Login to Access Video or Poster Abstract: MP18-12
Sources of Funding: none

Introduction

Due to the differences in tumor behavior and prognosis, discriminating different subtypes of renal cell carcinoma (RCC) is important. The purpose of this study was to investigate the role of diffusion weighted imaging (DWI) derived apparent diffusion coefficient (ADC) maps in differentiating clear cell renal carcinoma (ccRCC) from papillary subtype (pRCC).

Methods

ADC maps from 97 renal cell carcinoma lesions (20 papillary and 77 clear cell type) from 42 patients were segmented for volumetric and pixel based histogram analysis. Mean, standard deviation, skewness and kurtosis and different quantiles of the histogram were calculated for the segmented lesions. The standard of reference for diagnosis of different types of RCC was histopathology of surgical specimens. Receiver operating characteristic (ROC) analysis was performed for each of the extracted features.

Results

Amongst all the above mentioned features, the analysis of the quantiles yielded the highest sensitivity and specificity for differentiation of the two subtypes. The ADC range was between 258 to 3407 and 246 to 3686 mm2/sec for pRCC and ccRCC, respectively. Quantile 30 by ADC threshold of 1500 resulted in 96% sensitivity and 84% specificity that provided the highest sensitivity and specificity among the all features. Area under the curve (AUC) was 0.95.

Conclusions

Volumetric pixel based analysis of the ADC maps is an objective method that can accurately differentiate pRCC from ccRCC. _x000D_ _x000D_

Funding

none