A comparison between intrarectal ultrasound and CT scanning in staging of experimental rectal tumors.

The purpose of this study was to compare the accuracy of preoperative staging of experimental rectal tumors by digital rectal exam, intrarectal ultrasound (IRUS), and CT scanning with pathologic exam. Rectal tumor masses were induced in 10 mongrel dogs by submucosal injection of 2-3 cc of Freund's complete adjuvant. One week later, the animals underwent digital rectal exam, IRUS, and pelvic CT scans. Pelvic exenteration specimens were submitted for pathologic evaluation. Evaluations and interpretations were done in blinded fashion by independent examiners. The rectal "tumor" was detected in 9 of 10 digital exams, 10 of 10 IRUS exams, and 1 of 10 CT scans. Correct Duke's staging occurred in 70% of digital exams, 90% of IRUS exams, and 10% of CT exams compared to pathological staging. Lymph nodes were detected on pathologic exam in all animals (8.7/animal, range 3-16), on IRUS in all animals (6.4/animal, range 5-13), and in none of the digital or CT examinations. IRUS was significantly more accurate in detecting (P less than 0.0001) and locally staging tumors (P less than 0.0001), and in detecting and localizing lymphadenopathy compared to CT scan. Intrarectal ultrasound is a simple, highly accurate device for assessing depth of wall penetration of rectal tumors and in detecting pararectal lymph nodes and should be considered the preoperative staging procedure of choice for rectal cancer.