Blogger: Eugene Kim, Post Doctor
MR Cancer Group, Department of circulation and medical imaging

This image won the prize for Best Scientific Image at the 8th International PhD Conference in Medical Imaging 2016:

Inspired by Vincent van Gogh’s signature bold, sweeping brush strokes, this contemporary piece evokes a meteor shower back-dropped by the majestic Milky Way stretching across the midnight sky.

It is a false-color image of a human breast cancer tissue section in which blood vessel walls and proliferating (growing/replicating) cells were labeled using immunohistochemistry. A digital image of the tissue section was acquired with a microscope. Then, a computer algorithm was used to segment the image into different classes or types of cellular and tissue components. Each color represents a different class (e.g., orange = blood vessel and yellow = proliferating cell).

Technical details

The digital image of the human breast cancer tissue section was acquired at 40x with an Olympus VS120 slide scanner. K-means clustering was performed in MATLAB to segment the image into the different classes. This was part of a larger algorithm to automate the tedious task of counting proliferating blood vessels.

Bloggers:  Ailin Falkmo Hansen (PhD candidate) and May-Britt Tessem (Research Scientist), MR Cancer group

Movember is around with a “trøndersk” spirit on Facebook, Instagram, and the city is filled with mustaches in different shapes and varieties. The goal is increased awareness of men’s health and prostate cancer – a disease we in the MR cancer group want to understand better.

Prostate tissue samples are stored in liquid nitrogen prior to MR spectroscopy analyses (photo: Geir Mogen/NTNU)

How can cancer metabolism provide important information about prostate cancer?

Scientists around the world have shown that changes in metabolism are important characteristics of cancer. We have studied how metabolism is altered due to cancer and how metabolism is changed owing to cancer aggressiveness. Previously, we have found that the two molecules citrate and spermine may be markers for prostate cancer and also can reveal information about aggressiveness.

Prostate cancer is a heterogeneous type of cancer, and this is of importance for treatment and prognosis of patients. However, today there are no reliable methods for assessment of type of prostate cancer. Researchers world-wide are therefore searching for new methods that may provide diagnostic and/or prognostic information. Presence of the fusion gene TMPRSS2-ERG have been suggested to be a candidate method for risk stratification, and in a recently published study we investigated the link between prostate cancer metabolism and TMPRSS2-ERG.

What is a fusion gene and what role does it play in prostate cancer?

TMPRSS2-ERG is a fusion gene that occurs when two genes fuses together. Fusion genes may lead to altered properties of the cancer. For example, TMPRSS2-ERG has been shown to be linked to uncontrolled cell growth and development of prostate cancer. We therefore investigated how metabolism was affected by presence of the TMPRSS2-ERG fusion gene.

In this study, prostate tissue collected through the Regional Research biobank Central Norway, Biobank1 and samples from the MR biobank were used for validation. Comparing samples, with and without fusion gene, revealed several changes in metabolism. Especially interesting were changes in citrate and spermine – these were the two molecules that could distinguish aggressive from less aggressive cancer! The genetic data supported our findings, and our combined results suggested that patients with the fusion gene have a different metabolism profile compared to patients without the fusion gene. Changes could be seen both at gene and metabolism level, see the figure. The fusion gene was especially important for determining the metabolism in “low risk” prostate cancer patients.

Gene- and metabolic data suggests TMPRSS2-ERG to be linked with altered metabolism, e.g. lower levels of spermine. Lower levels of spermine have previously been linked to more aggressive prostate cancer. Spermine is part of the polyamine pathway and analysis of expression levels of genes in the polyamine pathway reveals alterations of several central genes. ODC1: ornithine decarboxylase 1, SRM: spermidine synthase, SMS: spermine synthase, SAT1: spermidine/spermine N1-acetyltransferase 1 (blue = down-regulated, red=up-regulated).

Our results were validated in another group of patients and levels of the most important changes in metabolites were confirmed in a small group of in vivo MRI patient examinations before surgery of the prostate. Also here, we could see the same differences as we detected in tissue samples, providing a translational potential.

Currently, there is an ongoing discussion whether patients with the fusion gene have a more aggressive prostate cancer than patients without this gene. Although the importance of TMPRSS2-ERG is debatable, our study suggests that patients with the fusion gene comprise an important subgroup of prostate cancer patients. Especially, among patients characterized as “low-risk” it seems interesting to know if the patient has the fusion gene or not.

We gratefully acknowledge the funding from the Cancer Society.

Reference: A.F. Hansen, E. Sandsmark, M.B. Rye, A.J. Wright, H. Bertilsson, E. Richardsen, T. Viset, A.M. Bofin, A. Angelsen, K.M. Selnæs, T.F. Bathen, M.-B. Tessem, Presence of TMPRSS2-ERG is associated with alterations of the metabolic profile in human prostate cancer, 2016.

Blogger: Siver Moestue, Associate Professor
MR Cancer Group, Department of circulation and medical imaging, NTNU

Cancer cells are notoriously difficult to deal with – partly because they have the ability to dodge the bullets we aim at them. They are inherently resistant to a lot of hostile conditions and therapies, but they can also activate compensatory mechanisms that rescue them when they are exposed to anticancer drugs. We have therefore tried to find a way to responsibly kill cancer cells (killing cancer cells without harming the normal cells in the body, that is) by identifying the crucial compensation mechanisms, and then applying a second drug that strikes the cancer cells where it hurts the most.

Essentially all the work that is performed inside our cells is carried out by proteins. Not all proteins work at the same time – most of them are responsive to cues from their surroundings that tell them when to work and when to take a break. One such cue is OGlcNacylation. Sounds difficult, but basically the cells put tiny “flags” consisting of amino sugars on proteins, thereby giving them new work orders (Figure 1).

Figure 1 Proteins can be activated, deactivated or otherwise change function when they are labelled with the amino sugar N-acetylglucosamine on specific amino acid residues. This process is called O-GlcNacylation, and may play an important role in cancer.

In cancer, it has been seen that an enzyme responsible for putting these amino sugar flags on proteins (OGT) is frequently over-activated. In a collaboration with Prof. Ian Mills (Norwegian Centre for Molecular Medicine) and Suzanne Walker (Harvard Medical School), we wanted to find out if a novel drug targeting OGT could be used to kill prostate cancer cells. To cut a long story short, they could – but only to a certain extent. The cells found a way to survive and continue to divide, just a little slower than before. To understand how the cancer cells escaped death, we took a deep dive into their metabolism using magnetic resonance spectroscopy.

Interestingly, we found that OGT inhibition made the cells reprogram metabolism. They stopped consuming glucose, instead turning to the common amino acid alanine as a new source of energy. Since OGlcNacylation is a process that normally allows cells to respond to starvation or altered supply of nutrients, this sort of makes sense. It also raised the question “What if we block the metabolism of the cells so they cannot use alanine anymore?”. Luckily, the well-known drug cycloserine (used to treat tuberculosis), does just that – by blocking the alanine-converting enzyme GPT2.

When we treated the prostate cancer cells with both the OGT inhibitor and cycloserine, the prostate cancer cells activated a self-destruct mechanism and died. Luckily, normal prostate cells tolerated the treatment very well.

Figure 2 Left: MR spectrum showing how cells treated with an OGT inhibitor (black line) have low levels of glutamate and are virtually depleted of alanine. Right: Diagram demonstrating how LNCaP prostate cancer cells treated with a combination of an OGT inhibitor + cycloserine (blue bar) are far less viable than untreated control cells or cells treated with each of the drugs alone (grey bars).

What did we learn from all this? First, we demonstrated an important principle: Cancer cells may display metabolic ”soft spots” leaving them in a vulnerable position once they are under attack from anticancer drugs. Finding these – and attacking them with a second drug – could be a way of improving cancer therapy in the future. Second, we proved that it is possible to learn an old dog new tricks: cycloserine was developed to make it hard for tuberculosis bacteria to make their cell wall. However, as a side effect it also blocks the ability of mammalian cells to use alanine as an energy source. This does not seem to be a problem for normal cells, because they don´t depend on this mechanism at all. But under given conditions (such as when they are under attack from drugs blocking OGT), prostate cancer cells are addicted to it – and we can take advantage of that when we try to kill them.

The project was carried out with financial support from the Norwegian Cancer Society and the Research Council of Norway.

Bloggers: Hanna Maja Tunset, PhD candidate, Eugene Kim, PostDoc,
MR Cancer group, Department of circulation and medical imaging

October is nearly over. The good news is that Christmas season is just around the corner. The bad news? This is the final post from the MR Cancer group in our blog series for breast cancer awareness month! If you missed the first two blogs, you can check them out here: A new cure for breast cancer? and Could breast cancer metabolism reveal new therapeutic targets?

Breast cancer is not just one disease, and different breast cancers need different treatment. Targeted treatment according to the type of tumor has increased breast cancer survival for a lot of patients.  However, around 15 % of breast cancer patient have what we call triple negative breast cancer (TNBC), which does not respond to currently available targeted therapies. This subtype also tends to be highly aggressive, develop at a younger age, and have poorer prognosis compared to other types. Going against the overall trend in breast cancer, survival rates for TNBC have not increased in the past few decades. So, new targeted therapies for TNBC are in high demand.

One way of finding a good therapeutic target is to look for something that is unique to that particular disease. We have previously identified that the inflammatory enzyme cytosolic phospholipase A2, or cPLA2, is overly active in TNBC as compared to other breast cancers. We wanted to see if inhibiting cPLA2 could be an effective treatment.  So we teamed up with Berit Johansen, a professor at the NTNU Department of Biology and CSO of Avexxin AS, a Trondheim-based start-up company developing a line of cPLA2 inhibitors intended to treat inflammatory diseases like psoriasis and rheumatoid arthritis.

We tested one of their drugs, AVX235, in a mouse model of TNBC, and the results look promising. Tumors in mice treated with the drug only reached 1/3rd of the size of untreated control tumors after 19 days of treatment.

But how does AVX235, an anti-inflammatory drug, inhibit tumor growth?

Targeting the link between inflammation, blood vessels, and cancer

Cytosolic phospholipase A2 is highly involved in both acute and chronic inflammation in the body. In 1863, Rudolf Virchow, the father of cellular pathology, theorized that cancer was caused by chronic inflammation. Over the years, research has revealed close links between inflammation and cancer. American pathologist Harold Dvorak  went as far as to describe tumors as “wounds that do not heal”. A major component of inflammation and wound healing is angiogenesis –  the growth of new blood vessels. Angiogenesis is a hallmark of cancer and is observed in most solid tumors. As a tumor grows, it produces blood vessels that deliver the oxygen and nutrients necessary to sustain it.

Our study suggests that cPLA2 plays an important role in angiogenesis in TNBC. We found that AVX235 reduced the amount and size of blood vessels in the tumors. High-resolution CT images showed that large portions of tumors treated with AVX235 did not have a blood supply.

CT-based images of representative tumors from control (Ctrl) and treated (Tx) groups, color-coded for vessel caliber (VC), i.e. vessel diameter – the more yellow or red, the larger the vessel. Note the lack of vessels in large areas of the treated tumor (arrows).

We also examined tumor tissue under the microscope, after staining for certain factors that could reveal if blood vessel cells were dividing. Treated tumors had fewer actively growing blood vessels. This indicates that blocking cPLA2 activity hindered the cancer’s ability to produce new blood vessels, resulting in tumors that were under-fed and unable to grow as they normally would.

Next steps towards improving treatment of TNBC

We are still in the early phases, but our initial results show that inhibition of cPLA2 may be a strategy for targeted therapy of TNBC. Continuing our collaboration with Avexxin, we are currently trying to find out more about how this may work. By using breast cancer cells grown in a flask as a model system, we can detect responses on a cellular level, like alterations of metabolism, how fast the cells grow, and their ability to invade new tissues.

Over the past decade, there has been an increasing interest in cPLA2 as an important player in the progression of various cancers, including breast cancer. We hope our research in this field will contribute towards providing a much-needed targeted therapy for triple negative breast cancer patients.

The results from this study was published in BMC Cancer.

This work was funded by the liaison committee between the Central Norway Regional Health Authority and the Norwegian University of Science and Technology (NTNU), the Norwegian Cancer Society, the Research Council of Norway, and Avexxin AS.

From the back (left to righ): Astrid Jullumstrø Feuerherm, Siver Moestue, Jana Kim, Eugene Kim, Berit Johansen and Hanna Maja Tunset

Bloggers: Trygve Andreassen, Senior engineer, and Siver Moestue, Associate professor,
The MR Cancer Group, Department of circulation and medical imaging

Scientists all over the world are strenuously pursuing new ways of killing cancer cells. They look for so called “drug targets” – proteins the cancer cells need for rapid and uncontrolled proliferation. In collaboration with Karolinska Institutet in Sweden, we have discovered such a protein called CHPT1. Based on this discovery we think it may be possible to develop new drugs for treating breast cancer.

A needle in a haystack

Before we can develop a new cancer drug, we must have some idea about what the drug should do. We have to identify a “drug target” – a protein the drug must recognize, bind to and alter the function of. It is not as simple as it might sound – we have about 20,000 different protein coding genes in our DNA, and it has been estimated that they can produce up to 100,000 different proteins (Savage, Nature 2015). These proteins communicate with each other in various ways, which means that we are trying to find a needle in a pretty big haystack.

Cancer drugs are often designed to block processes that are important for cell division, or the copying of DNA that happens prior to every cell division. In our research group, we are interested in the metabolism of cancer cells (in order to grow and divide fast, the cancer cells will reprogram their metabolism), and therefore we have searched for possible target proteins in the biochemical pathways.

In collaboration with Karolinska Institutet in Sweden, we started out with some known characteristics of breast cancer cells: They often grow faster in the presence of oestrogen; and they have an abnormally high turnover of the molecule choline. In order to study the choline turnover we apply a technique called MR-spectroscopy.

Main suspect: CHPT1

First we identified about 18,000 oestrogen binding domains in DNA, and found that these domains control the expression of about 2500 genes. Next we found how many of these genes that could be involved in choline turnover – hereby reducing the number of candidate genes to 19. By interpreting MR-spectra of cancer cells grown with and without oestrogen, we found which metabolic changes that could be derived from oestrogen stimulation. This allowed us to further reduce the number of proteins – and eventually we suspected the protein CHPT1 of being heavily involved in transferring the stimulating effect of oestrogen to the metabolic machinery of cancer cells.

To make sure that our findings are not only valid for cancer cells grown in the lab, we investigated tumour tissue from 70 breast cancer patients. Here we found higher expression of CHPT1 in tumour tissue compared to normal breast tissue. Even more interesting, we could see that tumours classified as oestrogen sensitive had significantly higher CHPT1 expression than the other tumours. This implies that oestrogen actually activates CHPT1 in breast cancer patients.

The next step in proving that CHPT1 is a relevant drug target was to show what happens if we switch off this gene. In the laboratory we have molecular tools which allow us to stop the production of the CHPT1-protein in cancer cells, and we found that cancer cells grew much slower when CHPT1 was switched off (Figure 1).

Figure 1: The two pictures to the left show to different types of breast cancer, seen in a microscope when growing freely in a dish. In the pictures to the right (marked siCHPT1) we see the same cells, but here we have “switched off” CHPT1. There are clearly fewer cells, which indicates that the cells need CHPT1 to grow and divide.

Killing cells in a dish at the lab is fairly easy, so the next step was to see if we could influence tumour growth in a living organism. Our choice of organism was zebrafish (they are quite transparent which enables us to observe the cancer cells directly). By switching off CHPT1 we could see that both tumour growth, as well as the cancer cells’ ability to spread to other tissue and organs, was reduced (Figure 2).

Figure 2: Above we see a zebrafish larva with implanted oestrogen sensitive MCF-7 cancer cells. They have developed a tumour at the injections site (red colour), and daughter tumours (metastases) have formed along the spine (marked with arrows (6). Below we see a zebrafish larva injected with the same type of cancer cells, but with CHPT1 switched off (siCHPT1). Here, there is no big tumour at the injection site, and there are clearly fewer daughter tumours along the spine (marked with arrows (3)).

All in all, we have presented several indicators suggesting that CHPT1 may be an important drug target for the treatment of oestrogen sensitive breast cancer. But there are still many things we haven’t found out yet: Is it possible to design drugs that block CHPT1? What are the side effects? Are they superior to existing drugs?

We have published our findings and hope that someone will take up the gauntlet and continue working on our ideas. And hopefully be able to develop a new drug based on our basic research.

The project has been partly funded by the Norwegian Cancer Society and the Research Council of Norway.

Blogger: Tonje Husby Haukaas, Senior engineer
MR Cancer Group at the Department of circulation and medical imaging

Could there be a better day to write my first blog? Especially when it’s a “pink” one! The 1^st of October marks the beginning of Breast cancer awareness month. This is a month that has the aim to improve knowledge among the population, show compassion for breast cancer patients, and raise funds for breast cancer research.

In the coming days and weeks, the streets will be filled with pink ribbons, balloons, and accessories, and a variety of pink activities will be organized throughout the country. With this blog I hope to teach you something you didn’t know about breast cancer. I also want to show you what research funding can contribute to.

Breast cancer affects many lives, not only the people who are diagnosed. It is the most common cancer among women in Norway, and just today eight women will be diagnosed with breast cancer. Luckily, due to early detection and improved treatment, 90 % of women with breast cancer are still alive 5 year after the diagnosis. But no two tumors are exactly alike, which makes it hard to predict who will be in the 10 % that respond poorly to the current treatments and have short survival times. Also, even with a good response to treatment the patient can experience post treatment adverse effects and a poorer quality of life.

We wish to improve this.

To provide targeted and optimal treatment that has been tailored to each patient, we first have to learn as much as possible about the cancers’ properties and potential weaknesses. Some important targets are already well established in the clinic, like drugs that attack cancer cells that depend on the hormones estrogen or progesterone to grow. But there are still some cancers that don’t have established targets and researchers are working hard to uncover new ways to treat them.

Important differences can be hidden in cancer metabolism

One of the aims of the MR Cancer group has been to reveal potential treatment targets by exploring cancer metabolism.

Cancer cells grow and divide uncontrollably, which increases the need for energy and building blocks compared to normal cells. This can be observed when MR spectroscopy is used to study which small molecules, called metabolites, are present in tissue samples. The result of such an experiment is called a ‘spectrum’, where the peaks originate from different metabolites (see figure). Some of the metabolites are more common than others, such as glucose and lactate. Others are perhaps less known, such as glycerophosphocholine and glutathione, but are still important to study. The metabolites can tell us something about which processes are ongoing at the time the sample is taken.

Previous studies have shown that the cancer tissue’s metabolic fingerprint, meaning the metabolites present, is related to the tumor grade (how aggressive the cancer is), potential for metastasis, and 5 year survival.

By studying the metabolic fingerprint, it is possible to increase the knowledge about which types of breast cancer exist, how aggressive they are, and at the same time look for more biological markers that can identify new targets for treatment.

The figure shows some of the metabolites we can observe by performing MR spectroscopy on breast cancer tissue.

The Oslo2-study: A large-scale breast cancer study from multiple biological levels

In the Oslo2-study, a large-scale breast cancer study at the Oslo University Hospital, sample material from 228 patients has been sent to Trondheim and analyzed by MR spectroscopy. Based on the samples’ metabolic fingerprints we have identified three new subgroups of breast cancer. These three groups displayed differences in metabolism; understanding what makes them different helps us to identify more precise drug targets specific for each of the three groups.

Three new subgroups of breast cancer, called Mc1, Mc2 and Mc3, was found by comparing their metabolic fingerprint.

The unique feature of this study is that the sample materials from the same group of patients have been analyzed with multiple methods. Both the expression of genes and level of breast cancer related proteins have been analyzed. This gives us the possibility to look for new relationships between genes, proteins, and metabolites in breast cancer. By combining the levels of data, we saw that one group in particular expressed big differences in genes and proteins compared to the two other groups.

Tissue samples are stored in liquid nitrogen prior to MR spectroscopy. The samples are kept frozen while they are cut to fit the sample tubes. Photo: Geir Mogen/NTNU.

Based on this, we suspect that this group has a more aggressive cancer type, and would therefore benefit from a different type of treatment than the two other groups.

We will soon acquire follow-up data from the patients in this study, which will provide us with new information on survival and relapse. Based on our metabolic findings, the hypothesis is that the patient survival is different depending on which metabolic subgroup they belong to. If this hypothesis is correct, it means that the metabolic classification can indicate which patients need extra follow-up, and possibly more intensive treatment to get well.

Finally: Why not wear something pink today to show your appreciation for Breast cancer awareness month?

The research is financed by the K.G. Jebsen Center for Breast Cancer Research.

Reference:

• Haukaas TH et al. Metabolic clusters of breast cancer in relation to gene- and protein expression subtypes. Cancer and Metabolism (2016)

Dr. Christian Doeller is head of the Doeller research group at the Kavli Institute for Systems Neuroscience

Christian Doeller at the Kavli Institute for Systems Neuroscience has been awarded the Radboud Science Award for his research on how the brain links memories of different events to form one coherent memory. To answer this question, he and his team used pictures and videos of the computer game “The Sims” to create stories. They then showed these stories to participants lying in an MRI scanner and recorded brain activity while people remembered events. They found that the brain forms memory networks of related events which are encoded hierarchically in a brain structure called the hippocampus. How these memory hierarchies are organized resembles what is known about how space is encoded in the brain. “Our findings might point towards a more general code for cognition” says Christian Doeller. “Our memories are what defines our personality and improving our understanding of these mechanisms will be crucial in understanding cognition and neural breakdown in neurodegenerative diseases”.

Screenshots from the computer game showed to participants while recording their brain activity in an MRI scanner

Blogger: Jana Cebulla, PostDoc
MR Cancer Group, Department of Circulation and Medical Imaging

Just as every human being is a unique individual, every cancer has its own characteristics, and this is increasingly being recognised in cancer treatment. Targeted therapies are being developed that attack cancers on the molecular level and treatment strategies are tailor-made for each patient.

But how do we know that a treatment actually works?

Typically, treatment response is measured as a change in tumour size using anatomical MRI or CT images, or tumour markers found in body fluids. However, these changes usually occur quite late in the course of the treatment. This is especially true of new, targeted therapies that do not directly kill cancer cells, but cause more subtle structural changes in the tumour tissue. These new therapies aim for a better treatment effect with less side-effects, but because the changes are so subtle, it also means it could be more difficult to see if a treatment is effective or not.

We therefore need different methods to assess such structural changes in the tumour.

In the MR Cancer group at NTNU, we are testing magnetic resonance imaging (MRI) modalities that can visualise structural changes in the tumour that occur on the cellular level rather than changes in tumour size.

Figure 1: Left: MRI of brain tumours (Anne-Line Stensjøen). Middle: blood sampling. Right: “water diffusion map” of a tumour (Jana Cebulla).

One of these methods is “diffusion MRI”, where the brightness of the image depends on the random movement of the water molecules in the tissue, which is called diffusion. In previous blogs by Siver Moestue and Jose Teruel, we have already described how we can use diffusion MRI to diagnose cancer: usually water diffusion is more restricted in malignant (bad) tumours than in normal tissue or benign (good) tumours. But how can we use diffusion MRI to monitor treatment response?

When tumours respond to treatment, the tumour cells and their surrounding cells change in their function or structure. For example, if a tumour cell dies, its cell membrane becomes porous and more permeable, and water can diffuse more freely.

Figure 2: Left: densely packed tumour cells where the water diffusion is restricted especially because of cell membranes.
Right: Dying cells with permeable membranes that restrict water diffusion much less. (Cebulla, Doctoral thesis at NTNU; 2015:174)

And we can measure this with diffusion MRI.  Tumours responding to treatment can be distinguished from non-responding tumours by an increase in ADC, which stands for “apparent diffusion coefficient” – the average distance that a water molecule travels within a certain time. This is illustrated in Figure 2 which shows how much further and more freely the water molecules can move among dying cancer cells than within a ‘healthy’ tumour. This method was used in one of our recent studies, where we characterised the effects of a drug targeting a specific pathway in the cells (PI3-Kinase pathway).

Figure 3: The left tumour shows no obvious difference in the anatomical image before or after treatment, while the increase in ADC suggests that the tumour responds to treatment. The right tumour does not show response to treatment (Cebulla et al. Br J Cancer 2014).

Diffusion MRI is already being used for tumour detection. However, monitoring treatment response using these techniques is not yet part of routine. It is difficult to decide how much the ADC has to change in the tumour in order to consider it a real treatment response. Further international standardisation of the method is therefore necessary for the technique to be used routinely.

The MR cancer group at NTNU will continue to explore how we can detect response to cancer treatment at early time points using diffusion MRI and other advanced imaging technologies. The goal of this research is to provide more personalised cancer treatment that is tailored to the individual patient.

Blogger: Jose Teruel
PhD candidate at MR Cancer group

In a previous post in this blog I described how a new magnetic resonance imaging (MRI) technique called ‘Diffusion weighted MRI” can be applied for the detection of breast cancer due to the random movement of water molecules, or diffusion, which is different in cancer compared to healthy tissue. Therefore, using diffusion MRI we can measure the extent of diffusion and detect cancer. But, can we take advantage of this technique for benign breast lesion assessment as well?

… nowadays there is an excess of what we call ‘unnecessary biopsies’

Differentiation of breast lesions is another important step in the clinic. Let’s imagine the situation; a woman comes into the clinic feeling a lump in her breast, and a routine mammography or ultrasound confirms the presence of a suspicious lesion. The next step in the clinic would be to obtain a biopsy of the lesion to properly establish if the lesion is a malignant tumor, a threat for the patient’s life, or if it is a benign tumor that needs no further treatment.

A needle biopsy is categorized as an invasive procedure. The procedure consists in introducing a needle through the patient breast into the lesion, to obtain a sample (or several) of the tumor tissue that will be analyzed by a pathologist to examine if there is a presence of cancerous cells.

Having explained what a biopsy is, it is important to note that nowadays there is an excess of what we call ‘unnecessary biopsies’, i.e., biopsies that could have been avoided if it would have been possible to characterize the lesion as benign by a different ‘non-invasive’ procedure. Biopsies are in the best case not pleasant for the patients, and the procedure consumes time and resources.

… differentiating malignant and benign lesions without the need of an invasive procedure

So, what could the role of diffusion MRI be in this matter? Diffusion MRI can characterize different tissues based on their microstructure, and this microstructure it is known to be different in malignant and benign lesions. Our preliminary results in the use of this technique suggest a very high accuracy in differentiating malignant and benign lesions without the need of an invasive procedure. This is possible due to the fact that water movement (diffusion) is much more restricted in malignant lesions compared with benign ones and this difference can be measured by different parameters obtained from a diffusion MRI examination.

Left: Benign tumor exhibiting clearly higher values of apparent diffusion coefficient (ADC), a measure of the extent of water diffusion within tissue, compared to much lower values found for cancer on the right panel.

Furthermore, the particular technique used in our work, known as Diffusion Tensor Imaging (DTI), provides information not only in the extent of the water diffusion but also about the directionality of the diffusion. The detailed results of this work performed at NTNU and St. Olavs Hospital, is now in press and will soon be published in the Journal of Magnetic Resonance Imaging under the title “Diffusion weighted imaging for the differentiation of breast tumors: From apparent diffusion coefficient to high order diffusion tensor imaging”.

Further research and validation of our results may strengthen the current opinion that a non-invasive MRI examination can be used before biopsy to confirm if the biopsy is really necessary or if malignancy can be directly ruled out.

… diffusion MRI could save in the future a great extent of unnecessary biopsies

In summary, the use of diffusion MRI could save in the future a great extent of unnecessary biopsies avoiding invasive diagnostic procedures to patients with suspicious breast lesions.

I would finally like to thanks the dedicated volunteers that take part in our study and the funding we receive from the local health authorities and The Norwegian Research Council.

Blogger: Debbie Hill,
Post doctoral fellow at MR Cancer group

As winter approaches us here in Trondheim, the art of keeping warm becomes more important; after all, the well-known Norwegian saying is ‘there is no bad weather, only bad clothes’. So, it is quite convenient that the shortening and cooling of days coincides with the month of November, or Movember as it is affectionately known, where people all around the world insulate their upper lips from the onset of winter to increase awareness of prostate cancer (and to raise some money too!).

Here in the MR Cancer group at NTNU, the ‘prostate team’ are no exception; roughly half way through the month we have cultivated some impressive moustaches.

The MR-Cancer Group Prostate Team.

Amongst all of the fundraising and amusing ‘facial knitwear’, it is important to acknowledge the other furry creatures that are crucial in our fight to improve prostate cancer treatment. I speak of the research animals. While we do our utmost to minimise the use of animals in research, they are currently essential for cancer studies to help us to understand how the disease develops and how we can best treat it.

A furry creature, crucial in our fight to improve prostate cancer treatment.

It is notoriously difficult to correctly diagnose aggressive, metastatic prostate cancer from the less-aggressive kind; leading to the possibility of over- or under-treatment.

Read more about how we are tackling this problem at NTNU and St Olavs Hospital:

• Read about early markers of a potentially dangerous type of Prostate cancer
• Read about treat or not treat? The role of PET/MRI in prostate cancer

The development of a mouse model of prostate cancer, called the TRAMP model, is helping us to understand cancer progression from early-stage to late-stage (aggressive) disease. The TRAMP mice develop prostate cancer as they reach adulthood, and the characteristics of the disease closely match what we see in the clinic. We have been using the TRAMP model to help us to develop good techniques for detection of cancer-onset, to monitor disease progression, and test new drug treatments.

In the MR cancer group, we use Magnetic Resonance Imaging (MRI) to diagnose and monitor cancers. For pre-clinical research, we have an MR scanner that is specially designed for small animals, like the one in this picture.

The MRI machine at NTNU for scanning small animals. Photo: Geir Mogen

We use the machine to take pictures, in this case of the prostate, and have been investigating which types of images are most sensitive to detection of the cancer onset. A technique called Diffusion Weighted Imaging (DWI) has been very useful in detecting prostate cancer before there is any change in the size of the prostate.

DWI is also used in the clinic to help detect cancer; it is sensitive to the movement of water within the tissue of interest. Cancer cells grow quickly and are usually densely packed, which makes it harder for water to diffuse in the tissue, therefore cancer looks different to healthy tissue in the images.

Left: MRI Image axially across the abdomen of the mouse (belly at top). Cancer in prostate is indicated with red arrow. Right: Map of the prostate from a diffusion weighted image, where dark area (red arrow) indicates less diffusion of water, indicative of cancer.

Using this technique, we can non-invasively monitor the progression of disease from the early-stages through to the late-stage. So far, we have characterised how TRAMP mice develop cancer without any intervention, and these results will act as a reference for future experiments. We are currently investigating the effectiveness of new drug treatments, as well as looking into lifestyle effects, such as diet and exercise, to see if having an active lifestyle could delay the onset or severity of prostate cancer.

Our research is ultimately directed towards improving clinical practice to help with our ongoing fight against cancer, and has been made possible by funding support from the Cancer Society and the Liaison Committee between Central Norway RHA and NTNU.

If you are inspired by our work (or by our moustaches!) please visit our medical faculty Mo-Bro team page to donate in support of men’s health issues, including prostate cancer research.