous studies described above except that the Olympus UM-1W was removed and Fujinon SP-101 was replaced by Fujinon SP-501. The detailed characteristics are described in Table 1. During the period from April 1992 to October 1996, a total of 54 cases with possible diagnosis of esophageal carcinoma were examined by one of the miniature probes and all of the cases underwent surgical excision followed by histopathologic examination. In a small number of cases, conventional EUS was also carried out for comparison. The examination procedure is identical to that of conventional EUS except 1 hr before each examination, the patient is asked to drink 200 ml of degassed water. Then the endoscope is inserted in the left decubital position and after visualization of the lesion, an adequate amount of degassed water is infused to distend the esophageal or gastric wall and to submerge the lesion under the water level for scanning. Placement of the miniature probe is

Table 4. Correlation between endoscopic ultrasonographic staging and histopathologic diagnosis of esophageal carcinoma (miniprobe scanner). Histopathologic diagnosis EUS staging ep m sm mp a1-a2 a3

ep

m

sm

1

6 3

1 15

—

6/9 67%

mp

1

5 2

15/17 88%

5/7 71%

a1-a2

a3

10 2 10/12 83%

1 —

Overall accuracy: 37/47 (79%); determination of cancer limited in ep-sm: 29/30 (97%). ep: depth invasion confined in intraepithelium (ep) of the esophagus.

made under direct visual control using the optics of the scope and the scanning is started. Staging of Esophageal Cancer by Microtip Transducers Table 3 describes the guidelines for diagnostic criteria of esophageal carcinoma adopted by the Japanese Society for Esophageal Disease in 1992, and Table 4 indicates the accuracy of cancer staging of esophageal carcinoma examined by the Olympus UM2R/3R miniprobe scanner. The highest rate of agreement between EUS staging (miniprobe) and histopathologic examination was obtained in the sm cancer group as 88% of the cases examined. In the m cancer cases in which tumor invasion remained in the mucosal layer, the accuracy of tumor staging was 67%. The overall accuracy of the combined use of 2 probes was 79% (Table 4); the accuracy of discrimination between the superficial or early (ep ⫹ m ⫹ sm) and advanced (pm cancer further, T2) carcinomas, the rate of determination in cancer cases limited in the ep-sm layers, was as high as 97% (see Tables 3 and 4 for definitions of abbreviations). Furthermore, adequate imaging of advanced tumors of considerable size also could be accomplished without difficulty by miniprobe scanners. However, the most important potential of the high frequency miniprobe scanner is the precise characterization of small, early esophageal carcinomas, the T1 stage cancer, and the m and sm cancers with flat or localized polypoid tumors. As demonstrated in Figures 8 and 9, the images of the m and sm cancers by the miniprobe scanners are surprisingly clear and the difficulties in interpreting the tumor staging were rather infrequent compared with conventional EUS, in which compression onto the mucosal surface by the inflated balloon together with lower resolution of the transducers resulted in significant degradation of the mucosal images, particularly for the m cancers, which should be essential findings in diagnosing superficial carcinoma. Figure 10 shows that the sonogram obtained from 1 case with a preoperative EUS diagnosis of a1-a2 tumors happened to be an early carcinoma, sm stage cancer. The hypoechoic solid mass-like lesion was demonstrated in the mid-portion of the esophagus by the miniprobe scans. Histologic examination of the operated specimen revealed a chronic fistula of the esophageal wall and sm stage cancer was found at the same location as identified by EUS examination.

222

World J. Surg. Vol. 24, No. 2, February 2000

Fig. 8. Three cases with superficial esophageal carcinoma, the mucosal cancer, demonstrated by miniprobe scanning. The arrows indicate the small carcinoma confined to the mucosal layer. All cases underwent radical operation and the diagnoses were confirmed histologically. Fig. 9. Four cases with superficial esophageal carcinoma, the submucosal cancer, demonstrated by miniprobe scanning. The arrows indicate respective flat or polypoid tumors. Infiltrations into the echogenic submucosa are demonstrated, however, the muscularis propria (pm) layers are intact. Postoperative histologic examination verified the endoscopic ultrasonographic diagnosis.

Lymph node enlargements were detected in 2 cases of m stage cancer and in 10 cases with sm cancers examined by the miniprobe scanners. Postoperative histologic examination has revealed that ⬍20% of cases of the sm cancers interpreted as lymph node swelling positive were found to have cancer metastasis to lymph nodes, however, the rest of the lymph node swellings were due to lymphadenitis or reactive fibrosis possibly due to concurrent inflammation; none of the m stage cancers had lymph node metastasis, thus indicating the difficulty in predicting metastatic nodes even in the presence of proven lymph node swelling by miniprobe scans. In advanced esophageal carcinoma, on the other hand, characteristic tumor formations and infiltrations accompanied by swollen regional lymph nodes are found (Figs. 11, 12). The lymph nodes invariably exhibit characteristic features of metastatic nodes: increased size, round contour, clear demarcations, and echotexture changes of swollen lymph nodes. However, here again, as is shown in Figure 13, considerable lymph node swelling can occur in the absence of cancer invasion. Hence direct biopsy

under US guidance is often needed to decide the nature of the accompanying lymph node swelling. Discussion Since the introduction of the catheter-type probe for EUS scanning, numerous diagnostic trials have been reported to date especially in the application of the imaging of GI tract malignancies and also pancreatobiliary diseases [15–20]. The definite advantages of the use of microtip transducers in GI tract malignancies may be summarized as the increased resolution of the US image by higher frequency probes; the precision of imaging based on direct visual guidance in steering the scan head to minute lesions; and the thinness of the scan head which allows free access even to areas where conventional EUS is difficult to approach. There are, however, a number of disadvantages, such as the relatively short length of the focal range and the limitation of sound penetration due to rapid attenuation of higher frequency US especially in imaging of carcinoma of an advanced stage. In

Fukuda et al.: EUS of Esophagus

Fig. 10. A case of esophageal fistula complicated by superficial esophageal carcinoma. The endoscopic ultrasonographic diagnosis of this case was advanced esophageal carcinoma of the a3 stage because of the presence of mass-like lesions continuing from the esophagus to the thoracic aorta and the biopsied specimen positive for esophageal carcinoma. Most confusing was the location of the small tumor (T, arrow) right next to the granulation tissue of the chronic fistula (F, arrow).

the latter case, it is necessary to use conventional EUS equipment in parallel or to use the microtip transducer with lower frequencies to cover the nonvisible ranges due to sound attenuation. In the latest models, the sensitivity has been improved substantially so that the use of miniprobe scanners is replacing conventional EUS in a significant percentage of cases examined. From the technical viewpoint, it was found to be adequate to use water ingestion before endoscanning as it had markedly reduced the examination time spent for water infusion and also detachment of adhered mucus. Another technical merit of miniprobe scanning was the precision in the placement of the minute scan head to smaller lesions. To achieve maximal image quality, however, it is essential to place the scan head in the focal zone (within 10 mm) at the perpendicular position over the respective lesions. Furthermore, simultaneous recording of US images on VCR during examination followed by meticulous checking of images right after the examination is mandatory to avoid overlooking subtle changes on the esophageal wall. The combined use of 2 microtip transducers with different US frequencies was expected to give better results compared with the use of a single probe with a fixed frequency. Due to improved imaging of miniprobe scanning, the swollen

223

Fig. 11. Massive infiltration of esophageal carcinoma of the lower esophagus. T shows the thickened, hypoechoic esophageal wall replaced by the tumor tissue. Arrows indicate the swollen regional lymph node by metastasis.

regional lymph nodes frequently found in the vicinity of the T1 tumors in either mucosal or submucosal carcinoma pose difficult differential diagnostic problems. Catalano and others have reported the echo characteristics of metastatic nodes by conventional EUS and the specific features characteristic of metastatic lymph nodes as size larger than 10 mm in diameter, sharp demarcation, rounded shape, and echo-poor texture [21–23]. The present series of esophageal carcinoma examined by the miniprobe system has shown some of the characteristics cited, however, a significant percentage of lymph nodes were found to be tumor cell free after surgery as shown in Figure 13. In some cases, however, distant metastases were found in lymph nodes located in infradiaphragmatic or supraclavicular regions, indicating that thorough examination is needed and that FNA is definitely necessary to decide on the N staging, hence altering the therapeutic regimen to be selected preoperatively. Accuracy in diagnosing early and advanced esophageal cancers by the microtip transducer may appear to be not significantly different from that obtained by current EUS equipment. However, closer analysis of the images obtained from very early mucosal cancer revealed definite superiority using a microtip transducer of higher frequencies compared with conventional EUS images obtained at 7.5–12 MHz frequency ranges. In fact, the precise staging of very early mucosal or superficial gastric cancer as well as esophageal cancer is only possible through detailed

224

World J. Surg. Vol. 24, No. 2, February 2000

Fig. 12. Series of miniprobe sonograms obtained from the lower third of the esophagus. Thick arrows indicate the tumor and periesophageal infiltrations; small arrow indicates lymph node metastasis.

Fig. 13. Lymph node swelling visualized by miniprobe scans. The upper 2 images are from the same pm cancer case and the lowest sonogram is from a case of a1 esophageal cancer. The arrows indicate the swollen lymph nodes. Postoperative histology revealed no cancer invasion and all were reactive lymphadenopathy. Distant lymph node metastasis was found in the celiac axis of the first case.

examination of the layering structures of the mucosa by miniprobe scans. The most important feature of the technique is that every endoscopic examination can be made in combination with miniprobe examinations in a very efficient manner by the endoscopist or surgeon himself to identify suspicious lesions on the spot. The EUS images thus obtained may even be analyzed later by an EUS specialist or by an experienced radiologist if necessary. In Japan and Asia, the majority of esophageal carcinomas are of squamous cell origin, whereas in Western countries the reports on adenocarcinoma complicating Barrett’s epithelium in relation to regurgitation esophagitis are increasing [24 –26]. Due to our limited experience with EUS on this particular category at present [27], the discussion will have to await future developments, however, the use of the miniprobe scanner will extend the possibility of early detection of adenocarcinoma complicating esophageal dysplasia of this particular disease entity. Due to increased detection of early mucosal carcinoma of the esophagus, the mm and sm cases, the number of cases treated by the least invasive method, endoscopic mucosal resection (EMR) or endoscopic mucosectomy, are also rapidly increasing [18, 20, 28]. A 1996 report by Makuuchi [29] of Tokai University Hospital of Japan described that 154 lesions of esophageal carcinoma found in 126 cases during the period from 1988 to 1995 had been treated by esophageal endoscopic mucosal resection (EEMR). The group of patients include 83 cases with m1-m2 cancer, 35 cases with m3-s1 cancer invading the mm layer and inner one third of the sm layer, and 8 cases with sm2-sm3 cancer invading the deeper layers of the sm. Five-year survival rates of the respective groups were 94.5% for m1-m2 cancer (n ⫽ 83), 71.2% for m2-m3 cancer (n ⫽ 35), and 30% for sm2-sm3 cancer (n ⫽ 8); the mean survival rate of the total was 83.2% of 126 cases. Analysis of the causes of the 12 cases of death after EEMR has disclosed only 1

Fukuda et al.: EUS of Esophagus

death due to the recurrence of original esophageal cancer and 11 cases due to malignant tumors of the other organs. Comparison of prognosis on those surgically operated and those treated by EEMR over the last 10 years has disclosed almost equal survival ratios when the disease stages are classified accordingly. These results must await further thorough evaluations on a much larger scale, however, the early detection of esophageal carcinoma followed by an effective staging and adequate treatment selection will undoubtedly bring about a more hopeful outcome to the disease, which has a very desperate prognosis at present.

225

5.

6. 7. 8. 9.

Re´sume´ L’e´choendoscopie est une technique accepte´e dans le staging pre´ope´ratoire des tumeurs digestives sus et sous me´socoliques. C’est dans le cancer œsophagien, en particulier, que l’utilisation de l’e´choendoscopie est conside´re´e comme la me´thode de choix car elle est facile `a re´aliser; l’imagerie endoluminale `a haute re´solution permet une pre´cision dans le staging. Cette revue comple`te couvre les appareils d’e´choendoscopie disponibles actuellement, les caracte´ristiques des images du cancer de l’œsophage et les images obtenues par la tomodensitome´trie sur mini-sonde, re´cemment introduite sur le marche´. On souligne le ro ˆle dans le diagnostic de cancer superficiel de l’œsophage combine´ au traitement potentiel par re´section mucosale endoscopique de cette maladie particulie`re. Resumen La sonografı´a endosco ´pica es una te´cnica de aceptacio ´n general en la estadificacio ´n preoperatoria de los tumores malignos del tracto gastrointestinal superior e inferior. Particularmente en el caso del carcinoma de eso ´fago, la ultrasonografı´a endosco ´pica (USE) es considerada como el me´todo de escogencia, en virtud de la facilidad para realizar el procedimiento y la precisio ´n en la estadificacio ´n por medio de ima´genes ultraso ´nicas de alta resolucio ´n tomadas desde la luz del eso ´fago. Se presenta una revisio ´n comprensiva de los instrumentos de USE actualmente en boga y de las caracterı´sticas del carcinoma esofa´gico, y se presentan ima´genes por el escano ´grafo de minisonda recientemente introducido al mercado. Se discute el rol de este me´todo en el diagno ´stico del carcinoma superficial del eso ´fago, junto con su posible tratamiento mediante la reseccio ´n mucosa endosco ´pica.

10. 11.

12. 13. 14.

15.

16.

17.

18.

19.

20. 21.

References 1. Fukuda, M., Hirata, K., Saito, K., Terada, S., Inaba, M.: On the diagnostic use of echoendoscope in abdominal disease. I. Diagnostic experiences with a new type of echoendoscope on gastric diseases (in Japanese). Proc. Jpn. J. Med. Ultrasonogr. 37:409, 1980 2. Fukuda, M., Nakano, Y., Saito, K., Hirata, K., Terada, S., Urushizaki, I.: Endoscopic ultrasonography in the diagnosis of pancreatic carcinoma. The use of a liquid filled stomach method. Scand. J. Gastroenterol. 19(Suppl. 94):65, 1984 3. Fukuda, M.: Use of the echoendoscope and echolaparoscope in the diagnosis of intraabdominal cancer. Intraluminal scanning. In Ultrasonic Differential Diagnosis of Tumors, G. Kossoff, M. Fukuda, editors, New York, Igakushoin, 1984, pp. 186 –199 4. Fukuda, M.: History and future of endoscopic ultrasonography. Pro-

22. 23. 24. 25. 26. 27.

ceedings of the 11th International Symposium on Endoscopic Ultrasonography, Kyoto, Japan, January 10 –11, 1998, pp. 62– 63 Fukuda, M., Hirata, K.: Endoscopic ultrasonography in gastroenterology. In Current Trends in Digestive Ultrasonography. Frontiers in Gastrointestinal Research, Vol. 24, L. Gandolfi, M. Fukuda, editors, Basel, S. Karger, 1997, pp. 182–202 Classen, M., Strohm, W.D., Kurtz, W.G.: Pancreatic pseudocysts and tumors in endosonography. Scand. J. Gastroenterol. 19 (Suppl. 94):77, 1984 Caletti, G., Bolondi, L., Lago, G.: Ultrasonic endoscopy. The gastrointestinal wall. Scand. J. Gastroenterol. 19 (Suppl. 102):5, 1984 DiMagno, E.P., Buxton, J.L., Regan, P.T., Hattery, R.R., Wilson, D.A., Saures, J.R., Green, P.S.: Ultrasonic endoscopy. Lancet 1:629, 1980 DiMagno, E.P., Regan, P.T., Clain, J.E., James, E.M., Buxton, J.L.: Human endoscopic ultrasonography. Gastroenterology 83:824, 1982 Hisanaga, K., Hisanaga, A., Hibi, N., Nishimura, K., Kambe, T.: High speed rotation scanner for trans-esophageal cross-sectional echocardiography. Am. J. Cardiol. 46:837, 1980 Wild, J.J., Reid, J.M.: Progress in the technique of soft tissue examinations by 15 Mc pulsed ultrasound. In Ultrasound in Biology and Medicine, Washington, D.C., American Institute of Biological Sciences, 1957, pp. 30 – 45 Wiersema, M.J., Kochman, M.H., Cramer, H.M., Lang, C.T., Wiersema, L.M.: Endosonography-guided real-time fine-needle aspiration biopsy. Gastrointest. Endosc. 40:700, 1994 Vilmann, P.: Endoscopic ultrasonography-guided fine-needle aspiration biopsy of lymph nodes. Gastrointest. Endosc. 43:824, 1996 Wegener, M., Adamek, R.J., Wedmann, B., Pfaffenbach, J.B.: Endosonographically guided fine-needle aspiration puncture of paraesophagogastric mass lesions: preliminary results. Endoscopy 26:586, 1994 Hirata, K., Tatsuguchi, H., Mitani, M., Fukuda, M.: Assessment of gastric cancer invasion using a new system of endoscopic ultrasonography, sonoprobe system, SP101 with a miniature probe scanner of 20 MHz. Gastroenterol. Endosc. 35:693, 1993 Fockens, P., van Dullemen, H.M., Tytgat, G.N.J.: Endosonography of stenotic esophageal carcinomas: preliminary experience with an ultrathin, balloon fitted ultrasound probe in 4 patients. Gastrointest. Endosc. 40:226, 1994 Maruta, S., Tsukamoto, Y., Niwa, Y., Goto, H., Hase, S., Yoshikane, H., Hayakawa, T.: Evaluation of upper gastrointestinal tumors with a new endoscopic ultrasound probe. Gastrointest. Endosc. 40:603, 1994 Yoshikane, H., Tsukamoto, Y., Niwa, Y., Goto, H., Hase, S., Shimodaira, M., Maruta, S., Miyata, A., Yoshida, M.: Superficial esophageal carcinoma: evaluation by endoscopic ultrasonography. Am. J. Gastroenterol. 89:702, 1994 Kakegawa, A., Akiyama, H., Iizuka, N., Isono, K., Endo, M., Kakegawa, A., Shirakabe, H., Sugimachi, H., Nabetani, K., Mitsumi, T., Mori, S., Fujimaki, M.: Guidelines for the Clinical and Pathologic Studies on Carcinoma of the Esophagus (in Japanese), edited by Japanese Society for Esophageal Disease, Tokyo, Kanehara Publishers, 1992, pp. 1–93 Yasuda, K.: Endoscopic ultrasonic probes and mucosectomy for early gastic carcinoma. Gastrointest. Endosc. 43:S29, 1996 Murata, Y., Suzuki, S., Hashimoto, H.: Endoscopic ultrasonography in the diagnosis of esophageal carcinoma. Surg. Endosc. 1:11, 1987 Rice, T.W., Boyce, J.A., Sivak, M.V.: Esophageal ultrasound and the preoperative staging of carcinoma of the esophagus. J. Thorac. Cardiovasc. Surg. 101:536, 1991 Catalano, M.F., Sivak, M.V., Rice, T., Gragg, L.A., Dam, J.V.: Endosonographic features predictive of lymph node metastasis. Gastrointest. Endosc. 40:442, 1994 Falk, G.W., Catalano, M.F., Sivak, M.V., Rice, T.W., van Dam, J.: Endosonography in the evaluation of patients with Barrett’s esophagus and high grade dysplasia. Gastrointest. Endosc. 40:207, 1994 Srivastava, A.K., Vanagunas, A., Kamel, P., Cooper, L.: Endoscopic evaluation of Barrett’s esophagus: a preliminary report. Am. J. Gastroenterol. 89:2192, 1994 Riddell, R.H.: Early detection of neoplasia of the esophagus and gastroesophageal junction. Am. J. Gastroenterol. 91:853, 1996 Itatsu, T., Miwa, H., Murai, T., Terai, T., Ohkura, R., Sorimachi, S.,

226 Yang, S.W., Ogihara, T., Watanabe, S., Sato, N., Hirai, S.: Multiple early esophageal cancer arising from Barrett’s esophagus, and a review of cases of early adenocarcinoma in Barrett’s esophagus in Japan. J. Gastroenterol. 32:289, 1997 28. Murata, Y., Suzuki, S., Mitsunaga, A., Izuka, Y., Uchiyama, M., Uchida, K., Nakamura, S., Hayashi, K., Yoshida, K., Toki, F., Ide,

World J. Surg. Vol. 24, No. 2, February 2000 H.: EUS diagnosis and mucosal resection for early esophageal cancer. Proceedings of the 11th International Symposium on Endoscopic Ultrasonography, Kyoto, Japan, January 10 –11, 1998, pp. 31–32 29. Makuuchi, H.: Endoscopic mucosal resection for early esophageal cancer—indications and techniques. Dig. Endosc. 8:175, 1996