Glucotrack implantable CBGM
Glucotrack
The most direct attack on sensor lag anyone is attempting — and years from proving it.
Every sensor you can buy today reads interstitial fluid — the fluid between your cells — which is why your CGM runs minutes behind your actual blood glucose. Glucotrack is putting the sensor in a vein instead: a fully implantable continuous *blood* glucose monitor (CBGM) with a three-year design life, measuring glucose directly in blood. A 10-person first-in-human study in Brazil reported a 7.7% MARD with 99% data capture and no procedure- or device-related serious adverse events, and the company submitted an FDA IDE in May 2026 to start a US study. No pivotal trial has run and no regulator has cleared it.
The scorecard
A 7.7% MARD with a 99% data-capture rate in the first-in-human study — a number that would sit at the top of the CGM field if it held up. But this was 10 people in a single small study in Brazil, not a pivotal trial, and accuracy in the low range has not been reported. Scored as a strong early signal, not a validated result.[1]
Sensing inside a vein should, by physics, remove the interstitial delay that limits every CGM on the market — that is the entire premise of the device. But no human lag figure has been published, so we do not credit it with a lag advantage it has not yet demonstrated. Scored at the category norm; we will move it when a measured comparison exists.[1]
Nothing loops with it today. It has no regulatory designation (no iCGM clearance), and no pump, AID or app partners have been announced.[2]
Designed as a three-year implant — far beyond the 10–15 days of a disposable patch or the 365 days of Eversense. A peer-reviewed in-vitro study published in April 2026 showed the electrochemical sensor holding its performance over one year on the bench; three years in a human body is a target, not a demonstrated result.[3]
Calibration requirements have not been published. Measuring blood directly should in principle reduce the need for fingerstick correction, but until the company reports a calibration schedule this is unknown — scored in the middle to reflect that, not to imply a benefit.[1]
No alerting or predictive-alarm system has been described publicly. A continuous blood signal is a good foundation for trustworthy alerts, but none has been shown.[1]
The sensor goes into the subclavian vein, near the collarbone, in a minimally invasive procedure — a vascular implant is a genuinely bigger ask than a patch you stick on yourself, or even the under-the-skin arm implant of Eversense. For some people three years without touching a sensor will be worth it; for many it will not.[1]
Nothing on the market anywhere. Glucotrack submitted an Investigational Device Exemption to the FDA on 7 May 2026, with a US clinical study targeted for the second half of 2026. A pivotal trial and regulatory review would still have to follow, so access is effectively zero — hence the floor score.[2]
Editor’s take
Every CGM you can buy solves sensor lag by *predicting around it* — smarter algorithms guessing where your blood glucose already is. Glucotrack is the only company we track trying to delete the problem instead, by measuring blood itself. The early human numbers (7.7% MARD, 99% data capture, no serious adverse events) are genuinely good, and a three-year implant would change what wearing a sensor means. But this is 10 people, one small study, and a device that asks you to accept an implant in a vein near your collarbone. The IDE submission in May 2026 is the thing to watch: it is the point at which the claims start meeting a US regulator. Until a pivotal trial reports, treat every number here as a promise, not a product.
The full picture
Every continuous glucose monitor you can buy — Dexcom, Libre, Eversense, all of them — measures glucose in interstitial fluid, the fluid sitting between your cells. Glucose has to move out of the bloodstream and into that fluid before the sensor can see it, which is why your CGM reads a few minutes behind what is actually in your blood, and why it is worst exactly when you need it most: during a fast drop, or after a rapid rise. The industry's answer so far has been smarter prediction — algorithms guessing where your blood glucose already is.
Glucotrack is attempting something different: put the sensor in a vein. Its implantable continuous blood glucose monitor (CBGM) is placed in the subclavian vein, near the collarbone, in a minimally invasive procedure, and measures glucose directly in blood rather than in interstitial fluid. It is designed to stay there for three years.1
How well does it work? Ten people know. In a first-in-human study in Brazil (n=10), the device reported a MARD of 7.7% — a number that, if it held, would sit at the very top of the CGM field — with a 99% data-capture rate and no procedure- or device-related serious adverse events.12 That is a genuinely encouraging first result. It is also ten people in a single small study. There is no pivotal trial, no published low-range accuracy, no head-to-head against a marketed sensor, and no regulator has reviewed any of it.
The lag claim. Reading blood directly should, by physics, remove the interstitial delay that constrains every CGM on the market. That is the entire premise of the device and the reason we list it against the sensor-lag gap. But we have not seen a human lag figure published — no measured comparison showing how much faster it tracks a real fall than a patch sensor does. Until that exists, treat "no lag" as a design claim, not a measurement.
Three years, and a bench test. A three-year implant would be a different category of thing from a patch you change every 10–15 days, or even from the one-year Eversense implant. In April 2026 Glucotrack announced a peer-reviewed publication supporting the long-life sensor design — but read the fine print: that study tested electrochemical sensor stability over one year in vitro (on the bench), not in a body.2 Bench durability is a necessary condition for a long-life implant, not proof of one.
What it works with. Nothing, yet. There is no iCGM clearance, so it cannot legally drive a pump or an automated insulin delivery system, and no pump, AID or app partners have been announced.3 There is no ketone sensing.
The form-factor question. This is the honest trade-off, and it is a big one. Eversense already asks people to accept a sensor inserted under the skin of the arm by a clinician. Glucotrack asks for a sensor inside a vein. It is described as a minimally invasive procedure, but "minimally invasive" is not "nothing," and a vascular implant carries a different set of risks and a different psychological weight than a patch. Three years of never touching a sensor will be worth it to some people and a non-starter for others.
Where it actually is. Glucotrack submitted an Investigational Device Exemption (IDE) to the FDA on 7 May 2026 — the application that would let it run a US clinical study — with the study targeted for the second half of 2026.31 An IDE submission is not an approval and not even a clearance to market; it is permission to test. Beyond that lie a pivotal trial and a full FDA review. The company also showcased the technology at the ADA 2026 Scientific Sessions in June, though that was a presence announcement rather than a release of new clinical data.4
Our read. This is the only serious attempt we track to fix sensor lag at its physical root rather than paper over it with prediction. That makes it the most interesting CGM programme in the pipeline — and also one of the furthest from your arm. Watch the IDE outcome and the first US study; those are the moments the promise starts colliding with evidence.
References
-
MedTech Dive. Glucotrack plans US trial of implantable glucose monitor this year — fully implantable CBGM, sensor placed in the subclavian vein near the collarbone via a minimally invasive procedure, three-year design life, measures glucose directly in blood rather than interstitial fluid; first-in-human study in Brazil (n=10) with MARD 7.7% and no procedure- or device-related adverse events; IDE planned for Q2 2026 with a US study to begin in the second half of 2026. https://www.medtechdive.com/news/glucotrack-plans-us-trial-of-implantable-glucose-monitor-this-year/816006/ ↩ ↩2 ↩3
-
Glucotrack. Glucotrack Announces Publication of Peer-Reviewed Study Supporting Long-Life Sensor Design for Implantable Continuous Blood Glucose Monitoring Technology (press release, 21 Apr 2026) — peer-reviewed study of electrochemical sensor stability over one year in vitro; the release also restates the first-in-human MARD of 7.7% and 99% data-capture rate with no procedure- or device-related serious adverse events. https://www.globenewswire.com/news-release/2026/04/21/3277876/0/en/glucotrack-announces-publication-of-peer-reviewed-study-supporting-long-life-sensor-design-for-implantable-continuous-blood-glucose-monitoring-technology.html ↩ ↩2
-
Glucotrack. Glucotrack Announces IDE Submission for U.S. Clinical Study of its Novel Implantable Continuous Blood Glucose Monitoring Technology (press release, 7 May 2026) — Investigational Device Exemption application submitted to the FDA; a submission, not an authorization. https://www.globenewswire.com/news-release/2026/05/07/3290388/0/en/Glucotrack-Announces-IDE-Submission-for-U-S-Clinical-Study-of-its-Novel-Implantable-Continuous-Blood-Glucose-Monitoring-Technology.html ↩ ↩2
-
Glucotrack. Glucotrack to Highlight Implantable Continuous Blood Glucose Monitoring Technology at the American Diabetes Association's 2026 Scientific Sessions (press release, 26 May 2026) — a showcase announcement, not new clinical results. https://www.globenewswire.com/news-release/2026/05/26/3301032/0/en/Glucotrack-to-Highlight-Implantable-Continuous-Blood-Glucose-Monitoring-Technology-at-the-American-Diabetes-Association-s-2026-Scientific-Sessions.html ↩
Coming soon
ETA · US clinical study targeted for H2 2026 following the May 2026 IDE submission. No approval anywhere; commercial availability is years away.
- →FDA decision on the Investigational Device Exemption submitted 7 May 2026 · 2026
- →First US clinical study of the implantable CBGM · targeted H2 2026
Sources
- [1]Glucotrack plans US trial of implantable glucose monitor this year · news — MedTech Dive. Describes a fully implantable continuous blood glucose monitor with the sensor placed in the subclavian vein near the collarbone via a minimally invasive procedure, a three-year design life, and direct measurement of glucose in blood rather than interstitial fluid. First-in-human study in Brazil (n=10) reported a MARD of 7.7% with no procedure- or device-related adverse events; an IDE was planned for Q2 2026 with the US study to begin in the second half of 2026.
- [2]Glucotrack Announces IDE Submission for U.S. Clinical Study of its Novel Implantable Continuous Blood Glucose Monitoring Technology · manufacturer · 2026-05-07 — Company press release announcing submission of an Investigational Device Exemption application to the FDA on 7 May 2026 for a US clinical study of the implantable CBGM. A submission is not an authorization — the FDA had not ruled at the time of the release.
- [3]Glucotrack Announces Publication of Peer-Reviewed Study Supporting Long-Life Sensor Design for Implantable Continuous Blood Glucose Monitoring Technology · manufacturer · 2026-04-21 — Company press release announcing a peer-reviewed publication on the electrochemical stability of the sensor design over one year in vitro (bench testing, not in humans). The same release restates the first-in-human results: "a highly-accurate MARD of 7.7% and a 99% data capture rate, with no procedure- or device-related serious adverse events."
- [4]Glucotrack to Highlight Implantable Continuous Blood Glucose Monitoring Technology at the American Diabetes Association's 2026 Scientific Sessions · conference · 2026-05-26 — Company announcement of its presence at ADA 2026 Scientific Sessions to showcase the implantable CBGM technology. A presence/showcase announcement, not a report of new clinical results.