Microneedle Array Sensor Innovation

A microneedle array sensor reads glucose through a patch of very short needles that stop in the upper layers of skin. Conventional CGMs from Dexcom and Abbott work differently: they don't use microneedles, and they sample fluid deeper, below the dermis [1]. As of October 2026, the Biolinq Shine is a microneedle glucose sensor with FDA authorization. The agency granted it as a prescription device on September 23, 2025, and cleared it for over-the-counter use on September 18, 2026 [2][3]. PKvitality's K'Watch, a watch-style microneedle monitor that often comes up in the same conversation, has no record in the FDA's 510(k) or premarket approval databases [4][5].

How intradermal microneedle sensing works

The skin's outer layer, the stratum corneum, is its main barrier [1]. Below it lies the dermis, which holds interstitial fluid: the fluid around cells that trades glucose with nearby capillaries and tracks blood glucose closely [6]. Microneedles are projections just long enough to cross the stratum corneum and reach that fluid. One review puts their length at up to 1 millimeter, another at under 1.5 millimeters, often with hundreds arranged in a single array [1][6].

Their appeal, as reviewers describe it, is that a needle that short usually stays above the nerves and blood vessels deeper in the dermis [6]. That is a design rationale from the research literature. We didn't find a pain or comfort claim in the FDA's decision summary for Shine [7].

The chemistry will sound familiar if you know how CGMs work. In Shine, the enzyme glucose oxidase reacts with glucose at each microsensor, and the electrode reads the resulting current, which rises with the amount of glucose in the fluid [7]. Some research prototypes coat the same enzyme onto platinum electrodes in the needle array [1]. What changes is where the sensor sits, and that shift drives most of the engineering problems covered below.

Biolinq Shine's FDA status

The FDA granted Biolinq's De Novo request for the Biolinq Shine Autonomous Time-in-Range Microsensor on September 23, 2025 [2]. The grant classified Shine as a Class II "glucose range monitoring system" under 21 CFR 862.1359, with prescription-only labeling [7].

A year later, on September 18, 2026, the FDA cleared 510(k) K262599, which names the De Novo device as its predicate and adds over-the-counter use [3]. Biolinq's 510(k) summary says the sensing principle, hardware and software are unchanged; the edits were to the indications and labeling, supported by human factors testing [3]. We couldn't confirm from a primary source when the over-the-counter version will be sold. Our overview of the over-the-counter CGM market covers the other nonprescription sensors.

Here's what surprised us when we read the decision summary: Shine shows no glucose numbers on the wearable at all. Seven microelectrodes sit in the papillary dermis, the top layer of the dermis, on the inner forearm. A light on the device changes color with the glucose range, and an optional app shows narrower ranges [7]. The sensor measures from 70 to 400 mg/dL, shows a red light below that range and lasts up to five days, with no alerts or alarms [7].

Shine is labeled for adults 22 and older who don't take insulin. The labeling says it must not be used by people with type 1 diabetes or with automated insulin delivery systems [7]. Both FDA documents say users shouldn't take medical action based on its output without a qualified healthcare professional [3][7], so any change to treatment is a decision for your care team.

What the pivotal study showed

Biolinq's pivotal study ran at six US centers from February 29 to June 10, 2024. It enrolled 247 adults, and 202 of them wore Shine [7][8]. Sensor output was compared with venous blood measured on a laboratory analyzer, drawn every 15 minutes during clinic sessions [7].

When the wearable showed blue (70 to 180 mg/dL), the lab value fell in that same range in 79.8% of paired readings [7]. Of the primary sensors, 84.7% lasted the full five days [7]. About two-thirds of participants had type 1 diabetes, a group the device isn't labeled for; the FDA says they were included to collect enough data across the display range [7].

PKvitality's K'Watch has no FDA authorization

The K'Watch comes from PKvitality, a company based in France. A peer-reviewed survey of glucose monitoring products, with searches current to December 17, 2020, described it as reading interstitial fluid glucose through microneedles with a glucose oxidase reaction. The same survey listed it as neither FDA-cleared nor CE-marked [9].

The trail since then is thin. PKvitality's one study on ClinicalTrials.gov was an early San Diego trial that tested a series of design iterations. It still shows an "unknown" status and was last updated in May 2022 [10]. As of October 2026, searches of the FDA's 510(k) and premarket approval databases return no PKvitality records [4][5]. A listing that presents the K'Watch as sold in the US doesn't match anything we found in FDA records.

Open technical challenges

Shine shows the approach can clear FDA review for a narrow use. The reviews we read point to the same problems standing in the way of broader use:

• Breakage. Silicon microneedles are mechanically strong, but parts of them can break off under the skin [6].
• Biofouling. Proteins and cells build up on the sensor tip and weaken the signal. One review calls for better antibiofouling protection to stretch microneedle wear toward 15 days, in line with common CGM platforms [1]. Shine is labeled for five [7]. Our article on biofouling mitigation for dermal sensors goes deeper.
• Protecting the enzyme. The enzyme layer has to survive insertion and storage, and sterilization methods still need careful assessment [1]. For Shine, Biolinq validated e-beam sterilization and a six-month shelf life [7].
• Interference. In bench testing reported to the FDA, acetaminophen, hydroxyurea and xylose pushed Shine's readings high [7].
• Lag. The delay between blood and interstitial glucose differs from person to person. It also shifts with activity and with whether glucose is rising or falling, so a single lag figure from a manufacturer can't be relied on [9]. One dermal microneedle prototype showed about 10 minutes of lag in testing [1].
• Skin and fit. Short studies of microneedle sensors reported minor swelling and redness that cleared without treatment [9]. One review argues that designs need to account for differences in sex, body mass index and ethnicity [6].

The largest open question is reach. Shine's authorization covers adults not on insulin, five-day wear and color ranges instead of numbers [7]. Moving microneedle arrays beyond that will take large clinical studies, which one review names as a requirement before these sensor arrays reach practical medical use [1].