Background: Accurate preoperative staging is the key to correct selection of rectal tumors for local excision. This study aims to assess the accuracy of endorectal ultrasound (ERUS) at our institution.
Patients and methods: Retrospective analysis was carried out of patients treated by transanal endoscopic microsurgery (TEM) from 1996 to 2008. TEM was considered the treatment of choice for uT0-1/N0 lesions located between 2 and 12 cm from the anal verge. It was also proposed in selected uT2-3 patients. Preoperative staging was compared with histopathologic findings.
Results: Eighty-one patients (46 males, mean age 66 years) underwent TEM. Mean distance of the tumor from the anal verge was 6.6 cm (range 2-12 cm). ERUS staged 15 of 27 adenomas (55%) as uT1. Of 54 carcinomas, 5 were pT0 because TEM was performed to remove resection margins of a malign polyp already snared. Five of 19 pTis (26%) were overstaged uT1, while 7 of 17 pT1 (41%) were understaged. Overall, ERUS enabled distinction between early and advanced rectal lesion with 96% sensitivity and 85% specificity, giving accuracy of 94% (65/67). Thirteen patients had advanced lesions (eight pT2 and five pT3). Only in two of them (15%) was depth of invasion underestimated by ERUS (one uT0, one uT1) and thus was subsequent salvage surgery necessary.
Conclusions: ERUS is useful to confirm the diagnosis of adenoma and predict depth of mural invasion in early rectal cancer. Differentiation between T0/is and T1 lesions remains challenging, however this does not usually influence surgical strategy.