Editor's note: Irene Tracey, MA, DPhil, is professor of Anaesthetic Neuroscience in the Nuffield Department of Clinical Neurosciences and Pro-Vice Chancellor (without portfolio) at the University of Oxford, UK. She is also warden of Merton College, Oxford. In 1997, she helped to co-found the now world-leading Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB) at the University of Oxford and was its director from 2005 until 2015. Over the past 18 years her multidisciplinary research team has contributed to a better understanding of pain perception, pain relief, and nociceptive processing within the injured and non-injured human central nervous system using advanced neuroimaging techniques. More recently, they have been investigating the neural bases of altered states of consciousness during anesthesia.
In this interview, Tracey chats with PRF Correspondent Isobel Parkes, a PhD student at University College London, UK, to discuss brain imaging biomarkers of pain, research priorities for mental health in the COVID-19 context, her appointment as president-elect of the Federation of European Neuroscience Societies (FENS), and much more. Below is an edited transcript of their conversation.
In a 2019 perspective in Neuron that you co-authored with Clifford Woolf and Nick Andrews, you placed emphasis on the notion of mechanistically driven biomarkers of pain, and you argued for composite biomarker signatures for pain. What did you mean by that?
We wrote the review because we felt it was a good time to bring together quite a broad array of research approaches to try to understand pain. Clifford works very much on the molecular/cellular side of things, and I’m very much focused on how systems are behaving. In science, we often work in our little camps, and there’s not much opportunity to bring them together. But because pain is an example of a complex phenomenon where things go wrong right from the level of the receptor all the way through to brain mechanisms, we are actually starting to think about how we can pool our data and have a more integrated understanding of what’s gone wrong, particularly in the chronic pain state.
We have a huge number of tools now – you might want to call them objective – that can help you understand the mechanisms for the underpinnings of pain. Having run an imaging center, I see lots of different diseases of the central nervous system, and it’s evident from many other conditions – let’s take dementia and Alzheimer’s as examples – where bringing together different pieces of information into a composite measure gives you far more predictive power in determining a disease trajectory or response to treatment.
The opioid crisis and the NIH HEAL Initiative have emphasized the fact that we haven’t got simple, usable biomarkers for pain. I don’t think, however, that there’s a single biomarker for pain, so we wanted to emphasize this composite biomarker approach rather than a Holy Grail of a single biomarker with a single technique that is going to answer everything. It’s just not that easy, sadly.
How nuanced are our current state-of-the-art objective measures? Hospitals routinely use MRI scanners with field strengths of 1.5 Tesla or 3 Tesla; however, ultra-high-field scanners are on the rise, with 7 Tesla scanners cleared for clinical use in 2017, for example. You argue for the use of neuroimaging as an objective biomarker of pain, and yet even a 7 Tesla scanner only gives a resolution of about 25,000 to 50,000 neurons.
It all comes back to the question you’re asking. So if, for example, your question is very detailed about understanding something that requires single neuron resolution in a particular location in the brain, then you wouldn’t use imaging; you’d use electrophysiology or another technique. Imaging with a 7 Tesla isn’t going to give you that single neuron resolution, but that doesn’t mean to say it’s bad; it’s using the tool for what it’s powerful at doing, which in the case of imaging is very much looking at the entirety of the whole organ – the brainstem and the spinal cord. Imaging can be incredibly nuanced if your question is looking into, say, how cognition or mood influence pain mechanisms at a system network level. It can then reveal things that you wouldn’t necessarily get from the limited ways we might verbally interrogate somebody’s pain experience.
Virginia Woolf once wrote: “The merest schoolgirl, when she falls in love, has Shakespeare or Keats to speak her mind for her; but let a sufferer try to describe a pain in his head to a doctor and language at once runs dry.” Pain is by definition a subjective experience. What is your opinion on the conversation about these objective measures replacing self-report as our primary measure of a patient’s pain?
If there is a capacity for somebody even with limited language and vocabulary to give a report – we’ll come back to cases where you just don’t have the luxury of subjective report – that is the experience that they’re having and that clearly has purpose and relevance. The job for the physician, the diagnostic job, or for the scientist trying to understand pain is to work out why they’re having that particular experience and why it is that they’re not capable perhaps of experiencing relief in response to a treatment. But even with more and more sophisticated questionnaires, it is very hard to unravel completely why "it is hurting."
This is where imaging can help. You’re marrying the two; imaging is not replacing self-report, and it’s not a surrogate for it. That is often the mistake made by non-imagers. The pain rating and the person’s description of pain are of primary importance because that’s their private experience, and at a philosophical level you cannot objectify a subjective state. The subjective and objective give you two pieces of valuable information.
And of course there are instances where you haven’t got the luxury of self-report, such as elderly people with dementia, those who are comatose or anesthetized, and babies. This is where you get into more challenging dilemmas as to what decisions you make based on what you’re seeing with your metrics, and then inferring what they might be experiencing. We have to be honest about those limitations. It does get a bit philosophical, and at one level you’ll never know what people are really feeling and perceiving – as that is their experience – but we can try to better understand what contributes to and makes up that experience in terms of brain processes and mechanisms.
Can you say a little bit about your work relevant to the transition to chronic pain?
In the definition of chronic pain, there’s a three- to four-month period of transition after normal tissue healing time when things go wrong. There’s a lot of interest now in pre-vulnerability – that there might be something already predisposing one to chronic pain. We now appreciate that there is more contributing to that likelihood of a persistent chronic pain state than just this magical three- to four-month transition.
Several central mechanisms are coming through as possibly dominant, for example, the descending modulatory systems – descending facilitation and descending inhibition – which no doubt appear to be pivotal in controlling whether you’re going to stay in a chronic pain state or not. It’s a normal response for these systems to produce hypersensitizations in reaction to an injured state. But for people who end up in a persistent chronic pain state, these systems appear to be altered or imbalanced, so the capacity to drive healthy descending inhibition is removed. You get this double whammy: You’re losing the ability to benefit from endogenous analgesia, and you’re driving the facilitation. Why these systems become disengaged is an area of really interesting research, and it might well be they are already imbalanced before the person developed persistent pain. Further research is needed.
So this imbalance seems to be really important and is centered largely in the brainstem. We have spent a lot of time developing imaging to see these modulatory systems, as the brainstem is quite hard to image, and then verifying that the systems are being recruited in central sensitization, in experimental models of hyperalgesia. We have then validated and proved that these systems are relevant across different chronic pain conditions such as painful diabetic neuropathy, musculoskeletal pain, and osteoarthritis and rheumatoid arthritis. It’s a nice story, with basic science developing imaging tools, proving that the mechanism is relevant in experimental medicine models, and then translating it into the clinic. In parallel, we are also doing pharmacological studies showing that if you’ve got drugs that are reasonably good at treating, say, neuropathic pain, this seems to be a system that we can see a corresponding change in, meaning it could be used as a biomarker of effective therapy.
Another area of interest is the Bayesian brain: The idea that you have a prior probability of expectation about what things are going to feel like, such as pain relief subsequent to taking a pill. We know that these “priors” – a nice way of computationally describing expectations – very powerfully shape what pain or pain relief you’re going to get, more powerfully than I would have ever predicted. The challenge, then, is when these priors are coupled with all the modulatory systems being out of kilter, we have a potentially far more complex problem on our hands when it comes to treating chronic pain. A temporal mismatch might occur. That is to say, although you might therapeutically fix certain mechanisms, these priors are likely to influence or bias the patient’s perception when it comes to pain. The challenge will be: How do you fix people’s priors? How do you get people to shift that prior back to normal? Not easy.
Shifting gears, you’ve talked in the past about the concept of “hedonic flipping” – making pain pleasant. What research have you done in this area?
That’s quite early research and was led by one of my many wonderful students, Siri Leknes, who is now a professor at the University of Oslo and is still carrying on work related to this topic. I haven’t pursued it much subsequently, but it’s still an area that interests me because of rethinking about all the old data and old experiments we did, but now with a Bayesian brain framework.
You generally seek things that are nice, and avoid things that are not nice, and every day, in fact, that’s what you’re doing, making these tiny little decisions. The analogy I often use is that I’ve been running marathons over the past few years, and after a long training session leading up to the big day, you’re pretty stiff and sore, but it’s “nice” muscle pain because I am interpreting it as a positive thing; I’ve done that long training run, I’m hitting my target, it was a great session. It’s a reinterpretation, a reappraisal of what the pain means.
We did a range of experiments [see here, here, and here] to look at the analgesic effect of distraction, and how the quality of pain can change when you change the context in which you’re experiencing it. All that person is really doing in that context is reinterpreting and reframing the meaning of the pain. So in these experiments, we showed that moderate pain feels very different if you’re expecting high pain but receive moderate pain – you actually like it because it could have been worse. However, if you thought you were going to get low pain and you receive the moderate pain, you don’t like the moderate pain so much. It seemed like we were just doing fun experiments on a Friday afternoon, but it actually told us something about how powerful these systems in the brain are – that they can change the quality even of something as powerful as pain. You really can make it pleasant in some circumstances.
There’s still a lot to be unraveled, but what we found was that a lot of these reward networks were routed through the descending system. Interestingly, the behavioral experiment we published showed that if you give a lot of naloxone, an endogenous opioid blocker, you can stop the more distraction-type pain, but you can’t stop the more reappraisal-type pain. So it’s obviously through a different modulatory system, which is really interesting.
A lot of these experiments, I hope, are of interest and relevance to people using psychological interventions because they might in part explain what’s going on with those interventions in terms of distraction versus reappraisal. It may help elucidate different strategies that patients might be more disposed to benefit from or not, based on how their brain is wired.
Moving beyond pain research for a moment, in April of last year you were a co-author on a position paper in Lancet Psychiatry comprehensively outlining current research priorities for mental health during the COVID-19 pandemic. How did this position paper come about, involving such a diverse range of researchers?
I’m a trustee for a mental health charity, MQ, which I’m very proud to be involved with, raising money to support mental health research. I’m also a fellow at the The Academy of Medical Sciences in the UK. When the COVID crisis hit, we were very quickly learning, from some of the symptoms, that there were neurological issues. The Academy of Medical Sciences wanted to discuss and swiftly disseminate information on what some of the neurological and psychological repercussions might be, and they joined forces with MQ, asking several scientists to come together. It was felt it would be important to very quickly write a position paper that was a sort of “call to arms” highlighting various issues about the virus' effects on the central nervous system and the consequences of lockdown on mental health.
On the neuroscience side, for those patients who were dying, we needed to be thinking, morbidly: Are we biobanking their brains? Are we collecting tissue samples? This needs to be thought about and coordinated at a national level so that we don’t miss the boat on capturing samples to understand the effects of the virus on the brain. Are we capturing the data from tests ongoing in the wards that might tell us more about the virus and its impact? What are we implementing, on the psychological side of things, with regard to the mental health problems of impending lockdown, from increased suicide, abuse, and loneliness? We wanted to outline many things that would be important to alert both the funding agencies and the scientific communities to, as well as patient groups and different charities – the idea being that if you’ve got patient access, let’s do some quick surveys, and let’s find out how people are coping. What might we do to mitigate problems arising?
We didn’t capture everything, but we did as much as we could during the several weeks of the Easter vacation. Lancet Psychiatry was very keen to publish the position paper, and we hope it’s useful. It was, in my opinion, particularly helpful for the funding agencies at that time, who were focused on fast-tracking what research they were going to fund, to make sure they didn’t forget the mental health and the neurological side, whilst obviously they were focused on the viral infection itself and getting vaccines out.
To bring it back around to pain, one of the key research priorities identified in the paper is a need for real-time mental health monitoring. Given that depression and anxiety, for example, are very common co-morbidities in patients with chronic pain, do these research priorities have potential benefit outside of pandemic-related contexts?
I think that is the direction we’re moving in. The use of neurotechnologies and wearables can be beneficial. In particular, with pain, being able to capture things at an ecological level about a patient, measuring how the person is doing in terms of movement and enjoyment of life, in addition to exactly what they’re perceiving, can be useful in helping us understand what intervention might be working. Patients may not even be necessarily aware of it, but, for example, you’re noticing with the wearables that people are actually walking longer each day, or they’re enjoying things a bit more. This use of technology, when done well, can be very powerful because of the amount of data it can give you and the different windows into the person’s health it provides.
The problems with mental health changes, in terms of anxiety and depression, are so prevalent in pain, and so we absolutely have to be capturing and understanding this better. We’ve shown powerfully that these are huge neural amplifiers in the brain that make the situation, the pain, worse, so we’ve got to address them.
Congratulations on your recent appointment as the president-elect of the Federation of European Neuroscience Societies (FENS). What are your priorities during your presidency, and what are you hoping to achieve?
I’m really honored to be elected. For the next two years I am president-elect and then I take over, so it’s been great to start to work with the steering committee. FENS is an extraordinary organization that brings together all the different neuroscience societies across Europe. FENS is all about making sure that within Europe we support and value neuroscience by helping our member societies and championing neuroscience wherever possible, from blue-sky discovery through to clinical translation and other areas of societal translation where neuroscience has “reach.” We are an advocate for the value of neuroscience, and so partly it’s about making sure we have secured funding and more funding.
Part of what I want to champion during my presidency alongside this core work of FENS is to improve diversity in neuroscience, as well as better support career development. How are we looking after and supporting researchers, particularly in a post-COVID world where it’s going to be challenging, and, specifically, how are we supporting our younger-career researchers? I’m passionate about research culture, and the models by which we do science and the cultures that we’ve created around science. It’s important that we think and reflect deeply on whether this is a good culture that we’ve created for scientists and for new people to develop their scientific careers within. Are there things that we should be tweaking, altering, or dismantling in order to create an environment that maximizes the impact a scientist can make, irrespective of background?
Your appointment was fantastic news to British science and research, given that the Brexit transition period has now ended. What implications do you think Brexit might have on pain research, and research more generally, in the UK?
I was not in favor of Brexit, like most scientists; I’m firmly European, and I’ve been European most of my life. It’s worrying. We’re already seeing, both at an undergraduate and a graduate level, a reduction in the number of European students choosing to attend some of our universities. We saw these aftereffects of the referendum very quickly, and I think we’re still seeing them. Our desire to attract and retain the very best from Europe to want to come to Britain is still important, not just for the scientific opportunities, but also for what is, hopefully, an enjoyable place for them to be culturally, too.
Brexit is still a highly divisive issue. Of course, we’ve always been very global with our science, and it becomes more global as you go from undergraduate to graduate to postdoc to faculty. Such diversity is what has kept British science excellent. Will it be the case that we’re going to become less and less diverse? We’re certainly seeing an erosion at the undergraduate and graduate level coming in from Europe. Are we necessarily seeing more people coming from other countries? Too early to say.
At a European funding level, we’ve been given a lot of assurances, but we will just have to see how that actually plays out in the fullness of time. All of us share a concern that this might have a very significant impact. Although recently, it’s been extraordinary, the realization of just how important science is in Britain, and with UKRI [UK Research and Innovation] putting in mitigation plans and huge amounts of extra funding, and obviously with the Wellcome Trust, we’re still very well supported for science. Nonetheless, we have been very successful with European funding and we just hope that opportunity will remain.
To conclude, in the spirit of one of my favorite interview podcasts, Desert Island Discs, but to adapt it for the bookworms out there, what three books would you take with you to a deserted island?
Blimey – that’s tough as I’m a real bookworm and read every day. I am never without a book on the go. Right now, I’m just about to finish the third and final book on Thomas Cromwell by Hilary Mantel – she’s unbelievable and her writing is so beautiful. I have loved being immersed in that period of Britain. But for a lifelong stay on a deserted island, I should take books I’ve not read or have read parts of but need more time to complete and have the time to reflect. So, as I love Shakespeare for the sheer brilliance of storytelling, intuition about people and their relationships as well as for the beauty of expressed language, my first choice would be his collected works.
I am also a fan of poetry and so I’d take an anthology of poetry; please, that must include my favorite poets drawn from past and current times. I’m also a massive fan of Bill Bryson, who I have been very fortunate to meet and who is just wonderful. His observational abilities of people and situations are unparalleled and his writing always makes me laugh out loud. So I’d like to take his book, Notes from a Small Island, and the sequel, too, The Road to Little Dribbling: More Notes from a Small Island, to crack me up and to remind me of home.
PRF Correspondent Isobel Parkes is a PhD student at University College London, UK.
Additional Reading
Tracey I. Pain: A Ladybird Expert Book. Penguin Michael Joseph, 2020.