1. What an implantable sensor is, in three points

An implantable sensor is a sensor placed entirely beneath the skin that goes on measuring a substance in the body for months (our commentary). This article is limited to a confirmable example in FDA public documents. The approved product we could confirm is a fluorescence-based implantable CGM (Eversense). How CGM works in general is covered in our explainer on continuous glucose monitoring.

• How it differs: a transcutaneous sensor is a thread inserted through the skin and replaced every 10 to 15 days. The Eversense sensor is inserted under the skin by a healthcare provider and left in place. The labelled wear period was up to 90 days at first approval and was later extended to up to 180 daysSourced
• How it measures: not with an enzyme and electrode but by fluorescence. The FDA documents say the sensor detects glucose in interstitial fluid using fluorescence and measures every 5 minutesSourced
• Is there a battery: the FDA documents describe the implanted sensor as powered by radiofrequency (RF) from the transmitterSourced. It is a design that avoids a second operation to change a battery (our commentary)

2. A confirmable example: the fluorescence-based implantable CGM — structure, power and communication

Eversense is a fluorescence-based implantable CGM approved by the FDA through PMA (premarket approval). The summary of safety and effectiveness (SSED) describes the system as four componentsSourced [Reference 1].

View the original SVG schematic, source notes and numerical details

All Sourced (PMA P160048 SSED [Reference 1], description at first approval; reportable range 40 to 400 mg/dL).

On power and communication, the SSED's test tables give the following specificationsSourced [Reference 1]. The sensor antenna operates at 13.56 MHz; NFC (near-field communication) reading must work at a distance of up to 12 mm; the sensor's electronics must communicate using the ISO 15693 protocol and be able to excite the fluorescent glucose indicator and detect the fluorescent light it emits; and the transmitter must communicate with the smartphone over BLE within 10 m. In effect, the sensor in the body works like a small wireless tag that draws power from outside and operates only briefly (our commentary).

3. A materials engineer's view (1): what is inside a fluorescence sensor

4. Foreign body reaction and capsule formation: what the body does

Anderson and colleagues call the body's response to a material placed in it the foreign body reaction, and describe it as the end-stage response of inflammation and wound healing, made up of macrophages and foreign body giant cellsSourced [Reference 5]. Ward's review lays out the sequence of the response to a subcutaneous implant from the viewpoint of sensorsSourced [Reference 6].

• The first weeks: Anderson and colleagues state that the period in which a material's surface properties modulate the reaction is the first two to four weeks after implantation, but that the foreign body reaction at the tissue–material interface is present for the whole life of the deviceSourced [Reference 5].
• The capsule: Ward describes how, in response to macrophage-derived signals, fibroblasts become myofibroblasts, and collagen and other extracellular matrix form a dense, hypocellular fibrous capsule. The capsule can become impermeable or hypopermeable to many compoundsSourced [Reference 6].
• Can it be avoided: the same review states that thousands of biomaterials have been tested, but that, as of writing, it had not been possible to create a stealth material that evades the mammalian ability to distinguish foreign from selfSourced [Reference 6].
• The role of an anti-inflammatory drug (a research example): Klueh and colleagues implanted glucose sensors subcutaneously in mice and reported that daily intraperitoneal dexamethasone extended function to as long as 2 weeks, whereas control sensors usually lost function within the first 24 hours. The result indicates an important role for inflammation in loss of sensor functionSourced [Reference 7]. This is a short-term study in mice and does not show effects in humans or the performance of any product.

The Eversense silicone collar is designed to elute dexamethasone acetate locally. The SSED says that risks from local and systemic exposure are “not expected to occur”, while noting that the clinical studies did not include subjects taking dexamethasone or other glucocorticoid medicationsSourced [Reference 1]. The labelling also contraindicates use of the system in people for whom dexamethasone or dexamethasone acetate may be contraindicatedSourced [Reference 1].

5. Lifetime: from 90 to 180 days, and what changed

The value of an implantable is that it needs replacing less often. So how long it keeps working is at the centre of the design.

All Sourced (FDA PMA database and SSEDs [References 1 to 3, 10]). Further supplements to the same PMA have followed since 2022; their content was not checked for this article.

The study that supported the extension to 180 days is PROMISE. The FDA summary and a peer-reviewed paper state the followingSourced [References 3, 4].

• Design: sensors were inserted in 181 adults at 8 US sites (85 with one sensor, 96 with one in each arm), from December 2018 to May 2020. 170 (94%) completed day 180. The reference was plasma from centrifuged blood measured on a YSI 2300 bedside glucose analyser.
• Accuracy: the overall MARD of the original sensor was 9.1% (49,613 pairs, 40 to 400 mg/dL), with 85.6% within 15%/15 mg/dL and 92.9% within 20%/20 mg/dL. Accuracy was lowest on day 1; MARD by visit was 11.0% on day 1, 7.7% on day 60 and 10.4% on day 180.
• Sensor survival: the Kaplan–Meier estimate of the share of original sensors still functioning was 98% at day 90, 90% at day 120, 74% at day 150 and 65% at day 180.
• The modified sensor (E3): 43 participants wore the modified sensor as a second sensor. MARD through 180 days was 8.5% (12,034 pairs), and survival was 96% from day 30 to 120, 94% at day 150 and 90% at day 180.

As residual uncertainty, the SSED records that the E3 sensor was evaluated in a relatively small group of 43 subjects, that the effect of repeated insertion and removal over a long 180-day wear has not been evaluated, and that the rate of adverse events in non-adjunctive use was not evaluated because the E3 sensor's output was blinded in the studySourced [Reference 3].

6. Our calculation: the size of the sensor, and how to read the lifetime data

7. Surgical procedure: what happened on insertion and removal

For an implantable, the procedure is part of the product. Here are the procedure-related facts from the FDA summariesSourced [References 1, 3].

All Sourced (PMA P160048 SSED [Reference 1] and P160048/S016 SSED [Reference 3]). These events are observations in the studies described in the review documents and are not a basis for judging the safety or suitability of any individual procedure.

8. Biocompatibility, sterilisation and storage: what the review document looks at

• Biocompatibility: for the sensor, tests were done for cytotoxicity, sensitisation, intracutaneous reactivity, acute systemic toxicity, pyrogenicity, 4-, 13- and 26-week subcutaneous implantation in rats, genotoxicity (three tests), chemical characterisation (ISO 10993-17 and -18) and particulates (ISO 14708-1), all reported as passingSourced [Reference 1].
• Animal study: the sensor with its steroid-eluting collar and an approved pacing lead eluting the same drug from silicone were implanted subcutaneously in rats, and tissue was examined at 30 and 90 days; no adverse tissue reactions were seen with eitherSourced [Reference 1].
• Sterilisation: the sensor is sterilised in its holder with ethylene oxide to a sterility assurance level (SAL) of 10⁻⁶. Residuals of ethylene oxide and ethylene chlorohydrin are reported to meet the limits of ISO 10993-7Sourced [Reference 1].
• Storage: the sensor at original approval was to be stored at 2 to 8 °C with a labelled expiry of one month, with shelf-life studies ongoingSourced [Reference 1]. The current storage conditions were not checked for this article.

Here, too, there is a difference from the previous article (continuous glucose monitoring). Some transcutaneous disposable sensors are electron-beam sterilised (Libre). This implantable sensor, by contrast, is reported to be sterilised with ethylene oxide. Different sterilisation methods place different demands on the tolerance of the hydrogel and the indicator (our commentary).

9. Interfering substances and restrictions on use

Even with fluorescence, substances other than glucose can shift readings. The paper explains the mechanism as followsSourced [Reference 4]. Signal is raised by substances that bind to the indicator hydrogel or that emit light within the sensor's operating spectral range, and reduced by substances that absorb the excitation light or absorb or quench the fluorescence. What follows reports what the review documents and the paper state; it is not a basis for any individual's decision about use.

All Sourced (FDA SSEDs and a peer-reviewed paper). The paper states that the interference from vitamin C and acetaminophen seen with some other systems is not seen with this sensor, but this article has not confirmed that point in a review document. Labelling can change with version and region, so any actual decision rests on the latest instructions for use and a healthcare professional.

Physical conditions were also evaluated. In the study before first approval (PRECISE II), accuracy visits included upper-arm exercise sessions and 30-minute compression sessionsSourced [Reference 1]. They reflect the environment of use, in which a sensor under the skin is subject to external forces (our commentary).

10. FDA classification: implantables go through PMA (approval)

Most transcutaneous CGM, in which a thread-like sensor is inserted from outside the body, is treated as Class II (iCGM) through De Novo or 510(k) (see the CGM explainer). An implantable, by contrast, is classified in the FDA product classification database as Class III and is subject to PMA (premarket approval)Sourced [References 10, 11, 12].

All Sourced (FDA product classification database [References 11, 12]; both are device class 3, with no individual 21 CFR regulation number assigned).

FDA PMA is an “approval”, a separate procedure from a 510(k) “clearance” or a De Novo “grant”. The original approval of Eversense (P160048) was on 21 June 2018, the extension to non-adjunctive use (S006) on 6 June 2019 and the 180-day E3 (S016) on 10 February 2022Sourced [References 1 to 3, 10]. Each approval is a judgment for a specific intended use; it does not mean the device is safe and accurate in any use.

11. A materials engineer's view (2): the price of staying a long time

12. Open problems, and what this article could not confirm

(1) Implantable sensors other than Eversense

The approved implantable glucose sensor this article could confirm in official documents was Eversense. The approval status of other products, or of implantable sensors that measure something other than glucose, was not confirmed and is not covered.

(2) The specific composition of the indicator, hydrogel and housing materials

The SSED does not specify the chemical structure of the fluorescent indicator, the composition of the hydrogel or the material of the housing. What this article could write is limited to boronic acid, VPBA and the cylindrical dimensions from the paper, and the silicone collar and the epoxy joining of the end cap from the SSED.

(3) Wear beyond 180 days, and supplements since 2022

The FDA PMA database shows further supplements to the same PMA after the E3 approval (2022). This article checked only the SSEDs for the original approval, the extension to non-adjunctive use (S006) and the 180-day E3 (S016), and did not check the content of later changes, including any further extension of wear.

(4) Histological data in people, such as capsule thickness

No primary source in the documents checked shows capsule thickness, tissue structure or the state around the sensor in people. The account of foreign body reaction rests on reviews [References 5, 6] and animal studies [References 7, 8] and is general.

13. Glossary

14. References (primary sources)

1. U.S. Food and Drug Administration (FDA) “PMA P160048 FDA Summary of Safety and Effectiveness Data (Eversense Continuous Glucose Monitoring System)”, approved 21 June 2018 — www.accessdata.fda.gov
2. FDA “PMA P160048/S006 FDA Summary of Safety and Effectiveness Data (extension to non-adjunctive use)”, approved 6 June 2019 — www.accessdata.fda.gov
3. FDA “PMA P160048/S016 FDA Summary of Safety and Effectiveness Data (Eversense E3 CGM System, up to 180 days)”, approved 10 February 2022 — www.accessdata.fda.gov
4. Garg SK et al. “Evaluation of Accuracy and Safety of the Next-Generation Up to 180-Day Long-Term Implantable Eversense Continuous Glucose Monitoring System: The PROMISE Study”, Diabetes Technol Ther 24:84 (2022) (PMC) — pmc.ncbi.nlm.nih.gov
5. Anderson JM, Rodriguez A, Chang DT. “Foreign body reaction to biomaterials”, Semin Immunol 20:86 (2008) — doi.org
6. Ward WK. “A review of the foreign-body response to subcutaneously-implanted devices: the role of macrophages and cytokines in biofouling and fibrosis”, J Diabetes Sci Technol 2:768 (2008) — doi.org
7. Klueh U et al. “Inflammation and glucose sensors: use of dexamethasone to extend glucose sensor function and life span in vivo”, J Diabetes Sci Technol 1:496 (2007) — doi.org
8. Gough DA et al. “Function of an implanted tissue glucose sensor for more than 1 year in animals”, Sci Transl Med 2:42ra53 (2010) — doi.org
9. Hatada M, Okuda-Shimazaki J, Sode K. “Glucose Oxidoreductases for Glucose Monitoring: A Bird's Eye View of Glucose-Sensing Enzymes II”, J Diabetes Sci Technol (2026) (PMC) — pmc.ncbi.nlm.nih.gov
10. FDA PMA database record for P160048 (Senseonics, Incorporated; Eversense) — www.accessdata.fda.gov
11. FDA Product classification database, product code QCD “Continuous Glucose Monitor, Implanted, Adjunctive Use” (via the openFDA API) — api.fda.gov
12. FDA Product classification database, product code QHJ “Sensor, Glucose, Implanted, Non-Adjunctive Use” (via the openFDA API) — api.fda.gov

15. Claim-to-source audit

Last updated 3 October 2026. Sources are limited to primary material (FDA PMA summaries and the product classification database, peer-reviewed papers and reviews). No market-size or market-share estimates are used. Because the article includes structural readings and materials-design interpretations, those are marked as Commentary and kept separate from sourced fact. The approval status of implantable sensors other than Eversense, the specific composition of the indicator, hydrogel and housing, the content of supplements after E3, and histological data on capsules in people are not covered, because they could not be confirmed in published primary sources. Figs. 1, 2 and 3 are concept diagrams, Fig. 4 includes our calculation, and the hero image and Fig. 5 are AI-generated images; none of them shows a real product, cross-section, micrograph or measured data. This article explains technology. It gives no advice on diagnosis, treatment, whether to undergo a procedure, or how to read a measured value, and it does not recommend any product.