Detection of Bladder Tumors with Dynamic Contrast-Enhanced MDCT

Abstract

OBJECTIVE. In a small pilot study, we assessed whether early-phase dynamic contrast-enhanced MDCT can be used to detect bladder tumors and whether thin reconstruction improves the detection rate. SUBJECTS AND METHODS. Thirty-six patients (30 with 59 cystoscopy-proven bladder cancers and six with normal bladders) underwent dynamic contrast-enhanced MDCT of the pelvis and abdomen. Images were obtained from the symphysis pubis to the diaphragm 70 seconds after injection of 100 mL of contrast medium. McNemar test was used to compare sensitivity per patient, segment, and tumor and specificity per patient and segment for each of three reconstruction methods: 5-mm sections with no overlap (i.e., 5-mm axial images), 2.5-mm sections with 1.25-mm overlap (i.e., thin-section axial images), and 2.5-mm sections with 1.25-mm overlap and multiplanar reformation (MPR) (i.e., thin-section axial images with MPR). RESULTS. MDCT with a combination of thin, overlapped sections and MPR depicted all but one of 47 bladder tumors larger than 5 mm but only five of 12 tumors 5 mm or smaller. There were no false-positive findings. Per-tumor sensitivity was significantly better with thin-section images with MPR (90%) and thin-section images alone (86%) than with 5-mm axial images (80%) (p < 0.05). Per-segment sensitivity was significantly better with thin-section images with MPR (95%) and thin-section axial images alone (87%) than with 5-mm axial images (79%) (p < 0.05). Per-patient sensitivity and per-patient and per-segment specificity did not differ with the three methods. CONCLUSION. Dynamic contrast-enhanced MDCT of the pelvis shows promise for the detection of bladder tumors. Use of thin-section images with MPR and thin-section axial images alone had a significantly better rate of detection of bladder tumors than use of 5-mm axial images.