October 15, 2019 | About half of patients don't take their medication as prescribed. It's a serious problem in healthcare, and one that Jeremy Frank, PhD believes we've been approaching all wrong.

It's incorrect to assume that poor adherence is the patient's fault, he argues. "Accepting a new diagnosis, struggling against a terminal disease, or managing the polypharmacy and side effects associated with your drugs, all while trying to live your life, will challenge anyone to maintain any kind of regimen!" Frank says.

Frank is Senior Vice President of Digital Medicine at Proteus Digital Health. He believes that medications coupled with tiny ingestible sensors and data analysis can powerfully bridge the gap between patients' lives and their healthcare needs.

"We take a very user-centered design approach to developing digital medicines," he explains. "By adding measurement, feedback and behavioral cues, we can deliver a significantly improved therapeutic experience that fits into a patient's life and fits into how providers are practicing medicine."

On behalf of Clinical Research News Ben Lakin, Cambridge Innovation Institute, recently spoke with Frank about Proteus, his role there, the company's key technology, and the field of digital medicines.

Editor's Note: Frank is speaking at the Sensors for Medical Device and Implantable Applications track at the upcoming Sensors Summit in San Diego, Calif., December 10-12. His conversation with Lakin has been edited for length and clarity.

Clinical Research News: Jeremy, would you mind please giving a quick overview of Proteus Digital Health and your role at the company?

Jeremy Frank: Proteus Digital Health has created digital medicines, which is a new category of pharmaceuticals that measures medication treatment effectiveness, which helps physicians improve clinical outcomes and patients reach their health goals. This is based on a technology that we invented—this ingestible sensor—that was approved by the FDA in 2012. The first digital medicine incorporating that sensor was Otsuka's ABILIFY MYCITE, which was approved by the FDA in 2017. And we have a pipeline of additional digital medicines in development in beachhead therapeutic areas, including oncology, infectious diseases and mental health.

These digital medicines include drugs that communicate when they've been taken and involve wearable sensors that capture that a medication has been taken, and a host of other physiologic metrics and responses to that drug. Finally, mobile and web-based applications that support the patients' self-care and physician decision making and data analytics that serve the needs of patients, doctors, health systems, payers.

I lead our digital medicine organization. Our organization is responsible for the creation of the development methods and high-volume manufacturing processes that allow us to make any drug digital: tablet or capsule. We've built these capabilities to serve our own digital medicine programs as well as our pharmaceutical partners in digitizing their portfolios of drugs.

Excellent. It sounds like you are doing some very exciting work at Proteus. I know that prescription medicine has been a challenge in healthcare for a while. You talked a little bit about this, but can you expand on what technology innovations in particular allowed Proteus to develop this new approach to addressing the problem of medication adherence?

The biggest innovation has to be the now-ubiquitous presence of smartphones; patients now have a computer in their pocket that also connects them to the Internet. That connectivity allows our product to essentially close the loop between patient behavior and what physicians and their caregivers see or can understand.

Beyond that, there are a host of advances in the integrated-circuit industry. The integrated circuits are in both the ingestible sensor as well as the wearable sensor. There's been broad innovation on the wearable side, driven by the explosion of wearable devices, both medical grade and consumer, that exist today.

And then over the course of the company, the skin-adhesive technology has advanced dramatically, largely again driven by the overall wearables industry. And so we're able to take advantage of that as well.

I can imagine that, given the nature of an ingestible sensor, it must have to meet many rigorous design requirements to be safe and effective. What would you say were some of the greatest engineering or technical challenges that had to be overcome to develop such a small ingestible sensor?

That's a great question. From the beginning of this product development journey, we knew that the ingestible sensor had to meet three main criteria. First, the sensor had to be safe. Secondly, the data had to be secure. Finally, the sensor had to be sensitive enough to allow a physician to make decisions based on accurate data over the course of a patient's treatment. So, safe, secure, sensitive.

When you think about safety, you think about two main criteria. One is operating at a very low voltage because we certainly can't disturb a patient's body in any way, and also it has to be biocompatible. The ingestible sensor is powered in the same way as, a potato battery. It has bio-galvanic materials deposited on the integrated circuit that briefly turn it on when it hits the patient's stomach fluids.

Those bio-galvanic compounds were essentially selected based on the label of a Centrum multivitamin, so we knew that we could only pick materials that were already considered essential daily minerals in the human diet. It was a very constrained material set. We identified two compounds, copper and magnesium, that created just enough power to power the integrated circuit for only a matter of minutes.

To keep the patient's data secure, we knew that the communication scheme between the medication and the wearable worn on the patient's torso had to be contained within the patient's body. And so this in-body communication scheme required certain other design decisions be made so that the signal was not broadcast or radiated beyond the patient's body. You can only detect the existence of the medication if you're touching or if you're in close contact with the patient's torso (like an ECG electrode). So from a security standpoint, you cannot pick it up the way you might pick up, say, an RFID-based technology.

And then of course it has to be sensitive. It has to be accurate enough to give the physician confidence that they understand a certain patient's adherence, and then they can make treatment decisions based on those data. We invested significantly on the sensitivity of the detector. That's the wearable device on the patient's torso. Our clinical data shows that we can detect over 97% of all ingestions taken, and in practice we see even higher detection rates.

Was that sensitivity requirement necessary for FDA approval? Or, was that an internal standard that you set prior to wanting to go forward with the product?

We take a user-centered design approach to everything that we do. Physicians need to have confidence in our technology and therefore expect a high degree of accuracy. In early studies and early discussions with physicians, rates above 95% of all ingestions detected was deemed an appropriate benchmark from users. And then we just continued to work on iterating further, making  10ths of a percentage point improvements on that detection accuracy over time.

I imagine, in addition to the technical or engineering requirements and challenges to developing this sort of a product, there must also be some other challenges: maybe more people-oriented challenges. What challenges around training, physician buy-in, or payer approval had to be overcome prior to the product launch?

These are the challenges that we are confronting now as we commercialize in this space. Healthcare is a really risk-averse industry, and it isn't always quick to embrace new technology or new products. But there are creative payers and providers out there that are open-minded and are looking for better ways to treat their patients. Identifying those, both on the payer and provider side, has been critical to getting these early commercial proof points demonstrated.

We have the ability to leverage some pretty elegant technology for both automation and digitization. And we have now the clinical data that shows that digital medicine can be very effective in delivering significantly improved outcomes with patients that use them. We have focused on creating the data that physicians and payers want to see, and then making sure that we identify the right early partners as we commercialize.

Having those two key approvals, for lack of a better term—the physicians and the payers—that would definitely be critical.

Yes. And they often want different things. And so we have to make sure that we meet the needs of all our stakeholders.

How receptive have patients and physicians been to incorporating this kind of technology into their lives?

Well, it's pretty well known now that 50% of patients don't take their medication as prescribed. And as a result, there's an aching need for providers, caregivers, and payers essentially to understand what is the ground truth about a particular patient's adherence to their medication.

What's incorrect is to assume that those 50% of patients and their poor adherence is somehow their fault. Accepting a new diagnosis, struggling against a terminal disease, or managing the complexity of polypharmacy and side effects associated with your drugs, all while trying to live your life, will challenge anyone to maintain any kind of regimen.

You have to embrace that reality. And when you do, you actually realize that we don't have defective patients. Instead, we have defective products, which don't meet the needs of patients. We take a very user-centered design approach to developing digital medicines. By adding measurement feedback and behavioral cues, we can deliver a significantly improved therapeutic experience that fits into a patient's life and fits into how providers are practicing medicine.

As a result, we've been getting very high levels of satisfaction from patients and providers that use our products. Specifically, we receive net promoter scores that are equal or better to the beloved consumer electronics brands that are known worldwide for delighting users.

Thinking a little bit into the future, what do you consider some other promising applications for ingestible sensors?

Right now all our ingestible sensors do is mark the time of an ingestion of a medication, but it's very easy to add other functionality to the ingestible sensor itself. Think patient stomach pH or stomach contents, perhaps even GI transit time. So, there's a host of additional  capabilities, or functionalities, that you can build into the ingestible sensor given a specific product application.

Beyond the sensor itself, the areas of fitness and wellness are also very attractive to us. More and more focus is being given to the recovery period in between training sessions. And this isn't just for elite athletes. The ingestion of supplements can help athletes avoid injury is a pretty interesting market in itself. We have this product prototype sitting on the shelves waiting for the right time for us to be able to focus on it.

Are you thinking hydration monitoring or more sophisticated than that?

It starts with hydration and then goes beyond that.

As an industry scientist, I'm sure you're always keeping your eye on the latest technology or innovations that are out there. What new sensor technologies or innovations are on your radar? Are there any really exciting topics in the field of medical sensors right now that have your attention?

Definitely. The smart-textile space has finally moved beyond just gluing wires into T shirts, and there's really great innovation going on by a few different companies in this space. When you think about developing truly functional clothing, I think you're going to see a step change in the availability and the type of data that's collected from consumer and medical device applications. That's a space that we watch quite closely.

The compliment to that is all the innovation in the power space. All of these medical devices require electricity to work, and there's been advancement in the passive or wireless charging space that will start to reduce a patient's need to either replace batteries or charge batteries. Back to that point of fitting into a patient's life, the more you can do that, the more successful you'll be. Having things that charge themselves effectively, or power themselves, is an important space that we watch and we've seen some great advancements there as well.

Would powering themselves be through the locomotion of a person walking around? Or would it be harvesting some other kinetic energy in the body?

It's a broad range. Kinetic-energy harvesting is an interesting space. It seems a bit early and still academic, given the power requirements for what we believe most wearable devices have. But certainly RF-based charging, or wireless charging, is something that's quite a bit more near-term. You can already inductively charge your phone now.  Imagine how that can expand over time and influence the broader wearables space.