The accuracy of MRI, endorectal ultrasonography, and computed tomography in predicting the response of locally advanced rectal cancer after preoperative therapy: A metaanalysis

Background. To perform a metaanalysis to determine and compare the diagnostic performance of MRI, endorectal ultrasonography (ERUS), and computed tomography (CT) in predicting the response of locally advanced rectal cancer after preoperative therapy. Methods. All previously published articles on the role of MRI, CT, and/or ERUS in predicting the response of rectal cancer to preoperative therapy were collected. We divided the objective in 3 parts: the accuracy to assess (i) complete response, (ii) to detect T4 tumors with invasion to the circumferential resection margin (CRM), and (iii) to predict the presence of lymph node metastasis. The pooled estimates of sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated using a bivariate mixed effect analysis. Results. Forty-six studies comprising 2,224 patients were included. (i) The pooled accuracy to assess complete tumor response were (a) 75 % for MRI, (b) 82% for ERUS, (c) and 83 % for CT (ii) Pooled accuracy to detect T4 tumors with invasion to the CRM were (a) 88% and (b) 94% for ERUS. Pooled accuracy to predict the presence of lymph node metastasis was (a) 72% for MRI, (b) 72% for ERUS, (c) and 65 % for CT. Conclusion. MRI, CT, and ERUS cannot be used to predict complete response of locally advanced rectal cancer after CRT. In addition, the positive predictive value for these imaging techniques is low for the assessment of tumor invasion in the CRM. The accuracy of the modalities to predict the presence of metastatic lymph node disease is also low.