FDA advisory panel recommends CGM as replacement for SMBG

An FDA advisory panel recommended a change in intended use for Dexcom’s G5 Mobile Continuous Glucose Monitoring System, a move that, if formally approved, would allow patients with diabetes to use the device to make diabetes treatment decisions without confirmation from a blood glucose meter.

The Clinical Chemistry and Clinical Toxicology Devices Panel of the Medical Devices Advisory Committee voted 8-2 that there is “reasonable assurance” that the Dexcom G5 Continuous Glucose Monitoring System (CGM) is safe for the proposed indications for use, and 9-1 that the system is effective for the proposed indication. In addition, the panel voted 8-2 that the benefits of the proposed indications outweigh any risks.

“This recommendation is a big milestone for people with diabetes,” Kevin Sayer, DexCom president and chief executive officer, said in a press release. “The diabetes community turned out in force to support this decision. We commend the FDA for bringing this important subject into a public forum, and thank the panel members, as well as the public speakers, for their willingness to participate. We look forward to continued positive discussions with the FDA as we seek the agency’s approval of our application.”

Finger stick alternative

Currently, the Dexcom G5 CGM system is indicated for detecting trends and tracking patterns in blood glucose in patients with diabetes as an adjunctive device to complement — but not replace — information obtained from standard home glucose monitoring devices. Dexcom is seeking to modify the current indication to state that the G5 CGM system is designed to replace finger-stick blood glucose testing for diabetes treatment decisions (a non-adjunctive indication), including determining an insulin dose, ingesting carbohydrates, or assessing when to wait before dosing or eating.

The change to the indication will not require changes to the system components, and the system will still require calibration with a finger stick taken by self-monitoring of blood glucose (SMBG) every 12 hours, according to meeting briefing papers prepared by Dexcom. The primary modification will be in the Instructions for Use.

Following discussions with the FDA, Dexcom conducted two simulations: the first evaluated the safety and effectiveness of CGM relative to SMBG in patients with type 1 diabetes over a 2-week period; the second evaluated the risks of hypo- and hyperglycemia in type 1 diabetes with meal-time insulin decisions based on CGM vs. SMBG. A Human Factors Usability Study evaluated the revised Instructions for Use.

Safety concerns

The reviewers noted that human-factors related risks of non-adjunctive CGM use may vary based on user group and their level of training and requested the panel’s input regarding Dexcom’s role in providing appropriate labeling or training on using trend information.

Voting member David W. Cooke, MD, associate professor of pediatrics at Johns Hopkins University School of Medicine, who voted against the question of safety but in favor of the question of efficacy, said a lack of clinical trial data in a real-world setting was his greatest concern.

“The anecdotal experience, the modeling experience are reassuring, but not reassuring enough without clinical trial data,” said Cooke, who added that the meter accuracy is the driving force for efficacy. “My concern for safety outweighs the efficacy benefit. A clinical trial powered to ensure reasonable safety, non-inferiority against the current standard of care, would be what I look for before being comfortable with the safety.”

Voting member Lisa M. McShane, PhD, chief of the biostatistics branch of the NIH National Cancer Institute, also expressed concern for a lack of data and requested that a postmarketing study be conducted.

“I would like to believe that the information collected by these devices could, in fact, be brought together and really give us tons of valuable information,” McShane said after the votes. “I fully appreciate the feelings of people who think it’s terrible I voted ‘no’ to all of these [questions] ... We just need to go a little further before we give it approval.”

Several other voting members also expressed their concern regarding the lack of randomized controlled trial data, but ultimately voted to recommend the change in use indication.

Real-world use

“I can sit here and discuss and propose many different ways to do more studies ... but the horse has been out of the barn for years,” George Grunberger, MD, FACP, FACE, president of the American Association of Clinical Endocrinologists, said during panel deliberations before the votes. “This meeting should have taken place 5 years ago. So do we discuss the optimal trial designs, which would satisfy the scientists, or do we go with the flow?”

“We’re dosing off of our CGM data,” Anna McCollister-Slipp, a non-voting patient representative with type 1 diabetes, said during panel deliberations. “That’s the reality of clinical practice today. The critical question we need to address is whether or not we want to company to give us guidance. Right now, they can’t. They have to pretend this doesn’t exist. I don’t think that is helpful to anybody.”

Addressing the panel during the open public hearing, a steady stream of patients, health care providers and speakers on behalf of patient advocacy groups cited data suggesting that many patients using CGM technology already make treatment decisions without consulting a blood glucose meter.

“You can’t think of another disease in which the person with the disease is doing laboratory-based science ... and then deciding on a therapeutic dose [of insulin] with a very narrow therapeutic window,” said Robert E. Ratner, MD, FACP, FACE, chief scientific and medical officer for the American Diabetes Association. “The question of non-adjunctive use, with all due respect, is silly. The question is whether the education and the access is available to allow them to do it well.”

Michael Dempsey MD, FACE, vice president of the mid-Atlantic chapter of AACE, said the organization supports the action to change the intended use, and that a wider adoption of CGM technology that would likely follow could lead to improved diabetes outcomes for patients.

An approved change in use “would be the first step to provide much-needed coverage ... for Medicare patients on intensive insulin regimens,” Dempsey said while addressing the panel.

Wider implications

In an interview following the votes, Janet B. McGill, MD, FACE, professor of medicine at Washington University School of Medicine in St. Louis and an investigator in Dexcom’s ongoing Diamond study, called the panel’s recommendation a crucial decision to move diabetes technology forward.

“This decision will help future efforts in the development of the bionic pancreas,” McGill told Endocrine Today. “It will also help our efforts to get Medicare to approve CGM for type 1 diabetes patients who are now on Medicare. Many of these patients are at high risk for hypoglycemia, either because of age, insulin sensitivity, kidney failure or other disability.”

“The committee made the right decision,” she added.

The panel’s recommendations are nonbinding, although the FDA often follows its suggestions. – by Regina Schaffer

Disclosure: Dempsey and Ratner report no relevant financial disclosures. McGill reports travel expenses for the meeting funded by Dexcom.