Sensor Bio-fouling and Biocompatibility

Sensor Bio-fouling and Biocompatibility in Continuous [Glucose Monitoring](/)

A CGM sensor under the skin gets coated with proteins and surrounded by immune cells, and both can slow the glucose that reaches its sensing chemistry. Researchers have called this biofouling effect a serious obstacle to implanted sensors meant to last a long time [1]. Here is what we learned about how it happens, which materials engineers have tried, and why wear time is tied so closely to the tissue around a sensor.

Two terms, two questions

Bio-fouling is the buildup of biological material, first proteins and then cells, on a sensor's surface. Biocompatibility is the broader question of whether a material and the body's tissues can coexist. A 2008 review of more than 50 years of long-term implants concluded that, in most cases, it has meant one thing: the material does no harm to the surrounding tissue, achieved by being chemically and biologically inert [2]. A glucose sensor needs more. It also has to keep glucose flowing to its sensing layer for as long as it is worn.

What happens after a sensor goes in

Implanting any device sets off an inflammatory and wound-healing response that ends in the foreign body reaction. Macrophages, a type of immune cell, gather on the surface, and some fuse into large foreign body giant cells [3]. The first step is protein adsorption, meaning proteins from blood and tissue fluid sticking to the material. A surface's chemistry and texture shape the reaction most strongly in the first two to four weeks, though the reaction lasts as long as the device stays in [3].

Proteins arrive fast. Researchers implanted working needle-type glucose sensors in rats for 24 hours and then identified what had stuck to them: mainly serum albumin, the most plentiful blood protein, largely as fragments [4]. They concluded that such fragments work their way into the sensor membrane and slow glucose diffusion, lowering sensitivity [4].

Then comes the capsule. Over time the reaction builds a dense collagen layer, called a fibrous capsule, that blocks mass transport, meaning the movement of molecules such as glucose, between an implant and the body [5].

Fouling or tissue: which slows glucose more?

Here's what surprised us. A Duke University study in rats tracked glucose moving through three membrane materials over 8 days and separated the protein film from the surrounding tissue. The film stuck to the probe accounted for only about 12% to 24% of the total resistance to glucose transport; the tissue accounted for much more [6]. The authors suggested future materials work should focus on the tissue at the implant site, not only on the sensor surface [6]. Our article on biofouling and foreign body response mitigation covers that tissue side.

Membrane strategies

Most designs put a protective membrane between the sensing chemistry and the body. A 2000 review sorted the approaches into nine groups, among them hydrogels, Nafion, surfactants, naturally derived materials and surface texture [1].

Hydrogels

A hydrogel is a water-swollen polymer network, and hydrogels head that review's list [1]. Zwitterionic hydrogels carry both positive and negative charges along their chains. In mice, an ultra-low-fouling zwitterionic hydrogel resisted capsule formation for at least 3 months under the skin, and the tissue around it grew new blood vessels [5]. That was hydrogel samples in animals, not a glucose sensor in people. Our article on zwitterionic polymer coatings follows that research further.

Nafion

Nafion is a perfluorinated ionomer, a fluorine-based polymer with charged groups. In a 1994 study, a heat-cured outer coating of DuPont's Nafion blocked interference from ascorbic acid and urea, and several needle-type sensors kept working for at least 10 days under the skin of dogs without losing sensitivity [7]. The weakness showed up later: Nafion membranes implanted in rats for 4 weeks cracked extensively, apparently from calcium phosphate deposits, a kind of mineral buildup [8].

Textured membranes

Another approach changes the surface's shape instead of its chemistry. Dexcom's US patent 7,192,450 B2, granted in 2007, describes a two-layer membrane: an outer layer meant to let tissue grow in, interrupt the contracting forces of the foreign body response and encourage blood vessels, and an inner layer that keeps cells out but lets glucose pass [9]. A patent shows what was claimed, not whether any sensor sold today uses that design.

Drug release: calming the reaction at its source

Dexamethasone is a corticosteroid, a steroid that suppresses inflammation. In a mouse model where untreated sensors usually stopped working within 24 hours, daily dexamethasone injections extended sensor life to as long as 2 weeks, and the tissue around treated sensors showed almost no inflammatory cells [10].

Releasing the drug from the coating itself is more practical, but it has a catch. In rats, dexamethasone-releasing hydrogel coatings reduced inflammation at 2 weeks, yet by 6 weeks they looked no different from plain coatings; the effect faded once the drug ran out [11].

The Eversense 365 brings the idea into a cleared product. A healthcare provider inserts this fluorescence-based sensor under the skin, and it is indicated to measure glucose for up to 1 year in adults 18 and older [12]. FDA's review describes a silicone collar on the sensor that releases 1.75 mg of dexamethasone acetate locally to minimize the foreign body reaction, and it compares the 365 with an earlier Eversense system rated for up to 6 months [12]. The system is contraindicated in people for whom dexamethasone or dexamethasone acetate may be contraindicated [12], and whether that applies to someone is a question for their care team. Our article on implantable fluorescence sensors explains the optics.

Wear time and the tissue clock

Mechanics matter too. A 2011 review argued that motion and pressure at the sensor-tissue interface, often tiny micromotions, create stresses that change the surrounding tissue [13]. Brief forces can disturb the signal for a while, while lasting forces can alter the foreign body response itself [13].

Sensors worn on the skin run on a much shorter clock. As of October 2026, FDA's review of the Dexcom G7 15 Day lists a wear period of up to 15 days, against up to 10 days for the Dexcom G7, each with a 12-hour grace period [14]. Set beside the two-to-four-week window in which surface properties most strongly shape the reaction [3], a 10- or 15-day sensor spends its whole life inside that early stretch. A sensor meant to stay in for a year has to work through the full reaction, which is the job Eversense gives its steroid collar [12].

Not every short-wear sensor finishes, either. In the G7 15 Day pivotal study, 73.9% of 131 sensors lasted the full 15 days, and 19.1% were shut off early by the sensor's own self-checks [14]. The review doesn't say how many early endings came from tissue response, so we don't read that figure as a biofouling rate.

How regulators judge biocompatibility

FDA's final guidance on ISO 10993-1, the international standard for biological evaluation of medical devices, was issued in September 2023 [15]. It helps manufacturers determine whether a device's materials could cause an unacceptable adverse biological response, with a risk-based approach to deciding whether testing is needed [15]. The Eversense 365 review, for example, lists parts of the ISO 10993 series among its referenced standards, including tests for cytotoxicity, local effects after implantation and skin sensitization [12]. That covers the "do no harm" half of biocompatibility. Whether a sensor keeps reading accurately as tissue responds shows up instead in accuracy and sensor-life data like the G7 15 Day figures [14].

What is still unanswered

The best evidence that a coating can prevent a capsule still comes from mice [5], and the drug-release studies show the benefit lasting about as long as the drug [11]. Eversense's one-year label shows a slow-release steroid can carry an implanted sensor far past a few weeks [12]. How long a sensor could last on a coating alone is a question the studies we read don't yet answer.