Researchers analyzed 301 randomized clinical trials of patients with type 2 diabetes, and found that, metformin outperformed some other drug classes for its effect on hemoglobin A1c levels, there were no significant differences in mortality -- including when placebo was included as a drug class.
Of all the classes, sodium-glucose cotransporter-2 (SGLT2) drugs were associated with the lowest odds of hypoglycemia when added to metformin (odds ratio 0.12, 95% CI 0.008-0.18; risk difference -22%), but when when added to both metformin and sulfonylurea, glucagon-like peptide-1 receptor agonists had the lowest odds (OR 0.60, 95% CI 0.39-0.94; risk difference 10%), according to lead author Suetonia Palmer, PhD, at the University of Otago Christchurch in New Zealand.
Palmer and colleagues published their findings on Tuesday in the Journal of the American Medical Association.
"A central finding in this meta-analysis was that despite more than 300 available clinical trials involving nearly 120,000 adults and 1.4 million patient-months of treatment, there was limited evidence that any glucose-lowering drug stratified by coexisting treatment prolonged life expectancy or prevented cardiovascular disease," the authors wrote.
Four classes of drugs were associated with higher HbA1c levels when compared to metformin:
In addition, of all the drugs, sulfonylurea (OR 3.13, 95% CI 2.39-4.12; risk difference 10%, 95% CI 7%-13%) and basal insulin (OR 17.9, 95% CI 1.97-162; RD 10%, 95% CI 0.08%-20%) were associated with the highest risk of hypoglycemia.
Most (177) of the trials were of drugs given as monotherapy; in 109 trials the drugs were given as an add-on to metformin, and 29 trials of drugs added to metformin plus sulfonylurea. There were nearly 120,000 patients included in the trials in total.
There was no evidence of differences in the association between the drug classes with serious adverse events, myocardial infarction, or stroke. However, "Considerable uncertainty about the association of drug treatment with cardiovascular mortality existed within trial evidence, largely because of few events in most available studies," the authors warned.
All of the trials included in the study were randomized and lasted at least 24 weeks. In addition to the classes already mentioned, meglitinide was also included, as was placebo. Only trials that had publicly available data as of March 21, 2016 were included in the meta-analysis. Cardiovascular mortality was the primary end-point, and all-cause mortality, myocardial infarction, stroke, HbA1c levels, and treatment failure were secondary end-points. Safety end-points were also included.
All of the monotherapies were associated with lower HbA1c levels versus placebo, but thiazolidinedione, DPP-4 inhibitors, and a-glucosidase inhibitors were associated with higher HbA1c levels compared with metformin. Placebo was most associated with treatment failure (OR versus metformin 3.83, 95% CI 2.88-5.10; RD 11%), and GLP-1 receptor agonist monotherapy was associated with a lower body weight compared with metformin.
There were no data for basal insulin and GLP-1 receptor agonists as a monotherapy for cardiovascular mortality. For dual therapies, all classes were associated with lower odds of hypoglycemia than metformin plus sulfonylurea; the GLP-1 class was ranked best in avoiding hypoglycemia, and thiazolidinediones were ranked worst.
The risk of bias was classified as either high or unclear for most of the trials for reasons including random sequence generation (208 trials), concealment of treatment allocation (232 trials), masking outcome assessment (281 trials), and selective reporting of outcomes (172 trials). In 63% of the trials, the trial sponsor was involved in authorship, data management, or both.
The authors wrote that their findings are consistent with guidelines from the American Diabetes Association, which - like the algorithm from the American Association of Clinical Endocrinologists - recommend that metformin monotherapy be used for the initial treatment of patients with type 2 diabetes. "Based on this review, clinicians and patients may prefer to avoid sulfonylureas or basal insulin for patients who wish to minimize hypoglycemia, choose GLP-1 receptor agonists when weight management is a priority, or consider SGLT-2 inhibitors based on their favorable combined safety and efficacy profile," the authors wrote.
Two recent studies, EMPA-REG and LEADER, have shown two diabetes drugs to be associated with lower mortality. But the authors added that neither of those trials analyzed treatment as a monotherapy or added to metformin. "Future trials might prioritize comparisons of SGLT-2 inhibitors against metformin or added to metformin to compare specific dual-therapy regimens," wrote Palmer and colleagues.
Relatively few studies reported cardiovascular mortality and most of those had zero or very few events, which is a limitation to the meta-analysis, they added. The authors didn't examine triple therapy treatments, and the analysis was not adjusted for baseline kidney function. Most of the studies were relatively short-term and were conducted in higher-income countries.
Return to News Home