Purpose: To compare staging of advanced primary cervical carcinoma (pT2b-pT4a) by conventional and pharmacokinetic magnetic resonance imaging (MRI) with the giant cross section specimen and histopathological findings.
Materials and methods: Seventeen patients with biopsy-proven cancer of the cervix and clinically suspected invasive cancer (FIGO IIB-IVA) were prospectively examined by conventional (T2 and contrast-enhanced T1-weighted spin echo images) and pharmacokinetic MRI. All MRI findings were compared with the giant cross section specimen and histopathology as the standard of reference. For pharmacokinetic MRI, a saturation recovery TurboFLASH sequence was used with a high temporal resolution of 13 s per ten sections. Signal time changes were analyzed using a pharmacokinetic model and the computed parameter values were visualized by color-coded overlay.
Results: Analysis of parametrial invasion on T2-weighted images resulted in an accuracy of 85% and 73% on contrast-enhanced T1-weighted images and on pharmacokinetic MR images respectively. Accuracy of analysis of bladder and/or rectal wall invasion was significantly (P < 0.05) higher on pharmacokinetic MR images (88%) than on T2-weighted images (67%). Contrast-enhanced T1-weighted spin-echo images improved staging accuracy compared with T2-weighted images (76% vs 67%).
Conclusion: At present, conventional T2-weighted SE images are superior to contrast-enhanced T1-weighted SE and pharmacokinetic MR images in depicting infiltration of the parametrium. However, suspected infiltration of the bladder and/or rectum (pT4a) is diagnosed more accurately on pharmacokinetic images than on conventional MR images.