Please can you give a brief introduction to your research?
The main objective of our research is to improve and individualize cancer diagnostics and cancer treatment. We try to achieve this through the integrated use of MR technology and the development of data-driven tools to analyze tumors on both a functional and molecular level.
Understanding Cancer at a Molecular Level
One important feature about the research facilities we have here is that my research group has access to equipment for analyses on the molecular level through translational pre-clinical research, but we also perform clinical imaging studies. This creates a framework which is interesting in the sense that you can translate basic findings over to more clinical settings.
What is specific about Norway's scientific center here in Trondheim and also in the context of your research?
I think Trondheim is really the technology capital in Norway. There has been a long tradition of collaboration between the technology environments at the university and the clinical staff at the university hospital. This is also the case within our lab, which is located within the hospital campus.
For decades, we have had a strong collaboration with clinicians and I think this is tremendously important because they meet the clinical challenges in their everyday life. They know what cases need solutions, and this dialogue and collaboration are important to ensure that the research we do is viable and useful for the future.
Do you have an example of a recent clinical NMR metabolomics study that was of significant impact on your work and for the future direction of this research?
Yes, I think one of our important studies has been working with tissue samples from breast cancer patients who had received chemotherapy prior to their surgery. The patients in this group have large tumors that need to be reduced before they can have surgery. We know that not all treatments work equally well for all of the cancer types and being able to either predict patient response or outcome at an early stage would be a benefit in terms of planning and monitoring treatment.
The study found that all of the breast cancer patients had a huge metabolic response to the treatment, but we couldn't differentiate between the good responders and those who didn't respond. However, when we had access to the survival data for this patient cohort, we realized that, based on the metabolic profile, we could differentiate between those who survived longer than five years and those who passed away before this time.
I think there is an increasing interest in research towards personalized medicine or precision medicine as it is also referred to , in clinical cancer care.
In addition to focusing on clinical NMR metabolomics, you also have pre-clinical MRI imaging systems in your lab infrastructure. Do you have any examples of research studies where both come together in a synergistic way to provide answers to complex questions?
I think there are many benefits of this in treatment studies. When new drugs are tested for cancer treatment, we follow the response with pre-clinical MRI in animal models. We then use imaging techniques to look into vasculature and micro-environmental properties. Since we also have access to metabolomics equipment or NMR spectroscopy, we can also analyze the tissue samples to see how this treatment affects the metabolism and how that may differ between tumor types or between different treatments.
One example that could show the benefit, is one of the clinical projects where we work on optimizing MR-guided biopsies. A biopsy is taken while the patient with suspected prostate cancer is within the scanner. In this way, we can obtain a biopsy from the prostate whilst knowing the exact location from which this biopsy is taken. We know what the radiologist said about the image, and we get to know what the pathologist determined after the metabolomics analysis, so we know whether it's a cancer or not. In this way, you can bridge the molecular information you can obtain from this biopsy by applying metabolomics to the clinical images that were obtained from the same patient.
What kind of function and metabolic properties of cancer have been revealed through NMR in recent studies?
It's recently become more evident that we need the metabolic level to understand how tumors develop and progress. One molecular level, for example, the gene expression level, is not enough. Our studies really indicate that a metabolic profile is able to subtype breast cancer in a similar fashion to how gene expression does.
We know that a metabolic profile can predict hormonal status, which is an important prognostic and predictive factor in breast cancer, for example. We know that the metabolic profile is related to the long-term outcome of the disease.
How important is NMR for the future of cancer research?
I think NMR is also a driving force for future research within cancer because metabolism or metabolomics is giving us so much information on a molecular level that wasn't really easily accessible before. The non-destructive nature of NMR, also when analyzing tissue samples, gives unique possibilities for combining information from several molecular levels, which is important for understanding underlying molecular mechanisms, an important basis for identification of relevant biomarkers and the development of new drugs and treatments.
When looking at NMR metabolomics now, what would you say were the key innovative steps that really enabled these technologies to enter a more clinical hospital research environment?
I think there have been several improvements over the last decades. One example is the accessibility to 600 megahertz systems, as a kind of a standard system for metabolomics analysis. The magnets now have much better shielding, which makes their placement into the lab or into a clinical environment simpler. We have much better possibilities for high throughput of samples and I think that's also tremendously important.
The software has also progressed, which has contributed to this, especially the simpler user interfaces. You don't need to be an NMR specialist anymore to run a spectrometer. Although you require knowledge to really understand the underlying physics, you don't need to be an NMR specialist to analyze the samples.
We usually analyze samples that are already collected in biobanks or in biobanks that we collect together within clinical environments here at the hospital. We try to analyze as many of these samples belonging to one study as quickly as possible. The responsible student or researcher then prepares the samples according to strictly defined protocols.
If the samples are biofluids, automation is much simpler and has come much further. In that case, it's simply a matter of putting the samples in the magnet and they will be analyzed automatically according to already defined protocols. For the tissue analysis, the work is more manual and more time consuming, although there have also been advances there, which has made it simpler.
Do you see biobanks playing an increasingly important role in your research arena?
I think there is an increasing awareness of the importance of good and large biobanks. I think this is necessary when it comes to solving some of the huge societal challenges. For example, in the case of cancer or cardiac heart disease, such biobanks will make it easier to validate the findings we have or the indications that we now have for disease biomarkers. This will be an important aspect in terms of taking current research a step further, for the validation process.
Do you think NMR metabolomics could contribute towards a personalized healthcare approach?
I think NMR spectroscopy and metabolomics definitely play a role in personalized medicine. For example, we can look at the genetic subtypes of breast cancer that was identified more than ten years ago and has since been reproduced and validated several times. These genetic subtypes aren't regularly used in the treatment and follow-up of breast cancer patients yet. There is increasing evidence surrounding this, to show that you can't just stick to one molecular level; you need the whole cascade, where metabolomics is the downstream level of this, close to the phenotype that you are investigating. In this sense, I think NMR spectroscopy has a central role in terms of further research and potential clinical use in personalized medicine.
Can you tell us a little bit more about treatment stratification?
Within cancer treatment, there is now an increasing focus on targeted treatment, which means you really need to know what treatment would be most beneficial to the patient before the patient receives treatment.
In a cohort that we recently investigated, we were able to define three naturally occurring metabolic clusters across the breast tumors. The interesting part here was that these clusters were not the same as the genetic subtypes that were defined for the same patients. This means that the information that NMR brings is supplementary to the genetic information. This means you can achieve an even better characterization of tumors when you combine the genetic and metabolic information.
What's also interesting is that these three metabolic clusters that we identified expressed certain metabolic features that could be used to direct additionally targeted treatment, which may differ depending on what metabolic cluster your tumor is characterized as.
I think personalized/precision medicine or individualized therapy is an important concept. For example, new cancer drugs can be incredibly expensive. We know that it will not be effective for all patients, so knowing upfront or at least being able to detect early on in the treatment course which patients respond and which patients don't, is important.
This is important not only because of the costs related to the drug, but also for making sure that you direct the patient towards the most effective treatment and save them from ineffective treatment that will have side effects.
What further research and strategies are needed to improve diagnostic tools to stratify patients to treatments?
The research from metabolomics projects has clearly shown or identified biomarkers that could be used in a clinical setting. What is currently lacking is the possibility to test these biomarkers in large cohorts, perhaps in multi-center settings, to be able to really validate their usefulness.
One of the prerequisites for performing such validation studies, is the need to standardize how samples are collected and how they are treated before you analyze them - acquisition protocols. In addition, it is very important to standardize how you analyze the NMR spectra and present the results in the end.
It's also important that data is shared. When you have analyzed and published data, the raw data should be shared. This has already been the standard for the genetic community for years; the data are shared and can be used by other researchers later. I think this is really necessary in order to move forward with research. We need to be less protective about our data and share them, so that other researchers can contribute to take them further.
You mentioned the importance of sharing data. Is there also a significant importance in the collaboration with your industry partners to push the research to the next level?
I think collaboration with industry is important because these may be the vendors who convert new findings into commercially useful products. Therefore, this is also definitely an aspect of research which is important for ensuring that what you do can be brought for use.
Where do you see this research area in about five to ten years from now, in the context of cancer diagnostics?
When it comes to understanding cancer on a molecular level and understanding the mechanisms of development and progression, I think we're still scratching the surface.
However, I think the clinical efforts now being made within metabolomics to clinically apply the technology could lead to methods to stratify patients to certain treatments, to monitor their response at an early time point in the treatment course or even to say something about their prognosis.
About Professor Tone F. Bathen
Professor in medicine (MR Technology) Tone F. Bathen has extensive experience within medical applications of MR Technology. She is head of the MR Cancer Group at the Norwegian University of Science and Technology (NTNU).
This interdisciplinary group consisting of 25 researchers and engineers covers a wide range of research from molecular and translational work to clinical applications, mainly within breast and prostate cancer.
The long-term objective is to improve and individualize cancer diagnostics and treatment by developing integrated MR-based methods and data-driven tools for functional and molecular assessment of tumors.