TRANSRECTAL SONOGRAPHY IN STAGING RECTAL-CARCINOMA - THE ROLE OF GRAY-SCALE, COLOR-FLOW, AND DOPPLER IMAGING ANALYSIS

OBJECTIVE. The purpose of this study was to evaluate the efficacy of c ombining gray-scale sonography with color-flow imaging and pulsed Dopp ler transrectal sonography in the staging of rectal carcinoma. SUBJECT S AND METHODS. Thirty-nine patients with primary rectal carcinoma unde rwent transrectal sonography. The rectal masses were staged T1-T2 or T 3-T4 on the basis of gray-scale imaging. The local nodes were classifi ed as benign or malignant on the basis of size and echogenicity. In 22 patients, color-flow imaging and pulsed Doppler imaging of the rectal mass and of the local lymph nodes were performed. The peak systolic v elocity (PSV) and end diastolic velocity were documented, and the resi stive index was calculated. RESULTS, Gray-scale imaging alone was used to stage T1-T2 masses with 88% sensitivity and 82% specificity. T3-T4 masses were staged with 82% sensitivity and 88% specificity. Overall accuracy was 85%. Gray-scale imaging of lymph nodes using a discrimina tory size of less than or equal to 5 mm for benign nodes and greater t han 5 mm for malignant nodes yielded a sensitivity of 100%, a specific ity of 28%, and an accuracy of 52%. Using receiver operating character istic curve analysis, we determined that a size of greater than or equ al to 7 mm was optimal for characterizing nodes. Such a size provided an accuracy of 83%. PSV of less than 25 cm/sec distinguished T3-T4 fro m T1-T2 rectal masses with 75% sensitivity, 80% specificity, and 77% a ccuracy. A PSV of greater than 20 cm/sec classified a node as malignan t with 100% sensitivity, 62% specificity, and 76% accuracy. A resistiv e index of greater than 0.61 classified a node as malignant with 71% s ensitivity, 85% specificity, and 80% accuracy. CONCLUSION. Color-flow imaging and pulsed Doppler imaging are useful additions to gray-scale transrectal sonography in staging primary rectal carcinomas. The combi nation has most value when evaluating perirectal nodes.