Pragmatic trial design
LIBERATES was a prospective, multicenter, parallel group, randomized controlled trial, explained the study’s statistician Deborah Stocken, PhD, professor of clinical trials research at the University of Leeds. There was “limited ability to blind the interventions,” so it was an open-label design.
“The patient population in LIBERATES was kept as inclusive and as pragmatic as possible to ensure that the results at the end of the trial are generalizable,” said Dr. Stocken. Patients with type 2 diabetes were recruited within 5 days of hospital admission for ACS, which could include both ST- and non-ST elevation MI. In all, 141 of a calculated 150 patients that would be needed were recruited and randomized to the flash monitoring (69) or SMBG (72) arm.
Dr. Stocken noted that early in the recruitment phase, the trials oversight committee recommended that Bayesian methodology should be used as the most robust analytical approach.
“Essentially, a Bayesian approach would avoid a hypothesis test, and instead would provide a probability of there being a treatment benefit for continuous monitoring. And if this probability was high enough, this would warrant further research in the phase 3 setting,” Dr. Stocken said.
What else was shown?
“We had a number of prespecified secondary endpoints, which to me are equally important,” said Ramzi Ajjan, MD, MMed.Sci, PhD, associate professor and consultant in diabetes and endocrinology at Leeds University and Leeds Teaching Hospitals Trust.
Among these was the TIR at days 16-30, which showed a 90-minute increase per day in favor of flash monitoring over SMBG. This “seems to be driven by those who are an insulin,” Dr. Ajjan said, adding that “you get almost a 3-hour increase in time in range in people who are on insulin at baseline, and you don’t see that in people who are on sulfonylurea.”
Conversely, sulfonylurea treatment seemed to drive the reduction in the time spent in hypoglycemia defined as 3.9 mmol/L (70 g/dL) at 3 months. For the whole group, there was a 1.3-hour reduction in hypoglycemia per day with flash monitoring versus SMBG, which increased to 2 hours for those on sulfonylureas.
There also was a “pattern of reduction” in time spent in hypoglycemia defined as less than 3.0 mmol/L (54 g/dL) both early on and becoming more pronounced with time.
“Flash glucose monitoring is associated with higher treatment satisfaction score, compared with SMBG,” Dr. Ajjan said.
Although A1c dropped in both groups to a similar extent, he noted that the reduction seen in the flash monitoring group was associated with a decrease in hypoglycemia.
There was a huge amount of data collected during the trial and there are many more analyses that could be done, Dr. Ajjan said. The outcome of those may determine whether a phase 3 trial is likely, assuming sponsorship can be secured.
The LIBERATES Trial was funded by grants from the UK National Institute for Health Research and Abbott Diabetes Care. None of the investigators were additionally compensated for their work within the trial. Dr. Stocken had no disclosures in relation to this trial. Dr. Ajjan has received research funding and other financial support from Abbott, Bayer, Eli Lilly, Johnson & Johnson, and Novo Nordisk.
SOURCE: Ajjan R et al. EASD 2020. S11 – The LIBERATES Trial.