This paper gives a summary of more than 40 presentations about digestive endoscopic ultrasonography (EUS) given during the last European Meeting of Gastroenterology. The responsibility of the data only involves the authors. As EUS technology improves, innovative systems and clinical trials will be opening up more and more windows in this field. NEW IMAGING TECHNOLOGIES 3D-Endosonography is one of the potential future developments. The 3D EUS images are reconstructed by the combination of radial and linear images and can be moved on the screen and viewed from any angle. Because of short 3D reconstruction time at the site of the examination, results of 3D-EUS are instantly available to guide treatment decisions. Additional information may be gained by slicing the 3D-block at different angles. The method may be particularly suitable for volumetric measurements in order to assess response to chemo-radiotherapy and to guide treatment decisions. 3D-EUS is also surely useful for diagnosing the depth of early gastric cancer, especially in case for EMR. 3D-EUS seems to be a promising tool in assessment of portal hypertension. 3D-EUS images can indeed easily assess the architecture and the hemodynamic of varicose in case of portal hypertension . Three types of varicose could be classified by 3D images, and the indications for endoscopic treatment either by endoscopic varicose ligature or by endoscopic varicose sclerotherapy could be suggested according to variceal hemodynamics. 3D-EUS is then useful for the selection of the two different endoscopic treatments for esophageal varicose. Another advantage of EUS is that it can distinguish between blood vessels inside and outside the wall of the digestive tract and can compare the changes of varicose extramural collaterals after endoscopic varicose ligature (EVL) and even after endoscopic sclerotherapy (EIS), focusing on the extra gastric veins at the lesser curvature of the cardiac region which have direct inflow to varices and paraesophageal veins. With the analysis of extramural collaterals before and after treatment of esophageal varices, EUS can clarify the difference in the esophagogastric hemodynamics after EVL and EIS ; EIS could shut down effectively the blood-supplying route for varices and reduce the blood flow of collaterals, whereas EVL did not have significant effects on the perigastric varices nor parietal esophageal varices. This may be a reason that explains the easy recurrence after EVL and long-term effectiveness of EIS. EUS texture analysis in pancreatic disease has a diagnostic value. The sensitivity of visual interpretation of EUS images in pancreatic diseases does not exceed 70-80%, especially for discriminating cancer and pancreatitis. For discriminating the normal pancreas from the pathological pancreas (pancreatitis + cancer), texture measurements obtained a sensitivity of 93 %, a specificity of 80 %, and a diagnostic accuracy of 87.7 %. For discriminating pancreatic cancer from non-cancer pancreas, sensitivity was 90.5 %, specificity was 84.6 % and diagnostic accuracy was 86.3 %. INTERVENTIONAL EUS " By progressing from a pure imaging modality, to a modality that provides therapy, the clinical impact of EUS will continue to increase ". Celiac block under EUS guidance is a safe technique, easier than US or CT guidance and it is possible during the same examination to perform, in case of pancreatic cancer, the local staging, an EUS guided biopsy and the treatment of the pain. A randomised trial (celiac block EUS guidance vs CT-Scan) in patients with painful chronic pancreatitis showed better results in the EUS arm than the CT-Scan guidance arm. Cystogastrostomy can be entirely performed under endosonography guidance using a single step device and allows more accurate drainage of the pancreatic pseudocyst or pancreatic abscess without extrinsic compression with a lower risk of perforation and bleeding. EUS helps for guiding and monitoring of endoscopic interventional procedures. High-resolution mini-probe EUS (20 MHz) enables a precise morphometric measurement of thermal lesions by Argon Plasma Coagulation. The depth of the thermal lesions measured by EUS in the esophageal mucosa is exactly correlated to the histological findings and it is possible to define exactly the border between coagulated and normal tissue. EUS controlled thermal ablation improves precision and security of high frequency and Laser therapy. This may be important for the treatment of Barrett's esophagus and the definition of the correct depth of penetration. As regards minimally invasive surgery : early colon cancers (ECC) when restricted to mucosa (m) or submucosa (sm), are usually treated with an endoscopic mucosal resection technique. Because the metastatic rate of sm ECC is of about 13 % and there are no metastases in sm1 tumors, high frequency EUS could differentiate the borderline lesions between sm1 and sm2 . When ECC invade massively the submucosa, surgical resection is indicated. EUS-GUIDED FINE NEEDLE ASPIRATION Endoscopic ultrasound guided fine needle aspiration (EUS-FNA) has increased the diagnostic capability of endosonography with respect to lesions adjacent to gastrointestinal tract. Limitations of this technique are that it is often difficult to obtain adequate tissue for diagnosis. EUS-FNA can be applied to evaluation of mediastinal pathology and is of significant diagnostic value with the ability to alter work-up. This new role of EUS-FNAB may decrease the need of mediastinoscopy or thoracotomy for the diagnosis of idiopathic mediastinal tumors. Detection and evaluation of mediastinal lymph node metastasis in lung cancer is very important. EUS and EUS-FNA are of significant value in patients with node negative on CT. EUS alone is overstaging lymph node metatases and the use of FNA is therefore mandatory. Lung Cancer staging may become a new main indication for EUS and EUS guided FNA. The use of a 21-gauge needle can provide both cytology and core biopsy specimen. A multicenter assessment has shown an overall adequate tissue sampling ratio ranging from 56 to 85 % (mean 80 %), and an accuracy ranged from 38 to 81 % (mean 74 %). Cytology specimens were obtained with a sampling ratio ranging from 0 to 100 % (mean 70 %) and an accuracy ranging 0 to 100 % (mean 70 %). Histology specimens were obtained with a sampling ratio ranging from 56 to 83 % (mean 77 %) and an accuracy ranging from 38 to 75 % (mean 67 %). Sometimes it is difficult to differentiate between pancreatic carcinoma and focal or diffuse pancreatitis. EUS-guided biopsy (EUS-GB) is today the best technique to obtain the histology of a solid pancreatic mass with a sensitivity of 87 %. The EUS-Fine Needle Aspiration Biopsy technique has greatly improved the ability of EUS to differentiate a malignant lesion. EUS-GB had a direct impact on the decision making in around 59 % of cases. The sensitivity, the specificity and the accuracy of EUS-GB for the diagnosis of malignancy has been shown to be respectively 87.9 %, 93.3 % and 92.6 %. It has equally an impact on the therapeutic decision making, notably in case of ADKP not diagnosed by CT-Scan. In case of a stage advanced carcinoma, EUS-GB is being currently important due to the development of pre-operative radio-chemotherapy for pancreatic carcinoma. The reported sensitivity of this method is around 87 %, with a relatively high incidence of non diagnostic reports. The detection of the K-ras gene mutations seems to be a marker of pancreatic adenocarcinoma, able to increase diagnostic accuracy. Unfortunately, the study of molecular analysis of K-ras in cytological smears of pancreatic carcinoma does not appear to improve the efficacy of cytological diagnosis alone. Cytological diagnosis alone had a 100 % specificity and 85.7 % sensitivity in the diagnosis of pancreatic carcinoma, while K-ras mutation was detected only in 25 % of cases with proved carcinoma, with a sensitivity rate of 57 %. The overall complication rate for EUS and EUS-FNAB was 0.22 %, for EUS without FNAB 0.18 %, and for EUS-FNAB 0.52 %. No deaths occurred in relation to EUS and EUS-FNAB. ENDOBILIARY AND ENDOPANCREATIC EUS WITH MINIPROBES Recently developed miniprobes allow intraductal ultrasound (IDUS) scanning of the pancreatic and biliary ducts and the papilla with a very high resolution. IDUS shows the bile duct wall, papilla of Vater, adjoining parts of the cystic and pancreatic ducts and their entrances into the bile duct and adjacent parts of arteries and veins. The site and extension of strictures and tumours, possible tumour involvement of adjacent vessels and tissues and enlarged local lymph glands can be diagnosed. In cholangiocellular carcinoma, very small " skip areas ", which were not visible at cholangioscopy, were identified. IDUS was superior to EUS, CT and MR, and not least in small Klatskin cancers. High-resolution intraductal ultrasound has shown to be very sensitive (92-100 %) in the detection of even small neoplastic lesions in the bile and pancreatic duct with a specifity for malignancy ranging from 50 % to 92 %. IDUS has been shown to have a 100 % accuracy for the detection of invasion of the portal vein and the right hepatic artery. IDUS has a higher sensitivity for the detection of small stones and stone fragments in the bile ducts in comparison to ERCP. IDUS is useful for the diagnosis of localized stenoses of the main pancreatic duct or to determine if pancreatic stones are located in the main duct before pulsed dye-laser lithotripsy. Intraductal ultrasonography is a valuable clinical asset because it gives more important information in many biliary disorders than EUS, CT and MR. GALLBLADDER WALL LESIONS EUS is a promising method especially in differentiating gallbladder cancer from benign lesions. Following EUS findings were considered to be predictive for cancer : wall thickness over or equal to15 mm, irregular surface, destruction of layer structure of the wall, heterogeneous echo pattern, absence of small cystic lesion. When postulated that gallbladder lesions with three or more above findings could be diagnosed as cancer, the sensitivity, specificity, and overall accuracy of the EUS diagnosis of cancer were 77 %, 93 % , and 87 %, respectively. COMPARISONS OF IMAGING METHODS For evaluation of pancreatic adenocarcinoma, EUS and PET have excellent sensitivity and are superior to helical CT. The techniques are complementary: EUS is excellent for local vascular staging for resectability and FNA tissue diagnosis, PET identifies unsuspected metastases and clarifies ambiguous lesions visualized on CT. The value of EUS has been compared to helical CT in diagnosing lymph node metastasis in non-small-cell lung cancer patients according to postoperative histopathological evaluation of dissected lymph nodes. EUS evaluation has shown a sensitivity of 72 %, a specificity of 85 %, a diagnostic accuracy of 80 %, comparing to helical CT- 65 %, 80 %, 80 % respectively. EUS appears then to be an efficient, safe and well tolerated method of staging in patients with non-small-cell lung cancer. EUS AND EARLY CHRONIC PANCREATITIS The diagnosis of early chronic pancreatitis (ECP) remains difficult and EUS features are still controversial. Nevertheless, EUS appears to be even more sensitive than Endoscopic Retrograde Pancreatography (ERP) for the diagnosis of chronic pancreatitis (CP) in early stages. The role of minimal parenchymal changes has been emphasized with this imaging technology. EUS is able to visualize both parenchymal and ductal changes in the course of CP. In patients with initial normal ERP but with EUS-changes, a CP was diagnosed in 53 % of cases in a median follow up of 18 months. In all patients with abnormal ERP (Cambridge 1-3) EUS-changes of CP were visible. A long term follow-up of the patients with ECP diagnosed at EUS has been done. The ECP were classified as group A (possible ECP with either parenchymal signs or ductal signs), as group B (probable ECP with both parenchymal and ductal signs) and as group C (certain ECP with pancreatic stones).The positive predictive value of EUS signs of ECP was 64 % in case of possible or probable ECP and 75 % in case of certain ECP. SUBMUCOSAL TUMORS EUS is the first choice technique for the study of submucosal pathology of the accessible digestive tract, since it allows a more sensitive and specific approach than other currently used techniques. EUS allows differentiation between submucosal tumour and extrinsic compression in more than 95 % of cases. EUS provides information on size, origin and echographic characteristics of the submucosal tumour/extrinsic compression, which is very useful for the selection of treatment. The endoscopic resection (ER) is an alternative for surgical resection by thoracotomy or laparotomy, but indicated only in tumors limited to the submucosal layer. ER should be based on the EUS confirmation. Myogenic tumors of the stomach can be accurately diagnosed by EUS, but differentiation between leiomyosarcoma (LS) and leiomyoma (LM) by EUS has been difficult. The presence of 7 factors on EUS images has been described and they were precisely studied : myogenic tumors detected can be diagnosed according to EUS criteria as leiomyosarcoma with an accuracy rate of 73 % (sensitivity : 100 % and specificity : 64 %). Patients diagnosed as LM by EUS and EUS guided biopsy could be followed without surgery. However, as endosonography cannot replace histological examination, a deep biopsy must be performed if malignancy is suspected. Some authors have suggested to puncture submucosal lesions under endosonographic guidance. Unlike the above-cited sampling techniques, FNA provides only a cytological specimen, which is inadequate for assessing the histological nature of a submucosal tumor. This technique can be helpful if just a particular spot of the lesion is suspected to be malignant. For the other types of submucosal tumors as lipoma, the EUS aspects are typical. The future development of large EUS needle (19 gauges) will increase the sensitivity of EUS guided biopsy in case of submucosal tumor and large gastric folds. CLINICAL IMPACT OF EUS EUS provides an accurate preoperative staging of esophageal carcinoma. EUS has the best accuracy not just in staging, but also in changing treatment management. Because patients staged T4 by EUS have a very poor prognosis, regardless of further therapy, EUS permits to exclude from surgery patients (T4) whose survival is not influenced by this treatment. The presence of lymph node metastasis (N), and of an obstructive stenosis significantly influenced the survival (p < 0.05 for both), on the contrary to parietal extension (T) and to UICC classification. EUS has a role in therapeutic decision making for esophageal carcinoma. It can enlarge radiotherapy fields to enclose the lymph nodes not seen by CT. EUS has changed stages in 75 % of cases and therapeutic plans in around 39 % of cases. After radiotherapy and chemotherapy, knowing the residual tumor volume and its distal extension in the esophageal wall by EUS, it could be possible to localize better the isodoses of HDR-BRT and personalize fractions and doses. In assessing submucosal tumors, EUS can replace 2/3 of other diagnostic tests, and among those also potentially hazardous procedures such as large particle biopsies in extraluminal compression (ELC) caused by vessels or normal organs. EUS is also cost-effective in the diagnosis of SML. In the histologically confirmed cases EUS had a sens/spec. of 92 %/100 % in the differential diagnosis SMT vs. ELC and of 64 %/80 % in the differential diagnosis malignant/ benign SMT. EUS reduced the rate of further diagnostic tests from 93 % to 47 %, whereby 2 tests were replaced in 38 % and 1 test in 72 %. 97 % of all CTs, 70 % of all endoscopic controls and 64 % of all large particle biopsies were replaced by EUS. EUS is a high accurate method for detection of benign bile ducts abnormalities, especially CBD stones. It is particularly important for the patients with a suspicion of choledocholithiasis, but with negative findings of conventional US. Exactly in that group of patients EUS allowed to verify diagnose with appropriate accuracy and reasonably refused ERCP in almost half of these patients. Total accuracy of EUS in diagnosis of all benign abnormalities of CBD was 93.7 %, the sensitivity 90.2 % and specificity 97.4 %. Accuracy of EUS in diagnosis of CBD stones and papillary stenosis was 96.9 % and 88.7 % respectively. EUS is a valuable test in the perioperative evaluation of cholecystectomy patients. EUS can play a role in assessing pancreatobiliary disease before and after cholecystectomy. In patients undergoing EUS prior to cholecystectomy, EUS frequently ./imagesd information that either resulted in the preoperative detection of CBD stones or an indication to proceed to cholecystectomy. EUS was also very helpful in ./imagese evaluation of postcholecystectomy patients, usually avoiding the need for more invasive tests in this group of patients. Additionally, EUS provided useful information in cases where ERCP was technically unsuccessful.