Which criteria can be used to predict benignity in solid renal lesions lower-equal to 2cm?

Abstract

To evaluate magnetic resonance imaging (MRI) criteria of solid renal lesions lower-equal to 2cm to differentiate benign and malignant tumors, using histopathology as gold standard. Three radiologists independently evaluated objective and subjective MRI criteria of focal renal lesions. A total of 105 nodules of patients who had MRI and histopathological results in our institution were included. Subjective criteria evaluated were signal on T2-weighted imaging, presence of microscopic and macroscopic fat, hemosiderin, hemorrhage, central scar, segmented inversion enhancement and enhancement type; objective criteria were gender, ADC value, heterogeneity on T2-weighted imaging and proportion of enhancement in late post-contrast phases. Finally, the readers classified the lesions in probably benign or malignant. Interobserver agreement was evaluated by the Gwet method, and the quantitative variables by intraclass correlation coefficients. To adjust the predictive model, the logistic regression model was used considering the benignity variable as outcome. A total of 26 nodules (24.5%) were benign and 79 (75.2%) were malignant, with size ranging from 7 to 20mm (median: 14mm). The most frequent subtype was papillary renal cell carcinoma (RCC) (35.2%), followed by clear-cell RCC (24.8%) and oncocytoma (12.4%). The univariate and multivariate analysis showed, among all categories evaluated, that microscopic fat (p: 0.072), intermediate (p: 0.004) and hyper-enhancement (p: 0.031) and female sex (p: 0.0047) had the best outcome for benignity, within odds ratios of 4.29, 5.75, 4.07 and 2.86, respectively. In small solid renal lesions lower-equal to 2cm, microscopic fat, moderate and hyper-enhancement and female sex were associated with benignity.