What can be expected from the new 12 and 20 MHz miniprobes in visualizing these additional echo-layers?
R. Lambert (Lyons)
The instrumentation adapted to the endosonographic study of the digestive wall still requires more evolution in biomedical engineering.
The first variable element is the size of the instrument. The diameter of the conventionnal echoendoscope ranges from 9 to 12 mm; in this instrument the endoscopic vision is not separated from the ultrasonic probe. In the most recent models the optical fiber bundle is replaced by a chip camera. The instrument has an operative channel allowing the passage of a needle catheter a biopsy forceps, as well as the instillation of water for the acoustic transmission. The adhesion of the balloon to the esophageal wall is effective and homogeneous, but a distention (and thinning) of the esophageal layers occurs. Thin or mini probes [1-7] are in the range from 3 to 5 mm in diameter. The probe, dissociated from the endoscope,is introduced through the operative channel of the instrument. An advantage of the thin probe is that exploration does not requires the distention of the esophageal lumen; therefore the esophageal layers are well delineated. On the other hand the acoustic transmission is more unstable, the probe must be placed in contact with the mucosa and the adaptation of a balloon is not easy.
The second variable element is the ultrasound scanning system. Radial instruments, require a mechanical drive of the transducer (an element of fragility) and the needle aspiration for cytology is not possible with efficacy. On the other hand, they offer more facilities in diagnosis with the transverse section of the esophagus and surrounding structures. Sectorial or linear instruments require no mechanical drive and are more resistant. The ultrasound processor is compatible with other sectorial probes (percutaneous exploration) and with the doppler. The needle aspiration for cytology is adequately performed. On the other hand the interpretation of the sectorial image is not as easy as that of the radial scan. Ultrasound processors for radial scan and sectorial linear array are different. Both techniques being useful, there is a need for an instrumentation with the double potential. This has been proposed with a conventionnal echoendoscope using the same processor with two different tubes having different mechanical drives. It has also been proposed with the thin probe and a switch for either the radial or linear array system.
The third variable element is the frequency of the transducer. Low frequency transducer operate at 7.5 to 12 MHz. They produce a 5 layer diagram of the esophageal wall. The hyperechoic layers 1, 3 and 5 correspond respectively to the mucosal frontier with the balloon, the submucosa and the adventicia; low frequency transducers are usually placed in conventional echoendoscopes. The high frequency transducers operate at 20 or 25 MHz, and eventually higher frequencies. They are usually placed in miniprobes. With an increased frequency the depth of penetration is less, while there is an increased resolution and a multiplication of layers in the digestive wall. The classical diagram is showing 9 layers. The 5th one is hyperechoic and corresponds to the submucosa, while the 4th one hypoechoic corresponds to the muscularis mucosae. However, the actual number of layers is still questioned and 11 layer diagrams are also proposed. There is still uncertainty in the respective attribution of each layer to a precise anatomical structure. The identification of the muscularis mucosae is relatively reliable in the esophagus where the anatomical structure is thick, while it is seldom identified in the stomach. The high frequency thin probes should be adapted to the high resolution analysis of the mucosa and submucosa in the esophagus, i.e. the site of the early neoplastic lesions. However, lower frequencies are required for the complete staging of tumors.
Finally, the resolution of the endosonographic probe does not depend only on the frequency of the transducer; the design of the processor is just as important. It is not clear whether an improved separation of the anatomical structures in the esophageal wall shall result from progress in the processor or from the use of higher frequencies.
In the application of endosonography to the esophagus the anatomy of the superficial layers of the esophageal wall must be recalled. The normal lining is a squamous epithelium, the muscularis mucosae is very thick. Squamous cell cancer is called '"arly" when invasion is limited to the mucosa: m1 cancer or severe dysplasia is intraepithelial, in m2 and m3 cancer (micro invasive) the muscularis mucosae is invaded. Submucosal cancer (sm1 to sm3) is considered as an advanced cancer. In the columnar lined esophagus, the metaplastic mucosa is the healing process of peptic esophagitis; the muscularis mucosae may be partly destroyed. The distinction is limited to the intra mucosal or submucosal stage of glandular neoplasia in thickened digestive layers more or less altered by inflammation.
Thin endosonography probes, adapted to the exploration of the early stages of neoplasia, have been designed by Aloka, Fujinon and Olympus. A study of efficacy, in balance with that of conventionnal instruments, has been conducted on various models including the esophageal wall modified by peptic esophagitis or motility disorders such as achalasia [8-11]. Of course the objective is whether the thin probe may detect occult dysplasia in a normal mucosa at endoscopy, and whether the distinction between dysplasia and cancer is safe.
An increased thickness of the esophageal wall is a consequence of reflux. This has been observed with the radial scan in reflux esophagitis and in Barrett's esophagus as well. The normal thickness of the esophageal wall being 2.6 mm, it increases at the gastroesophageal junction to 3.64 mm in peptic esophagitis and to 7 mm in the Barrett's; but the difference is much less when the probe is more proximal in the esophagus. Is this thickening helpful, in the separation of simple metaplasia from dysplasia or high grade dysplasia from cancer? The response is no, and the evaluation is not considered as cost effective with the conventional echoendoscope [12-15]. No further progress is expected from high frequency thin probes in the sonographic detection of dysplasia, on the other hand an improved reliability in the separation of intramucosal from submucosal cancer is progressively accepted.
In conclusion, just as with a conventional echoendoscope, there is no use of the thin probes in the early diagnosis of dysplasia or intramucosal cancer in the esophagus. On the other hand, high frequency thin probes should be preferred in the staging of superficial cancer and imaging of the intramucosal neoplasia as an hypoechoic thickening. The instrumentation is still in development; its reliability in separating intramucosal neoplasia (m1 to m3) from submucosal neoplasia (sm1) should improve in the near future. The tendency to endoscopic curative treatment of severe dysplasia or intramucosal cancer in the esophagus justifies the clinical application of thin probes .
6. Yanai H, Fujimura H, Suzumi M. Delineation of the gastric muscularis mucosa and assessment of depth of invasion of early gastric cancer using a 20 megahertz endoscopic ultrasound probe. Gastrointest Endosc 1993;39,505-512.
7. Yanai H, Yoshida T, Harada T, et al. Endoscopic ultrasonography of superficial esophageal cancers using a thin ultrasound probe system equipped with switchable radial and linear scanning modes. Gastrointest Endosc 1996;44:578-582.