Cancer Imaging Research Program

The Cancer Imaging Research Program is focused on two areas of research: (1) developing new medical imaging acquisition and analysis tools, and (2) developing new molecular “probes” targeted to detecting specific types of cancer. Both research foci are aimed at providing new information for cancer diagnosis, for cancer treatment and to help develop new cancer treatments.

Molecular Imaging Probes

The aim is to adapt medical imaging methods to improve the specificity – ensuring that the imaging probe preferentially reaches the correct target cancer tissue, and sensitivity – improving signal levels so that smaller tumours can be detected. BIRC researchers and colleagues are adapting magnetic resonance imaging and computed tomography and developing targeted probes for positron emission tomography to improve earlier detection of tumours of the prostate and liver, when tumours are small and therapy can be more effective. They also are using novel MR methods to track single cells of the immune system, which is aimed at using the body's own cells to enhance immunity to target and attack the tumour.

BIRC has an extremely strong group of scientists based at Lawson working on molecular probes for a variety of purposes. There is more about their research on Page 50 of this report.

Breast Cancer

Breast cancer is the most common cancer afflicting Canadian women and the second leading cause of cancer death. It is estimated that one in eight Canadian women will develop breast cancer during their lifetime.

The cancer imaging program is developing novel parallel Magnetic Resonance (MR) imaging approaches, particularly for younger women for whom conventional X-ray mammography is often inconclusive due to dense tissue. MR imaging does not use ionizing radiation and provides excellent soft tissue contrast and multi-planar cross-sectional images, which have been shown to be useful for distinguishing malignant and benign lesions using quantitative dynamic gadolinium-enhanced MR imaging. Unfortunately, the temporal demands placed on dynamic MR imaging methods results in a loss of spatial information (i.e. coverage), which can result in a missed lesion or an incomplete bilateral breast examination, particularly important for directing a breast biopsy. To address these issues, Sentinelle Medical Inc. (Toronto) developed a system that navigates to a biopsy target using ultrasound while displaying the co-registered 3D MRI image. BIRC scientists have been working with this equipment on a preliminary clinical study to assess the overall accuracy of the targeting guidance.

A 3D ultrasound-guided breast brachytherapy device is also being developed by BIRC scientists.

Investigators within BIRC also are examining optical methods to better detect and identify breast cancer.  One team is developing a transportable 3D photoacoustic imaging system suited to imaging the human breast. Photoacoustic imaging offers a method to visualize blood rich tissues such as the vessels surrounding tumours and measure the oxygen levels nearby. By comparing biopsy results to photoacoustic measures of blood content and oxygen levels, the BIRC team plans to determine if 3D photoacoustic imaging improve the accuracy of identifying breast abnormalities. 

BIRC researchers are also working together with researchers from Simon Fraser University to develop a hyperspectral optical imaging technique to examine thick tissue sections obtained during breast conserving surgery. The goal of the project is to determine if tumour can be detected within surgically removed tissues during surgery. This potential capability will provide the surgeon with better confidence that the complete tumour mass has been removed and reduce the need for subsequent surgeries.

Prostate Cancer

Imaging plays an important role in early diagnosis, treatment selection and therapy delivery for prostate cancer, the second leading cause of cancer death in men. The cancer imaging program is developing new imaging probes to improve the sensitivity of detection of prostate cancer using magnetic resonance imaging (MRI). In one approach, probes are bound or trapped within the cancer cells. The probes increase the signal intensity of nearby water molecules enabling detection of the tumour. In a second approach, hyperpolarized carbon-13 contrast agents provide a unique signal that does not have to compete with background signals in the body, and offers the potential to enable detection of aggressive cancers at a much earlier stage. This probe enables direct imaging of elements (e.g. pyruvate) in the metabolic pathways within cells, which are upregulated in aggressive cancer long before the tumour has reached significant size. In a third approach currently being tested in patients, sodium and hydrogen within the prostate are imaged using a single integrated MRI coil. Our scientists have also identified a receptor, GHS-R, which differentiates prostate cancer cells from benign prostatic hyperplasia, and are developing targeted PET probes based on peptide analogues that bind to the receptor. Additionally, our scientists are key participants in the CIHR Team in Image Guidance for Prostate Cancer, developing integrated multi-modality clinical imaging for prostate cancer diagnosis and therapy.

Prostate MRI (left) showing cancers confirmed on an aligned microscopic gold standard (right).

Brain Cancers

Conventional MRI and CT imaging of the brain provide excellent anatomic detail of primary tumours of the brain but are limited in their ability to discern important biologic characteristics of the tumour and to separate tumour from treatment effects. In addition image-guided therapies can provide improved outcomes for patients with cancer metastatic to the brain. BIRC investigators are investigating novel forms of multi-parametric MRI imaging, hyperpolarized 13C imaging and Na imaging, dynamic contrast enhanced imaging and PET imaging to provide novel biologic information regarding brain tumors. High field (7T) MRI imaging is being utilized to detect vascular effects potentially associated with treatments like radiotherapy. New forms of brain metastases imaging and therapy are being developed to detect metastatic cancer earlier, help guide therapies and predict response.

Liver Cancer

Life expectancy for patients with cirrhosis of the liver and hepatocellular carcinoma is short, with fewer than half the patients surviving two years. The current standard treatment is surgical resection and this has the highest success rate of all treatment methods for the primary and metastatic liver cancer. However, only 8-30 percent of patients diagnosed with HCC are candidates for surgery. A factor contributing to this poor rate of curative surgery is the inability to detect the cancer when the tumour is less than two centimetres in diameter, when patients have a better chance of survival following surgery or a transplant. The aim of this research is to develop more sensitive detection of blood flow in the diseased liver using computed tomography (CT). Hepatocellular carcinoma progresses from cirrhotic nodules to pre-cancerous regenerative nodules before becoming malignant. The blood supply for the normal liver is mainly from the portal vein that drains nutrient rich but partly deoxygenated blood from the gut, whereas in the progression from cirrhotic nodules to pre-cancerous regenerative nodules to hepatocellular carcinoma, the blood supply is coming more from the aorta and its branches with their fully oxygenated blood than the portal vein.

For non-resectable liver tumour, focal ablative therapy including trans-arterial chemoembolization (TACE) and radiofrequency ablation (RFA) has become the treatment of choice. The basis for TACE treatment is that liver tumour has high arterial blood flow supplied by arterial routes as discussed above. When these routes are cut off by embolization, the tumour will be starved of oxygen and nutrients and die. However, if the tumour has high portal blood flow via portal venous routes like normal liver parenchyma, then it will survive the procedure and treatment will fail. Based on these considerations, our hypotheses are that: (1) high portal blood flow before TACE is predictive of treatment failure; while (2) low baseline portal blood flow plus post treatment decrease in arterial blood flow compared to baseline is predictive of good treatment response. We are developing dynamic contrast enhanced CT (DCE-CT) liver scanning techniques to derive imaging biomarkers that can be used to predict and monitor the effect of TACE. The technical objective of the project is to develop DCE-CT scanning methods to measure arterial and portal blood flow in liver tumours. 

Arterial phase CT, (total) blood flow, arterial blood flow and portal blood flow prior (baseline) to and at 2nd and 4th week post TACE with lipiodol. This patient is a responder to TACE treatment. The patient’s imaging data shown here supports our working hypothesis that low portal blood flow at baseline portends good response to treatment.  Arrows (yellow) point to liver tumor which had been embolized with Lipiodol. Baseline portal blood flow was low (red arrow). After embolization, blood flow, arterial blood flow and portal blood flow all remained low.

BIRC researchers have developed a free-breathing axial shuttle scanning protocol to enable current clinical CT scanners whose detector width is only 4 cm wide to repeatedly scan an 8 cm section of the liver by moving the patient couch back and forth between two contiguous 4 cm wide locations during 2 min of scanning without breath hold. This protocol can be easily incorporated into routine clinical CT liver scanning studies without requiring special hardware/software or patient cooperation; thus, the adoption of this scanning protocol in clinical practice is greatly enhanced. In addition, we have extended our CT Perfusion software, which has been licensed to GE Healthcare, to calculate arterial and portal blood flow from the DCE-CT image data. A clinical trial has been started on the application of DCE-CT derived arterial and portal blood flow in TACE.

Other research on hepatocellular carcinoma (HCC) is the development of a real-time 3D-ultrasound based guidance system to facilitate targeting malignant tumours for radiofrequency (RF) or microwave (MW) ablation procedure, thereby improving outcome (reducing the recurrence rate) while reducing the overall time the procedure requires. RF and MW convert absorbed radiofrequency and microwave energy, respectively, into heat that coagulates and destroys tumours. For this purpose, an antenna must be implanted accurately into the liver tumour. As the tumour moves with respiration, real-time imaging is required to localize the tumour during the insertion of the antenna. Currently, CT scanning is used for the image guidance, but this is not cost effective as it uses this imaging resource for a significant period of time. Our real-time 3D-ultrasound system is designed to replace CT scanning for image guidance in RF and MW ablation procedures, and has the advantages that it can be used at the bedside of patients and does not require the use of x-rays as in CT scanning.

BIRC scientists in another research project are focused on the development of a 3D-ultrasound based system to facilitate targeting malignant tumours and improve outcome of percutaneous ablation procedures while reducing the overall time required. Image-guided focal ablation is rapidly growing as a treatment option for localized liver cancers deemed inoperable. These techniques allow better conservation of hepatic reserves with fewer side effects. As a result, treatment guidelines have recently adopted ablation as the first-line treatment in early stage tumors when transplantation is not an option. This in particular is important, as better diagnostic imaging techniques and screening programs are detecting more cancers at earlier stages.

Currently focusing on guiding microwave and radiofrequency ablation procedures, the developed system provides all components for a thermal ablation procedure from planning to intra-operative guidance and treatment follow-ups. The system consists of automatic tools for pre-operative imaging and multi-modality planning, as well as automatic or semi-automatic localization of targeted tumours and ablation probes for intra-operative guidance.

Following successful laboratory and in vivo animal testing, the developed system is now used in two phase 1 clinical trials (IRB approved) with more than 50 patients enrolled since 2011. Further development is currently ongoing based on previous findings with the addition of new features and improvements to the existing techniques.

Assessment of expected microwave ablation margins using 3D ultrasound to ensure complete tumor coverage with no additional damage to the normal tissue.

Cell Tracking

There is considerable interest in harnessing the body’s own immune system to target and attack cancer, and harvested immune cells can be re-introduced into the body (cell-based immunotherapy) to boost the immune system. There is early evidence that these cell-based vaccines are technically feasible in humans and are non-toxic; however, significant barriers remain to be overcome before the use of cell-based vaccines is reliable. The aim of this research is to analyze the trafficking and accumulation of human immune cells in vivo using high resolution MRI and iron oxide based cell labels. For dendritic cell therapy, the team plans to use this strategy to monitor the ability of dendritic cells to boost the immune system. For natural killer cell therapy, they aim to use cell tracking to optimize the accumulation and invasion of these cells in tumours.

Image-Guidance and Image Analysis for Cancer and Clinical Trials

Imaging is used to improve and guide prostate, breast, and liver biopsies, allowing early diagnosis at a stage when the cancer is curable and the disease can be staged accurately, and to develop new technology to improve minimally invasive therapy. In addition, the team is creating new ways to look at cancer treatment responses with a special focus on lung cancer, the largest cancer killer of men and women in Canada.

3D Ultrasound-Guided Prostate Biopsy

Definitive diagnosis of localized, early stage prostate cancer has a significant false-negative rate (15 to 25%), which means that men who actually harbour curable prostate cancer are not detected on the first biopsy. The physician then is faced with a difficult challenge, requiring imaging with other modalities and a second and sometimes a third biopsy. Researchers in the cancer imaging research team have invented a 3D ultrasound-based system to improve planning and recording of the exact 3D biopsy coordinates in the prostate, which will help to resolve these issues, especially when suspicious tissue results require sampling from the same location as the initial biopsy. In addition, this new technique can make use of images from other imaging modality (e.g., MRI or PET/CT) to guide the biopsy needle to the target.

3D Ultrasound-Guided Prostate Therapy

Managing the increasing number of men with early stage cancers has generated a great deal of debate with a growing belief that aggressive therapy may not be justified for early stage disease, and that minimally invasive therapy, such as cryosurgery, brachytherapy, thermal therapy, or photodynamic therapy is a viable option. These strategies promise to reduce patient morbidity, recovery time, hospital stay, and overall cost, while preserving or increasing clinical efficacy. Imaging technology for real-time treatment guidance and monitoring is critical to the accurate delivery of these therapies and to their safety and effectiveness.

Our researchers have pioneered a 3D prostate US imaging for minimally invasive prostate cancer therapy that uses robotic aids and innovative real-time image guidance and verification tools, allowing all aspects of the procedure to be carried out intra-operatively for use in saturation biopsy, low dose-rate brachytherapy, high dose-rate brachytherapy, cryotherapy, thermal therapy, and photodynamic therapy.

3D Multi-modality Imaging and Pathology for Prostate Cancer Diagnosis and Treatment

The CIHR Team in Image Guidance for Prostate Cancer within BIRC seeks to generate co-registered 3-dimensional, multi-modality, pre-operative prostate cancer images (MRI, CT, US, PET) with post-operative 3-dimensional digitized pathology images to develop predictive models of prostate cancer growth and aggressiveness to aid the planning and delivery of minimally invasive therapy. The development of accurate 3-dimensional predictive models of prostate cancer based on non-invasive, clinically available imaging techniques will be integrated with other technology platforms being developed within BIRC among the CIHR Team partners (3D ultrasound-guided biopsy, minimally invasive therapies such as cryotherapy, laparoscopic surgery, and image-guided radiotherapy) to improve outcomes for men with prostate cancer. The capabilities being developed through the CIHR Prostate Cancer Team (accurate co-registration and statistical analysis of three-dimensional imaging and digital pathology) provides capabilities that can be leveraged for other cancers and diseases.

MR-Guided Focal Prostate Therapy

There is a growing belief among many clinicians that we are over-treating PCa with radical treatments and creating unnecessary morbidity particularly for low-risk disease where patients and physicians want treatments offering excellent tumour control with minimal morbidity. This therapy approach could provide definitive management with least treatment related morbidity. Although focal PCa therapy is being used in a few centres, its general acceptance is hampered as current planning and guidance techniques are still susceptible to variability and inaccuracy due to factors such as: inadequate lesion localization, inability to accurately guide the needle delivering the therapy to the target, the patient’s anatomy (pubic arch interference), and anatomical changes during the procedure, such as prostate motion.

The image-guidance group at BIRC is developing a 3D ultrasound-guided transperineal focal prostate therapy platform integrating robotic aids, multi-modality image fusion, and real-time image processing. The goal is to develop and validate an accurate and precise focal prostate therapy system that will operate in the bore of an MRI scanner and will include planning with pre-procedural MR image registration, monitoring of the procedure and prostate changes, and optimal needle guidance.

Image-Guided Prostate Resection

The most widely used treatment technique for prostate cancer is radical prostatectomy, with more and more procedures being performed using the daVinci surgical robot.

However, during surgery the surgeon only has access to the laparoscopic view of the prostate, but is unable to “see” the distribution of the tumour within the organ. While the objective of the surgery is to remove the entire organ, care must be taken to spare sensitive tissue such as the neurovascular bundles close to the prostate capsule. Unfortunately, when there is suspicion that the cancer location is close to, or even erupting through, the capsule, a larger resection margin must be allowed to ensure all the cancerous tissue is removed. This may also mean that the neurovascular bundles must be sacrificed, resulting in erectile dysfunction.

This research will integrate an image of the cancer distribution obtained from MRI with the image the surgeon sees in the operating room via the robot’s camera. With this knowledge, the surgeon will be able to adopt a resection strategy that spares the important nerve bundles responsible for erectile function and urinary control when there is no danger of cancer cell invasion in the surrounding regions.

Currently, these structures are often destroyed bilaterally in an effort to create a “safe” margin around the entire prostate to maximize resection of extra-capsular disease.  

Image-Guided Adaptive Radiotherapy of the Prostate

Advances in technology during the past decade have resulted in more precise targeting of cancer with conformally-shaped radiation beams. Multimodality imaging (CT, MRI, SPECT, PET) has improved treatment planning, and online CT imaging in the radiotherapy room has become commonplace for targeted intensity-modulated radiotherapy (IMRT). Advances in online CT image-guidance allow precise adjustments of the daily patient setup over periods of weeks. CT enables soft tissue visualization to correct for tissue plasticity and dose deformation.

BIRC researchers in this area are studying to what extent radiation dose distributions delivered to individual patients (in vivo) diverge from the planned dose distributions (in silico), and if there is a robust schedule of CT image guidance, with or without in-flight dose re-optimization that will mitigate these deviations to an acceptable level. They are using indicators of the 3D dose distribution compliance, including dose-volume histograms (DVH), with probabilities of tumour control (TCP) and normal tissue complications (NTCP) as surrogates for expected clinical outcomes.

Ultrasound and MRI-Assisted Laparoscopic Cancer Surgery

In many cases where cancer tumours are either resected from one organ (e.g. kidney), or removed by radical excision of the organ itself (prostate), it is important to know the location of the tumour relative to the incision. In the former case, this allows maximal sparing of the decisions regarding the extent of the required margin in relation to the neurovascular bundles. The group is addressing this problem by integrating ultrasound and pre-operative MR information into the space visualized by the laparoscope, allowing the surgeon to “see through” the organ surface to visualize the tumour.

Better Imaging of Lung Cancer for Better Lung Cancer Treatment and Survival

The prognosis for patients with lung cancer who are not candidates for surgery is dire, with no significant improvements in response and survival over the last 30 years. In other words, the promise of highly precise, computerized, and improved therapy, imaging, and radiation methods has not been realized for patients with lung cancer. To directly address this issue, the team is developing new ways to measure lung tumour treatment response from diagnostic images. These methods are rapid, automated, and independent of specialists, and have the goal to include these measurements in clinical trials of new treatments. Another goal is to provide a rapid measure of the tumour before therapy to help target personalized treatment, and to evaluate changes in tumours in response to treatment.

Many outstanding questions remain related to the best methods used to measure tumour response to therapy. What happens to tumours when they respond to therapy? What differentiates those tumours that respond over longer periods of time from those that do not? These questions are particularly pressing in the measurement of lung cancer, because the efficacy of conventional therapies is mediocre. Improved methods to measure lung tumour response to new molecular-targeted agents are required, and the development of such techniques will facilitate similar measurements for liver, brain, and other cancers.

Optimized Treatment for Oropharyngeal Cancers

The incidence of oropharyngeal cancers is rising dramatically and is affecting younger individuals who are non-smokers and non-drinkers. Traditional therapies with radical surgery or chemotherapy and radiation carry significant side effects. BIRC investigators, in combination with members of the Department of Otolaryngology and Head and Neck Surgery and Oncologists at the London Regional Cancer Program, are conducting a unique clinical trial comparing state-of-the-art minimally invasive surgery using Robotic Assisted Transoral Surgery against state-of-the-art image-guided, intensity modulated radiotherapy with chemotherapy. Outcomes being monitored include cancer control and side effects. The randomized trial is funded by the Canadian Cancer Society and is the only such trial in North America.

Improving Outcomes for Patients with Cancer Metastatic to Brain

Cancer metastatic to brain is a common cancer problem and one where combinations of high dose radiation or surgery with whole brain radiotherapy can provide improved cancer control and prolong life. BIRC investigators are conducting a unique Phase II trial examining a novel “simultaneous boost technique” that escalates the dose to individual brain metastases while simultaneously sterilizing microscopic brain metastases using lower dose radiation to the whole brain. This OICR funded trial will demonstrate if this technique, which is more resource efficient and convenient for patients, is as effective as historical results with surgical or radiosurgical results. BIRC investigators are also examining the imaging appearance of brain metastases to allow better prediction of response to radiation treatments, and with collaborators at the VU, Amsterdam, are conducting comparative effectiveness research of this novel radiation technique against traditional treatments for brain metastases.

Treating Oligo-Metastases

Over the past 20 years, there has been increasing interest in the ‘oligo-metastatic state’, which describes patients who have a limited number metastases (e.g. 1-5), in distinction to patients who have more widespread metastatic disease. In theory, patients who have oligo-metastases may enjoy long-term survival if all metastatic deposits can be eradicated, either with surgery or stereotactic radiation. Image-guided stereotactic ablative radiation (SABR) can achieve high rates of control of metastatic deposits. BIRC researchers are leading an international randomized clinical trial called SABR-COMET, which is attempting to measure the benefit of a comprehensive approach of eradicating all metastatic deposits using image-guided SABR, for patients with oligometastatic disease. 

Connecting our Researchers

Jointly hosted by BIRC and the Centre for Translational Cancer Research (CTCR), as a means to best exploit the strengths in imaging and cancer research citywide, a Citywide Cancer Imaging Seminar series has been initiated. Through web conferencing, investigators at the Robarts and the London Regional Cancer Centre Program are connected in real time. For each seminar a topic is discussed by a clinician and an imaging scientist.   Topics presented include: Stereotactic body radiotherapy for liver tumor ablation, image guided lung cancer treatment, prostate radiotherapy for targeting prostate cancer and the use of hybrid PET/MRI imaging in cancer.

Mark. A. Anastasio, PhD, Image Reconstruction
Robert Bartha, PhD, Magnetic Resonance Imaging
Jerry Battista, PhD, Radiation Oncology Physics
Glenn Bauman, MD, Radiation Oncology
Muriel Brackstone, MD, Surgical Oncology
Gord Campbell, PhD, Phantom Development
Ann Chambers, PhD, Oncology
Joseph Chin, MD, Uro-oncology
Aaron Fenster, PhD, 3D Ultrasound Physics
Paula Foster, PhD, Molecular and Micro-imaging
Neil Gelman, PhD, Breast Cancer Imaging
Jonathan Izawa, MD, Uro-oncology
Anat Kornecki, MD, Radiology
James Lacefield, PhD, Ultrasound Imaging
Ting-Yim Lee, PhD, CT Perfusion
Stuart Gaede, Radiation Medical Physics
George Rodrigues, Radiation Oncology
Leonard Luyt, PhD, PET Imaging Probes
Charles McKenzie, PhD, MRI and Spectroscopy
Ravi Menon, PhD, fMRI
Giulio Muscedere, MD, Radiology
David Palma, MD, MSc, PhD, Radiation Oncology
Grace Parraga, PhD, Respiratory Imaging
Rajni Patel, PhD, Robotics/Mechatronics, Haptics
Terry Peters, PhD, Image guided Interventions
Frank Prato, PhD, Nuclear Medicine Physics
Cesare Romagnoli, MD, Interventional Radiology
Giles Santyr, PhD, Hyperpolarized MRI probe Development
Tim Scholl, PhD, MRS Imaging Probes<
Keith St. Lawrence, PhD, NIR Spectroscopy
Robert Stodilka, PhD, Hybrid Imaging
Don Taves, MD, Radiology
Jake Van Dyk, MSc, Radiation Oncology Physics
Eugene Wong, MD, Radiation Oncology Physics
Aaron Ward, PhD, Image Analysis for prostate, lung, and brain cancer

Masoom Haider, MD, Cancer Imaging (Toronto)
John Trachtenberg, MD, Prostate Cancer (Toronto)
Martin Yaffe, PhD, Cancer Imaging (Toronto)
Leonard Marks, MD, Urology (UCLA)
Baowei Fei, PhD, Cancer Imaging (Atlanta)
Purang Abolmaesumi, PhD, IGST, Image Processing (Vancouver)
Parvin Mousavi, PhD, Ultrasound RF Signals (Kingston)
Gabor Fichtinger, PhD, Real-time Image Guidance (Kingston)

Theo van de Kwast, MD, Genitourinary Pathology (Toronto)