Physicians frequently use CT and MRI for cancer assessment. Dr. Bindu N. Setty and colleagues at Massachusetts General Hospital in Boston have analyzed these techniques in a bladder cancer setting and have made recommendations for their use. They also have examined novel imaging methods for diagnosing the condition.

CT’s wide availability and effectiveness in showing where cancer has spread make it a common choice. Multidetector-row CT, in comparison, provides improved spatial resolution and wider coverage. However, CT alone cannot assess the depth of bladder wall invasion. The CT should be performed before interventions such as cystoscopically guided biopsy, which can cause formation of fibroses that can have a tumor-like appearance.

With multiplanar capability and high tissue contrast, MRI enjoys an advantage in primary carcinoma staging, and the addition of endoluminal coils can enable more accurate assessment of tumor invasion. Ultrafast scanning methods with stronger magnets also improve the quality of images; however, differentiation of tumors from cells killed by radiation and infection can prove difficult using MRI, making a biopsy necessary for confirmation. Early acquisition of images — ideally within 90 seconds after intravenous contrast injection — offers increased accuracy.

More recently, virtual endoscopy with CT or MRI has emerged as an option. Although a potential substitute for invasive cystoscopy, ongoing CT-based endoscopic monitoring poses radiation concerns. MRI with ultrasmall, superparamagnetic iron oxide particles may improve characterization of lymph nodes but has not gained FDA approval. PET and PET/CT with 18F-fluorodeoxyglucose are not very useful for the detection of primary tumors because the compound is excreted from the kidneys into the bladder, but the tracer can help track early and distant metastases.

Cross-sectional imaging, the authors conclude, comprises a valuable, noninvasive tool that will play an increasingly useful role in bladder cancer diagnostics. Emerging methods likely will contribute to more accurate diagnoses. (Current Problems in Diagnostic Radiology, March/April 2007, pp. 83-96.)