Qualitative and quantitative CECT features for differentiating renal primitive neuroectodermal tumor from the renal cell carcinoma and its subtypes

Objective:To identify important qualitative and quantitative clinical and imaging features that could potentially differentiate renal primitiveneuroectodermal tumor (PNET) from various subtypes of renalcell carcinoma (RCC).Methods:We retrospectively reviewed 164 patients, 143 with pathologically proven RCC and 21 with pathologically proven renal PNET. Univariate analysis of each parameter was performed. In order to differentiate renal PNET from RCC subtypes and overall RCC as a group, we generated ROC curves and determined cutoff values for mean attenuation of the lesion, mass to aorta attenuation ratio and mass to renal parenchyma attenuation ratio in the nephrographic phase.Results:Univariate analysis revealed 11 significant parameters for differentiating renal PNET from clear cell RCC (age, p = <0.001; size, p =< 0.001; endophytic growth pattern, p < 0.001;margin of lesion, p =< 0.001; septa within the lesion, p =< 0.001; renal vein invasion, p =< 0.001; inferior vena cava involvement, p = 0.014; enhan...