摘要
Background There is no consensus regarding the performance for endorectal ultrasonography （ERUS） at every stage of rectal cancer. Thus, the purpose of our study was to further assess the value of ERUS in the preoperative staging of rectal cancer. Methods A retrospective study was performed with 44 consecutive patients （mean age： （63.3±10.2） years） who underwent surgical treatment for endorectal carcinoma and were preoperatively evaluated using Biplane ERUS between September 2008 and December 2010. We compared the ERUS staging with the pathological findings based on surgical specimens. Results ERUS staging agreed with the histologic staging in 39 of the 44 （88.6%） patients： the agreement on the depth of transmural invasion was good （k= 0.73; 95% CI： 0.60-0.86, P =0.000）. The detection sensitivities of rectal cancer with ERUS were as follows： T1 85.7%, T2 87.5%, T3 88.9%, and T4 100.0% with specificity values of T1 97.3%, T2 92.9%, T3 96.2%, and T4 97.6%. ERUS correctly staged patients with T1 95.5%, T2 90.9%, T3 70.5 %, and T4 97.7%. The positive predictive value of ERUS was lowest for T4 （75%）, but highest for T3 （94.1%） followed by T2 （87.5%） and T1 （85.7%）; the negative predictive values of ERUS from high to low were ordered as T4 （100%）, T1 （97.3%）, T2 （92.9%）, and T3 （92.6%）. The percentage of total over-staged cases was 4.5% and the under-staged cases was 6.8%. The extent of perirectal lymph node metastases was determined with a sensitivity of 68.4% （13/19）, specificity of 80.0% （20/25）, and diagnostic accuracy of 75.0% （33/44）.
Background There is no consensus regarding the performance for endorectal ultrasonography （ERUS） at every stage of rectal cancer. Thus, the purpose of our study was to further assess the value of ERUS in the preoperative staging of rectal cancer. Methods A retrospective study was performed with 44 consecutive patients （mean age： （63.3±10.2） years） who underwent surgical treatment for endorectal carcinoma and were preoperatively evaluated using Biplane ERUS between September 2008 and December 2010. We compared the ERUS staging with the pathological findings based on surgical specimens. Results ERUS staging agreed with the histologic staging in 39 of the 44 （88.6%） patients： the agreement on the depth of transmural invasion was good （k= 0.73; 95% CI： 0.60-0.86, P =0.000）. The detection sensitivities of rectal cancer with ERUS were as follows： T1 85.7%, T2 87.5%, T3 88.9%, and T4 100.0% with specificity values of T1 97.3%, T2 92.9%, T3 96.2%, and T4 97.6%. ERUS correctly staged patients with T1 95.5%, T2 90.9%, T3 70.5 %, and T4 97.7%. The positive predictive value of ERUS was lowest for T4 （75%）, but highest for T3 （94.1%） followed by T2 （87.5%） and T1 （85.7%）; the negative predictive values of ERUS from high to low were ordered as T4 （100%）, T1 （97.3%）, T2 （92.9%）, and T3 （92.6%）. The percentage of total over-staged cases was 4.5% and the under-staged cases was 6.8%. The extent of perirectal lymph node metastases was determined with a sensitivity of 68.4% （13/19）, specificity of 80.0% （20/25）, and diagnostic accuracy of 75.0% （33/44）.
