Prospective evaluation of diffusion-weighted MRI of the bladder as a biomarker for prediction of bladder cancer aggressiveness

Objectives: To prospectively test the utility of diffusion-weighted magnetic resonance imaging (DW-MRI) for bladder cancer (BCA) staging and grading. Methods and materials: We prospectively enrolled 51 consecutive patients (median age = 64 y, range: 30-90 y) with suspected BCA with or without gross hematuria who received 3-T DW-MRI before transurethral resection of the bladder. Parametric apparent diffusion coefficient (ADC) maps were automatically calculated from DW-MRI. ADC values (given in x10(-3) mm(2)/s) of bladder lesions were independently measured by 2 radiologists blinded to histopathology. Comparisons of ADC values with histopathologic features were performed using unpaired t tests. Diagnostic performance was calculated by means of receiver operating characteristics (ROC) statistics. Results: We excluded 8 patients: 1 presenting with metastatic melanoma to the bladder, 1 who had an incomplete examination, and 6 without BCA. In the 43 remaining patients (median age = 68 y, range: 41-85 y), the ADC values were lower in high-grade (n = 19, ADC = 0.787) compared with low-grade (n = 24, ADC = 1.233) tumors (P < 0.0001) and in muscle-invasive tumors (n = 10, ADC = 0.759) compared with non muscle-invasive tumors (n = 33, ADC = 1420, P = 0.0004). The area under the ROC curve was 0.884 for prediction of muscle invasion and 0.906 for prediction of high grade by using ADC values. Rule-in ADC criteria for high-grade lesions and rule-out ADC criteria for muscle invasion were identified by ROC analysis. Conclusions: ADC measurements obtained by DW-MRI are a promising imaging biomarker for prediction of BCA stage and grade providing high sensitivity and specificity. Further studies are necessary to establish the value of this test in guiding decision making in clinical practice. (C) 2014 Elsevier Inc. All rights reserved.