Stromal invasion by carcinoma of the cervix: assessment with dynamic MR imaging

OBJECTIVE: The purpose of this study was to evaluate dynamic MR imaging in assessing the depth of stromal invasion by carcinoma of the cervix and to compare dynamic MR imaging with T2-weighted and contrast-enhanced T1-weighted MR imaging. SUBJECTS AND METHODS: Forty-one patients with carcinoma that was clinically considered to be confined to the cervix were examined with T2-weighted, dynamic, and contrast-enhanced T1-weighted MR imaging before surgery. We evaluated enhancement patterns of the cervix and tumor and assessed the degree of stromal invasion with MR imaging. The degree of stromal invasion was divided into two groups: superficial disease (no stromal invasion or invasion of < or = 3 mm) and deep invasion (> 3 mm of stromal invasion). Then we compared these MR findings with histologic results for the depth of stromal invasion. RESULTS: With dynamic MR imaging, cervical carcinoma with deep invasion was seen as a focal enhanced area in the early dynamic phase. The cervical epithelium and stroma enhanced less vividly. In distinguishing deep invasion from superficial disease, we found the accuracy of T2-weighted MR images, dynamic MR images, and contrast-enhanced T1-weighted MR images to be 76%, 98%, and 63%, respectively. In particular, the detectability of 3.1-5.0 mm of stromal invasion with dynamic MR images was significantly higher than that with the other techniques: with T2-weighted MR images, we saw 3.1-5.0 mm of stromal invasion in 23% of patients; with dynamic MR images, in 92%; and with contrast-enhanced T1-weighted MR images, in none. Superficial disease was not revealed with any of the three MR techniques. CONCLUSION: We believe that dynamic MR imaging is superior to T2-weighted MR imaging and contrast-enhanced T1-weighted MR imaging when assessing the depth of invasion of cervical carcinoma.