Performance of diffusion-weighted MRI post-CT urography for the diagnosis of upper tract urothelial carcinoma: Comparison with selective urine cytology sampling.

OBJECTIVES

To evaluate the usefulness of adding diffusion-weighted MRI (DW-MRI) to CT urography (CTU) for diagnosing upper tract urothelial carcinoma (UTUC).

MATERIALS AND METHODS

We retrospectively evaluated 102 high-risk patients with UTUC who underwent both CTU and DW-MRI. The diagnostic performance of CTU and DW-MRI was evaluated in all the patients. Then, the diagnostic performance of DW-MRI in the patients with positive CTU findings was assessed based on the types of CTU findings (mass formation, wall thickening, and small filling defects). The diagnostic performance of DW-MRI and selective urine cytology sampling in these patients was also compared.

RESULTS

The sensitivity of DW-MRI (92%) was less than that of CTU (98%) (P = 0.25), since DW-MRI could not detect any small papillary tumors. The specificity of DW-MRI (91%) was greater than that of CTU (78%) (P = 0.065), since DW-MRI could discriminate some benign conditions from UTUC. Among the 59 positive CTU lesions, DW-MRI correctly diagnosed all 41 mass-forming lesions and improved the accuracy from 36% using CTU alone to 79% for 14 wall-thickening lesions but could not add any additional information for 4 small filling-defect lesions. Among the 44 patients with positive CTU findings who underwent selective urine cytology sampling, the sensitivity of DW-MRI (95%) was significantly greater than that of selective urine cytology sampling (56%) (P < 0.001).

CONCLUSIONS

The addition of DW-MRI would be useful for both mass-forming and wall-thickening lesions. DW-MRI has the potential to reduce the frequency of selective urine cytology sampling for such lesions.