关键词: 直肠肿瘤, 磁共振成像, 术前评估

Abstract: Objective  To evaluate the value of rectal MRI combined with intraluminal filling coupling agent in preoperative evaluation of rectal cancer. 
Methods  A total of 76 cases of rectal cancer were reviewed. All patients underwent rectal MRI examination and MRI examination after intraluminal filling of coupling agent. The T stage, N stage, mesangial vessel invasion and circumferential margin of the tumors were evaluated respectively. The results were compared with the pathological results after operation. 
Results  Rectal MRI was consistent with postoperative pathology in 58 patients with rectal cancer T stage, the overall coincidence rate was 76.3%. N stage was consistent with postoperative pathology in 55 cases, the overall coincidence rate was 72.3%(Kappa=0.572, χ2= 4.366, P＜0.001). EMVI analysis showed that MRI was consistent with pathology in 55 cases, EMVI(+) was misdiagnosed as EMVI(-) in 9 cases and EMVI(-) was misdiagnosed as EMVI(+)in 12 cases. The overall coincidence rate was 72.3%(Kappa=0.564, χ2= 23.677, P < 0.001). According to CRM analysis, 68 cases were consistent with pathology, including 4 cases of CRM(+) and 4 cases of CRM(-)were misdiagnosed . The overall coincidence rate was 89.5%(Kappa=0.754, χ2= 8.556, P=0.004). Sixty-six cases of rectal cancer T staging were consistent with postoperative pathology after filling the rectum cavity with coupling agent, the overall coincidence rate was 86.8%, and the coincidence rate with rectal MRI was analyzed by chi square test(Kappa=0.736, χ2= 3.102, P=0.043). N staging was consistent with postoperative pathology in 55 cases, with an overall coincidence rate of 72.3%. The coincidence rate with rectal MRI was examined by chi square test(Kappa=0.569, χ2= 0.000, P=1.000). EMVI analysis showed that MRI was consistent with pathology in 61 cases, among which EMVI(+) was misjudged as EMVI(-) in 7 cases and EMVI(-) was misjudged as EMVI(+) in 8 cases. The overall coincidence rate was 80.3%. The coincidence rate with rectal MRI was verified by chi square test(Kappa=0.612, χ2= 1.314, P=0.252). According to CRM analysis, 65 cases were consistent with pathology, including 4 cases of CRM(+) and 7 cases of CRM(-) were misdiagnosed. The overall coincidence rate was 85.5%. The coincidence rate of MRI and rectal MRI was analyzed by chi square test(Kappa=0.712, χ2= 0.543, P=0.465). 
Conclusion  Rectal MRI combined with intraluminal coupling agent can improve the accuracy of preoperative T staging and mesangial vessel invasion in rectal cancer, but has no significant influence on the accuracy of N staging and circumferential margin.

Key words: rectal neoplasms, magnetic resonance imaging, preoperative evaluation