Firas El-Khatib, PhD, Coinvestigator and Senior Research Scientist of the Bionic Pancreas Team, Boston University, addressed home and outpatient use of a bihormonal bionic pancreas, by patients with type 1 diabetes at the 76th Scientific Sessions of the American Diabetes Association, June 10-14, in New Orleans.
An 11-Day Study of Home Use of a Bihormonal Bionic Pancreas
Dr. El-Khatib and colleagues began by stating that the safety and effectiveness of continuous, multiday, automated glycemic control using insulin and glucagon have not been tested in a home-use setting. They evaluated a bihormonal pump vs conventional insulin pump therapy in adults with type 1 diabetes living at home and performing their normal activities without restrictions on their diet or exercise.
"The study was a natural next step for us," Dr. El-Khatib said, "to test the bionic pancreas in a real home setting, where volunteers drove to work and slept at home, with no restrictions on their activities or diet, and with only minimal real-time data monitoring." They performed a random-order crossover study that compared glycemic regulation with the bihormonal pump vs a conventional insulin pump over 11 days each.
The bihormonal bionic pancreas was associated with a reduction in mean glucose level (162 ± 29 vs 141 ± 10 mg/dL, P<.0001) and reduced percent time <60 mg/dL (1.9 ± 1.7 vs 0.6 ±0.6%, P<.0001) vs the conventional insulin pump. Over the entire study period, the bihormonal bionic pancreas was associated with a reduction in the mean number of symptomatic hypoglycemia events per day (0.59 ± 0.56 vs 0.90 ± 0.64 events per day, P=.023). The mean total daily dose of insulin delivered by the bihormonal bionic pancreas varied widely between participants and was ~6% greater than with the conventional insulin pump (0.66 ± 0.15 vs 0.62 ± 0.18 units per kilogram daily, P=.01).
"The study results were very positive," Dr. El-Khatib said, "and in addition to paving the way to future longer home studies, the results help make the case for removing real-time data monitoring in future studies."
Bihormonal Bionic Pancreas Outcomes
Dr. El-Khatib concluded by explaining that the bihormonal bionic pancreas was associated with a reduction in the mean glucose level and in hypoglycemia vs the conventional insulin pump in adults with type 1 diabetes living at home and participating in their normal daily activities. "The bihormonal bionic pancreas has been validated," he said. "And the results effectively pave the way toward final studies to qualify the bionic pancreas in the way it will be used once it becomes available."
Dr. El-Khatib and colleagues explained that the bihormonal bionic pancreas was shown to regulate glucose effectively and with minimal hypoglycemia. His team hypothesized that an insulin-only configuration of the device using the same insulin dosing algorithm as the bihormonal device could also achieve effective glycemic control, but that a higher glucose target than the 100 mg/dL value used in previous bihormonal bionic pancreas trials could be used to minimize hypoglycemia.
"The study was intended," Dr. El-Khatib noted, "as a first test of the efficacy and safety of our insulin-only bionic pancreas in a real home setting with no restrictions on activities or diet. As such, a conservatively high glucose target was first adopted, and the study included a comparison with the bihormonal bionic pancreas at that same glucose target."
In a random-order crossover study, the investigators compared two insulin-only configurations of the device (glucose targets of 130 and 145 mg/dL) with a bihormonal configuration (glucose target 130 mg/dL), and usual care (patient managed insulin pump therapy). Subjects went about their daily routines with no limitations on their diet or exercise during each 3-day test period. Coprimary outcomes were the mean continuous glucose monitor glucose level and time <60 mg/dL during days 2-3 of each arm.
No significant difference in aggregate mean continuous glucose monitored glucose level and time <60 mg/dL was observed between the bihormonal 130 mg/dL target configuration (156 ± 12 mg/dL, 0.6 ± 1.0%); the insulin-only 130 mg/dL target configuration (161 ± 17 mg/dL, 0.8 ± 1.4%); and usual-care (158 ± 31 mg/dL, 1.4 ± 2.6%).
The mean continuous glucose monitored glucose level for the insulin-only 145 mg/dL target configuration was higher (174 ± 23 mg/dL) than both 130 mg/dL target configurations and usual care (P=.0014, <.0001, and =.034, respectively) with no significant reduction in time <60 mg/dL (1.0 ± 1.5%).
No significant differences in insulin total doses were observed between the two configurations. The insulin-only and bihormonal configurations of the device with targets of 130 mg/dL both performed similarly to usual care with low levels of hypoglycemia.
The mean glucose level was increased with a glucose target of 145 mg/dL without a compensatory decrease in hypoglycemia. Dr. El-Khatib and colleagues recommended evaluation of the insulin-only bionic pancreas configurations with glucose targets lower than 130 mg/dL.
"The study," Dr. El-Khatib concluded, was of great significance as it signified the beginning of testing and ultimately qualifying the insulin-only configuration of our system, which is likely to be released on the market before the bihormonal system."
He continued, "Additionally, even when the bihormonal bionic pancreas subsequently becomes available, its everyday use will inevitably include periods where glucagon administration is (temporarily) not available and where an insulin-only system is effectively in place."
"Finally," Dr. El-Khatib added, "in addition to setting the stage for future trials to ultimately qualify an insulin-only system, the study served to emphasize the additional safety that a bihormonal bionic pancreas brings forth."
"The results motivate repeating the comparison between insulin-only and bihormonal configurations at lower glucose targets, which is ongoing. The results also suggest that our insulin-only system would be effective and safe to use at lower glucose targets, which we are also investigating."
Elkhatib F, Buckingham BA, Buse JB, Harlan DM, et al. 77-OR/77 Home Use of a Bihormonal Bionic Pancreas vs Conventional Insulin Pump Therapy in Adult with Type 1 Diabetes: A Multicenter, Randomized Clinical Trial. American Diabetes Association's 76th Scientific Sessions, New Orleans, LA, June 10-14, 2016.
F. Elkhatib: None. B.A. Buckingham: Research Support; Author; Medtronic, Inc. J.B. Buse: Consultant; Author; PhaseBio Pharmaceuticals, Inc.. Research Support; Author; AstraZeneca, Boehringer Ingelheim GmbH, Eli Lilly and Company, GI Dynamics, Inc., Johnson & Johnson, Lexicon Pharmaceuticals, Inc., Novo Nordisk Inc, Orexigen Therapeutics, Inc., Bristol-Myers Squibb Company. Stock/Shareholder; Author; Phase Bio. Other Relationship; Author; AstraZeneca, Dance Biopharm, Eli Lilly and Company, Elcelyx Therapeutics, Inc., GI Dynamics, Inc., Lexicon Pharmaceuticals, Inc., Merck & Co., Inc, Metavention, Novo Nordisk Inc, Orexigen Therapeutics, Inc., vTv Therapeutics. D.M. Harlan: None. K. Magyar: Other Relationship; Author; Tandem Diabetes Care, Inc. T.T. Ly: Research Support; Author; Medtronic, Inc. M. Kirkman: Research Support; Author; Novo Nordisk Inc. S. Malkani: None. M.J. Thompson: None. J. Lock: None. L. Ekhlaspour: None. P. Clinton: Consultant; Author; Johnson & Johnson Diabetes Institute. J. Diner: None. M. Dezube: None. C. Hartigan: None. C. Balliro: None. R. Selagamsetty: None. A. Esmaeili: None. M. Sinha: None. M. Hillard: None. D. Mondesir: None. E.R. Damiano: None. S.J. Russell: Advisory Panel; Author; Tandem Diabetes Care, Inc., Campanion Medical. Consultant; Author; Sanofi U.S.. Other Relationship; Author; Sanofi U.S., Dexcom, Inc., Tandem Diabetes Care, Inc., Eli Lilly and Company, Abbott Diabetes Care Inc., Insulet Corporation, Medtronic MiniMed, Inc., International Biomedical.