TypeZero CEO Chad Rogers: NIH Funded $12.7 million International Diabetes Closed Loop Trial designed with a PMA submission in mind – January 6, 2016

Startup TypeZero announced on Monday that its inControl AP system will be used in the NIH-funded $12.7 million International Diabetes Closed Loop Trial (IDCL) slated to begin at ten sites in Q2 or Q3 of this year. The study will compare 160 patients on inControl AP (a commercial-grade version of UVA’s DiAs) to sensor-augmented pump therapy. Each person will stay in IDCL for six months, and the entire study is planned to last 2.5 years. We first heard about this major artificial pancreas grant during Dr. Frank Doyle’s IDF presentation last month, though a conversation with TypeZero CEO Chad Rogers has shared additional details about the study, planned devices (Dexcom CGM, unknown pump, smartphone running the algorithm), and plans for commercialization.

Most notably, we’ve learned that IDCL was conceived and designed with an FDA PMA submission in mind, and data from the first cohort (after the first six months of the study) could be used to submit the regulatory filings. FDA provided input on the trial design, and a future regulatory submission for the inControl AP system could come from either TypeZero or another party. The party that will lead the PMA submission is still under discussion; we assume it would be TypeZero or a yet-to-be-named insulin pump company. The obvious model would have TypeZero license its algorithm to a pump partner, who would then embed the algorithm in a pump and work with TypeZero and Dexcom to submit the PMA.

The news is encouraging for whatever partners are ultimately involved in the commercial closed-loop device, as it basically means NIH is funding a pivotal closed-loop study. The study could also de-risk bringing a closed-loop system to market for Medtronic competitors like Animas, Insulet, Roche, and Tandem. Tandem and Insulet could be the most likely pump partners, as both are partnered with Dexcom, both have pumps with Bluetooth capabilities (Tandem’s has been used in previous DiAs studies; Insulet’s is in its next-gen PDM), and both have publicly committed to closed-loop programs in their latest quarterly calls (here and here). Animas has long talked about closing the loop (and is also partnered with Dexcom), but has not shared any real progress in over a year. Though Roche’s Bluetooth-enabled pump has been used in past DiAs studies, the company is developing its own CGM, which presumably makes it an unlikely partner for the Dexcom-integrated inControl AP. Medtronic is of course leading the pack to hybrid closed loop commercialization, as its MiniMed 670G is already in a three-month pivotal study, slated to wrap up by May.

No matter who the TypeZero’s pump partner is, we look forward to seeing data from IDCL and learning more about how this system will come to market, what the final product will look like (e.g., algorithm in a pump, on a smartphone, or a separate proprietary device), and when it might be submitted. The academic-developed DiAs system has gathered a very impressive 183,708+ hours of clinical trial use since 2008 (!), offering experience and data to optimize a commercializable closed-loop system. That experience is a major asset for the TypeZero team. Can the system cross the chasm from academic research trials to a successful commercial product with PMA-level manufacturing, customer support, and a business model?

• The International Diabetes Closed Loop Trial is a randomized controlled trial, comparing sensor-augmented pump (SAP) therapy to the use of the inControl AP system running on a smartphone. The study is expected to have 240 adults with type 1 (160 AP patients vs. 80 SAP) and is scheduled to begin in Q2 or Q3 of 2016. Each person will stay in IDCL for six months, and the entire study is planned to last for 2.5 years. An FDA submission could be based on the first six months of the study. Recruitment will begin shortly, and interested parties should contact artificialpancreas@virginia.edu. Based on the press release, the trial will aim for a reduction in average blood sugar level and less risk of hypoglycemic events (we’re not sure if it will be powered for severe hypoglycemia events).
• A second six-month study will follow IDCL and look at a Zone-MPC control algorithm in 180 patients. This algorithm has been the work of Dr. Frank Doyle’s team, now at Harvard. We’re not sure how this would factor into the product design or PMA submission of inControl AP.     
• UVA’s Dr. Boris Kovatchev is the principal investigator for this study, and the highly-respected clinical trial sites are: the University of Virginia; Harvard University; Mount Sinai School of Medicine; the Mayo Clinic; Stanford University; the University of Colorado; the University of Padua, Italy; Regional University Hospital Center of Montpellier, France; the Academic Medical Center, Amsterdam; William Sansum Diabetes Center, Santa Barbara; with coordination by the JAEB Center for Health Research in Florida.
• IDCL is designed with two technology update points during the trial for new CGMs, pumps, and smartphone or controller hardware. This study is designed to run the commercial version of TypeZero’s Artificial Pancreas system (inControl AP) and will pass through other studies prior to its involvement in IDCL.  The utilization of the UVA research platform DiAs will not be apart of the IDCL or other large scale studies; going forward, it will be replaced with the TypeZero commercial platform. We’re not sure how different the research and commercial versions are, but presumably there have been updates to the user interface, system setup, and perhaps the algorithm.
  
  • The current implementation of inControl AP utilizes any number of Android OS smartphone platforms or a proprietary Android OS controller that is connected to a variety of insulin pumps and a Dexcom sensor. For the Dexcom sensor, TypeZero is still utilizing the G4AP Share receiver, but implementation timing for G5 and/or G6 is being discussed. As for the insulin pumps, TypeZero can still only disclose the utilization of the Roche and Tandem pumps in current studies; more details about future pump parties will be “forthcoming shortly.”
  • Future studies to establish equivalence to IDCL may be required depending on the implementation of the final device submitted to the FDA. We assume this will be needed if the commercial device chooses to embed the algorithm in a pump, or if it uses a different pump or algorithm or CGM than the one tested in the study. We assume such a bridging study would be short.
• As of the most recent company update last June, TypeZero had raised seed round funding (amount not specified). The startup originally licensed the core technology from the University of Virginia in 2013, and has worked over the past two years to translate it into FDA approvable solutions. TypeZero also has ambitions to move beyond closed-loop work to provide cloud-based, prescriptive analytics software to provide action, guidance and support to patients and clinicians across the entire spectrum of treatment (e.g., SMBG alone to CGM/pump). The company’s webpage currently lists three products:
  
  • inControl AP (described above).
  • inControl Advice (mobile based advisory system that generates real-time recommendations for meals, basal rates, bolus calculations, and exercise decisions. The application utilizes CGM, pump, or a smart-pen to give optimal advice, risk prediction, and notifications that answer the question, “What should I do now?” Sounds amazing!).
  • inControl Cloud (a cloud based system that provides real-time monitoring and notifications for caregivers and family, plus retrospective analysis of data to identify areas of risk or treatment inefficiencies).
• The eight-person TypeZero team includes CEO Chad Rogers and CTO Dr. Patrick Keith-Hynes (a noted technologist who originally developed the DiAs user interface on the phone). Dr. Boris Kovatchev is listed as the Chief Mathematician, though he is still fully employed at UVA. The rest of the team includes UVA mathematician Dr. Marc Breton (founder), Stephen Patek (founder), nurse and type 1 Molly McElwee-Malloy (patient engagement), Benton Mize (software engineer), and Antoine Robert (software engineer).
• NIH has shown impressive commitment to the artificial pancreas, which will continue in 2016 with its fourth workshop occurring from July 12-13, 2016. The meeting will be held on the NIH campus and the agenda is still in progress. Send us your ideas or contact Dr. Guillermo Arreaza-Rubin for more details.

Close Concerns Questions

Q: Who will TypeZero choose as its pump partner? Who will submit the PMA and how quickly? What will the commercial product look like?

Q: Assuming Medtronic is first to market with the MiniMed 670G hybrid closed loop, who will be second? Could Bigfoot leapfrog other pump companies? Will TypeZero and partners be able to submit inControl AP before the full 2.5 year study is complete?

Q: What is the appropriate comparator group in a pivotal closed-loop study? (sensor-augmented pump therapy vs. MDI vs. a combination)

Q: Should a closed-loop algorithm reside on a smartphone? Is the better user interface and flexibility worth it if patients have to carry an extra stripped down phone?

Q: Will closed-loop systems command a premium price from payers, assuming pivotal trials show reductions in A1c and/or hypoglycemia?

Q: What is the right revenue model for closed loop systems? Will the traditional pump model work?

Close Concerns’ Reflections on 2015+2016 – Theme on Automated Insulin Delivery

Closed-Loop Advances Towards Commercialization, led by Medtronic’s MiniMed 670G

• Last year (2015) made it clear that closing the loop is inevitable – it’s now a question of “WHEN” and “HOW,” not “IF.” JDRF’s Dr. Aaron Kowalski emphasized the field’s progress at ADA 2015, noting that technical feasibility is now proven in all six stages of the original JDRF roadmap – it’s now time for commercial development and regulatory studies to support approval (the “last 100 yards,” as JDRF’s ADA 2015 closed-loop night put it). Moving forward, Dr. Kowalski’s revised AP roadmap (Diabetes Care 2015) has just three steps: (i) low glucose suspend; (ii) predictive low glucose suspend; and (iiia) insulin-only closed loop or (iiib) multi-hormone closed-loop. The real challenges ahead are adoption and reimbursement – or as Dr. Kowalski suggested, appealing to people with diabetes (plus loved ones), clinicians, and payers via three key metrics: glycemic control, “burden”, and value. Encouragingly, the field seems quite comfortable with (and even excited about!) the regulatory path for artificial pancreas systems. This is a major testament to a forward-thinking FDA device division, and light years ahead from where the field stood in 2010! Business model, reimbursement, and product form factor questions, however, continue to perplex many in the field.
• See below for an overview of the automated insulin delivery landscape, as far as we aware. We acknowledge this list may be incomplete, as there may be other stealth startups or academic groups working to commercialize closed-loop technology. The list is ordered based on the expected time to a pivotal study.

Table 1: Automated Insulin Delivery Competitive Landscape

• The toughest questions surrounding the human and business sides of automating insulin delivery have yet to be answered. What’s the appropriate control group in a pivotal study (e.g., no control group (670G pivotal) vs. MDI vs. sensor-augmented pumps)? What is the incremental benefit of glucagon over insulin alone? How will patients trade off the level of automation (hybrid, full) vs. the level of glycemic control (low average vs. low hypoglycemia) vs. the device form factor (size, hassle factor, smartphone integration)? Will organizations have manufacturing in place to submit a PMA application? What’s the right business model? What do payers think? There’s no question that automated insulin delivery systems are going to look and function differently, and the full extent of the risk vs. benefit tradeoffs may not be answered in carefully controlled clinical studies, but in real world use and reimbursement. For more thoughts on the hard-hitting behavioral questions in closed loop, see Adam and Kelly’s diaTribe test drive following their experience in the DiAs closed-loop study.

-- by Adam Brown and Kelly Close