An update stemming from the exclusive collaboration between Boston Scientific and IBM Research yielded insight into how patient-reported outcomes and objective biomarker information, leveraging artificial intelligence, can provide insights around patient pain states, such as activity, alertness, medication, mobility, mood, pain, and sleep.

DocWire News spoke with Dr. Nilesh Patel, Vice President of Medical Affairs, Boston Scientific, about these exciting results. Dr. Patel explained the key role AI plays in understanding chronic care patients, and the importance of a personalized approach in treating patients with chronic pain.

DocWire News: Can you give us some background on yourself, and the company, Boston Scientific?

Dr. Nilesh Patel: Absolutely. It’s a pleasure to be here, Rob. I’m Nilesh Patel. I’m the Chief Medical Officer, or more correctly, the Vice President of Medical Affairs for the Neuromodulation Division. I have fellowship training from Cleveland Clinic where I also served on staff, and 26 years of clinical experience taking care of chronic pain patients before joining Boston two years ago.

Can you give us an overview of the data presented by Boston Scientific at this year’s NANS meeting?

Yes, we had an exciting NANS series of presentations with over 20 presentations combining real world and randomized control trial data. We are most excited about three therapies, fast-acting subperception therapy, about combination therapy, two year data that provided deep and durable relief, and relationship with IBM research that specifically uses artificial intelligence, machine-learning, and deep learning to better understand the patient and to potentially predict the patient’s pain and address the pain preemptively.

Can you talk more about the relationship with IBM?

Yes, this is our sixth year working with the IBM research team. Essentially, we are looking at studying the patient beyond just understanding the pain when the patient visits us. So as a pain physician, I would see the patient, ask the patient when he visits me, “What happens to the pain,” where the pain score is, and generally you only have a single data point. Then perhaps, you see the patient every three months, six months, and so forth. You don’t really know what happens to the patient in their home setting. So using a combination of questionnaires, as well as wearable information, you can understand the patients from multiple perspectives, including alertness, activities of daily living, medication use, mobility as measured objectively, mood, sleep, and pain, and you have multiple data streams that allow you to better understand the patient. Based on that, you can personalize the patient’s therapies going forwards. So as a clinician, I would understand the patient better. If you imagine that there is a fog in terms of what goes on in the patient’s lives outside of the clinic, that fog is now being lifted.

Talk to us about the importance of personalized care for chronic pain patients.

Yes, pain is complex and pain changes over time because of neuroplasticity. So no two patients are the same. Each patient needs a customized therapy and a therapy that is going to change and adapt as a patient’s pain changes. Having combination therapy as we do, we have been able to validate that the patients do need multiple options customized to their need and their time. Our two-year combination therapy data, which was the randomized control trial data that we presented at NANS shows that 85% of these patients are responders, so they’ll get at least a 50% pain relief, that there was a marked improvement in function with the measure of Oswestry Disability score, changing 25 points, and that the patients are satisfied as measured by PGIC, which stands for patient global impression and change, where 85% of the patients said, “I’m better,” or, “Very much better,” at two years.

This is corroborated by our publish work from last year with two years of real world data in consecutive patients. We have 420 patients in the real world trial, but we also need randomized control trial data. So your two years of robust randomized control trial data, and your two years of real world data that corroborates that combination therapy is superior to monotherapy. What goes on in the combination therapy, traditional monotherapy fails to provide long-term relief. So the traditional monotherapy, for example, is a single therapy. So if you combine that with the subperception therapy, combination therapy is combining a [tonics 00:04:51] therapy with a subperception therapy, then the outcomes will be much better or that’s the expectation. Now, we’ve been able to demonstrate indeed in real world and our CT, the outcomes are much better for these patients.

How big a role does artificial intelligence play in analyzing and understanding chronic pain?

So as a pain physician, if a patient comes to me, all I have is an understanding of what the patient tells me at that visit. I don’t really have an understanding of what goes into the patient’s life. The fact that you can collect data through questionnaires, as well as through wearables, and have both subjective and objective input allows you to put all of the data, and because of artificial intelligence, you can come with specific states and understand that particular patient better. Down the road, you can then predict the patient’s pain based on artificial intelligence collected states, and potentially have the patient dwell in a better state compared to where they have been. So this is collecting data on mass, but personalizing it to the patient.

Any closing thoughts?

Yes. We also have a new therapy called fast-acting subperception therapy. If you look at traditional subperception therapies, it takes between one in three days to wash in and to optimize. But with fast-acting subperception therapy, the patient can have relief within minutes. So before they leave a clinic, they will know that fast-acting subperception therapy is going to work. Our collaboration with Duke’s renowned neuroscientist, Professor Warren Grill, allowed to understand how this works. So the clinical data, as well as the bench signs clinical data to show that this therapy works and the bench signs to show how the therapy works is something that’s very exciting.