Search

BACKGROUND AND STUDY AIMS The prevalence of esophageal squamous-cell carcinoma in high-risk patients and the advantages of systematic Lugol staining during esophagoscopy have not yet been evaluated in a large prospective study. In this study we aimed to assess the prevalence of this type of tumor in high-risk patients, to examine the role of Lugol staining in endoscopic screening for esophageal squamous-cell carcinoma, and to establish whether it is possible to identify a particularly high-risk group which would benefit from systematic screening. PATIENTS AND METHODS A prospective study was undertaken in 62 endoscopy centers. A total of 1095 patients were enrolled, none of whom had any esophageal symptoms. These patients had presented with either a past history of or a recent head and neck or tracheobronchial squamous-cell carcinoma (group 1), with alcoholic chronic pancreatitis (group 2), with alcoholic cirrhosis (group 3), or were alcohol and tobacco addicts (group 4). The patients underwent a meticulous endoscopic examination of the esophagus, followed by Lugol staining. RESULTS The prevalence of esophageal squamous-cell carcinoma was 3.2 %. The group 1 patients showed the highest prevalence of carcinoma (5.3 %) and the highest prevalence of dysplasia (4.5 %). Of the 35 carcinomas detected in the 1095 patients, seven (20 %) were early lesions, and 20 % were only detected after Lugol staining (P = 0.02). High-grade dysplasia was only observed in group 1 patients and two-thirds of these lesions were only diagnosed after Lugol staining. The overall prevalence of low-grade dysplasia was 2.4 %, and 77 % of these were detected only after Lugol staining (P < 0.001). CONCLUSIONS Lugol dye staining increases the sensitivity of esophageal endoscopy for the detection of high-grade dysplasia and cancer. The prevalence of dysplasia and cancer reached 9.9 % in group 1, and we therefore believe that an endoscopic screening program could be justified for patients with head and neck or tracheobronchial cancer.

Background and study aims Granular cell tumors of the esophagus are rare tumors. A definite diagnosis is achieved by endoscopic biopsies in only 50% of cases. Endoscopic ultrasonography (EUS) is the best procedure in the evaluation of upper gastrointestinal tract submucosal tumors. The aim of this study was to describe the endosonographic findings of esophageal granular cell tumors. Methods From January 1989 to March 1994, 15 patients with 21 granular cell tumors which had negative biopsies were examined by EUS (Olympus GF UM3 or GF UM20,7,5 and 12 MHz). In five cases, the tumor was also studied with a 20 MHz Olympus miniprobe. The final histological diagnoses were obtained by subsequent endoscopic snare resection in 20 cases and surgically in one case. Results The endosonographic features (with the GF UM3 or GF UM20) of esophageal granular cell tumors were: a) a tumor size of less than 2 cm in 95% of cases; b) an hypoechoic solid pattern in 100% of cases; c) a tumor arising in the inner layers in 95% (second echo-poor layer n=15; third echo-rich layer n=5). In one case, the endosonographic finding was transmural malignant infiltration of the esophageal wall (histologically confirmed). Conclusion When a granular cell tumor of the esophagus is suspected, EUS can show the inner layer location of the tumor and thus contribute to planning the endoscopic resection or follow up. When the tumor also invades the outer layers, EUS can contribute to planning the surgical resection.

BACKGROUND AND STUDY AIMS: Little is known of the long-term outcome in patients with gastrointestinal bleeding of obscure origin, who undergo investigation by means of push enteroscopy. The aim of this study was to assess the rate of recurrent bleeding and its predictive factors in patients with gastrointestinal bleeding of obscure origin, after exploration by push enteroscopy. PATIENTS AND METHODS: 105 patients with gastrointestinal bleeding of obscure origin (iron-deficiency anemia: n = 56; overt bleeding: n = 49) underwent exploration by push enteroscopy from December 1994 to December 1998. They were classified into three groups according to enteroscopy findings: no lesion found (group A; 56 patients), arteriovenous malformations (group B; 18 patients), and other gastrointestinal lesions (group C; 31 patients). Actuarial rates of rebleeding during follow-up were calculated and factors associated with rebleeding were assessed by means of univariate and multivariate analysis. RESULTS: Follow-up data were obtained for 101 patients (96 %). The mean follow-up was 29 months (6 - 54 months). The 2-year actuarial rate of rebleeding was 31 % in the overall population, and 27.6 %, 56 % and 24 % in groups A, B, and C, respectively (P = 0.13). The number of previous bleeding episodes and the number of packed red cell units transfused were two independent factors predictive of recurrent bleeding. The modality of recurrent bleeding (anemia or overt bleeding) was similar to that of the initial episode in 94 % of cases. In group A, a gastrointestinal lesion was found after rebleeding in one of the 12 patients with iron-deficiency anemia, and in four of the five patients with overt bleeding. CONCLUSION: Recurrent bleeding occurs in about one-third of patients who undergo investigation by push enteroscopy for gastrointestinal bleeding of obscure origin, with a trend towards more frequent rebleeding in patients with arteriovenous malformations. Frequent previous bleeding episodes and transfusion requirements are predictive of recurrent bleeding.