ADA 2020
Highlights from the 80th Annual American Diabetes Association (ADA) Virtual Scientific Sessions.
- Real-world data find few differences in cardiovascular outcomes between SGLT2 inhibitors and GLP1 agonists
- Diabetes-related stress and A1c improved with visits to virtual diabetes clinic
- BEST: Bexagliflozin led to significant reductions in HbA1c and cardiovascular risk factors in patients with type 2 diabetes
- DPPOS shows long-term diabetes prevention with lifestyle and metformin intervention; preventing diabetes lowers risk of vascular complications
- Ketoacidosis rate reduced dramatically after initiation of sensor-based glucose monitoring
Real-world data find few differences in cardiovascular outcomes between SGLT2 inhibitors and GLP1 agonists
Presenter: Insiya Poonawalla, MD
Real-world data show higher treatment persistence with the sodium-glucose contransporter-2 (SGLT2) inhibitors compared with glucagon-like peptide-1 (GLP1 agonists) but equivalent cardiovascular outcomes between the two classes, with the exception of development of heart failure, which favors the SGLT2 inhibitors.
Investigators led by Insiya Poonawalla, MD, compared a series of outcomes between new users of SGLT2 inhibitors and GLP1 agonists, using medical and pharmacy claims data from the Humana Research Database, which includes data for approximately 13 million members.
The U.S. Food and Drug Administration has encouraged cardiovascular outcomes trials in an effort to enhance informed decision-making in the management of diabetes, said Dr. Poonawalla, a researcher at Humana Healthcare Research in Louisville, KY. The randomized controlled clinical trials show “a potential cardiovascular benefit with some of the agents in both of these therapeutic classes; however, population-based real-world evidence comparing these two classes on cardiovascular outcomes is limited, and also only focused on a few agents in both of these classes,” she said. Utilization and cost outcomes are even less studied.
Eligible patients were 18 to 89 years old, had a new claim for an SGLT2 inhibitor or GLP1 agonist from January 2015 to June 2017, had continuous enrollment pre- and post-index date for at least 12 months, and were treatment naïve to SGLT2 inhibitors or GLP1 agonists in the 6 months prior to the index date.
More than 13,000 patients met the inclusion criteria. Propensity score matching (1:1) was used to balance baseline characteristics between 5,507 patients in each treatment group.
The composite primary cardiovascular outcome of myocardial infarction or stroke or mortality occurred in 23.0% of the group taking GLP1 agonists versus 22.6% of those taking SGLT2 inhibitors. The composite secondary outcome of development of heart failure or death was achieved by 20.0% of the GLP1 agonist group compared with 17.9% of the SGLT2 inhibitor group (P = .005), driven by a significant reduction in occurrence of heart failure in patients taking SGLT2 inhibitors (18.8% vs. 16.3%; P < .001). Other individual components of the primary and secondary composite cardiovascular outcomes were not significantly different between the two treatment groups.
When accounting for time to event in an analytic model, no significant difference in cardiovascular outcomes was found between treatment classes when stratified by the presence or absence of established atherosclerotic cardiovascular disease at baseline.
Rates of treatment persistence were 40.4% and 36.1% (P < .001) among index users of SGLT2 inhibitors and GLP1 agonists, respectively. Treatment discontinuation was 15% more likely in those with index GLP1 agonist use (P < .001). The rate of inpatient stays was 11.9% in the SGLT2 inhibitor group versus 14.4% in the GLP1 agonist group (P < .001), and the rate of emergency department visits was 23.5% vs. 27.4% (P < .001), respectively.
Using the patients taking index SGLT2 inhibitors as the reference group, the mean difference in total per-person per-month (PPPM) cost was $179 higher (P < .001), the mean difference in medical PPPM cost was $70 higher (P < .001), and the mean difference in pharmacy PPPM cost was $108 higher (P < 0.001) in the GLP1 agonist group.
In addition to limitations characteristic of observational studies, Dr. Poonawalla cited lack of inclusion of newer agents in these classes (ie, semaglutide, ertugliflozin) due to a potential small sample and lack of sufficient follow-up.
Link to abstract: 36-OR: Comparative Effectiveness of SGLT2 Inhibitors vs. GLP-1 Agonists
Insiya Poonawalla, MD, reported nothing to disclose.
Diabetes-related stress and A1c improved with visits to virtual diabetes clinic
Presenters: William H. Polonsky, PhD, and Ronald F. Dixon, MD
Participation in a virtual diabetes clinic that incorporates a telehealth program to support people with type 2 diabetes mellitus (T2DM) leads to a reduction in diabetes-related distress and an improvement in hemoglobin A1c (A1c) level. Users of real-time continuous glucose monitoring (rtCGM) had a significantly greater reduction in distress than nonusers.
“In total, what we think these data tell us is that ongoing individualized lifestyle and clinic support. . .may be key contributors to explaining these reductions in diabetes distress,” said William H. Polonsky, PhD, from the Behavioral Diabetes Institute of the University of California San Diego.
Individuals with diabetes often feel overwhelmed by the numerous and often complex tasks required for daily self-care regimens, leading to diabetes-related stress, and more than one third of adults with T2DM meet criteria for elevated levels of distress. “Diabetes-related stress is a critical element of diabetes-specific quality of life,” he said. “Not only is it common, it’s linked to poor self-management adherence, suboptimal glycemic control, and a higher prevalence of complications.”
The retrospective exploratory analysis evaluated the change in diabetes distress among participants of the Onduo Virtual Diabetes Clinic (VDC) who reported at least moderate distress (score ≥ 2.0) on the diabetes distress scale (DDS17) at enrollment. The VDC is a telehealth program that incorporates a health coach, mobile app, connected blood glucose meter, rtCGM devices, and the opportunity for live consultations with board-certified endocrinologists. The program is intended to support people with T2DM in the primary care setting between office visits.
Of the 735 participants who had an initial DDS17 score ≥ 2.0, 228 completed a follow-up survey at ≥ 3 months. At enrollment, survey completers had a mean age of 51.8 years, 73.2% were women, mean body mass index was 36.8 kg/m2, and their mean DDS17 overall score was 3.0. During program participation, 94.7% used a connected blood glucose meter, and 33.8% used rtCGM intermittently, with both values being significantly higher (P < .001) than the survey noncompleters. Mean follow-up was 179.4 days.
The overall DDS17 score declined from a mean of 3.0 at baseline to 2.5 at follow-up (P < .001). Significant reductions in stress were recorded for each of the DDS17 subscales: regimen-related distress (P < .001), emotional distress (P < .001), interpersonal distress (P = .002), and physician-related distress (P = .006).
Greater reductions in the overall DDS17 score (–0.7 vs. –0.4; P = .012) and regimen-related distress (–1.3 vs. –0.7; P < .001) were reported by the 77 participants who were prescribed and used intermittent rtCGM compared with the 171 nonusers.
On multivariate linear regression, the DDS17 score (P < .001), older age (P = .013), and rtCGM use (P = .048) were significant predictors of improvement in DDS17.
Participation in the VDC for 6 months was also associated with significant improvement in A1c in all participants not meeting the American Diabetes Association treatment target, with greater improvement in those with intermittent rtCGM use, reported Ronald F. Dixon, MD, president and CEO of the Onduo Physicians Group PC.
Included in this analysis were 213 rtCGM users and 399 nonusers who participated in the VDC. Baseline A1c level was 7.8%, 33.0% were on insulin, and 23.5% were taking a sulfonylurea. After a mean follow-up of 6.1 months, A1c decreased significantly in CGM and non-CGM groups by 0.9% and 0.4%, respectively (both P < .001). When stratified by CGM use, among participants with baseline A1c > 9.0%, the change in A1c was –3.3% in the rtCGM users versus 1.7% in the nonusers (P < .001). The difference in the change in A1c level between users and nonusers of rtCGM was also significant for baseline A1c of 8.0% to 9.0% (P = .004) and baseline A1c of 7.0% to < 8.0% (P = .023). There was no significant difference in the change in those with baseline A1c < 7.0%.
Links to abstracts: Change in A1C with and without Intermittent Use of CGM in Adults with Type 2 Diabetes Participating in the Onduo Virtual Diabetes Clinic and Participation in a Virtual Diabetes Clinic and Reduced Diabetes-Related Distress in Individuals with Type 2 Diabetes
William H. Polonsky, PhD, disclosed advisory panel fees from Intarcia Therapeutics and Roche Diabetes Care; and consultant fees from Abbott, Dexcom, Inc., Eli Lilly and Company, Insulet Corporation, Novo Nordisk Inc., Onduo, Sanofi US, and Xeris Pharmaceuticals, Inc. Ronald F. Dixon, MD, is president and CEO of Onduo Physicians Group PC.
BEST: Bexagliflozin led to significant reductions in HbA1c and cardiovascular risk factors in patients with type 2 diabetes
Presenter: John J.V. McMurray, MD
The SGLT2 inhibitor bexagliflozin reduced HbA1c in patients at high risk of cardiovascular (CV) events and had beneficial effects on weight and blood pressure in the placebo-controlled multinational BEST trial. Bexagliflozin also proved to be noninferior to placebo in a prespecified analysis of CV safety, said John J.V. McMurray, MD, professor of medical cardiology and deputy director, BHF Cardiovascular Research Center, University of Glasgow, Scotland, UK.
In BEST, the efficacy and safety of bexagliflozin in controlling hyperglycemia, weight, and blood pressure was examined in patients with type 2 diabetes mellitus (T2DM) and established CV disease or multiple CV risk factors.
A total of 1,700 men and women > 40 years old with T2DM and a baseline HbA1c of 7.0% to 11.0% on stable medications were enrolled into BEST. Patients were enrolled into one of three groups: (1) history of atherosclerotic cardiovascular disease (n = 1,065), (2) history of New York Heart Association class II/III heart failure (HF; n = 246), or (3) age ≥ 55 years and ≥ 2 cardiovascular risk factors (n = 389). They were randomized 2:1 to bexagliflozin, 20 mg once daily, or placebo. Some 96% of the patients completed 24 weeks, at which time the primary endpoint of change in HbA1c was assessed. The median follow-up for assessment of CV safety was 123 weeks.
Approximately 30% of the overall cohort were women and 77% were white. Mean age of the study population was 64.4 years, mean body mass index (BMI) was 32.6 kg/m2, the mean systolic blood pressure (SBP) was 134 mm Hg, 19.6% had an estimated glomerular filtration rate < 60 mL/min/1.73 m2, 26.7% had a history of heart failure, and 69.2% had a history of coronary heart disease.
The mean HbA1c at baseline was 8.32% and the mean duration of diabetes at entry was 8.9 years. Neuropathy was present in 30.4%. The most common types of glucose-lowering medications were biguanides (76.6%), sulfonylureas (40.1%), and dipeptidyl peptidase-4 inhibitors (13.4%), and 53.1% of patients were taking insulin at baseline.
The placebo-corrected change in HbA1c at 24 weeks in patients randomized to bexagliflozin was –0.48% (P < .0001), despite nearly three times as many patients in the placebo group requiring rescue glucose-lowering therapy compared with the bexagliflozin group (19.6% vs. 7.1%). In the 834 patients on insulin at baseline, the placebo-corrected change in HbA1c at 24 weeks was 0.52% (P < .0001) in the bexagliflozin group.
The placebo-corrected decrease in body weight at 48 weeks in patients with baseline BMI ≥ 25 kg/m2 was 2.65 kg (P < .0001) in the bexagliflozin recipients.
The placebo-corrected decline in SBP at week 24 in participants with a baseline SBP ≥ 140 mm Hg was 2.96 mm Hg (P = .012) in those randomized to bexagliflozin. A reduction in SBP over the entire follow-up was also observed in the overall study population assigned to bexagliflozin versus placebo (P < .0001).
The hazard ratio (HR) for time to first hospitalization with worsening heart failure, an exploratory endpoint, was 0.63 in favor of bexagliflozin but did not achieve significance (95% CI, 0.34-1.18) because of the small number of events. Likewise, the HR for the time to the composite of CV death or HF hospitalization was 0.73 in favor of bexagliflozin, which was not significant because of the small number of events (95% CI, 0.46-1.15).
The proportion of patients with adverse events that led to treatment discontinuation was nearly identical in the two arms—8.4% in those randomized to bexagliflozin versus 8.5% in patients randomized to placebo. The only adverse event that occurred significantly more often in the bexagliflozin group versus the placebo group was genital mycotic infection (9.5% vs. 3.0%; P < .0001), “but the proportion of patients who had to stop study because of a genital mycotic infection was small; it was only 1.2% in the patients randomized to bexagliflozin,” said McMurray. No excess of urinary tract infection was observed in the bexagliflozin group, and only 0.9% of the bexagliflozin group versus 1.4% of the placebo group experienced serious hypoglycemia.
The composite CV outcome of CV death, myocardial infarction (MI), stroke, or unstable angina was assessed once 134 events occurred. This composite was numerically less in the bexagliflozin group (HR 0.77; 95% CI, 0.57-1.08), which met the criterion for noninferiority (P < .0001). Noninferiority was also established for a composite CV outcome that included CV death, MI, or stroke (P < .0001 using a 95% upper CI margin of 1.8; P = .0021 using a 95% CI margin of 1.3).
Link to abstract: 32-OR: The Bexagliflozin Efficacy and Safety Trial (BEST): A Randomized, Double-Blind, Placebo-Controlled, Phase IIII, Clinical Trial
John J.V. McMurray, MD, disclosed financial relationships with AbbVie Inc., Alnylam Pharmaceuticals, Amgen, AstraZeneca, Bayer Healthcare Pharmaceuticals Inc., Bristol-Myers Squibb, Cardurion, GlaxoSmithKline plc., Novartis Pharmaceuticals Corporation, and Theracos, Inc.
DPPOS shows long-term diabetes prevention with lifestyle and metformin intervention; preventing diabetes lowers risk of vascular complications
Presenters: David M. Nathan, MD; Mark Molitch, MD; Ronald B. Goldberg, MD; Brandy Heckman-Stoddard, PhD, MD; and Jose A. Luchsinger, MD, MPH
Long-term follow-up of 2,779 people enrolled in the Diabetes Prevention Program Outcomes Study (DPPOS) indicates a continued significant reduction in the participants’ risk of developing type 2 diabetes mellitus (T2DM) through intensive lifestyle intervention or use of metformin.
The DPPOS is the long-term follow-up of the Diabetes Prevention Program (DPP), a multicenter trial conducted from 1996 to 2001, in which lifestyle intervention aimed at achieving weight loss reduced the risk of development of T2DM by 58% and metformin use reduced the risk by 31%, both compared with placebo, after an average of 3 years, in persons considered to have pre-diabetes. The intensive lifestyle intervention was particularly effective in individuals ≥60 years old (71% reduction in risk), whereas metformin was particularly effective in younger, more obese patients (54% reduction in risk).
At the conclusion of the DPP, 88% of the surviving 3,234 participants enrolled in the DPPOS, including the DPP participants who had developed T2DM and those who still had not developed diabetes.
After an average of 22 years of study, 75% of the DPP participants who are still alive have continued to be evaluated, including those who have developed diabetes and those who have not. Some 47% of the total cohort and 61% of those who have developed T2DM are still on metformin. As of 2020, the mean age of the participants is 72 years, mean HbA1c level is 6.4%, mean systolic blood pressure is 124 mm Hg, mean level of low-density lipoprotein cholesterol is 93 mg/dL, and 56% are on statin treatment.
An overview of the latest findings from DPPOS was provided by DPPOS chair David M. Nathan, MD, director of the Clinical Research Center and Diabetes Center at Massachusetts General Hospital and professor of medicine at Harvard Medical School, Boston.
“The current results indicate that prevention of type 2 diabetes is possible and has important clinical benefits. The long-term benefits of the two DPP interventions on diabetes development, still present many years after they were started, are a testament to the power of these interventions and reinforce their importance in the reduction of diabetes,” said Dr. Nathan, in a statement prepared by the ADA.
New findings from DPPOS:
- Prevention effects in the original lifestyle and metformin treatment groups remain at 22 years, with a 25% and 18% reduction in the risk of developing T2DM, respectively, compared with the original placebo group, as presented by Dr. Nathan.
- The median time to T2DM development was delayed in the experimental arms: 9.5 years in the placebo arm versus 12 years in the metformin arm and 14 years in the intensive lifestyle arm.
- Among all participants and those with diabetes, HbA1c levels were lower in the intervention groups compared with placebo.
- Participants who did not develop diabetes had a significant 57% and 37% lower risk of developing retinopathy and kidney disease, respectively, and a 39% lower risk of major cardiovascular (CV) disease endpoints than those who developed diabetes, as presented by Mark Molitch, MD, professor emeritus of medicine (endocrinology), Northwestern University Feinberg School of Medicine, Chicago.
- The original group assignment did not affect the ultimate development of retinopathy over a mean follow-up of 20 years. There was also no significant benefit to lifestyle intervention or metformin on CV disease endpoints or the development of kidney disease, although favorable trends were observed with metformin on stroke and CV events in the subgroup of people who started the study before age 45. “Since avoidance of diabetes was associated with a lower risk for cardiovascular disease, the absence of benefit from the individual interventions is perplexing,” said Ronald B. Goldberg, MD, professor of medicine, Division of Endocrinology, Diabetes and Metabolism, University of Miami Miller School of Medicine.
- Diabetes status (prediabetes vs. diabetes) did not affect cancer incidence. Metformin was associated with a nonsignificant 12% lower risk of cancer compared with placebo. “If metformin does affect cancer risk, the effect is smaller than epidemiologic studies initially suggested, which would require either a larger study or longer follow-up to determine,” said Brandy Heckman-Stoddard, PhD, MD, chief, Breast and Gynecologic Cancer Research Group, Division of Cancer Prevention, National Institutes of Health.
- The intensive lifestyle intervention group had a long-term reduction in the development of frailty, as presented by Jose A. Luchsinger-Stuart, MD, MPH, professor of medicine and epidemiology, Columbia University Medical Center, New York City.
- The only long-term negative effect observed with either intervention was a modest increase in kidney disease with metformin, which appeared only in the oldest group of participants.
David M. Nathan, MD; Ronald B. Goldberg, MD; and Brandy Heckman-Stoddard, PhD, MD had nothing to disclose. Mark Molitch, MD, disclosed advisory panel fees from Janssen Pharmaceuticals, Inc., Merck & Co., and Inc., Pfizer Inc.; consultant fees from AstraZeneca; and research support from Novartis Pharmaceuticals Corporation and Novo Nordisk Inc. Jose A. Luchsinger-Stuart, MD, MPH, disclosed consultant fees from vTv
Ketoacidosis rate reduced dramatically after initiation of sensor-based glucose monitoring
Presenter: Ronan Roussel, MD, PhD
A dramatic reduction in the rate of ketoacidosis-related hospitalizations occurred in France in patients with type 1 or type 2 diabetes who started using a sensor-based glucose monitoring system (FreeStyle Libre), especially among those patients who previously reported a very low rate of self-monitoring of blood glucose.
The retrospective study, known as RELIEF, is the largest study to date on the impact of the FreeStyle Libre system on rates of hospitalization for diabetic ketoacidosis (DKA), said lead investigator Ronan Roussel, MD, PhD, chief of the endocrinology, diabetes, and nutrition department at Hôpital Bichat, Fédération de Diabétologie, AP-HP, Paris.
“We think that it is plausible that the use of the FreeStyle Libre system allowed, due to its convenience, people to detect and limit persistent hyperglycemia, and subsequently limit the occurrence of ketoacidosis,” he said. “This analysis has significant implications for patient-centered clinical care in diabetes and also for long-term health economic outcomes in the treatment of diabetes at a national level.”
Recent analysis of the Nationwide Emergency Department Sample and National Inpatient Sample from 2006 to 2015 shows that the rates of DKA and hospital admissions for hyperglycemic crises are increasing in the U.S. DKA is generally preventable through close glucose monitoring.
“This is a clinical issue but also the increased burden of hospitalization is associated with significant health care utilization and expenditure in several countries,” said Dr. Roussel.
The Freestyle Libre continuous glucose monitoring system is designed to overcome the pain and inconvenience associated with both strip and sensor-based glucose monitoring. Its use has been reimbursed in France since mid-2017 for patients requiring multiple daily insulin injections.
The investigators used the French national claims database to examine the risk of hospitalization related to DKA before and after access to the FreeStyle Libre system in France in patients with type 1 (T1DM) or type 2 diabetes mellitus (T2DM). The study included 74,076 patients who used the system between August 1, 2017, and December 31, 2017, and who had 1 full year of follow-up.
Eligible patients were then classified according to their use of self-monitoring blood glucose (SMBG) test strips in the 12 months before initiation of FreeStyle Libre.
The hospitalization rate for DKA fell by 52% among the cohort with T1DM and by 47% in the cohort with T2DM following initiation of FreeStyle Libre compared with the 12 months prior to initiation. The reduction was most marked in the subpopulation of patients who were not compliant with SMBG in the year prior to the FreeStyle Libre system—a 60% reduction for patients with T1DM and a 51% reduction for patients with T2DM. Patients with the most frequent use of SMBG testing (≥ 5 blood glucose test strips reimbursed per day) also experienced marked reductions in hospitalizations (59% for T1DM and 52% for T2DM).
The reduction in DKA-associated hospitalization rates was observed in patients treated with multiple daily injections (rates: 2.58% in the 12 months before FreeStyle Libre initiation vs. 1.8% in the 12 months after initiation) as well as those who were treated with continuous subcutaneous insulin infusion (2.89% vs. 1.32%).
A health economic analysis from this study will be reported in the future, said Dr. Roussel.