Search

Background: Endoscopic ultrasound (EUS) has been shown to be a reliable tool for staging rectal cancer. Nevertheless, the accuracy of EUS after chemoradiation remains unclear; therefore the purpose of the present paper was to compare the accuracy of EUS staging for rectal cancer before and following chemoradiation., Methods: Patients with rectal cancer undergoing EUS staging were stratified into two groups. Group I consisted of 66 patients who underwent surgery following EUS staging without preoperative chemoradiation. Group II consisted of 25 patients who had EUS evaluation following chemoradiation. The EUS staging was compared to surgical/pathological staging., Results: The accuracy of the T staging for group I was 86% (57/66). Inaccurate staging was mainly associated with overstaging EUS T2 tumors. The accuracy of the N staging for group I was 71% (47/66). The accuracy of EUS for a composite T and N staging relevant to treatment decisions in group I was 91%. In group II, the accuracy of T and N staging was 72% (18/25) and 80% (20/25), respectively. Overstaging EUS T3 tumors accounted for most inaccurate staging. The EUS staging predicted post-chemoradiation T0N0 stage correctly in only 50% of cases., Conclusions: Preoperative staging of rectal cancer by EUS is a useful modality in determining the need for preoperative chemoradiation. The EUS T staging following chemoradiation appears to be less accurate. Detection of complete response may be insufficient for selecting patients for limited surgical intervention.

Objectives: Two APC germline mutations, E1317Q and I1307K, have been linked to colorectal cancer (CRC) risk. Whereas the I1307K variant is almost exclusively encountered in (Ashkenazi) Jews, E1317Q is not restricted to certain ethnic populations. Data on its contribution to CRC risk in Jewish patients are sparse., Aims: To assess the contribution of E1317Q to CRC development in the Jewish population., Methods: A total of 538 consecutive Israeli Jewish CRC patients and 440 controls were genotyped for E1317Q. In addition, the rate of the I1307K APC missense mutation and the two predominant Jewish mutations in hMSH2, A636P, and 324delCA, associated with hereditary nonpolyposis colon cancer (HNPCC), were determined., Results: The E13117Q missense mutation was detected in 6/538 (1%) of CRC patients and 5/440 (1%) of controls. The I1307K variant was found in 8% of all patients and in 11% (35/322) of patients of Ashkenazi Jewish descent. Carriers and noncarrier CRC patients did not differ in age of onset or associated colonic adenomatous polyps. The carrier rate among controls was 5% among Ashkenazim and 1.6% among non-Ashkenazi individuals. The 324delCA hMSH2 mutation was not observed in this cohort, and 4 of 322 Ashkenazi patients (1.2%) displayed the A636P mutation., Conclusion: In Jewish CRC patients the E1317Q variant plays little if any role in colorectal cancer susceptibility and genetic testing for this variant is not warranted. The I1307K mutation is associated with a moderate excess risk for CRC, but age of onset seems not to be earlier and this variant is not associated with a multiple colonic polyp phenotype. Founder mutations in hMSH2 are rare in consecutive CRC patients.

Purpose: The aims of this study were to compare the performance of a mechanical radial endosonoscope and an endorectal electronic probe and to evaluate the value of hydrogen peroxide (H(2)O(2)) injection in the diagnosis of perianal fistulas., Patients and Methods: Sixty-one patients underwent clinical and endosonographic studies for suspected perianal fistulas or abscesses. Endosonography was performed using two instruments: a mechanical radial endosonoscope (Olympus GF-UM20) and an electronic endorectal probe (Hitachi EUP-R53). The patients were re-examined during and following H(2)O(2) injection using both systems., Results: Thirty-seven fistulous tracts were visualized with the electronic endorectal probe versus only 9 with the mechanical radial endosonoscope. Four patients had anal stenosis, precluding the use of the larger electronic probe. Three fistulas were detected in these patients using the mechanical radial endosonoscope. H(2)O(2) injection was not feasible in 26 patients (43%). Visualization of 11 (31%) fistulas improved after administration of H(2)O(2). Six fistulous tracts not detected before H(2)O(2) administration were clearly visualized during injection and for several minutes thereafter., Conclusions: Work-up of perianal fistula should be preferably performed using the endorectal electronic probe. However, the mechanical radial endoscope is preferred in patients with anal stricture. H(2)O(2) injection should become an integral part of every sonographic study of perianal fistulas., ((c) 2005 Wiley Periodicals, Inc. J Clin Ultrasound 33:226-232, 2005.)

Objectives: Crohn's disease (CD) has been reported to be more frequent among non-Ashkenazi Jewish patients suffering from familial Mediterranean fever (FMF). Interestingly, functional similarities between the CD susceptibility gene (NOD2/CARD15) and the FMF gene (MEFV) have been described: both belong to the death domain containing protein family, important in the regulation of apoptosis, cytokine processing and inflammation., Aims: To investigate the prevalence of MEFV mutations in Jewish non-Ashkenazi CD patients and its putative effect on CD presentation., Methods: Germline DNA of 105 Israeli CD patients of non-Ashkenazi and mixed Ashkenazi-non-Ashkenazi ethnic background was analyzed for three most common MEFV mutations: M694V, V726A, and E148Q. Five patients (4.7%) with a clinical diagnosis of FMF were included. Data obtained from each patient included: age of onset, disease location, and behavior, the presence of extraintestinal manifestations of CD and therapeutic regimens., Results: The overall prevalence of mutation carriers among non-FMF-CD patients was 13% (13/100). A stricturing disease pattern was observed in 56% (10/18) of all carriers, FMF-CD, and non-FMF-CD patients, and in 25% (22/87) of noncarriers (OR: 3.7, 95% CI: 1.3-10.5, p= 0.015). The prevalence of fistulas was comparable in both groups. Extraintestinal manifestations were significantly more frequent among carriers than noncarriers (65%vs 32%, OR 3.9, 95% CI = 1.3-11.5, p= 0.015). No differences were observed in disease location and disease severity., Conclusions: MEFV mutations are not associated with CD susceptibility, yet the presence of these mutations appears to be associated with a stricturing disease pattern and extraintestinal disease manifestations of CD.

Background: Relapse after an initial response to infliximab therapy poses a problem for maintenance treatment., Aim: To assess the effects of increasing the infliximab dosage in Crohn's disease (CD) patients who initially responded but flared during maintenance therapy., Methods: This was an observational study. Twelve CD patients with both inflammatory and fistulizing manifestations were included. All patients initially responded to 5 mg/kg of infliximab, relapsed during maintenance therapy, and were treated with 10 mg/kg. The Harvey-Bradshaw index, the fistula activity, and steroid use were assessed before and after treatment with the increased dose of infliximab., Results: The mean Harvey-Bradshaw index score after flare-up during treatment with 5 mg/kg of infliximab was 13.5+/-3.7. Treatment with 10 mg/kg, in a mean of 3.3 infusions, decreased the activity score to a mean of 8.8+/-2.5. Two patients were weaned off prednisone, and a reduced dose was possible in the other steroid-treated patients., Conclusions: Increasing the infliximab dose may be beneficial in CD patients who initially responded to therapy, but relapsed during maintenance with the lower dosage., (Copyright (c) 2005 S. Karger AG, Basel.)

Objective: The main objective is to study whether the month of birth is associated with the development of Crohn's disease (CD) in the Israeli Jewish population., Background: It was suggested that perinatal exposure to infectious agents may have a role in the pathogenesis of CD. Due to the seasonal nature of some infections, a linkage between birth dates and a risk to develop CD would support such a hypothesis. Previous studies that addressed this question were conducted in Europe and differed in their findings., Methods: Birth dates of 844 Jewish ulcerative colitis (UC) and CD patients from three medical centers representing the north, central, and the south of Israel were compared with the monthly rates of birth during the same period of time. The standard incidence ratio was used to define the risk to develop either disease according to the month of birth. The Score method was used for the evaluation of seasonality trends., Results: Birth during the winter period in Israel was associated with increased risk to develop CD, whereas birth during the spring was associated with a reduced risk. The Score method for seasonality showed a significant peak during winter time in these patients (z = 2.02, p= 0.021). No such seasonal variation was noted for UC patients., Conclusions: A seasonal pattern was observed in the risk to develop CD but not UC. The findings may support the involvement of environmental factors in the pathogenesis of CD.

Background & Aims: Allicin, the active substance of fresh crushed garlic has different biological activities and was implicated as an anti-inflammatory agent. Epithelial cells have an important role in intestinal inflammation. The aim of this study was to assess the immunomodulatory effect of allicin on intestinal epithelial cells., Methods: The spontaneous and TNF-alpha-stimulated secretion of IL-1beta, IL-8, IP-10 and MIG from HT-29 and Caco-2 cells was tested with, or without pretreatment with allicin. Cytokine secretion was assessed using ELISA and expression of mRNA was determined by an RNA protection assay., Results: Allicin markedly inhibited the spontaneous and TNF-alpha -induced secretion of IL-1beta, IL-8, IP-10 and MIG from the two different cell lines in a dose-dependent manner and suppressed the expression of IL-8 and IL-1beta mRNA levels. In addition, allicin suppressed the degradation of IkappaB. No effect on cell viability was noted., Conclusions: These observations indicate that allicin exerts an inhibitory immunomodulatory effect on intestinal epithelial cells and suggest that allicin may have the potential to attenuate intestinal inflammation.

Background: Screening for celiac disease is based on the sequential evaluation of serologic tests and intestinal biopsy; an optimal screening protocol is still under investigation. The screening policy of one of the main healthcare providers in Israel (Maccabi) consists of measuring total immunoglobulin A and tissue transglutaminase IgA antibodies and confirming positive results by endomysial antibodies. For IgA-deficient patients antigliadin IgG is measured., Objectives: To evaluate the use of tTGA as a first-level screening test in patients suspected of having celiac disease, Methods: The results of tTGA and EMA tests over a 3 month period were obtained from the laboratory computer. Letters were sent to the referring physicians of patients with positive tests, requesting clinical information and small intestinal biopsy results. tTGA was performed using an anti-guinea pig tTG-IgA enzyme-linked immunosorbent assay kit., Results: Overall, 2,505 tTGA tests were performed: 216 (8.6%) were tTGA-positive of which 162 (75%) were EMA-negative (group 1) and 54 (25%) EMA-positive (group 2.) Clinical information was obtained for 91 patients in group 1 and 32 in group 2. Small intestinal biopsy was performed in 33 (36%) and 27 patients (84%) in groups 1 and 2, respectively. Celiac disease was diagnosed in 4 biopsies (12%) in group 1 and 23 (85%) in group 2 (P < 0.0001). The positive predictive value was 45% for tTGA and 85% for EMA., Conclusions: Symptomatic patients with positive tTGA and negative EMA have a low rate of celiac disease compared to those who are tTGA-positive and EMA-positive. Confirmation with EMA is advised when tTGA is performed as a first-level screening for suspected celiac disease.

Ulcerative colitis (UC) and Crohn's disease (CD) are heterogeneous disorders characterized by chronic intestinal inflammation. Genetic predisposition is a major risk factor in both diseases. The CARD15 (NOD2) gene has been implied as a candidate gene in the pathogenesis CD. Our aim was to delineate the frequency of three missense and one frameshift variant of CARD15 in Israeli Jewish CD and UC patients. DNA was extracted from blood samples from 238 unrelated inflammatory bowel disease (IBD) patients, 68 with UC and 170 with CD. The DNA was genotyped for two missense mutations, R675W and G881R, and one frameshift mutation, 980FS981X. Mutations in CARD15 were observed with significantly greater frequency in CD patients (46/170, 27%) than in UC patients (7/68, 10%) (P = 0.005). Homozygous and compound heterozygous carriers were restricted to seven (4%) patients with CD as compared to none of the UC patients (P = 0.01). Similar rates in Ashkenazi and non-Ashkenazi Jewish patients were observed. Age-of-onset of disease was lower in Ashkenazi mutation carriers as compared to non-carriers of Ashkenazi origin (18.7 +/- 8.6 years vs. 25.8 +/- 13.4 years, respectively, P = 0.03). No other phenotypic characteristics could distinguish mutation carriers from non-carriers. We conclude that germline mutations in the CARD15 gene are more frequently found in CD than UC patients and appear to predict an earlier age-of-onset in Ashkenazi Jewish patients. No association could be demonstrated between CARD15 mutations and specific disease course or behavior., (Copyright 2003 Wiley-Liss, Inc.)

Background: Crohn's disease (CD) is a chronic transmural inflammatory disease of unknown etiology. Genetic background may influence the nature of disease. The Jewish population in Israel is composed of Ashkenazi and Sephardic Jews, who differ in their genetic background. Previous studies have reported that the prevalence of CD in Sephardic Jews is lower than in Ashkenazi Jews; however, no information is available regarding disease characteristics in these patient populations., Goals: To assess the demographic and clinical characteristics of CD in Israeli Jewish patients., Study: We studied 189 CD patients who were observed in our department. Demographic and clinical data were collected and analyzed., Results: Forty-eight percent of the patients had ileum-only disease. Fistulizing disease was more frequently seen in ileocolonic than in ileal ( p= 0.04) or colonic ( p= 0.01) disease, whereas strictures occurred more often in ileal than in colonic disease (22% and 3%, respectively, p= 0.007). No association was found between current smoking and disease severity. Sephardic patients more frequently had extraintestinal disease than did Ashkenazi patients (35% vs. 17%, respectively, = 0.019)., Conclusion: Jewish Israeli patients show a predilection for ileum-only disease. Disease severity appears not to be influenced by smoking. These differences in disease behavior may be related to genetic factors.

It is unknown which factors determine the severity of mucosal damage in gastroesophageal reflux disease (GERD). Our aim was to test whether the amount of esophageal acid exposure could predict the severity of esophageal injury in erosive reflux esophagitis. A total of 644 outpatients with symptomatic GERD underwent an esophagogastroduodenoscopy followed by esophageal manometry and 24-h pH monitoring. GERD was graded according to the endoscopic severity of mucosal damage as no erosions, single erosions, confluent erosions, esophageal ulcers, and strictures. A multiple linear regression was used to assess the joint influences of demographic characteristics, social habits, endoscopic anatomy, and various parameters of esophageal function tests on the severity of erosive reflux disease. No clear-cut association between the amount of acid reflux and the severity of erosive reflux esophagitis could be established. All individual parameters of esophageal pH monitoring, such as upright or supine acid contact time, frequency of all or only long reflux episodes, and an overall summary score of pH-metry, revealed no or only a weak correlation with the severity grade of erosive reflux esophagitis. Similarly, the pressure of the lower esophageal sphincter was only slightly more decreased in patients with extensive erosive esophagitis as compared to subjects without esophageal erosions. In the multiple linear regression, the presence of hiatus hernia was a stronger predictor of disease severity than any of the other parameters. In conclusion, factors other than exposure of the esophageal mucosa to acid must contribute to the development of erosive esophagitis.

Objective: The fear that colorectal adenomas were missed on initial colonoscopy or that new adenomas have developed is often a rationale for repeating a colonoscopic examination. The aim of this study was to delineate risk factors associated with recurrence of colorectal adenomas after an initial baseline screening colonoscopy., Methods: The study population comprised 875 subjects who underwent a baseline screening colonoscopy followed by a second examination 1-5 yr later. Multiple logistic regression was used to assess the influence of potential risk factors on the occurrence or recurrence of colorectal adenomas, the strength of the influence being expressed as an OR with a 95% CI., Results: Colorectal adenomas were detected in 484 of all patients (55%) at baseline colonoscopy. Within a 1- to 5-yr time interval, 181 patients (37%) had recurrent adenomas (adenomas were removed during the first colonoscopy) and 73 patients (19%) had newly developed adenomas (adenomas were absent on the first colonoscopy). The occurrence of adenomas at baseline screening colonoscopy was the only factor associated with an increased risk for the recurrence of adenomas at follow-up (OR = 2.51, 95% CI = 1.77-3.55). Recurrence was associated with multiple baseline adenomas (4.45, 2.98-6.64) and baseline adenomas larger than 1 cm (2.62, 1.99-3.11). Recurrence was not associated with histology type or family history of colorectal cancer. There was a significant trend for adenomas to recur in the same proximal or distal segment as the baseline adenomas (p = 0.02)., Conclusions: Colon adenomas tend to recur with greater frequency if the adenomas removed at baseline were either large or multiple. Although patients with large adenomas or multiple adenomas at baseline screening colonoscopy are at a 2.6- to 4.5-fold risk for recurrence of adenomas, the rate of de novo adenoma formation in patients without baseline adenomas may be large enough to warrant repeat colonoscopy at some time in the future. The exact timing of the follow-up colonoscopy needs to be determined.

One third of the general population may experience reflux symptoms, yet only a small fraction of patients with gastroesophageal reflux disease (GERD) have Barrett's esophagus. The aim of the present study was to compare the characteristics of GERD patients with and without Barrett's esophagus and identify potential risk factors for the appearance of Barrett's esophagus in reflux disease. Outpatients from a gastroenterology clinic who underwent upper gastrointestinal endoscopy, esophageal manometry, and 24-hr pH monitoring were recruited into a case-control study. A total of 256 case subjects with endoscopically and histologically proven Barrett's esophagus were compared to a control group of 229 subjects with nonerosive reflux disease. As compared to nonerosive reflux disease, Barrett's esophagus was strongly associated with more reflux episodes. Barrett's esophagus occurred more frequently among subjects with hiatus hernia and among subjects who consumed large amounts of alcohol or cigarettes. Frequent reflux episodes, hiatus hernia, smoking, and alcohol consumption were also risk factors for an increased length of Barrett's mucosa. Total esophageal mucosal acid contact time at pH < 4 was a significant risk factor for the length but not the presence of Barrett's esophagus. Intake of aspirin or NSAIDs was similar in patients with and without Barrett's esophagus. In conclusion, in comparison with nonerosive reflux disease, Barrett's esophagus is characterized by risk factors usually indicative of severe types of GERD. Mechanisms in addition to acid reflux must contribute to the development of Barrett's esophagus.

Objectives: Gallstones and hiatal hernia reportedly have been linked to similar dietary factors prevalent in western countries, and patients with cholelithiasis or previous cholecystectomy have been reported to have more duodenogastric reflux than healthy controls. Nonetheless, the contribution of duodenogastric reflux to the development of gastroesophageal reflux disease (GERD) remains controversial. The present study was aimed to assess the association between gallstone disease and GERD., Methods: Outpatients from general medical clinics who underwent upper GI endoscopy and abdominal ultrasonography were recruited into a case-control study. A case population of 790 patients with various grades of GERD was compared to a control population of 407 patients without GERD. In a multivariate logistic regression, the presence of GERD served as the outcome variable, whereas the presence of gallstones, hiatal hernia, social habits, and demographic characteristics served as predictor variables., Results: No associations were found between the presence of cholelithiasis or previous cholecystectomy and GERD or between the presence of cholelithiasis or previous cholecystectomy and hiatal hernia. The severity of GERD also remained unaffected by the presence of gallstones. The occurrence of GERD was influenced only by hiatal hernia (odds ratio [OR] = 3.15, 95% CI = 2.44-4.08), alcohol consumption (OR = 1.47, CI = 1.08-1.99), and not by cholelithiasis (OR = 1.02, CI = 0.68-1.51), or cholecystectomy (OR = 0.90, CI = 0.64-1.28). The frequency of GERD among hiatus hernia patients with gallstones (437/592 = 74%) was similar to the frequency of GERD among hiatus hernia patients without gallstones (168/220 = 76%, p = 0.516)., Conclusions: Neither cholelithiasis nor cholecystectomy poses a risk for the occurrence of GERD or hiatal hernia. Gallstone disease does not seem to influence the integrity of the esophageal mucosa through GERD.

Background: Current treatment for the eradication of Helicobacter pylori in patients with peptic disease is based on the combination of antibiotic and anti-acid regimens. Multiple combinations have been investigated, however no consensus has been reached regarding the optimal duration and medications., Objectives: To assess the efficacy of two treatment regimens in patients with peptic ulcer disease and non-ulcer dyspepsia, and to determine the need for gastric mucosal culture in patients failing previous treatment., Methods: Ninety patients with established peptic ulcer and NUD (with previously proven ulcer) were randomly assigned to receive either bismuth-subcitrate, amoxycillin and metrnidazole (BAM) or lansoprasole, clarithromycine and metronidazole (LCM) for 7 days. Patients with active peptic disease were treated with ranitidine 300 mg/day for an additional month., Results: Eradication failed in 8 of the 42 patients in the BAM group and in 2 of the 43 patients in the LCM group, as determined by the 13C urea breath test or rapid urease test (19% vs. 5%, respectively, P = 0.05). Five of these 10 patients were randomly assigned to treatment with lansoprazole, amoxycillin and clarithromycin (LAC) regardless of the culture obtained, and the other 5 patients were assigned to treatment with lansoprazole and two antibacterial agents chosen according to a susceptibility test. Eradication of H. pylori was confirmed by the 13C urea breath test. The same protocol (LAC) was used in all patients in the first group and in four of the five patients in the second group. The culture results did not influence the treatment protocol employed., Conclusions: Combination therapy based on proton pump inhibitor and two antibiotics is superior to bismuth-based therapy for one week. Gastric-mucosal culture testing for sensitivity of H. pylori to antibiotics is probably unnecessary before the initiation of therapy for patients with eradication failure.

Objectives: It is presently not fully understood which risk factors contribute to the occurrence of reflux esophagitis and how such factors might influence the severity of the disease. The aim of this study was to delineate the clinical epidemiology of erosive reflux esophagitis., Methods: Outpatients from a medicine and gastroenterology clinic who underwent upper GI endoscopy were recruited into a case-control study. A total of 1,533 patients with and 3,428 patients without endoscopically diagnosed reflux esophagitis were categorized as case and control subjects, respectively. Using multivariate logistic regressions for statistical analysis, the presence of esophageal erosions, ulcers or strictures, served as three separate outcome variables. Demographic characteristics, intake of nonsteroidal anti-inflammatory drugs (NSAIDs), consumption of alcohol and cigarettes, and the presence of hiatus hernia or peptic ulcer served as predictor variables., Results: Erosive reflux esophagitis tended to occur more frequently in Caucasian male patients. Hiatus hernia was associated with a strong risk for developing esophageal erosions, ulcers, and strictures. Although statistical significance was demonstrated only for esophageal erosions, in all grades of reflux esophagitis alike, gastric and duodenal ulcer exerted a protective influence. Consumption of NSAIDs increased the risk for esophageal ulcers only. Smoking and alcohol were not associated with an increased risk of developing any type of erosive reflux esophagitis., Conclusions: The results stress the critical role played by hiatus hernia in all grades of erosive reflux esophagitis. NSAIDs may act through a mechanism of topically induced esophageal injury. Our data also suggest that the presence of either gastric or duodenal ulcer exerts a protective influence against the development of reflux disease.

Background: About one-third of patients with severe ulcerative colitis do not respond to conventional therapy and require urgent colectomy. It was recently shown that cyclosporin is effective in some of these patients., Objectives: To review the current experience of six hospitals in central Israel that used cyclosporin in patients with severe ulcerative colitis., Methods: The files of all 32 patients treated with cyclosporin for corticosteroid-resistant ulcerative colitis were reviewed. Activity of disease was measured by a clinical activity, index colonoscopy and laboratory tests., Results: The average duration of treatment with intravenous cyclosporin was 12.7 days (range 9-28) after which the disease activity index dropped from an average of 14.22 to 4.74. The mean time for response was 7.5 days (4-14). Twelve patients (40%) required surgery within 6 months and another 6 patients (18.8%) were operated on after more than 6 months. Twelve patients (37%) maintained remission for at least 6 months and did not require surgery. In one patient treatment was stopped because of non-compliance and one was lost to follow-up. There were numerous side effects, but in only one case with neurotoxicity was treatment withdrawn., Conclusions: Cyclosporin is a relatively safe and effective treatment for severe ulcerative colitis. It induced long-term remission in 37% of the patients, and in those who required surgery the treatment resulted in an improved clinical condition before the operation.

Resistance to antibiotics is considered the main reason for failure to eradicate Helicobacter pylori (HP). Resistance rates are different in developed and developing countries and are not known for Israel. We studied HP resistance rates in 40 patients who underwent esophagoduodenoscopy for various indications and were found to have gastric HP colonies. Sensitivity was determined by E-test, using clarythromycin, amoxycillin, clindamycin, erythromycin and metronidazole. The resistance rate for metronidazole was up to 67% but that for clindamycin was only 10%. HP was very sensitive to both macrolide antibiotics, erythromycin and clarythromycin.

The major prerequisite for resection of a pancreatic tumor is non-involvement of large blood vessels. Preoperative assessment of blood vessel infiltration may prevent unnecessary surgery. The aim of our study was to investigate the accuracy of endoscopic ultrasonography (EUS) in diagnosis of pancreatic cancer and in preoperative staging. Thirteen patients (7 females, 6 males; mean age 64 years) with a pancreatic tumor, but no evidence of distant metastases, underwent EUS and computerized tomography (CT) in order to assess blood vessel involvement by the tumor. The results were compared with intraoperative findings in 12 patients and with postmortem findings in 1 patient. A tumor was demonstrated by EUS in 12 patients and was confirmed at surgery in all 12 patients. In one patient no tumor was demonstrated by EUS, although a tumor was visible by CT; no tumor was found at surgery. In two patients CT failed to demonstrate a pancreatic tumor that was demonstrated by EUS; at surgery a tumor was detected in both patients. EUS detected blood vessel involvement in seven patients, which was confirmed at surgery in six of them. In the other six patients surgery confirmed the EUS finding of no blood vessel involvement. CT detected blood vessel involvement in two patients only. The overall accuracy of EUS and CT for detecting the tumor was 100% and 77% respectively, and for blood vessel involvement 92% and 61% respectively. In conclusion, EUS is an accurate procedure for preoperative assessment of blood vessel involvement in patients with pancreatic cancer. This procedure may enable the selection of those patients who may benefit from surgery, and should be part of the evaluation of patients with pancreatic cancer who are candidates for curative surgery.

Surgery for esophageal cancer carries a high mortality rate and a low rate of resectability for cure. Accurate preoperative staging is therefore of utmost importance. Staging is based on computerized tomography (CT), and recently, the use of endoscopic ultrasonography (EUS). We performed EUS and CT on 10 patients with esophageal cancer. Tumors were staged according to the TNM classification. According to the CT results, seven patients had a T3 tumor, one T1-2 and two T0. All patients were diagnosed as T3 by EUS. One patient, who was treated by combined modality treatment with chemotherapy and radiotherapy, converted to T0. Six patients were operated on, and in five, pathological findings were of an invasive tumor. The T stage was predicted correctly in five patients by CT and in all six patients by EUS. N stage was correctly diagnosed in two patients by CT and in five by EUS. It is concluded that EUS is superior to CT for preoperative staging of esophageal tumors. EUS should be undertaken as a routine procedure prior to surgery for esophageal cancer.