Angiomyolipomes rénaux sans composante graisseuse : caractéristiques tomodensitométriques, histologiques et évolutives

Predictive Value of Chemical-Shift MRI in Distinguishing Clear Cell Renal Cell Carcinoma From Non-Clear Cell Renal Cell Carcinoma and Minimal-Fat Angiomyolipoma

OBJECTIVE

The purpose of this study was to evaluate the diagnostic performance of chemical-shift MRI in the differentiation of clear cell renal cell carcinoma (RCC) from minimal-fat angiomyolipoma (AML) and non-clear cell RCC.

MATERIALS AND METHODS

In this retrospective study, 97 patients with solid renal tumors without macroscopic fat and with a pathologic diagnosis of clear cell RCC (n = 40), non-clear cell RCC (n = 31), or minimal-fat AML (n = 26) who had undergone renal chemical-shift MRI were included. Size, location, morphology, and signal intensity (SI) of the tumors and the contralateral normal kidneys on T2-weighted and in-phase and opposed-phase images were recorded by readers blinded to the pathology. Percentage tumor-to-renal parenchymal SI drop (percentage SI drop) was calculated and correlated to tumor histology. The statistical analysis was done using Kruskal-Wallis, one-way ANOVA, chi-square, and Fisher exact tests.

RESULTS

The percentage SI drop was significantly higher in clear cell RCC compared with non-clear cell RCC and minimal-fat AML (p < 0.001). Percentage SI drop of greater than 20% had 57.5% sensitivity, 96.5% specificity, and 92% positive predictive value (PPV); and percentage SI drop greater than 29% had 40% sensitivity and 100% specificity for diagnosis of clear cell RCC within the cohort of clear cell RCC, minimal-fat AML, and non-clear cell RCC. A significant proportion of minimal-fat AML (46.2%) displayed homogeneous low T2-weighted SI as opposed to clear cell RCC (5%) and non-clear cell RCC (29%) (p < 0.001).

CONCLUSION

The percentage SI drop on chemical-shift MRI had high specificity and moderate sensitivity in predicting clear cell RCC over non-clear cell RCC and minimal-fat AML. A percentage SI drop greater than 20% in a renal mass without macroscopically visible fat has high PPV for clear cell RCC over minimal-fat AML and non-clear cell RCC. Among morphologic features, homogeneous low T2 SI favors minimal-fat AML over RCC.

Can Quantitative CT Texture Analysis be Used to Differentiate Fat-poor Renal Angiomyolipoma from Renal Cell Carcinoma on Unenhanced CT Images?

PURPOSE

To determine the accuracy of texture analysis to differentiate fat-poor angiomyolipoma (fp-AML) from renal cell carcinoma (RCC) on unenhanced computed tomography (CT) images.

MATERIALS AND METHODS

In this institutional review board-approved retrospective case-control study, patients with AML and RCC were identified from the pathology database: there were 16 patients with fp-AML (no visible fat at unenhanced CT) and 84 patients with RCC. Axial unenhanced CT images were contoured manually by two independent analysts. Texture analysis was performed for each lesion, and reproducibility was assessed. Texture features related to the gray-level histogram, gray-level co-occurrence, and run-length matrix statistics were evaluated. The most discriminative features were used to generate support vector machine (SVM) classifiers. Diagnostic accuracy of textural features was assessed and 10-fold cross validation was performed. Unenhanced CT images for each patient were independently reviewed by two blinded radiologists who subjectively graded lesion heterogeneity on a five-point scale. Differences in area under the receiver operating characteristic curve (AUC) between subjective heterogeneity ratings and textural features were evaluated by using the DeLong method.

RESULTS

There was lower lesion homogeneity and higher lesion entropy in RCCs (P ≤ .01). A model incorporating several texture features resulted in an AUC of 0.89 ± 0.04. The average SVM accuracy of textural features ranged from 83% to 91% (after 10-fold cross validation). An optimal subjective heterogeneity rating of 2 or higher was identified as a predictor of RCC for both readers, with no significant difference in AUC between readers (P = .06). Each of the three textural-based classifiers was more accurate than either radiologists' subjective heterogeneity ratings for the models incorporating a subset of the top three textural features (difference in AUC between textural features and subjective visual heterogeneity, 0.25; 95% confidence interval: 0.02, 0.47; P = .03).

CONCLUSION

CT texture analysis can be used to accurately differentiate fp-AML from RCC on unenhanced CT images.