CGM for 2020 – will new entrants reduce costs?

One of the highlights of the various conferences during 2019 has been the slew of new CGM providers that appear to have CE marks (European approval) and are therefore entering the previously uncompetitive market that is CGM.

Current Landscape

Let’s remind ourselves of the current environment, with the incumbent behemoth that is Dexcom with their G6, which is widely regarded as the most accurate system on the market. In addition, there is the annually underperforming Medtronic, with solutions that, when they work are okay, but that they have yet to figure out how to make their CGM manage without at least one, of not two calibrations per day.

In addition, Medtrum have come to the awareness of many, although my personal experiences with their A6 solution have been less than effective although the S7 has been given anecdotal thumbs up from users in central Europe, and we await their A8 with interest. Meanwhile Eversense has offered a slightly different approach to the traditional “change every 7-10 days” model, with their implantable, 90/180 days options, which still require some level of calibration.

On top of this, there’s the Abbott Freestyle Libre, and Libre 2 in some countries, that while technically not CGM, has sprouted a whole industry of add-ons and hacks to get CGM data out.

In the context of these, there seems to be a fairly significant opportunity for the right player at the right price. So what does that look like?

2020 – What’s coming next?

There are a few upstarts and options that have been displayed at a number of events this year, with the following appearing to be the lead examples. While they’ve displayed at EASD and other conferences, we’ve yet to see launch dates for them, and given that ATTD (the Diabetes technology conference) is taking place from 19-22 February, it’s not too unreasonable to suspect that this is when we’ll see some major launches taking place.

What does the line-up look like?

• Ascensia’s relaunch of PocTech
• Agamatrix’s Waveform Cascade
• Aidex from GlucoRX
• Infinovo’s Glunovo i3
• Nemaura’s Sugarbeat

What do we know about these providers?

To briefly look at all of these, we’ll review four points:

• What do we know about the system?
• How accurate is it?
• How much will it cost?
• Anything unique in the offering?

Ascensia with PocTech

What do we know about the system?

I’ve previously written about PocTech here. As far as we can see, Ascensia have picked up where other companies had left off and plan to release it as their CGM offering. We know anecdotally that there were “biocompatibility issues” in that a significant portion of those who tried it out in the past found it didn’t work well for them, and that the handset/phone app were considered to be fairly primitive, so here’s hoping that with Ascensia’s budget helping to get this product across the line, we’ll see something a little more robust and effective.

Highlights are supposed to be one calibration per day and a transmitter that allows the battery to be replaced, theoretically providing an unlimited life on the transmitter hardware, or at least, until the watertight seal fails. 

How accurate is it?

As noted in the previous Diabettech article, PocTech had published the following accuracy data, with no source information:

Should they be trusted then? I’d wait for further information from Ascensia’s rebrand before taking this as gospel.

How much will it cost?

There has so far been little communication from Ascensia in relation to cost or availability of this solution, so we await further indications from the company. 

Anything unique in the offering?

Poctech’s unique points are their four electrode sensor that is supposed to enhance accuracy by enabling better management of noise, and the parallel insertion which places the sensor parallel to the skin, reducing the depth to which it is inserted. 

Waveform Cascade from Agamatrix

What do we know about the system? 

Taking a step back, Agamatrix WaveForm’s CGM technology was acquired in early 2016 from iSense CGM Inc. and Bayer Healthcare, so isn’t especally new.

Aside from Waveform taking Dexcom to court over patent infringement, and Dexcom winning. they’ve published quite a bit more recently. According to the AgaMatrix Waveform Cascade CE mark approval press release, we know the following:

• Needle-free sensor insertion, which is virtually painless and reduces insertion complications observed with other systems, such as bleeding at the insertion site.
• Non-adjunctive use, which empowers users to make diabetes treatment decisions using data from the WaveForm CGM without the need to confirm their glucose level through a fingerstick test.
• Industry-leading “catch-up mode” to sync valuable glucose data from the past 24 hours from the on-body transmitter to the app, anytime a user was inadvertently separated from their app.
• Reusable sensor applicator and transmitters over a multi-year lifespan, which reduce plastic biological waste concerns associated with single-use systems on the market.
• Limited interference concerns, for example for those taking acetaminophen.
• Limited calibrations, 1 per day after initialization, to reduce the amount of fingerstick testing needed while optimizing CGM performance.
• Rapid warm up time, so that users can start to view glucose data in less than 1 hour.

They also talk about a very customisable app, especially when looking at alarms, which is often a common complaint with Dexcom’s app. 

It’s also worth noting that they’re talking about a 14 day sensor life.

How accurate is it?

There’s no data published directly on the press release, however, in 2018, there was an accuracy comparison published that had a small sample size of 15 users, and we assume followed the once daily calibration described as part of their CE mark. The outcomes?

• MARD of 11.0%
• 99% of results in zones A and B of the consensus error grid

There have also been a number of other publications (1, 2, 3) that suggest varying MARD values between 11.1% and 13.9%. More recent ones have resulted in the values at around 11.0%.

So while the available data suggests that it isn’t up there with Dexcom’s G6, it looks like it’s pretty good. It remains to be seen if the production version is better than these figures, and whether additional clinical trials will be published. 

How much will it cost?

So far, there’s no data relating to costs or release dates available on the web.

Anything unique in the offering?

Three things stand out from the above list of features – needle free insertion, reusable sensor applicators and the short warm up time. It’s not clear at this point how any of these things will work, given the description of the device, however it presents something pretty different. 

Aidex from GlucoRX

What do we know about the system?

Aidex is a new CGM product from GlucoRX which appears to be a rebranding of the MicroTech Medical CGM system from China. 

Those present at the Diabetes Professional Care conference mentioned that the inserter (which looks a lot like a Freestyle Libre inserter) is reusable, although the user manual that I’ve found suggests otherwise. 

The website also says that the sensors last for 14 days and are calibration free. 

How accurate is it?

Whilst there are no available MARD figures for the system yet, there is a consensus error grid, and the data suggests 100% of the results fall in zones A and B, which isn’t a bad result. For comparison, Dexcom G6 achieved 92% in zone A. Also, on the website, they state that the 96% of results were achieved against the consensus error grid zone A, which doesn’t align with the the image that I’ve copied from the website below. 

How much will it cost?

Once again, there’s no real information on this, and no date as to when it will be publicly available.

Anything unique in the offering?

This seems perhaps a little less unique than some of the other systems out there, so it remains to be seen.

Infinovo Glunovo i3 

What do we know about the system?

Infinovo are a chinese company. The marketing material tells us that they have a 14 day sensor with a once click inserter. The imagery available online suggests it perhaps looks the most like a Dexcom device. They  have a fully available suite of apps that work on both phones and tablets, but no receiver and their system takes glucose data recordings every 3 minutes. They claim that their transmitter will last for four years, and their website indicates that calibration is probably required for best accuracy, but doesn’t state how frequently. 

How accurate is it?

Once again, according to their website, they have recorded a MARD of 9.5% in a CFDA supervised clinical trial that contained 78 participants. 

How much will it cost?

Again, there is nothing offered that helps identify what the costs of this solution will be. It’s also not entirely clear that the system yet has a CE mark. 

Anything unique in the offering?

The key difference between this option and the others in this list is that it does not have a receiver. 

Nemaura’s Sugarbeat

What do we know about the system?

It’s a non-invasive patch that’s designed to be used throughout the day that we’ve looked at here over the last couple of years, is supposed to last 24 hours and is changed daily. In addition, it has a 25 minute warm up period and may or may not require calibration (in the submission to the FDA this hasn’t been clarified. It was supposed to launch in Q3 2019, but we’ve yet to see availability. It has a CE Mark.

How accurate is it?

The reported MARD is apparently 11.92% with one calibration daily (as mentioned in these slides) although in the low range (60-80mg/dl) this widens to 19.28%, so while you may have something non-invasive, it’s not that great when in hypo territory.

How much will it cost?

While UK pricing hasn’t been announced, in this press release they talked about 16 patches per month on a subscription basis costing $55 per month and including a transmitter. It’s not clear whether they’d offer 31 patches a month, but if so, it would bring the price in well below that of Dexcom’s G6.

Anything unique in the offering?

The key thing – it’s non-invasive. And while it might not have the greatest MARD, that’s a very big thing. Dependent on how you look at it, the model they provide may be useful for people that don’t want constant glucose monitoring, or need it in certain conditions, however, the performance outside of a normal glucose range may limit its use for some.

What does this mean for the UK market?

Given the reported performance of some of these systems, they would appear to be valid alternatives to some of the existing options, namely the Abbott Freestyle Libre, however, in order to really challenge the crown of this device, they’ll need to come in at an all important price point that sits at a competitive level.

WIth no pricing information available right now, it’s hard to tell where this is likely to end up. 

The other question that these raise is the validity of their performance data. We’ve seen before that unbacked up statements about accuracy can be wildly out when the devices reach those of us in the real world, and that the reliability of them can also be as big a challenge as the accuracy. Some of those listed here have already raised questions on that front and it will be interesting to see if the investment from deep pocketed companies has improved this aspect.

Ultimately though, for most people, the bigger question is whether true CGM will be made widely available on the NHS. For this to be the case, the systems named would have to be in discussion with the powers that be at NHS England and also the NHS BSA, and would have to be offering strongly competitive pricing and evidence that what they say they can do is real. That would be a question for Partha Kar, who has already hinted at discussions with Dexcom and maybe others via Twitter.

However you take it, the news looks positive for those with T1D. Increased competition with longer lasting and better performing sensors should lead to reduced costs and better access across the board. And if they are good enough, would the NHS to accept this and allow alternatives to Libre where the cost is no different or better and the potential data better?