The Value of Blood Oxygenation Level-Dependent (BOLD) MR Imaging in Differentiation of Renal Solid Mass and Grading of Renal Cell Carcinoma (RCC): Analysis Based on the Largest Cross-Sectional Area versus the Entire Whole Tumour

OBJECTIVES To study the value of assessing renal masses using different methods in parameter approaches and to determine whether BOLD MRI is helpful in differentiating RCC from benign renal masses, differentiating clear-cell RCC from renal masses other than clear-cell RCC and determining the tumour grade. METHODS Ninety-five patients with 139 renal masses (93 malignant and 46 benign) who underwent abdominal BOLD MRI were enrolled. R2* values were derived from the largest cross-section (R2*largest) and from the whole tumour (R2*whole). Intra-observer and inter-observer agreements were analysed based on two measurements by the same observer and the first measurement from each observer, respectively, and these agreements are reported with intra-class correlation coefficients and 95% confidence intervals. The diagnostic value of the R2* value in the evaluation was assessed with receiver-operating characteristic analysis. RESULTS The intra-observer agreement was very good for R2*largest and R2*whole (all > 0.8). The inter-observer agreement of R2*whole (0.75, 95% confidence interval: 0.69~0.79) was good and was significantly improved compared with the R2*largest (0.61, 95% confidence interval: 0.52~0.68), as there was no overlap in the 95% confidence interval of the intra-class correlation coefficients. The diagnostic value in differentiating renal cell carcinoma from benign lesions with R2*whole (AUC=0.79/0.78[observer1/observer2]) and R2*largest (AUC=0.75[observer1]) was good and significantly higher (p=0.01 for R2*largest[observer2] vs R2*whole[observer2], p<0.01 for R2*whole[observer1] vs R2*largest[observer2]) than R2*largest for observer 2 (AUC=0.64). For the grading of clear-cell RCC, both R2*whole and R2*largest were good (all > 0.7) and were not significantly different (p=0.89/0.93 for R2*largest vs R2*whole[observer1/observer2], 0.96 for R2*whole[observer1] vs R2*largest[observer2] and 0.96 for R2*whole [observer2] vs R2*largest[observer1]). CONCLUSIONS BOLD MRI could provide a feasible parameter for differentiating renal cell carcinoma from benign renal masses and for predicting clear-cell renal cell carcinoma grading. Compared with the largest cross-section, assessing the whole tumour provides better inter-observer agreement in parameter measurement for differentiating renal cell carcinoma from benign renal masses.