Memorandum

FDA proposes new expedited access program for medical devices that address unmet medical needs – April 28, 2014

Executive Highlights

  • Last Tuesday, the FDA announced an exciting new expedited access PMA program for medical devices that address major unmet clinical needs. The program will feature earlier and more interactive engagement with FDA staff, a collaboratively developed plan for collecting data to support approval, reduced premarket requirements, and a priority review.
  • We believe artificial pancreas (AP) devices may qualify for this program, particularly overnight closed loop systems with daytime hypo-hyper minimization.
  • The FDA has released two draft guidance documents, posted here and here, and is seeking public comment by July 23, 2014.

Last Tuesday, the FDA announced a new expedited access PMA program (“EAP”) for medical devices that address major unmet clinical medical needs. The new EAP program appears to be modeled on the FDA Innovation Pathway that was first announced in 2011. Similar to that 2011 effort (which explicitly identified “truly pioneering technologies with the potential of revolutionizing patient care or health care delivery”), the new EAP program seeks to expedite access to transformative new medical devices. The Agency will use a variety of avenues to do so: earlier and more interactive engagement with FDA staff (including senior management), a collaboratively developed plan for collecting data to support approval, a priority review, and many others. Notably, the new program is not just about reducing the time for the premarket review – it also seeks to reduce the time associated with product development. Devices must meet the following criteria:

1) Be intended to treat or diagnose a life-threatening or irreversibly-debilitating disease or condition (certainly the case for type 1 diabetes)

2) Represent one of the following: (i) no approved alternative treatment/diagnostic exists; or (ii) a breakthrough technology that provides a clinically meaningful advantage over existing technology; or (iii) offers a significant, clinically meaningful advantage over existing approved alternatives; or (iv) availability is in the patient’s best interest. 

3) Have an acceptable data development plan that has been approved by the FDA

Along with the announcement, the FDA has released two draft guidance documents, posted here (program background; 35 pages) and here (balancing pre-market and post-market data collection; 14 pages). The Agency is seeking public comments by July 23, 2014. Broadly speaking, we are very excited to see these guidance documents, and it’s clear that a lot of thought has gone into them. This appears to be a significant effort by FDA to respond to critiques from patient advocates, industry, and Congress that current regulatory practices are restricting the availability of life-saving/changing products for patients. Certainly, the guidance documents are filled with encouraging efforts to change this – use of surrogate/intermediate endpoints to allow for smaller trials, a shift to greater use of post-market data collection, reduced premarket manufacturing inspections, and involvement of senior FDA management. The devil will of course be in the details, but this appears like the real deal after a first read-through.  

The key question that comes to mind is whether any diabetes devices will qualify. We believe an overnight closed loop system with daytime hypo-hyper minimization would absolutely meet the “breakthrough” criteria (2ii) outlined above, as it would be “breakthrough” and a very meaningful advantage over current therapy. On the other hand, we do not expect marginal improvements in CGM or another sensor-augmented pump would qualify. The gray zone is more challenging – would predictive low glucose suspend qualify? Would an insulin pump advisory system qualify? Would a better bolus wizard qualify? The program will obviously be subjective in nature, and we expect the FDA will designate devices on a case-by-case basis.

  • The new expedited pathway outlines several new approaches to speed PMA approvals for certain products: 1) creation of a Data Development Plan with agreed upon data needed for PMA approval; 2) early PMA approval coupled to mandated and expanded post-market surveillance studies; 3) senior FDA management involvement (the opposite of the current environment, where more junior members are typically used early on); 4) an FDA case manager to work with the company on the Data Development Plan; and 5) use of surrogate endpoints to allow for smaller pre-market trials/shorter follow-up.
  • We would emphasize the key words “sufficiently advanced,” “breakthrough,” and providing “a clinically meaningful advantage.” With this understanding, the FDA will likely be faced with an optimization problem – the technology offered would need to be advanced enough to qualify for the EAP, yet tested enough to offer a degree of certainty about its risk/benefit ratio. Based on that, two conclusions follow:
    • The EAP may make it possible to short-circuit the “traditional” path of closed-loop development, which called for a gradual sequence of small steps. This original framework was proposed by Dr. Aaron Kowalski and called for increasingly advanced automated insulin delivery – beginning with pump shutoff (threshold or predictive), moving to a hypo-hyper minimizer, then to 24-hour closed-loop control with insulin only, followed by multi-hormone control. The EAP may make it possible to jump in the middle of this sequence, as the first couple of steps would not qualify as “breakthrough,” and the later steps would not be tested enough to offer a convincing risk/benefit ratio.
    • The EAP could be particularly well suited to an overnight closed-loop system accompanied by daytime safety hypo-hyper minimization. This type of technology has been tested in a number of studies in the US and overseas, and has shown qualitative jumps in glucose control without too many technology complications – that strikes us as exactly the kind of optimization and risk/benefit the EAP program is tailored for. As David Panzirer (Trustee, Helmsley Charitable Trust) said in our recent interview, “My immediate aspiration is to try to automate basal rates overnight, so that people can sleep through the night without fear and without running high to avoid severe hypos. I would like to take nighttime from a period of extreme worry to a period of pristine blood glucose control. I wholeheartedly believe that this is an attainable goal that would have an absolutely dramatic effect people’s outcomes.”
  • We’d speculate that this program may have the most impact within companies, who may now feel slightly more comfortable and less daunted by the path to commercializing automated insulin delivery devices (should such devices ultimately qualify). Medtronic had to navigate an onerous regulatory pathway to get the MiniMed 530G to market, which included two clinical trials, three-plus years of negotiation with the FDA, and plenty of post-market data collection. While we hope the pathway to more automated systems proves less burdensome, closed-loop commercialization is still a huge area of uncertainty for industry, particularly given the required upfront investment, post-market liability concerns, and uncertain reimbursement environment. Over the past year, we’ve heard many call for industry players to step up more enthusiastically in getting automated insulin delivery devices to market (e.g., Dr. Aaron Kowalski at the 2013 NIH/FDA/JDRF Workshop, David Panzirer in our recent interview) – hopefully this new FDA program will encourage more movement on this front.
  • We’d note that the FDA’s artificial pancreas team already does some of the activities included in this program (e.g., expedited review and approval of IDE submissions from academic groups), but having a formalized process (once fully implemented) seems like it could only strengthen this trend.
  • In an effort to get devices to market sooner, the FDA may require less pre-market data to support approval. This is a big shift from the current environment, where there is a burden on companies to do pretty much everything in the premarket setting. Of course, this is a very tricky balance to strike, as the guidance document notes: “The right balance of premarket and postmarket data collection facilitates timely patient access to important new technology without undermining patient safety.” The guidance document does have encouraging (though very general) language on this front, highlighting that the FDA may accept greater pre-approval uncertainty regarding risk-benefit if certain conditions are met. A more favorable risk-benefit determination is likely if one of 11 criteria is met: (i) premarket data demonstrate that the probability of serious harm is low; (ii) postmarket patient exposure to the device prior to the required submission of postmarket data to FDA will be small; (iii) the device is non-implantable; and eight other criteria listed on page 17 here. Broadly speaking, the criteria read encouragingly with automated insulin delivery devices in mind.
    • Postmarket data may be relied upon more in cases where FDA has “robust experience with the device type, typically a more mature technology.” We assume this would be the case for insulin pumps and perhaps for CGM, though it’s not as clear if their combination (automated insulin delivery devices) would qualify. We assume not, given limited commercial experience.
  • A gating factor in this program’s success will almost certainly be FDA resources. We took particular note of the words “as resources permit” in the following statement: “As part of this EAP program, FDA intends to provide, as resources permit, more interactive communications during device development and more interactive review of Investigational Device Exemption (IDE) applications and PMA applications.” The program will also enlist senior FDA staff early on and include a case manager, which may limit the number of devices that the Agency can take on. We have said for a long time that the Agency is under-resourced, given all that it is asked to do. Attracting and retaining strong talent – especially in light of the “thankless nature” of the job – is also a big challenge. The program’s goal of early and frequent communication is a no-brainer, but only possible if the FDA has the bandwidth to take it on.
  • Companies must justify that their device meets the Expedited Access PMA (EAP) criteria outlined above – the guidance document has explicit instructions (page 30 here) on what companies must discuss in their applications. In most cases, sponsors are expected to apply for an EAP designation prior to commencement of an IDE pivotal study. Consideration for the EAP program is not required and is entirely voluntary on the part of the sponsor. Notably, if the FDA believes that a device could be eligible, and the sponsor has not yet requested EAP Designation, the FDA intends to inform the sponsor of the program.
  • Notably, the FDA is willing to entertain multiple requests for expedited products from different manufacturers. This struck us as impressive flexibility and particularly important for closed-loop technologies, where there are likely to be a number of devices from different companies on the table under development at the same time.
  • The press release mentions that “most devices” that enter the new program “will be in the pre-clinical trial phase” – that could imply that much earlier stage products could stand to gain the most from the new program. Although the FDA has accepted the use of in silico modeling in place of many pre-clinical studies for the artificial pancreas, we believe this will still shorten development times for artificial pancreas or automated insulin delivery devices.

 

Close Concerns Questions

Q: Will automated insulin delivery devices qualify under this program?

Q: Where will the FDA draw the line on “breakthrough” technology?

Q: How much of a timing difference could this program make on ultimately getting devices to patients? 

Q: How much will constrained FDA resources limit the success of the program?

Q: Will this program incentivize companies to more aggressively bring automated insulin delivery devices to market?

 

--by Adam Brown and Kelly Close