Investigation of diffusion kurtosis imaging for discriminating tumors from inflammatory lesions after treatment for bladder cancer.

Background: The bladder wall may thicken resulting from chronic inflammation after initial treatment (transurethral resection [TUR] or neoadjuvant chemotherapy), which may mimic the feature of recurrent or residual bladder tumors (RBT). Therefore, it is critical to discriminate RBT from benign lesions after initial treatment.

Purpose: To investigate whether diffusion kurtosis imaging (DKI) could discriminate RBT from post-therapy bladder inflammatory lesions.

Study Type: Retrospective.

Subjects: Fifty patients diagnosed with bladder cancer underwent TUR or received neoadjuvant chemotherapy.

Field Strength/sequence: 3.0T MRI/conventional T -weighted imaging (T WI), T WI, and diffusion-weighted imaging (DWI) with nine b-values ranging from 0-2000 s/mm .

Assessment: Mean diffusion coefficients (MD , MD , and MD ) and mean kurtosis values (MK , MK , and MK ) were obtained from three different measurement methods. The region of interest (ROI) was placed 1) to encompass the entire portion of the thickening bladder wall or to portions that were the most restricted, with a b-value of 2) 2000 s/mm or 3) 1000 s/mm .

Statistical Tests: The independent-samples t-test was used to compare the differences between RBT and the inflammatory group. Differences in DKI parameters were analyzed by comparing the areas under the receiver-operator characteristic curves (AUCs).

Results: In patients with RBT, the MD (MD , MD , MD ) values were significantly lower and the MK (MK , MK , MK ) values were significantly higher than those in patients in the inflammatory lesions group (all P < 0.01). The AUC of MK (0.934) was significantly larger than those of MD , MK , and MK (0.793, P < 0.05; 0.694, P < 0.01; 0.719, P < 0.01, respectively).

Data Conclusion: MK obtained from DKI provided better performance than conventional DWI in distinguishing RBT from inflammatory lesions after bladder cancer treatment. MK calculated with high b-values setting provided better performance in differentiation.

Level Of Evidence: 1 Technical Efficacy Stage 3 J. Magn. Reson. Imaging 2017.