SUMMARY The value of multiple biopsies and brush cytology at oesophago-gastroscopy was assessed in relation to macroscopy and localization on 100 verified tumours in a prospective study. The cumulative accuracy achieved was 96 %. This was significantly better (P < 0.01) than that of biopsy (83 %) and of cytology (85 %). While the reliability of both procedures was not significantly different in malignancies of the oesophagus, the gastric body, and the antrum, cytology was significantly more accurate in cancers of the cardia (90 % and 55 % respectively, P < 0 05). Cytology was also more reliable in stenosing tumours (92 %/72%, P <0 05). In polypoid malignancies a positive but not significant trend was found in favour of multiple biopsies (94 %/64 %). One of the two early cancers was only diagnosed by cytology. The results confirm the high diagnostic accuracy of multiple endoscopic biopsies combined with brush cytology and demonstrate the value of cytology in stenosing tumours, especially in those of the cardia.The reported rate of accuracy of multiple endoscopic biopsies and of brush cytology is at great variance. False negative results were observed between 7% and 68% for biopsy (Dollinger, 1972;Hampel et al., 1974) and 4-2 %-56 % for brush cytology (Kobayashi et al., 1970;Dollinger, 1972). In the light of the difficulties in deciding whether an endoscopically visualized lesion is benign or malignant and the high value of a correct preoperative diagnosis, it was considered to be important to assess the reliability of both procedures in a prospective study. Also little is known of the value of both procedures in relation to the macroscopic aspect and the localization of an individual lesion and this question was therefore included in the protocol. MethodsTwo hundred and fifty-one patients with endoscopically suspect oesophageal (58) and gastric (193) lesions were investigated. The age of the 167 male and 84 female patients varied between 22 and 88 years (mean 57 8 years). Thirty-one had two ex-

No abstract

In a prospective study 96 patients with esophageal and gastric lesions were investigated by multiple endoscopic biopsy, brush cytology, and guided lavage cytology. The aim was to evaluate the diagnostic accuracy of each of these methods in a prospective study. 50 malignant and 46 benign lesions were found. The final diagnosis was proven by operation or autopsy in 53 cases, by multiple endoscopic follow-up in 31, and by clinical control in 8. No follow-up was possible in 4 patients. The malignant lesions were diagnosed by multilple biopsy in 86.0%, by brush cytology in 84.0%, and by lavage cytology in 50.0%. The diagnostic yield of lavage cytology was thus significantly lower (P less than 0.01). It is concluded that a combination of multiple biopsy and brush cytology is the most valuable aid in endoscopic diagnisis of esophageal and gastric malignancy with a diagnostic accuracy of 96%.

Percutaneous biopsy of the pancreas with a fine needle aimed by ultrasound was carried out on 26 patients with sonographically proven carcinoma of the pancreas. The accuracy of this method was correlated with other radiological techniques. Needle biopsy of the pancreas proved an informative and economical procedure, since the sonographic demonstration of a pancreatic mass otherwise requires additional radiological and endoscopic investigations.

No abstract