Prospective comparison of the value of brushings before and after biopsy in the endoscopic diagnosis of gastroesophageal malignancy

Abstract

The hypothesis that the cytodiagnostic results on endoscopic brushings obtained before biopsy may be superior to those brushings obtained after biopsy, but with the accuracy of the subsequent biopsy reduced, was examined for 300 consecutive patients, including 256 with esophageal or gastric carcinomas. Following stratification by site and endoscopic appearance of their lesions, the patients were randomized to undergo brushing either before or after forceps biopsy. The accuracy of brushing cytology in patients with carcinoma was significantly higher when the brushing was performed before biopsy than after biopsy (93.5% versus 82.6%; P less than .01). The diagnostic yield of the biopsy was not significantly different whether the lesions were brushed before or after the biopsy (92.7% versus 93.2%; P less than .5). The diagnostic superiority of brushings obtained before biopsy did not relate to the site or endoscopic appearance of the tumor. There were no false-positive cytologic or histologic reports. For all 256 carcinomas, the cumulative accuracy (brushing cytology plus biopsy) reached 98.8% and was significantly better (P less than .001) than that of biopsy alone (93.9%) or cytology alone (87.9%). Apart from reinforcing the belief that the combined application of brushing and biopsy is mandatory for achieving optimal results, this study indicated that the brushing should be performed before the biopsy.