When a patient begins to experience inflammation, a key driver of many human diseases, the proverbial horse has already left the barn.
“If you’re feeling that you’re inflamed, it probably means the inflammation has already taken off,’” Shana O. Kelley, PhD, an expert in the development of electrochemical sensors, told Inside Precision Medicine. “Ideally, to prevent disease, we want tools to track inflammation when it’s first getting going.”
The most recent developments in measuring inflammation involve the analysis of interstitial fluid, which has not been extensively used in clinical settings due to its difficulty in extraction and the fact that only small amounts can be examined externally. An alternative to interstitial fluid monitoring is the direct, in situ analysis of interstitial analytes using sensor-integrated microneedles or microdevices that can be adjusted to different form factors.
So, a little more than a year ago, Kelley and her research team at Northwestern University were brainstorming ways to track inflammation in real-time when they started looking into wearable glucose sensors.
“We went after developing a sensor that would work like a continuous glucose monitor (CGM), something where you could have a tiny needle you could put right under the skin,” said Kelley. “Instead of detecting glucose, if we could engineer it to detect the protein markers of inflammation, we thought that would give us a really powerful way to watch inflammation in real-time. But those sensors didn’t exist.”
Previously developed wearable or implantable sensors have primarily focused on small molecules, while real-time, continuous in vivo protein monitoring has remained challenging due to slow dissociation kinetics of affinity receptors like antibodies. Existing systems often lack the ability to dynamically track protein level changes in physiological media, leaving the need for real-time protein monitoring in live subjects largely unmet.
To solve the in vivo inflammation monitoring problem, Kelley’s group collaborated with engineers from the new Chan Zuckerberg Initiative (CZI) Chicago Biohub. Today, they’ve published an article in Science revealing the reagent-less, implantable protein sensor technology and proof-of-concept studies in diabetic mice.
The potential for a continuous inflammation monitor is massive when compared to the existing global CGM market, which was valued at approximately $6.32 billion in 2023. In the United States alone, there are an estimated 2.4 million users of CGMs, and approximately the same number of people annually suffer from inflammatory bowel disease (IBD), just one of many inflammatory conditions.
Another way to view the impact that a continuous inflammation monitor could have is to compare it to the number of Americans who take nonsteroidal anti-inflammatory drugs (NSAIDs) regularly—estimates suggest about 15% of the United States population takes an NSAID—along with sporadic users, more than 30 billion doses are taken each year.
Do a little dance
Because cytokine levels tend to rise with inflammation, Kelley’s group chose two cytokines that show more inflammation and used electrochemical sensors, a technology the lab already knew how to use, specific to these cytokines, IL-6 and TNF-α. But the tricky part, according to Kelley, is that when the cytokine protein levels rise, their electrochemical sensors bind the protein, detect it, and form a strong complex. This complex prevents the sensors from responding when the protein levels decrease. Kelley’s group discovered that vibrating the sensors at high frequencies could dissociate the complex, thereby continuously priming the sensors for subsequent measurements.
“We learned to shake the sensors on the surface of the electrode, and if you shake them hard enough, the protein comes off, and then you reset the sensor to be ready for taking another measurement,” said Kelley. “We do shake using a voltage change, and that attracts the sensor to the surface. Then, we reverse the process, oscillating back and forth, and since you’re doing this, the sensor is literally shaken off. So, you can go up, then back down and see trends downward in real-time. That wasn’t possible before.”
Upon demonstrating that these “active-reset” protein sensors worked in vitro using simulated interstitial fluid, Kelley’s group then made a microneedle-shaped housing for the wire-based electrodes that can go through soft tissues, which they implanted into rodents without any toxic effects. Upon injecting lipopolysaccharide (LPS) to trigger inflammation, the implanted microdevice’s active-reset mechanism effectively tracked sharp changes in cytokine levels, demonstrating agreement with ELISA results and highlighting its accuracy.
Time to port over
Having validated the sensor’s ability to monitor inflammatory biomarkers in vivo under various conditions, the next step is to translate the lab-made rodent design to a platform that could be used in clinical settings.
“If you think about the Freestyle Libre 3 or the Dexcom G7 CGM sensors, it’s really a matter of getting our sensor on a platform like that; it is otherwise ready,” said Kelley. “It uses the same kind of electrochemical sampling that the commercial products do. It uses the same electronics that the existing CGMs do. From a technology standpoint, we’re ready to go with the next phase.”
But Kelley said that the next step isn’t so much about retrofitting their tech from rodents to humans in terms of form factor—the physical specifications of computer hardware components; that part is relatively easy compared to addressing regulatory and product development hurdles, which is not trivial and requires capabilities that her group simply does not have or typically deals with.
“Our work ends when product development begins,” said Kelley of her group at Northwestern. “We do technology development. We make sure that it works and is robust. We do proof-of-concept studies where we go into an animal model and show what we can do. I do think it’s time for a handoff into a startup or a partnership with a larger company, whatever makes sense. Then it’s time to say, ‘What’s the platform we’re going to port this over to?’”
Kelley continued, “It could be an existing commercial platform or a completely new CGM-like device. I think we’re at that point because then it’s all about product development, which is about making it reproducible 100% of the time, making it incredibly accurate and precise, and having all the data you need to go to the FDA. That stuff is not what graduate students and postdocs are best equipped to work on. They’re best on the let’s try many different things to get the technology to work.”
Personalizing inflammation treatment
With the ability to examine inflammation in real time at far earlier time points than ever before in the human body, Kelley believes that the device can have a major impact on how inflammation is treated.
“We’re excited because right now, we treat late-stage disease most of the time,” said Kelley. “If you get diagnosed with something, it’s usually pretty far along. This focus on inflammation gives us the potential to go way upstream and look at things that are going wrong with the human body at a early time point where we can turn that around. The impact of that could be very profound. You could take away a huge amount of the disease burden if you could get in there and deal with inflammation, figuring out why it’s happening when it first presents. That’s why we’re excited about working on this problem.”
The real icing on the cake for a continuous inflammation monitoring approach is the potential for an improved understanding of the etiology of inflammation to identify how to treat individual patients best.
“If you could monitor and understand how your diet affects your levels of systemic inflammation, how your environment affects it, and how stress levels affect it, that would give us a window into what’s happening that may negatively affect your physiology,” said Kelley. “At the same time, there are immediate applications for this kind of technology in managing late-stage chronic diseases like heart failure and kidney failure. There are aspects of diabetes that we still need protein biomarkers to inform. We have to be able to monitor inflammation, and then we’ll learn how to prevent it. And for patients already diagnosed with late-stage disease, this gives us a way to keep a handle on them to make those diseases more manageable.”
Apart from extending the testing of this “active-reset” in vivo protein monitoring technique to other protein families, it will be intriguing to observe how it is modified for use in measuring different bodily fluids or in situations where real-time biomarker detection presents challenges, like neurodegeneration markers in cerebrospinal fluid. A widely applicable in vivo protein monitoring diagnostic could lead to a seismic shift in medicine.
