Comparison of EUS and magnifying colonoscopy for assessment of small colorectal cancers

Prediction of invasion depth and lymph node metastasis is mandatory when local treatment is considered for small colorectal cancer. The aim of this study was to compare the accuracy of EUS with a catheter probe (probe-EUS) and magnifying colonoscopy for prediction of invasion depth and lymph node metastasis for small colorectal cancer.

Small colorectal cancers were imaged by both probe-EUS and magnifying colonoscopy. Invasion depth by probe-EUS was determined by the presence or absence of distortion of the third sonographic layer. Findings by magnifying colonoscopy were divided into regular, distorted, and amorphous patterns. Histopathologically, depth of invasion was classified as intramucosa/slight or deep invasion. Findings by probe-EUS and magnifying colonoscopy were compared with respect to deep invasion and lymph node metastasis.

There were 22 small colorectal cancers with intramucosa/slight invasion and 28 with deep invasion. Four of 30 cancers had associated lymph node metastasis. Accuracy for depth of invasion was 91.8% for probe-EUS and 63.3% in magnifying colonoscopy, the difference being statistically significant (p = 0.0013). Negative predictive value of probe-EUS for deep invasion was higher than that for magnifying colonoscopy (respectively, 90.9% vs. 54.1%) in the population studied (prevalence deep invasion 56%). The accuracy for lymph node metastasis was 24.1% for probe-EUS and 72.4% for magnifying colonoscopy, the difference being statistically significant (p < 0.001). Positive predictive value for lymph node metastasis was higher when the amorphous pattern was noted by magnifying colonoscopy compared with the positive predictive value for deep invasion by probe-EUS (respectively, 33.3% vs. 8.7%) in the population studied (prevalence lymph node metastasis 13.3%).

Probe-EUS is superior to magnifying colonoscopy for determination of invasion depth in small colorectal cancer. Magnifying colonoscopy may be predictive of lymph node metastasis, thereby suggesting that the procedures provide complementary information with respect to the decision for local versus surgical therapy.