Memorandum

Ypsomed receives two-year JDRF funding to develop open-protocol, fully interoperable mylife YpsoPump – August 28, 2018

Executive Highlights

  • Ypsomed announced that it has received two-year funding from JDRF for the development and regulatory approval of an open protocol, fully interoperable mylife YpsoPump. The funding commitment (amount not disclosed) will support Ypsomed’s participation in JDRF’s Open-Protocol Automated Insulin Delivery Systems Initiative, as Ypsomed plans to enable seamless pump connectivity with third-party apps and devices (presumably via Bluetooth). Over the next year, Ypsomed plans to integrate CGM data into the existing mylife app and to establish two-way communication between the pump and app; currently, it can only read pump status, but not remotely control the pump.

  • This is a smart competitive move for Ypsomed: opening up its pump to third-party AID innovation (DIY included) may prove simpler than building its own AID system (and the required trials and partnerships that needs). This could also become a marketing advantage – especially for the DIY community – as YpsoPump is already in 16 countries and aims to launch in the US (2019; FDA review underway) and Canada (2018).

  • Ypsomed is the fourth (!) pump manufacturer to join JDRF’s open protocol AID initiative, following Roche (February), SOOIL (June), and SFC Fluidics (July). With an increasingly crowded AID space, having automation and/or an open-protocol pump will prove increasingly critical for competing.

Ypsomed announced today that it has received a two-year grant (amount not disclosed) from JDRF to accelerate development and regulatory approval of an open protocol, fully interoperable mylife YpsoPump. The funding commitment will support Ypsomed’s participation in JDRF’s Open-Protocol Automated Insulin Delivery Systems Initiative, as Ypsomed plans to enable seamless connectivity with “smartphone-based applications and/or other devices” (i.e., third-party AID algorithms and interoperable CGMs) via published communication protocols. The YpsoPump already has Bluetooth and sends data to the mylife app (e.g., to display pump status).

Ypsomed is the fourth pump manufacturer to join JDRF’s open protocol automated insulin delivery initiative following Roche in February, SOOIL in July, and SFC Fluidics last month. The momentum is certainly building for interoperability following this initiative’s launch last October. 

This is a very smart strategic move for Ypsomed, given the increasingly crowded AID landscape and Ypsomed’s lack of a CGM partner or control algorithm. (Plus, Dexcom just acquired the obvious algorithm choice, TypeZero.) With an open-protocol pump, Ypsomed can theoretically avoid the traditional AID path – formal partnerships, studies, and PMA submission – allowing for flexibility and an AID-ready pump without the associated hassle. This is speculation on our part; more on the regulatory implications below.

According to the press release, Ypsomed plans to integrate CGM data into the mylife app and to establish two-way communication between the pump and app over the next 12 months - i.e., getting secure app control of the pump is key for making this whole thing work. Dexcom is the most obvious CGM choice, though perhaps Senseonics would integrate too. The mylife app is currently available on Android and iOS – it reads pump status and includes a bolus calculator. However, the bolus calculation must be transferred over to the pump.

  • YpsoPump is currently available in 16 countries with plans to launch in eight additional countries, including Canada in 2018 and the US in 2019 (FDA review underway). This timing aligns with FY18 expectations for a Canadian YpsoPump launch in October 2018 and a US launch in the summer of 2019. Per the press release, FDA submission for the YpsoPump is “ongoing,” and management previously confirmed a May 30 submission. The US timing will place YpsoPump in competition with at least two AID systems – Medtronic’s MiniMed 670G and Tandem’s Basal.IQ with Dexcom G6 – alongside a broader launch of Omnipod Dash next year. (Tandem’s Control.IQ launch is also expected in summer 2019 – a very competitive product.) The Canadian launch is likely to be head-to-head with Tandem’s upcoming international expansion and will serve to fill the hole left by Animas’ exit. We’re not sure how much uptake Ypsomed’s YpsoPump has had in Europe thus far, though the company obviously has a big hole to fill with Insulet taking over Omnipod EU distribution in July.

  • The FDA path for an open protocol pump remains unclear and is an area JDRF is working on. Will Ypsomed submit as an iPump to FDA? At JDRF/HCT’s April workshop on interoperability, the possibility of iPumps and iAlgorithms came up, which would enable more seamless swapping of AID components. Right now, only iCGM exists on this front (i.e., Dexcom G6). Based on Tandem’s Basal-IQ as a case study, the current (non-iPump) framework would require submission of YpsoPump with iCGM compatibility and with some algorithm already incorporated (i.e., similar to Basal-IQ). At ADA in June, SOOIL Korea’s Justin Walker shared plans to submit the company’s smartphone-controlled Dana RS insulin pump to the FDA/CE Mark with an open communication protocol – an interoperable pump to meet the wishes of DIY users. Depending on when it goes in to FDA, this could be first-to-market in the US, and it is already available outside the US. YpsoPump is already under FDA review, and we assume it would be a separate submission for an open protocol version.

  • Ypsomed’s May FY18 call shared that it was in “the final discussions” with closed loop algorithm developers. We’re not sure if Ypsomed is still pursuing an algorithm partnership, and if so, Dexcom/TypeZero may still be the most likely candidate. Of course, an open pump that patients can buy brand-new for use in systems like Loop and OpenAPS is certainly compelling by itself. Again, the regulatory framework on this front is the big question.

  • A mylife YpsoPod patch pump is also in the works, slated for final conceptualization in FY19 (by March 2019) as of Ypsomed’s May FY18 call. This will fill the major hole left by Insulet’s Omnipod, though it is still quite far from market – last November’s more specific update expected YpsoPod launch in late 2020/early 2021.

Close Concerns Questions

Q: Who will submit to FDA first with an open-protocol pump? Will such pumps need an algorithm on board, or will FDA permit them to be open and allow users to choose their algorithm – i.e., Loop, OpenAPS?

Q: Which is a faster approach to getting AID to market – going the open protocol route (Ypsomed) or submitting for iCGM compatibility with an algorithm on board (Tandem Basal-IQ)? 

Q: Will open protocol pumps prove to be a meaningful competitive advantage? If so, over what time horizon will it take the field to catch up and full leverage the vision?

Q: How will companies weigh the potential advantages of open protocol with the potential tradeoffs – e.g., less end-to-end control over the entire system, more risk of commoditization, etc.?

 

--by Maeve Serino, Adam Brown, and Kelly Close