Comparison of two techniques of transrectal ultrasonography for the assessment of local extent of polypoid tumours of the rectum.

The capability of transrectal ultrasonography (TRUS) to differentiate between benign and malignant rectal polyps was evaluated. Digital palpation and sigmoidoscopy were followed by TRUS with a 7.5-MHz linear-array transducer, in a blinded design. In the second part of the study the procedure was extended with an acoustic window system (AWS), preventing compression of the polyp. Pathological examination after surgical resection was used for definitive diagnosis in 110 polyps. Digital examination detected 10 of 22 carcinomas with adenomas, compared with 19 by TRUS (P < 0.01). However, TRUS falsely indicated 20 of 49 adenomas to be carcinomas, compared with 4 out of 49 by digital examination (P < 0.001). TRUS combined with AWS detected 23 of 24 carcinomas, and gave a false positive result in only 3 of 26 benign polyps. The results suggest that the best way to discriminate between benign and malignant rectal polyps is to combine TRUS with AWS.