Characterization of Renal Tumors: Initial Experience Integrating Biomechanical and Morphological Assessment Using 3 Tesla Magnetic Resonance Imaging and Elastography (MRE)
Davide Prezzi1, Radhouene Neji2, James Stirling1, Sami Jeljeli1, Hema Verma3, Tim O'Brien4, Ben Challacombe4, Ashish Chandra5, Vicky Goh1, and Ralph Sinkus1

1Division of Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom, 2MR Research Collaborations, Siemens Healthcare, Frimley, United Kingdom, 3Department of Radiology, Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom, 4Urology Centre, Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom, 5Department of Histopathology, Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom


Incidentally detected renal tumors are overtreated surgically, as up to 15% of them are benign, most frequently oncocytomas. We hypothesize that integrating biomechanical with morphological MRI assessment can improve lesion characterization, precluding unnecessary surgery. Initial experience and pathological correlation in four resected renal oncocytomas and renal cell carcinomas (RCC) demonstrate that 30Hz MRE with shear modulus elastography parametric mapping is feasible, correlating spatially with gross pathology, with lower viscosity/elasticity (y) ratios [mean = 0.22] in malignant RCCs compared to oncocytomas [mean = 0.46], showing promise for clinical application.


· To assess the feasibility of 3T Magnetic Resonance Elastography (MRE) for evaluating renal tumors with surgical pathology as the gold standard

· To correlate the distribution of MRI shear modulus elastography maps to same-level MRI anatomical sequences including high-resolution HASTE and post contrast VIBE.

· To correlate the distribution of MRI shear modulus elastography maps to gross pathological morphology: cellular distribution, vascular architecture and areas of necrosis to determine how the underlying architecture affects biomechanical properties.


With the growing use of cross-sectional imaging, the incidence of incidental asymptomatic small renal tumors (≤ 4 cm) has risen steadily [1, 2]. Although most of these lesions represent renal cell carcinoma (RCC), up to 20% will be benign [3]. A recent estimate suggests that more than 5600 benign renal tumors are resected yearly in the United States [4]. Renal oncocytoma is the most common benign lesion, accounting for up to 15% of all renal tumors, and cannot be differentiated from RCC reliably by conventional imaging [5, 6]. On gross pathology oncocytomas have a characteristic macroscopic appearance of homogenous dark brown lesions with a frequent central scar and absence of necrosis. Microscopically they consist of tight cellular nests surrounded by myxoid or hyalinized stroma. Conversely, clear cell RCCs, the most common subtype, consist of soft yellow material alternating with fibrous or mucoid areas, and are frequently hemorrhagic and necrotic. Microscopically they are composed of clear cells, rich in lipids and glycogen, surrounded by an extensive capillary network.


We hypothesize that renal tumor MRE biomechanical properties are determined by the underlying tumor architecture, including their cellular distribution, vascular architecture and nature of the extracellular-extravascular space.


Following IRB approval, informed consent, and initial MRE optimization in 5 healthy volunteers, 5 patients with renal tumors (≤ 4 cm) scheduled for partial or total nephrectomy were recruited prospectively. In addition to standard MRI (T2 HASTE, DWI and post contrast T1 VIBE), MRE was performed on a 3T MRI system (Biograph mMR, Siemens Healthcare, Erlangen, Germany) based upon a prototype 2D multi-slice interleaved gradient echo sequence synchronized with the transducer's vibrations, TE = 7.38 ms, motion encoding gradient amplitude = 30 mT/m and GRAPPA acceleration factor = 2 [7]. Mechanical vibrations at 30Hz were transmitted from the mechanical transducer sited over the kidney of interest. Four consecutive breath-holds of 17 sec each (3 motion-encoding directions and one reference scan) provided MRE data within 6 consecutive slices of 128 x 88 pixels at 3 mm isotropic resolution and 4 wave phase offsets. Reconstruction of viscoelastic parameters used firstly the application of the curl operator for removal of the compressional component, secondly an iterative method comparing the measured with the simulated curl-field within a subzone of 11 x 11 x 5 pixels using the data as boundary condition on the surface of the subzone (similar to [8]). Viscoelastic parametric maps, including storage modulus (elasticity, kPa) and viscosity/elasticity ratio [y=2/π*atan(Gl/Gd)] were generated and regions of interest (ROI) drawn around tumors.


One patient was unable to tolerate the MRI. Four MRI datasets, including MRE, were successfully acquired. Mechanical vibrations showed good unilateral wave penetration (image 1). Histological assessment revealed 2 renal oncocytomas (images 2 and 3) and 2 clear cell RCC (images 4 and 5). The two oncocytomas showed homogeneous intermediate T2 HASTE signal with central hyperintensity and homogeneous contrast enhancement on T1 VIBE. RCCs showed spatially heterogeneous T2 HASTE signal and T1 VIBE contrast enhancement. Mean ADC and Minimum ADC values were not obviously discriminatory, ranging between 1258.16 - 1823.29 x 10-6 mm2/s and 381.00 - 861.00 x 10-6 mm2/s respectively across tumors. Areas of necrosis and the central scar were excluded from MRE analysis. Mean elasticities were 1.14 and 2.50 kPa; y ratios were 0.52 and 0.40 for each oncocytoma respectively. Mean elasticities for each RCCs were higher at 3.50 and 3.60 kPa; corresponding y ratios were lower at 0.3 and 0.15 kPa respectively.


Our initial exploratory data suggests that the biomechanical properties differ between clear RCC and oncocytoma, with higher mean elasticity and lower viscosity/elasticity ratios in RCCs, raising its potential for clinical application. Ongoing prospective recruitment with pathological correlation will provide confirmatory evidence.


The authors acknowledge financial support from the Royal College of Radiologists through the Clinical Radiology Pump Priming Grant scheme; from the Department of Health via the National Institute for Health Research Comprehensive Biomedical Research Centre award to Guy’s and St Thomas’ NHS Foundation Trust, in partnership with King’s College London and King’s College Hospital NHS Foundation Trust; and from the King’s College London/University College London Comprehensive Cancer Imaging Centre funded by Cancer Research UK and Engineering and Physical Sciences Research Council, in association with the Medical Research Council and Department of Health.


1. Hollingsworth JM, Miller DC, Daignault S, et al. Rising incidence of small renal masses: a need to reassess treatment effect. J Natl Cancer Inst. 2006; 98: 1331–1334.

2. Gill IS, Aron M, Gervais DA, et al. Clinical practice. Small renal mass. N Engl J Med. 2010; 362: 624–634.

3. Frank I, Blue ML, Cheville JC, et al. Solid renal tumors: an analysis of pathological features related to tumor size. J Urol. 2003; 170: 2217–2220.

4. Johnson DC, Vukina J, Smith AB, et al. Preoperatively misclassified, surgically removed benign renal masses: a systematic review of surgical series and United States population-level burden estimate. J Urol. 2015; 193: 30–35.

5. Pierorazio PM, Hyams ES, Tsai S, et al. Multiphasic enhancement patterns of small renal masses (<=4 cm) on preoperative computed tomography: utility for distinguishing subtypes of renal cell carcinoma, angiomyolipoma, and oncoctyoma. Urology. 2013; 81: 1265–1271.

6. Pedrosa I, Sun MR, Spencer M, et al. MR imaging of renal masses: correlation with findings at surgery and pathologic analysis. Radiographics. 2008; 28: 985–1003.

7. Garteiser P, Sahebjavaher RS, Ter Beek LC et al. Rapid acquisition of multifrequency, multislice and multidirectional MR elastography data with a fractionally encoded gradient echo sequence. NMR Biomed. 2013; 26(10): 1326-1335.

8. Van Houten EE, Miga MI, Weaver JB et al. Three-dimensional subzone-based reconstruction algorithm for MR elastography." Magn Reson Med. 2001; 45(5): 827-837.


Axial T2 HASTE and shear wave velocity map through normal renal cortex demonstrates good wave penetration; the renal parenchyma is clearly distinguishable from the surrounding fat. The simple cyst arising from the posterior renal cortex has not been included in the ROI.

Axial T2 HASTE, post contrast T1 VIBE, corresponding level cut specimen and parametric elastography maps of renal oncocytoma 1.

Axial T2 HASTE, post contrast T1 VIBE, DWI (b = 800), ADC map and parametric elastography maps of renal oncocytoma 2.

Axial T2 HASTE, post contrast T1 VIBE, DWI (b = 800), ADC map and parametric elastography maps of RCC 1.

Axial T2 HASTE, post contrast T1 VIBE, DWI (b = 800), ADC map and parametric elastography maps of RCC 2.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)