Comparison of the diagnostic accuracy and validity of biparametric MRI and multiparametric MRI-based VI-RADS scoring in bladder cancer; is contrast material really necessary in detecting muscle invasion?

Abstract

(1) To evaluate the accuracy and validity of the biparametric MRI (bp-MRI), including T2-weigthed image (WI) and DWI sequences, and the availability of an alternative to the multiparametric MRI (mp-MRI), for the muscle-invasiveness assessment of bladder cancer (BC). (2) To evaluate the diagnostic performance and agreement of readers with different experiences in the abdominal imaging of using both protocols. Preoperative bladder mp-MRI was performed on 128 patients with a initial diagnosis of BC. Two sets of images, set 1 (bp-MRI) and set 2 (mp-MRI), were independently evaluated by both readers. Descriptive statistics, including sensitivity, specificity, accuracy, and area under the curve (AUC), for VI-RADS scores were calculated using ≥ 4 as the cutoff for muscle invasion for each reader and image sets. Inter-reader agreement was evaluated using the Cohen's kappa coefficient. The sensitivity ranged between 90.3-93.5% and 87.1-90.3%, specificity ranged between 96.6-99.1% and 91.6-96.6%, accuracy ranged between 96-97.3% and 91.3-94.6%, and AUC ranged between 0.947-0.951 and 0.919-0.921, for bp-MRI and mp-MRI, and reader 1 and reader 2, respectively. No significant differences were shown in diagnostic performance for either reader between both the protocols (p = 0.238 and 0.318). There was excellent agreement among the readers in the VI-RADS scores, using both protocols. A bp-MRI protocol has a diagnostic accuracy comparable to an mp-MRI protocol for the detection of muscle-invasive BC using the VI-RADS criteria. Also, in both MRI protocols, the reader's experience does not appear to significantly affect diagnostic performance when using the VI-RADS criteria.