Time

Prog #

Junqian Xu¹, Peter A. Humphrey¹, Adam S. Kibel¹, Abraham Z. Snyder¹, Vamsidhar R. Narra¹, Joseph JH Ackerman¹, Sheng-Kwei Song¹

¹Washington University, St. Louis, Missouri, USA

Increased cellularity, and hence decreased luminal spaces, in peripheral zone PCa led to about 40% and 50% ADC decrease compared to benign peripheral zone tissues in vivo and ex vivo, respectively.  In contrast, no significant diffusion anisotropy differences were observed between the cancerous and non-cancerous peripheral zone tissues.  However, the dense fibromuscular tissues in prostate, such as stromal tissues in benign prostatic hyperplasia (BPH) in central gland, exhibited high diffusion anisotropy.  PCa identified in volume rendered diffusion sensitive MR images correlates well with histologically determined PCa foci.

Roberto Alonzi¹, Jane Taylor¹, Peter Hoskin¹, James Stirling¹, James d'Arcy², Ben Wilton², David Collins², Michele Saunders¹, Anwar Padhani¹

¹Mount Vernon Hospital, Northwood, UK; ²Institute of Cancer Research, Sutton, UK

Quantitative functional MRI techniques are increasingly being used in the assessment of the malignant prostate gland. For these techniques to have clinical application, their reproducibility needs to be established. 20 patients with prostate cancer that were due to be treated with neo-adjuvant androgen deprivation (AD) prior to radical radiotherapy underwent 2 paired MRI investigations; two prior to the commencement of and two scans after three months of AD. This study is the first to document the variability and reproducibility of T1-and T2*-weighted DCE-MRI and R2* for the benign and malignant human prostate gland.

Peter Gibbs¹, Gary P. Liney¹, Martin D. Pickles¹, Bashar Zelhof¹, Greta Rodrigues², Lindsay W. Turnbull¹

¹University of Hull, Hull, UK; ²Hull and East Yorkshire NHS Trust, Hull, UK

Prostate cancer rates have risen dramatically over the last few years and MRI is increasingly used in this area. With a diagnostic accuracy of only around 70% conventional high spatial resolution and dynamic contrast enhanced imaging has been supplemented with DWI and T2 mapping. However, there has been little investigation of the diagnostic potential of these parameters via correlation with pathological data. This work concentrates on comparing cell density determined from hematoxylin and eosin staining with ADC and T2 values.

Deanna Lyn Langer¹, ², Theodorus H. van der Kwast¹, ³, Andrew J. Evans¹, Laibao Sun⁴, Martin J. Yaffe, ²⁴, John Trachtenberg¹, Masoom A. Haider¹, ²

¹University Health Network, Toronto, Canada; ²University of Toronto, Toronto, Canada; ³Mount Sinai Hospital, Toronto, Canada; ⁴Sunnybrook  Health Sciences Centre, Toronto, Canada

The impact of normal peripheral zone (PZ) tissue components within prostate cancers on ADC and quantitative T2 values was investigated.  Whole mount sections were compared to MRI.  PZ cancers were classified as sparse or dense according to the percentage of normal PZ tissue in the tumor.  ADC and T2 values from ROIs in normal PZ, dense cancers and sparse cancers were compared.  There were no statistically significant differences seen between normal PZ and sparse cancers for absolute values or contrast (tumor to normal ratio) measurements in both ADC and T2, whereas normal PZ was significantly higher than dense cancer.  The percentage of normal PZ within a tumor may affect the detection and delineation of prostate cancer.

Roberto Alonzi¹, Anwar Padhani¹, Jane Taylor¹, James Stirling¹, Ben Wilton², James d'Arcy², David Collins², Michele Saunders¹, Peter Hoskin¹

¹Mount Vernon Hospital, Northwood, UK; ²Institute of Cancer Research, Sutton, UK

A study in 17 patients with prostate cancer evaluating the value of incorporating Dynamic Susceptibility Contrast (DSC) MRI based blood volume measurements with BOLD imaging to detect tumour hypoxia. Patients were imaged with BOLD and DSC MRI prior to radical prostatectomy. Hypoxia was determined using pimonidazole immunohistochemistry. The sensitivity and specificity for hypoxia detection was 80% and 77% respectively

Guang Jia¹, Xiangyu Yang¹, Yukihisa Takayama¹, Steffen Sammet¹, Zarine K. Shah¹, Ketul K. Shah¹, Pankaj Dangle¹, Wenle P. Wang¹, Rafael E. Jimenez¹, Vipul R. Patel¹, Michael V. Knopp¹

¹The Ohio State University, Columbus, Ohio, USA

Improving the non-invasive capabilities of cancer delineation might be achieved by dynamic contrast-enhanced MRI at high field. 27 patients with clinically proven prostate cancer were imaged at 3T without using an endorectal coil. 3 different methods including semi-quantitative parameters, adjusted Brix’s model and Larsson’s model were applied to the ROIs from tumor, non-cancerous peripheral zone, central gland. All parameters could differentiate tumor from the non-cancerous peripheral zone. However, only t_max, washout-score, k_ep could differentiate tumor from central gland. In conclusion, DCE-MRI at 3T is capable of non-invasively detecting prostate cancer especially from the central gland by selecting appropriate parameters.

Sophie F. Riches¹, Veronica A. Morgan¹, Geoffrey S. Payne¹, Cyril Fisher², S Sandhu², David J. Collins¹, Nandita M. deSouza¹

¹Royal Marsden NHS Trust & Institute of Cancer Research, Sutton, UK; ²Royal Marsden NHS Trust, London, UK

20 patients were imaged prior to prostatectomy using diffusion imaging, contrast enhanced MRI and 1H spectroscopy. Histopathology-defined tumour regions were transferred onto parametric maps to obtain values for normal and tumour tissue for T2 intensity, apparent diffusion coefficient, area under the contrast uptake curve, contrast volume transfer coefficient, reflux rate constant and the ratio of choline to citrate peak area. Significant differences were identified between the mean functional parameters of tumour and combined non tumour tissue, and between tumour and normal peripheral zone. No difference was found between tumour and normal central gland unless small tumour regions were excluded.

Bao Zhang¹, Steve Roys¹, Khan Siddiqui², Danielle Holanda, John Papadimitriou, James Borin, Michael Naslund, Eliot Siegel, Rao P. Gullapalli

¹University of Maryland, Baltimore, Maryland, USA; ²VA Maryland Health Care System, Baltimore, Maryland, USA

Conventional magnetic resonance (MR) imaging, along with 3D-MR spectroscopy imaging (MRSI) and dynamic contrast enhanced (DCE) MRI were performed on fourteen patients with biopsy proven prostate cancer prior to prostatectomy.  Parameters derived from these techniques were compared with step section histopathology. Results showed that both techniques are able to distinguish cancerous tissue from normal but they however lacked sensitivity to tumor grade. Receiver operating characteristic (ROC) based statistical analysis revealed a positive predictive value of 0.95 and 0.94 respectively and an overall accuracy of 72% and 65% respectively.

Gregory John Metzger¹, Carl J. Snyder¹, Can Akgun¹, Gregor Adriany¹

¹University of Minnesota, Minneapolis, Minnesota, USA

Previous work by our group addressed the challenges associated with performing ultra high field prostate imaging with external transceive surface arrays.  In this abstract, we pursue a preliminary study which investigates how the external coil array performs against a transceive endorectal coil (ERC) at 7 Tesla.  It was found that the ERC produced the expected gains in SNR along with improved anatomic detail on T2w TSE imaging.  While improvements in image quality were observed the challenges of managing the large transmit B1 homogeneities of a local transceive coil requires further investigation.