The validity of magnetic resonance imaging (MRI) in the staging of bladder cancer: comparison with computed tomography (CT) and transurethral ultrasonography (US).

The magnetic resonance imaging (MRI) of bladder cancer was accomplished in 26 patients who had histologically proven transitional cell carcinoma through transurethral resection. All 26 patients were simultaneously studied by computed tomography (CT) and 21 were also studied by transurethral ultrasonography (US) before surgery or preoperative irradiation. The evaluations of the MRI and CT were carried out by the same radiologist and urologist postoperatively; the evaluation of the US was carried out by several urologists preoperatively, and their results compared with the postoperative pathological findings. The staging accuracy for pT2 tumors and above for each imaging modality was as follows. MRI: sensitivity 83.3%, specificity 71.4%, accuracy 76.9%; CT: sensitivity 58.3%, specificity 42.8%, accuracy 50%; US: sensitivity 100%, specificity 70%, accuracy 85.7%. The results of the staging with regard to extravesical extensions were as follows. MRI: sensitivity 60%, specificity 85.7%, accuracy 80.8%; CT: sensitivity 60%, specificity 81%, accuracy 76.9%. On the other hand, the sensitivity of US was 25%, specificity 94.1% and accuracy was 81%. The higher the pathological staging, the more the diagnostic accuracy was likely to be increased with MRI and CT, but diminished for US. There were three cases of pT1 tumors being over-diagnosed by US, but two of them were accurately staged by MRI. The other 11 cases, diagnosed as T2 or above by US, were only minimally overstaged or understaged by MRI. MRI seemed able to overcome the disadvantage of US which tended to over-diagnose pT1 tumors and which was not suitable for use in diagnosing infiltrative tumors correctly. CT seemed useless for local staging.