Objective: To quantify first-pass enhancement of cervix carcinoma using fast dynamic MRI. To assess the accuracy of dynamic contrast-enhanced colour-coded MRI for determining tumour invasion into surrounding pelvic tissues.
Methods: Gadolinium enhanced dynamic MRI at one image every 2 s was performed in 47 patients with cervical carcinoma and five controls. First-pass contrast enhancement of cervix carcinoma and surrounding pelvic tissues was quantified. Automated colour-coded images were constructed using the dynamic parameters slope, amplitude and timing of enhancement. Of 47 patients, 28 underwent surgery and colour coded images were correlated with histological findings.
Results: First-pass contrast enhancement imaging of cervix carcinoma required a temporal resolution of dynamic MRI of one image every 3-4 s. Cervix carcinoma first-pass was more rapid and intense than that of other pelvic tissues (P<0.001) with the exception of normal myometrium (P>0.05). Binary colour coding, however, was not reliable for tumour delineation or for accurate assessment of tumour invasion into the parametria or the bladder wall. Overestimation of the extent of tumour invasion occurred in 15, 16 and nine out of 28 patients, respectively, using amplitude, slope and timing of enhancement as parameters.
Conclusion: Dynamic contrast-enhanced colour-coded MRI of cervix carcinoma has limited value for assessing the extent of tumour spread and tumour staging.