EAU 2018: MR Spectroscopy Distinguishes Histologically-diagnosed Kidney Tissue

Copenhagen, Denmark (UroToday.com) Current imaging modalities, including ultrasound, CT and MRI, unfortunately, are not effective at characterizing tumor histologic subtype or predicting clinical aggressiveness, outside of published clinical nomograms. Clinical staging still remains the gold standard for decision making.

However, as there is increasing interest in conservative management, knowing that there is a population of renal masses that appear to be indolent, exploration of novel imaging modalities to better characterize renal masses and help guide decision making is underway.

One such technology, which the current group explores, is MR Spectroscopy – a combination of MRI and mass spectrography to characterize lesions based on imaging appearance and clustering of known metabolites.

The specifics of the technology are as follows: “Two dimensional 1H homonuclear COrrelation SpectroscopY (COSY) spectra were acquired (Agilent Technologies MR400-DD2 NMR Spectrometer, running VnmrJ 4.2). Data were processed and analysed using FelixNMR and cross and diagonal peak volumes measured for 128 metabolites. The choline diagonal peak at 3.2ppm was the internal chemical shift reference. Metabolite peak volumes were normalized to tissue weight. Data were log2 transformed and quantitative data visualization and multivariate analyses conducted using GMine.”

The authors used 96 renal tissue samples, including tumor and non-tumor tissues, all of which were collected post nephrectomy. 20 samples were from small renal masses. Histopathology was used as gold standard diagnosis.

Final histopathology of the 96 specimens included 38 clear cell renal cell carcinoma (RCC), 7 chromophobe RCC, amongst other benign pathologies.

Principle components analysis showed clusters based on tissue type – malignant, benign or non-cancerous. Canonical correlation analysis (CCA) demonstrated that the clustering between clear cell and non-clear cell malignant histology was significant, p= 0.001. This held true even in small renal masses. In particular, it demonstrated that there was higher metabolite resonances (such as cholesterol) in clear cell RCC.

However, what the analysis did not show was discrimination between histologic subtypes other than clear cell RCC. This would be more useful than malignancy vs. benign tumor, though invariably this is also useful information. However, with renal tumor biopsy become increasingly safe and improved accuracy, the burden of evidence required for this new technology is quite high.

Speaker: S. Del Vecchio

Co-Author(s): Urquhart A., Krause L., Ellis R., Ng K., Samaratunga H., Gustafson S., Galloway G., Gobe G., Wood S., Mountford C.

Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, Twitter: @tchandra_uromd at the 2018 European Association of Urology Meeting EAU18, 16-20 March, 2018 Copenhagen, Denmark