Prostate Cancer (Clinical Studies)
Wednesday 5 May 2010
Room A7 16:00-18:00 Moderators: Jurgen J. Futterer and Amita Shukla-Dave

16:00 472.  

Delineation and Visualization of Prostate Cancer for Targeted Radiation Therapy (Rt)
Radka Stoyanova1, Raj Rajpara1, Elizabeth Bossart1, Victor Casillas2, Jill Palma1, May Abdel-Wahab1, Alan Pollack

1Radiation Oncology, University of Miami, Miami, FL, United States; 2Diagnostic Radiology, University of Miami, Miami, FL, United States

We present an application of Pattern Recognition technique for analysis of DCE-MRI data from patients with prostate cancer and after prostatectomy. Our analysis indicates that we can detect the area of tumor burden in the prostate as well as abnormalities suggestive of residual/recurrent tumor in the prostate bed. The constructed 3D maps can be directly imported into DICOM-RT ready format to the RT planning system for targeting of the contrast enhancing areas specifically in order to improve tumor control and limit toxicity.

16:12 473.  

A Study of Endorectal MRI and MRSI of the Prostate as Predictive Biomarkers of Biochemical Relapse After Radical Prostatectomy
Kristen Zakian1, Hedvig Hricak2, Nicole Ishill3, Victor Reuter4, Steven Eberhardt5, Chaya Moskowitz3, Amita Shukla-Dave, Liang Wang, Peter Scardino6, James Eastham, Jason Koutcher
1Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 2Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 3Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 4Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY, United States;  5Radiology, University of New Mexico, Albuquerque, NM, United States; 6Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY, United States

The purpose of this study was to determine whether pre-treatment endorectal MRI/MRSI can predict biochemical relapse (BCR)  after radical prostatectomy (RP). 130 of 202 patients who underwent endorectal MRI/MRSI in 2000-2002 followed by RP satisfied data quality criteria and were followed until Jan. 2009. MRI risk score was assigned based on local disease extent.  An MRSI index lesion comprised of voxels with elevated [Cho+Cr]/Cit volume was designated. MRI risk score, MRSI index lesion volume and the presence of high grade MRSI voxels correlated with time-to-biochemical failure after radical prostatectomy even when adjusted for clinical stage, biopsy Gleason score and PSA.

16:24 474

3D Proton MR Spectroscopic Imaging of Prostate Cancer: Accuracy Evaluation in Different Prostate Regions
Stefan Zbyn1,2, Martin Krssak2, Mazda Memarsadeghi2, Klaus Kubin2, Andrea Haitel3, Michael Weber2, Thomas Helbich2, Ewald Moser1, Siegfried Trattnig1,2, Stephan Gruber1,2
1MR Centre of Excellence, Medical University of Vienna, Vienna, Austria; 2Department of Radiology, Medical University of Vienna, Vienna, Austria; 3Department of Pathology, Medical University of Vienna, Vienna, Austria

Since spectral quality of prostate 3D-MRSI data vary dramatically within measured volume, the total inaccuracies in (choline+creatine)/citrate (CC/C) ratios using Cramér-Rao lower bounds were calculated to compare the accuracy of this method between different prostate regions and measurement resolutions. Our analysis suggests that voxels from prostate base and from periphery of the prostate suffer the most from inaccurate CC/C ratios. To prevent from misleading findings or time consuming manual inspection of spectral quality in each prostate voxel, the metabolic-quality maps, that combine the information of CC/C ratio and its accuracy in one image using the various voxel transparencies, are demonstrated.

16:36 475.  

MRI-Guided Transurethral Ultrasound Therapy with Real-Time Feedback – a Human Study
Masoom Haider1,2, Laurence Klotz3,4, Michael Bronskill5,6, Kashif Siddiqui3, Alexandra Colquhoun3, Linda Sugar7, Rajiv Chopra5,6
Medical Imaging, Sunnybrook Health Sciences Center, Toronto, Ontario, Canada; 2Medical Imaging, University of Toronto, Toronto, Ontario, Canada; 3Urology, Sunnybrook Health Sciences Center, Toronto, Ontario, Canada; 4Surgery, University of Toronto, Toronto, Ontario, Canada; 5Imaging Research, Sunnybrook Health Sciences Center, Toronto, Ontario, Canada; 6Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; 7Pathology, Sunnybrook Health Sciences Center, Toronto, Ontario, Canada

MRI-guided transurethral ultrasound therapy with real-time thermometry feedback has the potential to reduce morbidity of prostate cancer therapy. To our knowledge this is the first report of the use of this technology in humans. The procedure was performed immediately prior to prostatectomy. Ultrasound energy was delivered while MR thermography was performed. The rate of rotation and output power of the applicator were adjusted by computer control. Treatment times were 9-10 minutes. The maximum temperature distribution map 55°C boundary matched the histologic section showing necrosis. It is feasible to perform accurate spatial heating of the prostate in humans using MRI-guided transurethral ultrasound.

16:48 476.

Wash-Out Gradient Derived from Dynamic Contrast-Enhanced MRI Detects Cancerous Tissues and Predicts Gleason Scores in Prostate Cancer
Yu-Jen Chen1, Woei-Chyn Chu1, W-Y Isaac Tseng2,3
1Institute of Biomedical Engineering, National Yang-Ming University, Taipei, Taiwan; 2Departments of Medical Imaging National Taiwan University Hospital, Taipei, Taiwan; 3Center for Optoelectronic Biomedicine, National Taiwan University College of Medicine, Taipei, Taiwan

DCE MRI was reported to assess microvascularity of prostate cancer, and is potentially useful to predict clinical staging. However, there are few studies demonstrating weak association between DCE MRI parameters and Gleason score. In this study, we have retrospectively analyzed the DCE MRI parameters in pathologically confirmed PCA regions. We found that washout gradient values were capable of differentiating PCA from normal tissues and best correlated with Gleason score.

17:00 477

Validation of Multiparametric Magnetic Resonance Imaging and Spectroscopy (DWI/MRSI) to Assess Prostate Cancer Aggressiveness
Thiele Kobus1, Thomas Hambrock1, Christina Hulsbergen - Van de Kaa2, Jelle Barentsz1, Arend Heerschap1, Tom Scheenen1
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands; 2Pathology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands

To validate the use of diffusion weighed imaging (DWI) and magnetic resonance spectroscopic imaging (MRSI) for tumor aggressiveness assessment, 37 patients with prostate cancer had a magnetic resonance imaging, DWI and MRSI exam on a 3T system with endorectal coil prior to prostatectomy. Individual and combined DWI and MRSI methods were used to discriminate between high- and low-grade tumors using histopathology as gold standard. Combining DWI and MRSI with linear discriminant analysis to separate tumors gave a higher sensitivity and specificity than any of the techniques separately. This suggests that DWI and MRSI provide complementary information about aggressiveness.

17:12 478

Evaluating and Relating Contrast Mechanisms in Prostate Cancer: Heterogeneity Within Normal and Tumor Regions at 3.0 T
Sharon Giles1, Sophie F. Riches2, Veronica A. Morgan1, Catherine Simpkin1, Nandita deSouza2
1Cancer Research UK and EPSRC Cancer Imaging Centre, Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom; 2Cancer Research UK and EPSRC Cancer Imaging Centre, Institute of Cancer Research & Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom

We report T2, magnetisation transfer ratios, and apparent diffusion coefficients for prostate tissues at 3T. ADC showed a lower coefficient of variation for all prostate regions compared to other parameters, indicating it to be more reliable at differentiating tumor from non-tumor in the prostate. Negative correlation of ADC with MTR suggests that diffusion-weighted contrast may be linked to features other than cellularity, with presence of large macromolecules playing a role.

17:24 479

Does Quantification of T2 SNR Decrease After USPIO Administration Allow Differentiation Between Benign and Malignant Normal Sized Pelvic Lymph Nodes?
Johannes M. Froehlich1, Benedikt Rückriem1, Maria Triantafyllou1, Frederic D. Birkhaeuser2, Michael von Gunten3, Peter Vermathen4, Harriet C. Thoeny1
1Institute of Diagnostic, Interventional and Pediatric Radiology, University Hospital, Bern, Switzerland; 2Department of Urology, University Hospital, Bern, Switzerland; 3Institute of Pathology, University Hospital, Bern, Switzerland; 4Department of Clinical Research, University Hospital, Bern, Switzerland

Methodologically the staging of lymph nodes based on uptake of USPIO is judged on a qualitative level analyzing signal decrease and distribution on T2/T2*-weighted sequences. Quantification of SI/SNR-decrease in 320 lymph nodes comprising 20 malignant lymph nodes, 57 benign inguinal and 243 benign iliacal lymph nodes revealed significant differences (p<0.05) when comparing benign with malignant lymph nodes. Contrary to iliacal lymph nodes (24.8% ± 54.6%) inguinal ones presented limited SNR decrease (3.4% ± 55.4%). Substantial overlap of single data, limit the diagnostic potential of quantification. In practice morphological criteria, fatty content and localization of lymph nodes must be considered.

17:36 480.

Tissue Segmentation Improves Prostate Cancer Detection with Artificial Neural Networks Analysis of 1H MRSI
Lukasz Matulewicz1, Jacobus F. Jansen2, Herbert A. Vargas Alvarez2, Oguz Akin2, Samson Fine2, Amita Shukla-Dave2, James Eastham2, Hedvig Hricak2, Jason A. Koutcher2, Kristen L. Zakian2
1Memorial Sloan-Kettering Cancer Center, New York, NY, United States; 2Memorial Sloan-Kettering Cancer Center, New York, NY, United States

Artificial Neural Network (ANN) model was introduced for automatic detection of tumor voxels in the prostate from 1H-MRSI datasets with additional information about tissue segmentation. The ANN’s accuracy for automatic detection of tumor voxels in the prostate MRSI datasets was demonstrated. Applying tissue segmentation from MRI as an additional input to ANN improves the accuracy of detecting tumor voxels from MRSI.

17:48 481

Absolute Quantification in 1H MRSI of the Prostate at 3T
Paul Michael Walker1,2, Gilles Créhange3, Sébastien Parfait2, Alexandre Cochet1, Philippe Maingon3, Luc Cormier1, François Brunotte1,2
1University Hospital of Dijon, Dijon, France; 2LE2I, University of Burgundy, Dijon, France; 3Radiotherapy Department, CGFL, Dijon, France

Although, it is common in MRSI to use a Choline/Citrate ratio when evaluating PCa, the use of citrate (Cit) as a reference is questionable in the context of treatment such as hormonotherapy and radiotherapy, because Cit levels fall very sharply even in non-cancerous tissue. We have proposed an absolute quantification method at 3T and we observed significantly higher Cit in normal PZ than in CG tissue. However, in PCa, reductions in Cit were not accompanied by important increases in tCho, suggesting increases in tCho/Cit are primarily due to loss of Cit and not to a sharp rise in tCho.



Back to Main Meeting

Back to Home