Assessment of tumor infiltration depth in rectal cancer with transrectal sonography: caution is necessary

To assess the depth of infiltration of rectal cancer with transrectal ultrasound (US) (TRUS) and analyze interpretation errors. Fifty-five consecutive patients with rectal cancer who underwent TRUS were prospectively studied. The effect of different patient inclusion criteria and US criteria was evaluated retrospectively. Extensive overstaging of T2 tumors, partially caused by inflammatory (desmoplastic) reaction or retraction of the muscularis propria, resulted in a specificity of only 24% for detection of perirectal infiltration. Sensitivity was 97%, and accuracy was 64%. Considerable variation in staging accuracy was observed when different patient selection criteria were used. Variation of US criteria improved differentiation between T2 and T3 tumors only slightly. The effect of different US criteria on differentiation of T2 and T3 tumors is limited. Spontaneous or iatrogenic inflammation is a major limiting factor. The accuracy of TRUS in staging rectal cancer is affected by patient inclusion criteria