Pick up the trace - using PET to identify changes in cell metabolism

11 March 2016



Imaging tumours and metastases against the complex background of the brain is no easy task, but scientists at the Stanford University School of Medicine have developed a new method of highlighting tumour tissue using PET to identify changes in cell metabolism. Sarah Williams speaks to lead authors on the study, Dr Tim Witney and Dr Michelle James, about developing the technology, and what it could mean for brain cancer patients.


The rate at which cancer cells divide and replicate makes it vital to pinpoint the size and position of tumours as early as possible. When the tumour in question occurs within the brain, achieving this clear-cut image is especially challenging for radiologists, because the complex structures and activity of the surrounding healthy tissue can shroud it from view.

Now, scientists have found a way of locating brain tumours by essentially beating them at their own game - using a new PET ligand to trace the presence of an enzyme during the unique process by which cancer cells multiply.

A promising prognosis

Developed at Stanford Medical School, the technique uses a specially formulated radiotracer to highlight pyruvate kinase M2 (PKM2), an enzyme central to the metabolism of tumour cells, which differs to that of healthy cells.

The study, first published in Science Translational Medicine in October 2015, was co-led by Dr Tim Witney, former postdoctoral fellow at Stanford and now group leader at the Centre for Advanced Biomedical Imaging at University College London, and Dr Michelle James, instructor in the department of radiology and neurology. Professor Sanjiv Sam Gambhir, director of Stanford's molecular imaging programme, oversaw the project and is the paper's corresponding author.

The team's interest in the metabolic processes of cancer cells began with the well-trodden path upon which PET itself was developed. Known as the 'Warburg effect' after the Nobel prize-winning scientist who discovered it over 90 years ago, aerobic glycolysis is the process by which tumour cells obtain the energy needed to divide and grow.

For many women in developing countries, the choice does not lie between screening mammography and screening ultrasound: it lies between screening ultrasound and nothing at all.

One of the most commonly used PET tracers, fluorodeoxyglucose (FDG), targets this process by mimicking glucose. It is consumed by cells accordingly, and can therefore be used to highlight cancer cell growth (by identifying where glucose uptake is greatest) within many parts of the body during PET. But as James points out, its effectiveness is diminished when it comes to brain cancer due to FDG also accumulating in healthy brain tissue.

"The brain needs a lot of glucose, so it's difficult to pick up tumours above that huge background of the normal requirements for glucose," she says. "That was a big driving force for us to come up with a better approach."

With a focus on brain tumours and a desire to improve rates of early diagnosis for hard-to-detect, aggressive tumours such as glioblastoma, James and Witney set out to design a method that avoided the problems encountered with FDG.

"We thought it was very promising to continue looking at tumour metabolism, since tumours reprogram their metabolic pathways to support synthesis of new macromolecules, so that they can rapidly divide. Glycolysis is certainly up-regulated in cancer cells, so we wanted to look at other parts of that process."

Through their initial reading, the researchers became excited about the potential to target pyruvate kinase, a rate-limiting enzyme present in the very last stage of glycolysis within tumour cells. In cancerous tissue, it is predominantly the type M2 isoform of the enzyme (PKM2) that is expressed, while normal brain tissue expression was limited to type M1 (PKM1). As well as its potential to improve diagnosis, the idea of tracing PKM2 also appealed to the researchers because it might provide valuable insight into cell metabolism.

Witney, whose ongoing work includes looking into the mechanisms that drive resistance to cancer therapeutics, explains why this biomarker is so significant.

"PKM2 is really the master regulator of tumour glycolisis," he says. "It's an extremely important target because it allows tumour cells to switch between catabolic and anabolic metabolism. Catabolic metabolism is the ability to produce energy, and anabolic metabolism is creating complex molecules from more simple building blocks. Most other cells don't have that plasticity: the ability to switch between these two different states.

"PKM2 allows the cancer cell to adapt to its local environment so that it can switch between these two different states depending on its needs. So it's very important to understand its role in the body and in vivo, and there's really no way that we can interrogate this other than through imaging."

After verifying what they'd read by sampling healthy mouse brain tissue and cancerous tissue to test for the two kinds of isoforms, the team developed their approach. Taking inspiration from a new stream of therapeutics aimed at PKM2, James (who specialises in designing and evaluating novel PET imaging agents for brain diseases) created the molecular tracer. Named [11C]DASA-23, it is based upon a group of molecules known to bind to PKM2.

The team could then begin to test the tracer in practice to assess its efficacy, with Witney taking the lead in designing the models for evaluation.

"For all of these radiotracer development projects, we always start in isolated cells in culture, which, of course, is not really amenable to what we would see in vivo in animals, but it gives you an idea of the levels of uptake," Witney says. "It's very important to look at the specificity of the radiotracer, and by that I mean whether it will bind just to PKM2, or whether it will bind to other targets within the cancer cell."

Clinical trials

After proving the tracer's effectiveness in this regard, the researchers implanted the same cancer cells onto the backs of mice, growing small subcutaneous tumours. The mice were injected with the radiotracer and scanned using PET over the course of an hour, to test the distribution of the radiotracer over time and the degree to which it targeted the tumour.

"And fortunately, we again found that we got very good uptake within this tumour," Witney says.

Next, the researchers moved to an orthotopic implant model, in which tumour cells are seeded and grown within the mouse brain in situ, in order to recreate the local environmental conditions of the brain. Here, the significance of Stanford's method over conventional methods came into its own, with the team observing a very clear signal corresponding to the expressed PKM2 of the tumour, and no uptake in the surrounding healthy brain.

"I clearly remember that night in the lab," James says. "We were very excited, and I think the pinnacle for us was then using the newest type of tumour model, the 'patient-derived tumour'."

This model, in which a cell line taken directly from a cancer patient is injected into the mouse brain, is widely seen as the most clinically relevant, because they grow very differently to the cell lines developed in culture, says Witney.

"They have these protruding, infiltrating nodes that really dig deep into the brain of the mice, and this is what we see in patients as well," he says. "So it was great to see that we can pick these up as well with our PET imaging agent."

With the effectiveness of [11C]DASA-23 as a PET tracer proven in mice, the team now plans to trial the approach in human patients, once US FDA approval has been secured. While optimistic about the results and hopeful about the potential applications of the technique, Witney and James caution that it is a large transitional leap to make.

"When you look at a new imaging strategy in mice, it's not always indicative of what you see in people," James says. "The mouse is so small, it's pretty much a walking liver, so things get metabolised very quickly and all the organs are close to each other."

Look forward

Alongside the trials in the clinic, which could reveal wider uses for the technique in other parts of the body, another step will be revisiting and revising the radiotracer. One additional application for this PET tracer that could be significant is in the prostrate, believes Witney.

On whether or not a substantial number of patients will be being imaged with PKM2 within that ten-year timescale, I’m optimistic.

"PKM2 itself is very tightly associated with the grade of prostate cancer - how aggressive it is - which ties in with patient survival rates," he says. "It would be really nice to be able to stratify your patients into these grades more accurately through PKM2 images of the prostrate, so I think that's somewhere there are unexplored avenues."

For now, the major focus for the researchers, moving to the clinic, remains the brain, and the hope of proving the technique so it can help to improve patient outcomes as soon as possible.

"The radiotracer is not taken up by the normal healthy cells, so that's one big advantage," says Witney. "But probably more exciting, the second advantage is the fact that it can be used as a surrogate marker of response for new therapies against glioma.

"There are a number of drugs currently in development that are specifically targeting PKM2 in order to try to kill the cancer, and this radiotracer will be able to give us a marker of whether the drugs are actually binding to the brain cancer itself."

James agrees, pointing towards the sensitivity of the new tracer and its specificity to PKM2 as its major calling card. "I think, when we compare it with FDG, we're going to have a much better signal-to-background in the brain, which leads to greater sensitivity, so we can detect tumours much earlier," she says. "We can also monitor response to therapeutics at an earlier stage and pick up those small changes that you wouldn't otherwise pick up with FDG.

"There are some other tracers out there that people are evaluating, but because this particular tracer targets PKM2, the last rate-limiting step of glycolysis, it seems to be a lot more specific to cancers, especially when we're looking at the brain."

Asked when their method could be initiated in hospitals, the pair is hopeful that - providing its efficacy is fully proven in the clinic - patients could see the diagnostic benefits of the technique in the near future. To achieve this, collaboration with industrial partners will be key, says Witney, but his faith in the work of James, Gambhir and the truly multidisciplinary team at Stanford is at the forefront of his outlook.

"If ever there is a place to do it, Stanford is the best place to get this into patients as quickly as possible," he says. "So whether it will be in routine clinical practice around the world in ten years, maybe not - but on whether or not a substantial number of patients will be being imaged with PKM2 within that ten-year timescale, I'm optimistic."

A radioactive tracer is a chemical compound in which one or more atoms have been replaced by a radioisotope. It can be used in research for the better understanding of chemical reactions.
Dr Tim Witney is a Sir Henry Dale fellow and group leader at the Centre for Advanced Biomedical Imaging at University College London, UK. His group is concerned with discovery and development of new PET and MR methods for cancer imaging.
Dr Michelle James is an instructor in the department of radiology and neurology at Stanford University. She received her PhD from the University of Sydney. Her research is focused on designing and evaluating novel PET imaging agents to improve the way we diagnose, treat and understand brain diseases.


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.