A comparison of pelvic retroperitoneal pneumography and computed tomography in the assessment of extramural invasion of rectal carcinoma.

Pelvic retroperitoneal pneumography (PRP) and pelvic computed tomography (CT) were performed on 33 patients with rectal carcinoma in order to compare the usefulness of the two diagnostic procedures in the preoperative assessment of local malignant extramural invasion. Six PRP-negative patients in whom no free air was visualized in the retroperitoneal space surrounding the mass, were all assessed as having extramural invasion by CT scan and all had histologic evidence of invasion. Of 27 PRP-positive patients in whom free air was seen surrounding the mass, 18 were diagnosed as having extramural invasion on CT, 15 of whom had histologic proof of invasion. In the remaining 9 PRP-positive patients, there was no evidence of extramural invasion on the CT scans, but 5 patients showed evidence of invasion histologically. PRP, when positive, had an unacceptably high rate of being false positive and was therefore unreliable in assessing extramural invasion, whereas CT was able to detect, to some extent, extramural invasion which PRP failed to demonstrate. Based on these findings, we conclude that CT is more useful than PRP in the preoperative assessment of extramural invasion of rectal carcinoma, but is of limited diagnostic value when negative.