Habitats in Prostate Cancer

 

Project Summary:

Prostate cancer is one of the most common cancers in men and accounts for the second largest number of cancer related deaths amongst men (1). Although the 5 years survival rate remains high at 98.9%, if the disease is not diagnosed and treated could lead to metastasis leading to death. The 5 year survival of prostate cancer that spread to other parts of the body drops to 28%. Magnetic Resonance Imaging (MRI) is the conventional imaging modality used as a diagnostic aid to identify abnormalities and to target the tumor lesions for pathological evaluation.

We believe detecting the tumor in a MRI scan with certain level of confidence will provide a non-invasive option to the patients and the oncologist. Creating a risk assessment on these imaging would then provide a quantitative aid to the oncologist that will help decide between invasive biopsy and treatment options.

 

Figure 1. Prostate cancer patient with multiple sub-regions (habitats) identified on T2w (top left), ADC (top right), T1w (single phase) (bottom left) and time activity map (region) (bottom right) of the MRI scan.

Currently, image evaluation is used for detection purpose, although there are few clinical assessments based on genomics, and clinical factors that are used to aid the oncologist. We believe quantitative imaging will provide sufficient information to develop risk assessment that can be used to decide on patients active treatment or wait and watch for the disease progression.

It had been shown by various genomics studies that tumors are heterogeneous population composed of multiple clonal populations. Which is certainly visible on a MRI scan, most of the tumors appear with different shades. These populations are similar to ecological habitats that are seen on an aerial satellite image. We identify these sub regions based on statistical and mathematical functions on the MRI image.

 

In our lab we are in the pursuit to following research hypothesis.

  • We propose to use quantitative imaging (radiomics) metrics to quantify regions of prostate lesion seen on the multi-parametric Magnetic Resonance Imaging (mpMRI).
  • We propose to build statistical models based on the imaging features to provide prognostic and predictive risk assessment to the prostate cancer patients.
  •  We would like to evaluate perfusion characteristics of these MRI imaging as a risk factor to the disease progression.

References:

  1. Cancer Facts & Figures 2016, Altanta, American cancer society, 2016.
  2. RJ. Gillies, PE. Kinahan, H. Hricak, Radiomics: Images are more than picture, they are data, radiology, 278 (2), 2016.
  3. Y.Balaguruanthan, Y.Gu, H.Wang, W.Kumar, O. Grove, S. Hawkins, J. Kim, D. Goldgof, L. Hall, R. Gatenby, R. Gillies, Reproducibility and prognosis of quantitative features extracted from CT images, Tran. Onc, 7(1), 2014.
  4. R.Stoyanova, M. Takhar, Y. Tschudi, J, Ford, G. Solorzona, N. Erho, Y. Balagurunathan. S. Punnen, E. Davicioni, R. Gillies, A. Pollack, Prostate cancer radiomics and promise of radiogenomics, TCR, 5(4), 2016.
  5. G. Litjens, O Debats, J Barentsz, N. Karssemeijer, H. Huisman, Computer-aided detection of prostate caner in MRI, IEEE Trsn. Med. Img, 33(5), 2014.
  6. NA. Parra, A. Orman, K. Padgett, V. Casillas, S. Punnen, M. Abramowitz, A. Pollack, R. Stoyanova, Strahlebther Onkol. 193(1), 2017.

 

Funding Agencies

PQ Grant (NCI), R01/U01 (NCI)

 

Project Members and Collaborators

 

Yoganand Balagurunathan, PhD

Applied Research Scientist

Moffitt Cancer Center:

Urology, Radiology, Pathology, Cancer Physiology

University of Miami:

Radiation Oncology