Editor’s note: B. Duncan X. Lascelles, PhD, is professor in small animal surgery and pain management at North Carolina State University (NCSU), Raleigh, US. His research program, Translational Research in Pain (TRiP), develops methods to measure pain associated with spontaneous disease in animals, and seeks to understand the underlying neurobiology. His work improves pain control in companion animals and facilitates analgesic development in human medicine. He is also director of the Comparative Pain Research and Education Center (CPREC) at NCSU, and co-founder of AniV8, a company dedicated to developing innovative methods of measuring pain.
In this interview, Lascelles chats with PRF Correspondent Sherelle Casey, a PhD student at the University of Sydney, Australia, to discuss the latest advancements in veterinary pain management, the development of wearable sensors to improve animal health, and much more. Below is an edited transcript of their conversation.
What are the most exciting developments in veterinary pain management in recent years?
If you ask me to pick one, it’s the recent approval of anti-NGF [nerve growth factor] monoclonal antibodies. The first one has been approved for dogs in Europe, and it’s likely that cats will follow and that approvals will come to North America. It’s probably one of the most exciting developments in pain research and management across human and veterinary medicine. It’s a novel therapeutic mechanism – something different for the clinician’s toolbox.
But veterinary medicine has seen a number of novel therapeutics recently. Just in the last few years, we’ve had grapiprant [an antagonist of the prostaglandin E2 EP4 receptor]; similar to the anti-NGF antibodies, that therapeutic approach is not yet available in human medicine. We’ve also seen Nocita, the veterinary equivalent of Exparel [liposome-encapsulated bupivacaine]. The fact that those drugs have been approved tells you there’s activity in this space. After a number of years of not seeing innovation in veterinary pain management, suddenly things are really moving and providing us more tools to manage pain.
From a personal perspective, one of the most exciting things is the acceptance of naturally occurring painful disease in companion animals as a model for human conditions. It’s been gradually happening over the years, but now, instead of veterinary medicine and human medicine being viewed as two separate fields, you are starting to see these disciplines come together in the pain space, and quite rightly so. It’s pretty exciting because the payoff for veterinary medicine will be that we will learn more about measuring and managing pain in companion animals, and there’ll be more money for research. The payoff for humans is that research involving veterinary species will contribute to translational pain research, and facilitate better decision-making during therapeutic development – a topic I’ve recently written about.
Are human clinicians welcoming of this translational approach, or are they a bit aloof?
The people I find most open are the human pain researchers and clinicians who have the ability and time to think laterally. What I mean by that is, they’re in a good place in their career. The people who are less accepting of these ideas are early-career individuals who are rightly focused on being successful – the cycle of getting the next grant and getting the next paper out. If you’re a young faculty member and you want your research to be successful, you do what everyone tells you is the right thing to do, which is to submit grants for rodent research as models of human conditions. And you do not have the bandwidth to think laterally, because that diverts you from this accepted path to success.
That said, I have found many pain researchers and pain clinicians who are very open to and excited about the idea of embracing companion animals as a translational model, but they tend to be those who are already successful in their careers and have the luxury of being able to think both laterally and innovatively.
What is the role of rodent models in the development of new pain therapeutics?
Rodent models have been fundamental to advancing knowledge of pain mechanisms. However, with the general lack of translation of basic research into new therapeutics, questions have been raised about the predictability of rodent research. I don’t believe we should abandon rodent research, but the rodent models – the induced pain state – need to be refined to better reflect human conditions. The outcome measures also need to be refined to better reflect the aversive component of pain that humans suffer from. Classic outcome measures like von Frey and Hargreaves – reflexive assays of sensitivity – can be useful, but only up to a point. Assays that measure the aversive component of pain should be employed more often, coupled with refinement of the induced pain state, so that the model-outcome measure combination better reflects the human condition and problem.
Another facet that needs some attention is “target discovery.” Most often, putative targets are identified using induced models. Arguably, a better approach would be to identify targets using tissue from humans suffering from the naturally occurring painful disease, and then investigate these leads in rodent models. A similar approach could be employed using tissue from naturally occurring painful states in non-human animals – pain conditions that have fidelity with the human condition, such as osteoarthritis in pet dogs. Overall, whatever the approach, there has to be forward and backward translation; you need to keep making sure that what you’re doing in the rodent is relevant.
What advice would you give to general practitioner vets about management of chronic pain in cats and dogs?
There are many chronic pain conditions in cats and dogs, but the most common by far is osteoarthritis. For osteoarthritis, I’d say they should consider the four basic pillars of effective pain management: the basis for chronic pain management is an effective analgesic, which at the moment means nonsteroidal anti-inflammatory drugs, and soon this will also include anti-NGF monoclonal antibody therapies; then I would focus on exercise and weight modulation; and there’s probably a role for omega-3 fatty acids. Only once these four basic pillars have been addressed do you move to the other tiers of pain management like adjunctive drugs such as gabapentin and amantadine, manual therapies, physical therapies, laser, acupuncture, and so on.
In 2017, you held the Pain in Animals Workshop (PAW) with the aim of revolutionizing veterinary chronic pain management. You set a self-imposed timeline of 10 years. How’s it going now?
That was a very successful workshop, and we had a follow-up workshop in 2019. These workshops were focused on discussing how to optimize and improve outcome measures for chronic pain, and acute pain, respectively. Being able to measure the impact of pain on multiple dimensions is fundamental to advancing therapeutic development and pain management. Both workshops were held at the National Institutes of Health in Bethesda, Maryland, which was important because we wanted to have a conversation with both “human” pain researchers and the veterinary community. The idea behind the workshops was to advance measurement of pain in animals for the benefit of veterinary medicine, and to increase the utility of naturally occurring models of pain for human translational pain research.
Since 2017, we’ve had a couple of position papers published, and the ideas from the first two workshops have led to a third workshop being planned for later this year, where specific recommendations for capturing outcome measures and analyzing data will be developed. As we refine the outcome measures, we’re also building a network of centers that can perform clinical pain studies in veterinary species, and do them to a high standard and in a uniform manner.
In order to be taken seriously and to open the door to comparative pain research, we have to get our ducks in a row and elevate our standards. There’s a lot of work to do to answer important questions about measurement of pain, and also to drive forward a thorough understanding of how naturally occurring pain conditions in veterinary species can contribute meaningfully to human translational pain research. We should have that all wrapped up by 2027!
What are some of the hurdles that translational pain medicine faces?
The main hurdle is to better develop the outcome measures. We’ve come a long, long way, but we need to keep working on developing and refining these measures. Second, we need to develop molecular tools, like antibodies and reagents, in order to better utilize the tissue we have access to. Companion animals are euthanized frequently, and veterinarians have unprecedented access to tissue from these animals, as well as, with consent, access to tissue during surgery. If we know the pain phenotype, we can then use that tissue to figure out what’s driving the pain. The same goes for collecting serum, plasma, and other bodily fluids.
But we tend to be stymied by the lack of tools. Antibodies work well in mice and humans because they’re developed for those species, but we need antibodies and other molecular reagents for canine and feline tissue, too. We also need the canine and feline genome to be better annotated. With funding, these tools could be developed and made available on a public platform, and that would really catapult the utility of companion animals as translational models.
You are very interested in sensor technology. How can wearable sensors contribute to improving the health of humans and animals?
Great question – wearable and implantable sensors are going to be an important component of healthcare in the future. I have been most interested in wearable sensors – accelerometers and IMUs [inertial measurement units] – to measure activity in the clinical research setting. Such sensors provide an objective readout of activity, and pain absolutely impacts activity. However, it’s not as simple as just measuring total activity, since pain can both decrease peaks of activity while simultaneously elevating troughs of activity, or compromising restfulness.
Sensors can collect movement data at varying epochs, or intervals, from hundreds of Hertz [frequencies] to data points every hour, and different granularity of data provides different opportunities for understanding the effects of pain. Therefore, we have to think carefully about how to both collect and analyze the data in order to answer the question at hand. At the moment, we are focused on measuring patterns of activity and statistical approaches to appropriately analyze such high-frequency longitudinal data. Overall, the use of activity monitors as objective outcome measures is going to be of great help in using companion animals as translational models.
Another aspect of using activity monitors in companion animals is that if we can encourage pet movement through the use of activity monitors, this will encourage human movement, because pets don’t go out and walk themselves. So getting owners out to walk their dogs for the dogs’ health will also benefit humans.
The biggest hurdle to the routine use of activity monitors in companion animals will be keeping owners interested in the idea – you probably have a Fitbit sitting on your dresser at home. We’ve all bought these gadgets, and we’re obsessed with them for a few day or weeks, but then the battery runs out, and we just stop using them. Maintaining engagement with these sensors in the clinical setting may become easier as sensors move from being wearable to being implanted and self-charging from bodily fluids. We’ve also got to think about data transfer; depending on what information you want, there may be smaller or larger volumes of data to be transferred, and this needs to occur easily and seamlessly. Ultimately, sensors on animals will become a fundamental part of optimizing animal health in the clinic and will contribute to human health because of the engagement of humans with their pets.
I guess it’s a bit harder in cats. You can’t exactly take a cat for a walk.
Yes, we’ve been talking about dogs. But cats decide exactly how much they move so they’re a great readout of the effect of pain on spontaneous movement. If a cat is in pain, it’s going to move less or differently, but the cat is defining that, making cats a great model to look at the effect of pain on movement.
So things like jumping are better measures in cats? I saw you had a paper on that.
Right – pain can affect both how much movement occurs, as well as how and even whether particular activities occur. You’re talking about the paper where we defined a particular signature of acceleration of jumping. And we believe pain will impact that signature of acceleration in a specific, measurable way. The same is true for humans. For example, as you get older, you may have trouble getting out of bed; there’s a signature of acceleration associated with the difficulty in getting out of bed. Wearable sensors can be set up to detect those very specific activities, not only whether they’re performed but how they’re performed. That’s where sensor technology is really going to have an impact in human medicine and clinical research in the future.
At the moment, in people, we rely on self-report, but the field is moving towards utilizing objective measures of the impact of pain, in addition to self-report. In veterinary medicine, we’ve had to be focused on proxy measures, including objective measures, because animals can’t self-report. In some respects, I believe we’re a little bit ahead of the curve there.
Speaking of sensors, this year you co-founded AniV8, which is developing clinical sensor technology. Can you tell me a bit more about that?
AniV8 is a company that was born out of the idea of measuring the quality or smoothness of motion. The idea came about because, after 15 years of trying to use the amount of activity as a surrogate measure of pain, it hasn’t been what I thought it would be – the holy grail of measuring chronic pain in non-verbal species.
I then got the idea that maybe it’s not the total amount of activity, but how that activity is performed, or the quality of that activity. Think about two people in a room, an older person with chronic joint pain and a young, healthy teenager, both sitting down and wanting to get up and get a drink at the other side of the room. The teenager is going to get up smoothly and transition to moving across the room in one fluid motion. The older person with chronic pain will have difficulty getting out of the chair, and the movement across the room will be uneven – a stop/start pattern – as the joint pain inhibits and compromises muscle movement.
That intermittency or smoothness of movement is what AniV8 is measuring. AniV8 was formed around an algorithm that takes IMU data and measures how intermittent or smooth that motion is. It’s agnostic of the size and breed of the animal, its lifestyle, and how much it moves, because it’s simply measuring how movement occurs. It’s not the be-all and end-all, but I believe this approach is probably quite important.
You get a lot of data from these sensors. Is it hard to find what you actually want?
That’s where computing, artificial intelligence, and machine learning come in, because you can apply those techniques to the massive volumes of data that are collected.
One can also use targeted approaches to collect data. Let’s say you think a particular movement is important to measure, say, going up and down stairs. You can set up the sensor so that it wakes up when it detects going up and down stairs and captures that data, and then the sensor turns off when that movement stops.
Such sensor technology can amass huge volumes of data. The problem we are facing at the moment is that we don’t know what those data mean. So in the next few years, you’re going to see a lot of work trying to understand data that is collected very frequently, and longitudinally for days, weeks, and months at a time. We’ve been talking about activity data, but you can collect many other forms of data with sensors. The real power is going to come in combining all of those datasets to predict disease, or measure disease burden. Going back to Fitbit, they recently announced that they could detect COVID-19 two to three days prior to anyone feeling symptoms, with a combination of ECG [electrocardiogram], heart rate, activity, etc. I envision the use of sensors in the future to indicate the onset of chronic pain, and to measure the utility of therapeutics on an individual basis.
What’s your overall vision for AniV8?
To become the company that uses technology to measure pain and the impact of pain, and for this technology to be widely used for the benefit of companion animals and the benefit of translational pain research.
Sherelle Casey is a PhD student at the University of Sydney, Australia.