June 17, 2025 | Modern jet engines can generate a terabyte or more of data per hour that gets transmitted back to the engine manufacturer to ensure aircraft safety and prevent disasters from happening. Yet the human body, arguably the most important machinery of them all, gets checked only once a year or whenever something has already gone wrong, points out Tom Soh, Ph.D., professor of electrical engineering, bioengineering, and radiology at Stanford University.
Soh has spent the past 15 years pondering the absurdity of “reactive medicine,” where health issues get treated only after symptoms appear and patients seek care. “That model is not sustainable economically or for the quality of life that we wish for ourselves,” he says, “so what we want is a technology that can pick up problems super-early” to make diseases like cancer much easier to treat and help prevent the spread of infections.
The latest advance in this quest to flip the narrative came with a demonstration by Soh and his team led by Yihang Chen, a Ph.D. student in his lab. They created a modular biosensor for the real-time monitoring of molecules in flowing whole blood that operates stably for a record seven days (Nature Biomedical Engineering, DOI: 10.1038/s41551-025-01389-6). “Researchers including our group have figured out how to make molecular switches for many things, but what was really plaguing advancement in the field was [how to achieve] this prolonged operation in vivo.”
The technology—known as Stable Electrochemical Nanostructured Sensor for Blood In situ Tracking (SENSBIT)—can be thought of as a small microfabricated soft needle with nanostructured electrodes with many “nooks and crannies that are 10s of nanometers” says Soh. These can be placed pretty much anywhere, including into a vein or under the skin.
SENSBIT was used in the newly published study to track drug concentration profiles of kanamycin, an aminoglycoside antibiotic easily detected because it is not naturally produced by the body. Remarkably, the biosensor retained over 70% of its signal after one month in human serum and over 60% after a week implanted in the blood vessels of live rats. The previous limit for intravenous exposure for this type of device was around 11 hours.
The imagined clinical potential of SENSBIT goes well beyond measuring drugs, he says. It could, for example, be used for measuring important metabolites, hormones, and potentially cytokines in the future.
The breakthrough idea that has been driving the work in Soh’s lab was a molecular switch (sensor element) that can reversibly bind to the target to enable continuous molecular measurements—be they drugs or other biomarkers—using a synthetic antibody (an aptamer) that binds to that target and changes its shape, as first described in a 2009 paper (Journal of the American Chemical Society, DOI: 10.1021/ja806531z). Here, a microfluidic system was used to detect cocaine at low micromolar concentrations in blood serum.
The most prominent example of existing technology for continuous measurement of biomarkers in situ is the continuous glucose monitor (CGM), he notes. That achievement took about 40 years and $40 billion and was made possible by a naturally occurring enzyme called glucose oxidase that conveniently converts glucose into a measurable electrochemical signal.
“Without that enzyme available in nature, CGMs would not exist today,” Soh explains. The limitation with other important biomarkers is that they don’t have such enzymes, so his group set out to engineer synthetic antibodies that would bind to specific antigens and change their conformation.
The latest work addresses the second big problem of operational stability. Even CGMs last only about two weeks before they must be switched out for a new device, says Soh, because “our bodies don’t like foreign objects.” If immune cells such as macrophages can’t breakdown the foreign material, it will create a scar around the object to wall it off from the rest of the body.
That the measurement is happening in whole blood flowing through an artery for a week is the latest key advance. CGMs, in contrast to SENSBIT, go into the interstitial space (typically of the upper arm) that is filled with interstitial fluid, “which is a lot less messy than whole blood,” Soh says.
Although work continues to push the operation of SENSBIT beyond seven days, a week is generally long enough for most clinical circumstances, he adds. It’s also “longer by an order of magnitude compared to what was possible before.”
The nanostructures of SENSBIT have a nanoporous gold surface protecting it from the harsh environment of whole blood, says Soh. This protective coating is “bioinspired” by the microvilli lining the gut wall, biostructures that help shield the intestinal epithelium from degradation by bacterial cells and the passage of food. Microvilli have a glycocalyx on their surface that also acts as a barrier against harmful substances in the digestive tract.
Importantly, nanostructures serve as a “size exclusion filter” to block out large blood components like red blood cells and platelets, as well as big sticky proteins that can “gunk up the sensor,” he explains. The inert polymer selected to decorate the surface was hyperbranched PEG because it best mimicked what the gut does with glycocalyx but to detect chemical biomarkers in a harsh environment for a long period of time.
SENSBIT could eventually serve many different purposes, says Soh. “On the research side, we can only measure a handful of molecules continuously.” Opportunities abound in areas ranging from “basic cell biology to neuroscience to immunology.”
In the clinical space, his team hopes to start first-in-human trials using SENSBIT in the next 12 to 18 months. “Right now, we are trying to get IRB [institutional review board] approval to show that our sensor poses non-significant risk to healthy volunteers.”
In the future, a holy grail would be to create a biosensor that can continuously measure very low abundance biomarkers such as insulin. This is a very difficult problem because the concentration of insulin (~100 picomolar, is 50 million times lower in abundance than glucose (5 millimolar). “If we can measure insulin, then we could create a true artificial pancreas and solve diabetes.”
Another exciting possibility is “listening-in to the communication among immune cells” says Soh. When our body is infected by microorganisms, our immune cells communicate with each other through cytokines and other communication molecules to mount a “coordinated defense” that is different depending on whether it is a viral, bacterial, or fungal infection. “If we can listen into the conversation among our immune cells, then we can know what’s going on before any symptoms manifest themselves.”
This idea of listening to our immune system could potentially extend to many other diseases, including autoimmune disorders and cancer, says Soh. “When we have the technology to do that, then we can truly flip the script in healthcare from fixing problems after symptoms arise to keeping people healthy by identifying problems before disasters happen. That is a truly exciting vision of the future that we want to bring to reality.”