Unenhanced CT and MRI Parameters That Can Be Used to Reliably Predict Fat-Invisible Angiomyolipoma.

Abstract

The objective of our study was to determine unenhanced CT and MRI parameters that can be used to reliably predict fat-invisible angiomyolipoma (AML). A total of 152 patients with 155 histologically proven renal cell carcinomas (RCCs) and 16 patients with 18 histologically proven AMLs were included. No AML measured less than -20 HU on CT images. The tumor-to-cortex ratio was measured on unenhanced CT, T2-weighted MRI, and T1-weighted MRI to compare fat-invisible AML and RCC. On chemical-shift MRI, the signal intensity (SI) index and tumor-to-spleen ratio were calculated to compare these lesions. The unpaired t test and ROC with AUC were used for statistical analysis. The reference standard was pathologic examination. The mean tumor-to-cortex ratios on unenhanced CT of AML and RCC were 1.37 ± 0.66 (SD) and 0.83 ± 0.47, respectively (p < 0.001). However, the tumor-to-cortex ratio on unenhanced CT of AML was not different from that of chromophobe RCC (p = 0.093). The mean tumor-to-cortex ratios on T2-weighted MRI of AML and RCC were 0.75 ± 0.15 and 1.21 ± 0.41, respectively (p < 0.001). However, the tumor-to-cortex ratio on T2-weighted MRI of AML was not different from that of papillary RCC (p = 0.161). The tumor-to-spleen ratio on chemical-shift MRI, tumor-to-cortex ratio on T1-weighted MRI, and SI index on chemical-shift MRI of AML were not different from those of RCC (p = 0.104-0.670). The AUCs of the tumor-to-cortex ratio on T2-weighted MRI and the tumor-to-cortex ratio on unenhanced CT were 0.862 and 0.835, respectively. The tumor-to-cortex ratio on T2-weighted MRI and the tumor-to-cortex ratio on unenhanced CT are good parameters to differentiate fat-invisible AML from clear cell RCC. Nevertheless, percutaneous biopsy is necessary because these parameters are not reliable in differentiating fat-invisible AML from non-clear cell RCC.