SAN FRANCISCO – SGLT2 inhibitors did a better job than GLP-1 receptor agonists at preventing heart failure hospitalizations in elderly patients with type 2 diabetes, but at the cost of more strokes, myocardial infarctions, and deaths among those without preexisting cardiovascular disease, according to Harvard University investigators.
Using Medicare claims data and propensity scoring, they matched 43,609 elderly patients who started a sodium-glucose cotransporter 2 (SGLT2) inhibitor for type 2 diabetes, 77% of whom were taking canagliflozin (Invokana), to 43,609 who started a glucagonlike peptide–1 (GLP-1)–receptor agonist, 60% of whom were taking liraglutide (Victoza).
Patients were paired by age, comorbidities, diabetes severity, and dozens of other variables, more than 120 in all. The data window ran from April 2013 through December 2016.
The idea was to compare the drugs directly in order to help clinicians decide which class to choose for older patients as second-line therapy, an important consideration at a time when there’s not much guidance specifically for the elderly, and manufacturers are issuing dueling placebo-controlled trials.
Both classes have shown cardiovascular benefits, but studies were mostly in younger people with preexisting cardiovascular disease (CVD). “The comparative impact of these agents in the older population has not yet been established,” lead investigator, MD, DrPH, of Harvard University, Boston, said at the annual scientific sessions of the American Diabetes Association.
General themes are emerging from Dr. Patorno’s work; it seems that deciding between the two classes has a lot to do with whether the main concern is heart failure or cardiovascular events. Even so, she said, it’s too early to incorporate the observations into guidelines. The analysis is ongoing, and there are plans to compare impacts on renal disease and other problems.
In the meantime, she and her colleagues found that initiating an SGLT2 inhibitor versus a GLP-1 receptor agonist in the elderly was associated with a 34% decreased risk of heart failure hospitalization (2.5 fewer hospitalizations per 1,000 patient years), with an even larger drop among people who had preexisting CVD.
There was, however, a 41% increased risk of lower limb amputations (0.8 more events per 1,000 patient years) and a 62% increase in diabetic ketoacidosis (DKA, 1 more event), problems previously associated with the class.
Results were comparable – fewer heart failure hospitalizations but more amputations and DKA – when SGLT2 initiation was compared to initiation with dipeptidyl peptidase-4 (DPP-4) inhibitors, another second-line option for type 2 diabetes that includes sitagliptin (Januvia), among others.
There was a 25% increased relative risk of the composite primary outcome of myocardial infarction, stroke, and all-cause mortality when patients without baseline CVD were started on an SGLT2 inhibitor instead of a GLP-1 receptor agonist (3.7 more events per 1,000 patient years). There was no increased risk among patients who already had CVD.
SGLT2 initiation actually had a protective effect, compared with dipeptidyl peptidase-4 inhibitors, with a 23% decreased risk of the composite outcome (6.5 fewer events) among patients both with and without baseline CVD. The findings were all statistically significant.
The average age in the study was 71.5 years; 45% of the subjects were men; 40% had a history of cardiovascular disease; and 60% were on metformin and 24% on insulin at study entry.
The work was funded by the National Institutes of Health. Dr. Patorno disclosed research grants form Boehringer Ingelheim and GlaxoSmithKline. Other investigators reported relationships with numerous pharmaceutical companies.