Rectal endosonography can distinguish benign rectal lesions from invasive early rectal cancers.

OBJECTIVE: To determine whether an experienced ultrasound examiner, using good ultrasound equipment with high multifrequency probes, can discriminate between a high grade or low grade dysplastic adenoma (pT0) and very early invasive rectal cancers (pT1). SUBJECTS AND METHODS: Sixty consecutive patients with clinically possibly pT0 or pT1 rectal tumours referred for transanal local excision underwent endorectal ultrasound examination. Lesions where the endorectal ultrasound image showed the mucosal layer to be expanded but the submucosal layer to be intact (uT0) were considered to represent a low grade or high grade dysplasia adenoma (pT0). An irregularity or disruption of the submucosal layer (uT1) was considered to characterize early invasive rectal cancers (pT1). The ultrasound staging was compared with the histological staging made on the basis of the diagnoses in the excised specimens.
RESULTS: The histopathological diagnoses were: invasive rectal cancer (n = 18, 10 pT1, 4 pT2, 4 pT3 cancers); high grade dysplastic adenoma (n = 21); low grade dysplastic adenoma (n = 18); non adenomatous benign lesions (n = 3). Endorectal ultrasound incorrectly classified two of the invasive cancers (both pT1 tumours) as noninvasive lesions. Five of 42 pT0 tumours were overstaged as uT1 tumours. Overstaging was more common in patients who had undergone a previous excision and in tumours with peritumoral inflammation and desmoplastic reaction. The sensitivity of endorectal ultrasound with regard to invasive cancer was 89% (16/18), specificity 88% (37/42), positive predictive value 76% (16/21), negative predictive value 95% (37/39), and accuracy 88% (53/60). Among pT0 and pT1 tumours, the corresponding figures were 80% (8/10), 88% (37/42), 62% (8/13), 95% (37/39), and 87% (45/52).
CONCLUSION: Endorectal ultrasound can distinguish between noninvasive lesions and invasive rectal cancers clinically of stage pT0 or pT1.