Paul O’Gorman Building, University College London,
72 Huntley Street, London, WC1E 6DD
United Kingdom

The Witney Group at UCL's Centre for Advanced Biomedical Imaging. Specialising in the development of advanced cancer imaging tools.


Flexibility is a killer

Tim Witney

I'm often asked why, despite many $/£ millions have been spent on cancer research, it seems like we're no closer to a cure. This is quite a difficult question to answer, because there are a number of factors that contribute to cancer being such a tough disease to treat. I should point out though, that some really great progress has been made in recent years, exemplified by breakthrough immunotherapy drugs, where our own immune system is trained to kill the cancer; or new combination drugs that target specific genetic mutations found in many cancers. Also, as Cancer Research UK's very informative website states, cancer survival rates have doubled in the last 40 years. But this really isn't good enough. There are a variety of reasons why we're struggling to do better, but in my mind there are two predominant reasons why cancer is such a moving target:

1. Each cancer is different and specific to each individual patient. Cancer results from problems arising in our own DNA, known as gene mutations. The problem is that for each person, these 'problems', or mutations, occur in different places in the DNA, affecting different genes. Each mutation really needs a different treatment, meaning that we need a whole library of drugs to be effective. There are efforts to try and tailor each treatment to individual patients, known as personalised, or precision medicine

2. Cancers are adaptable. By their very nature, they are designed to grow and outcompete normal healthy tissue that surround them. They co-opt new blood vessels to provide them with the nutrients they need, and can change their behaviour in order to survive, say drug treatment, or to enable them to grow in harsh environments. 

Over the past couple of years at Stanford University, I've worked on trying to better understand how this adaptability gives cancer an unhealthy advantage over normal cells, and the organs they colonise and take over. One element of flexibility that cancer cells possess is the ability to choose how they use the sugar they take up. Sugar, in the form of glucose, is needed for all cells in order to create energy that keeps the cell alive. To produce this energy, glucose is taken up by cells and converted to other molecules in a process called 'glycolysis'. When oxygen is present (the reason why we need to keep breathing!), these molecules are further modified and energy is produced. Cancer cells have generated a unique ability to put the brakes on this process. On the face of it, this seems like a bad idea, but what this means is that the molecules of glycolysis usually used for creating energy can now be used to make other things, such as new bits of cells. The result is that more cancer cells are produced. Cancer cells can quickly switch between these two states: 1) energy production; or 2) the creation of new building blocks to create more cancer cells. 

    Fig. 1.  PKM2 controls the balance between the production of new building blocks or energy in cancer cells. Here, PKM2 is represented by an orange sphere.  


Fig. 1. PKM2 controls the balance between the production of new building blocks or energy in cancer cells. Here, PKM2 is represented by an orange sphere.  

We're now starting to understand what controls this flexibility inside cancer cells, and it's an enzyme called pyruvate kinase M2, or PKM2. PKM2 has been found in all cancer cells studied to-date, and acts as the master regulator of glycolysis. When the cancer cell requires energy, PKM2 binds to 3 other identical PKM2 enzymes, but when new building blocks are needed for the creation of new cells, PKM2 only binds to one other PKM2 enzyme, blocking energy production almost completely (Fig. 1). Most normal healthy cells lack PKM2, instead using an enzyme, PKM1, that always leads to energy production.

Fig. 2.  DASA-23 binding to a tumour in the brain of a mouse, indicated by the arrow.

Fig. 2. DASA-23 binding to a tumour in the brain of a mouse, indicated by the arrow.

In the lab of Sanjiv Sam Gambhir at Stanford, my colleagues and I worked on ways to try and detect this cancer-related enzyme in animal models of human brain cancer. We reasoned that if we could detect PKM2, we may be able to detect the cancer itself. This would be massively helpful to the brain surgeon when they came to remove the brain tumour as they would know its precise location within the brain. This is important, as the surgeon doesn't want to take too much healthy tissue, which may result in unwanted side effects, or be too cautious and leave some of the tumour behind. We developed a new positron emission tomography imaging agent (see earlier posts for an overview of this technique) based on a known drug, DASA-23, that specifically targets PKM2. By tagging DASA-23 with radioactivity and injecting it into mice, we could see its precise location within the animal and how that changed over time. As shown in Fig. 2, the tagged DASA-23 honed precisely to the PKM2-containing brain tumours and not to the healthy surrounding brain tissue. Moreover, we showed that by adding another drug that targeted PKM2 we could stop DASA-23 binding to the tumour cells. This showed that the new drug was going to the right place, meaning that DASA-23 might be useful for the development of new PKM2-specific drugs. Although exciting, it should be pointed out that these are very early findings and a lot more research is needed before we will be able to use DASA-23 in hospitals around the country, but early tests in humans are planned for next year.

If you're interested, this work was published as the front cover of this week's Science Translational Medicine

Cell death in disease

Tim Witney

I'm delighted to be the Lead Guest Editor of a Special Issue of BioMed Research International, entitled "Cell Death In Disease". This is an exciting opportunity to publish your research on this topic, which will be made freely available for all to read. Having worked in the field for nearly a decade, it's a subject close to my heart, and I look forward to putting together what promises to be a stimulating edition.

The call for papers can be found here.

30 days of Movember


I look ridiculous. It itches, and the majority of the comments I've received can't be repeated here. Nevertheless, I've survived 30 days of Movember; summed up in this video:

And all for a good cause. In 2012 alone, over 307,000 people were estimated to have died from prostate cancer. 30 die a day in the UK alone. Whereas the main risk factor for prostate cancer is age, testicular cancer is the most common cancer in men aged 25-49 in the UK. Luckily, cure rates are far higher with this form. Regardless, there is still far more that can be done. Please support this good cause on my Movember page here. Thank you to everyone that's done so already.
If you're interested, more cancer stats and info can be found on CRUK's webpage here.

1 week in...


Thanks to everyone who's donated so far. We've raised $161 already (!) and I haven't even really grown anything yet. It's only going to get worse as well. Facial hair really doesn't suit me, so the worse I look, the more you should donate! Please carry on supporting this great cause (prostate cancer, not my face). Thanks so much!

1 week and it's starting to grow...



This year I've decided to take part in Movember. For those still on dial-up, Movember is the annual ritual of growing a mustache over the 30 days of November. The idea is to try and raise funds and awareness of the issues surrounding prostate cancer, testicular cancer and men's mental health. In the imaging world, prostate cancer is a difficult one. We just don't have the right tools for the job. Whilst most men will die with, rather than of prostate cancer, there is a small, but significant proportion of aggressive prostate cancer that requires urgent treatment. The problem is knowing what type of prostate cancer you have. The solution is to remove the cancer by surgery - but this comes with major risk of damaging the near-by nerves, which can result in erectile disfunction and incontinence. So we don't want to remove all prostate cancers that are diagnosed. Imaging could play a great role in distinguishing the nasty from the not-so-nasty cancer. I hope we will have some success in developing such a test in my lifetime, but these things are harder to do in practise and the research is costly. So I'm going to look stupid for a month and talk to people I don't know about why we should be botherd about prostate cancer. And if you are bothered too, please donate on my Movember page here! They are already funding some great research by my fellow imaging scientists over at John's Hopkins: Thanks.
Me. Day 1: 1st November 2014.

Measuring fat breakdown to detect cancer


Last year I moved to the United States to take up a postdoctoral position at Stanford University in the lab of Prof. Sanjiv Sam Gambhir. I'm lucky enough to work in a very creative environment, nicely summed-up in this video. Living in the US however has brought some challenges I've been ill-equipped to deal with. Namely, how to avoid getting run over when crossing a road (walking seems to be prohibited here) and what to do to stop getting fat. High fructose corn syrup is added to pretty much everything and it costs nearly the same amount to eat out as to buy the constituent ingredients (eating out is cheap!). I don't stand much of a chance. Running across the road however may solve both issues.

Given my rapidly-expanding waistline, it seems pretty appropriate that my research during my last year at Imperial College London focused on measuring the breakdown of fat by cancer cells. As I mentioned in my last blog, cancer cells take up more glucose for energy production and storage. Additionally, tumours require increased levels of fat to make new cells and to create even more energy. The breakdown of fats at high rates to produce this energy sets cancer cells apart from most normal 'healthy' tissue. We have recently shown that by imaging fat breakdown, we can detect breast, prostate and brain cancer in preclinical (non-human) models, published this month in The Journal of Nuclear Medicine. This is important as existing techniques to identify and diagnose both brain and prostate cancer are not effective in all cases. Further tests, such as the one proposed in this research article, may provide additional information and eliminate the need for an invasive biopsy. By accurately detecting these cancers at an early stage, the chances of survival are greatly improved. We're still some way off evaluating this diagnostic test in patients, but I have high hopes for this new imaging technique, an example of which is shown below:

In other cool news this week, my academic mentor at Stanford, Prof. Sanjiv Sam Gambhir, is partnering with Google's secret research division, Google [x] - the division that's brought us Google Glass and those internet balloons. The idea behind this project, named Baseline, is to define and thoroughly characterise the genetic and molecular make-up of healthy adults (initially from 175 people, increasing to many thousands). By understanding the key features of good health, it's hoped that we may be better placed to understand and detect things that go wrong. Details of this project are light on the ground, but it's thought that some cool wearables, such as the 'smart contact lense', will be used to monitor those enrolled in the project 24/7. Let's hope this ambitious project results in a major scientific breakthrough.

Cancer's sweet tooth


Otto Warburg
One of the best things about being a scientist is discovering something no one has ever done or seen before. Whether it be the creation of a new man-made plastic, or the discovery of the Higgs boson, science is tirelessly expanding our collective knowledge. Sometimes however, we're so busy focusing on the horizon and the next scientific breakthrough that we forget to look over our shoulder and examine in sufficient detail what has come before.

Cancer detection through imaging
A German scientist, Otto Warburg made the discovery in the 1920s that cancer cells consume sugar in far greater amounts than normal healthy cells.  It's only recently though that we have started to use this discovery to our advantage.  By designing drugs that curtail this 'addiction' to sugar it is hoped that we can stop these cancer cells from growing. We also take advantage of cancer's sweet tooth during diagnosis. Following injection of a radioactive sugar into the bloodstream, clinicians can detect cancer using a scan that measures where that sugar is being used in the body.  An example is shown to the left, with the tumour indicated by the arrow.  It is now thought that cancer cells use the sugar to protect against harmful waste products, for energy, and to create building blocks to form new cells.

Following in the footsteps of Warburg, a team of French scientists made the discovery in the late 70s/early 80s that cancer cells save some of this sugar for a rainy day - when extra energy is needed, or to keep the cells alive when the supply of sugars from elsewhere runs out. This discovery is fascinating given that sugar stores are normally only found in the liver and muscle. Cancer cells that originate from say the breast or ovary seem to acquire the ability to store these sugars through, as of yet, unknown mechanisms. These findings have been largely ignored until now. In a research article published this month in Cancer Research, myself and my colleagues at Imperial College London further explore this phenomenon, some 30 years later.  We showed that cancer cells store more sugar when they stop growing and that we can detect these sugar stores through imaging. A picture of these sugar stores are shown below, indicated by the intense orange/yellow dots within the cancer cells. The identification of these stores has wider implications as cancer cells that grow more slowly are typically more resistant to traditional chemotherapy. It's hoped that this new imaging method might be able to detect these slower growing cells that we can then target with different drugs. Although this technique hasn't been tested in humans yet, we are hoping scans, similar to the one shown above, will be performed in the next few years.  There is also hope that this technique can be used to detect other sugar storage diseases such as diabetes.
Cancer's sugar stores

For more information, the research article, 'A Novel Radiotracer to Image Glycogen Metabolism in Tumors by Positron Emission Tomography' can be found here.

The smallest detail sometimes makes the biggest difference


This latest post comes courtesy of my great friend and fellow scientist, Dr Peter Canning.  We were lab partners at the University of Warwick some 10 years ago and were always the last to leave the lab - mostly because I was so slow!  Having completed his PhD in Structural Biology at Warwick, he is now working as a postdoc for the Structural Genomics Consortium at Oxford University... 

The smallest detail sometimes makes the biggest difference 

Looking at how two molecules "talk" to each other may provide the basis for new cancer treatments

The science of cancer imaging encompasses a wide range of different techniques and disciplines. Imaging technologies allow cancerous cells to be detected, characterized and monitored, or using different kinds of imaging methods, much smaller, molecular-scale events can be observed. In every cell of the human body, millions of times a second, biological molecules signal to one another, create things, destroy things, transport things and carry out thousands of individual tasks needed to keep a cell running. Various factors can cause these highly organised processes to break down, causing the cell to malfunction. These are the kinds of malfunctions that lead to the development of cancers and indeed other diseases.

Fortunately, cells come with a range of quality control mechanisms built in. They are capable of fixing all kinds of damage, or if the damage is too severe, they are even capable of activating a kind of self-destruct mechanism that destroys the cell before the problem gets too severe. Of course, the mechanism to control the self-destruct system is carefully controlled and monitored.

One molecule involved in the control of the self-destruct sequence is called p53, in fact it is more or less the control hub, the big red button. If a cell is damaged and on the path to becoming cancerous, p53 is activated and either shuts the cell down or destroys it for good. It has been a subject of great interest for some time to biologists, because in the vast majority of cancer cells, p53 itself has become damaged and is no longer able to destroy the damaged cells. For some unknown reason, p53 has evolved to be very fragile, and so damage to p53 happens all too easily. With this in mind, scientists are working to find ways to reactivate damaged p53, or alternatively to find a way to trigger the same response that p53 would normally activate, hitting the self-destruct button for the cancerous cells and causing them to destroy themselves.

Under normal conditions (a), when a cell detects that it is damaged, a signal is sent to p53, which activates a kind of "self destruct" mechanism to destroy the cell before it can do too much damage. If p53 malfunctions (b) then it is unable to trigger this response and cells are allowed to become cancerous, growing and multiplying unchecked.

I am currently a Postdoctoral Research Associate at the Structural Genomics Consortium (SGC), at the 
University of Oxford, in the Growth Factor Signaling Group ( The SGC is a not-for-profit organization with labs in Oxford and Toronto which looks to investigate biological molecules (proteins) involved in various diseases and study them on the atomic level using an imaging technique called X-ray crystallography, then put the information into the public domain free of charge. This enables further research by the global scientific community, in particular speeding up the lengthy and expensive process of discovering new drugs.

In a paper published in the Journal of Molecular Biology this month, we use X-ray crystallography to image a communication between two molecules at the the atomic level. We wanted to address the idea of self-destructing a cell in which p53 has failed and to do this we looked at a protein very closely related to p53 called p73. p73 is capable of standing in for p53 and destroying a bad cell, with the added bonus that it is far less fragile, but for some reason this is not a common occurrence in the course of normal cellular events. In our paper we not only look at the molecular structure of p73 and how it is subtly different to p53, but also how p73 is activated. We revealed that a protein known to activate both p53 and p73 called ASPP2 activates p73 in almost exactly the same way as p53. This finding raises some interesting questions. For instance, if the system of activation targets both proteins in the same way then how is one protein chosen over the other? However, it also provides useful information for scientists looking to find a way to get p73 to switch on, stand in for p53 and destroy cancerous cells.

These types of images are used to represent the molecular structure of proteins. Here, the ASPP2 protein molecule (red) is shown interacting with both the p73 protein (yellow) and the p53 protein (blue) in an almost identical fashion. 

If you’re interested, the paper is now available from the Journal of Molecular Biology’s website:

Or you can read more about the group’s activities on the SGC’s website:

Detecting cancer... the earlier the better


Christmas this year has been a very mixed affair. Having battled with cancer for over a decade, my wife's Grandmother, Mary, died as a result of the disease a few weeks back.  There aren't many people that I know that haven't been affected one way or another by cancer.  The same can be said too about dementia and heart disease.  All too often, scientists researching ways to prevent, treat, and cure the disease are reminded just how much more work is needed.

That's not to say that progress is not being made. Thousands of papers describing new drugs that target tumours, along with improved methods to detect cancer in the first place are published each month.  The problem is taking these great ideas and converting them into a product that will improve patient well-being and prolong survival.  This process requires collaboration between industry (pharma companies) and academia (Universities), roughly 10 years of research and development, and many millions of pounds.  High-profile failures have also resulted in Industry being less willing to stump-up the millions of pounds required when there is no guarantee that they will get a return on their money.

Professor Eric Aboagye
Given this environment, I am fortunate enough to be working at Imperial College London's Comprehensive Cancer Imaging Centre.  Here, under the expert guidance of Prof. Eric Aboagye (the smiliest man in Science!), we have the ability to take an idea, test this idea in cancer cells and other models of cancer, before trialing it in humans.  The same regulatory and strict peer review processes are in place, but we are lucky enough to have all the required cogs working under one roof, with funds available to facilitate a faster transition than is normally expected.  

This week we publish work in the Journal Clinical Cancer Research which describes an improved method for detecting cancer.  It has been well established that the earlier cancer is detected, the better.  If left untreated, cancer spreads to other parts of the body.  It's these secondary tumour sites, known as metastases, that normally result in fatality.  If cancer is detected early enough, the primary tumour can be removed by surgery or shrunk by a cocktail of drugs before it has a chance to spread.  The recent debate regarding breast cancer screening (article from the BBC here) has highlighted the need to develop improved ways to detecting cancer.  We want to make sure everyone with the disease has it detected, while others are not falsely diagnosed.  

An example of the type of scanner used for cancer diagnosis
In this article (found here), we describe refinement of an already existing technique used to detect cancer - monitoring how cancer cells use large amounts of choline, an essential nutrient which must be consumed in the diet for the body to remain healthy.  The amount of choline 'consumed' by cancer cells is far higher than the normal surrounding tissue, meaning that by measuring choline consumption, one can assess whether the patient has cancer or not.  The current method has proven particularly useful for detecting prostate cancer, where other more conventional methods have had less luck. 

Choline 'consumption' in tumours can be measured by injecting small amounts of radioactive choline into a vein and monitoring its accumulation into tissues in the body. This is all performed while the patient is in a scanner, an example of which is shown above.  Doctors are then provided with a 3D map of radioactive choline accumulation which is used to make a diagnosis.  In this article, we provide evidence that by slightly modifying the radioactive choline that is injected, a far more accurate 3D map is produced which should better discriminate cancerous tissue from healthy tissue.  Far more validation is required, with preliminary tests in humans scheduled for this year, but if these exciting initial results are shown to be robust, this test may be employed by the NHS and others in the not-too-distant future. 



Welcome all to my new Science blog!

My name is Tim Witney and I'm a biologist working at Imperial College London on new methods to detect and monitor Cancer.  Over the next few months I hope to give a brief introduction to how we can gain valuable information on one of the World's most deadly diseases using new and exciting imaging techniques.  Where a simple x-ray can be used to detect a broken bone, these new methods are being developed to detect cancer itself and subsequently whether the cancer responds to treatment.

I hope to try and explain some of the work that I have published in a simple fashion that is easy to understand. Traditionally, the science community has found this quite hard work, so please bear with me.

While I'm at it, I should point you in the direction of the Comprehensive Cancer Imaging Website: where you can find more information on the work we do, along with my own personal site: