AIM: To assess the efficacy and safety of combination therapy with a glucagon-like peptide-1 receptor agonist (GLP-1 RA) and a sodium-glucose co-transporter-2 inhibitor (SGLT2i) in patients with type 2 diabetes.
METHODS: We searched Medline, Embase, the Cochrane Library and grey literature sources up to 2 December 2019 for randomized controlled trials in adults with type 2 diabetes assessing the combination of GLP-1RA and SGLT2i, either as co-initiation therapy or as add-on to each other, against placebo or an active comparator. The primary outcome was change in HbA. Secondary outcomes included change in body weight, blood pressure and estimated glomerular filtration rate, and incidence of severe hypoglycaemia, all-cause mortality, cardiovascular mortality, myocardial infarction, stroke and hospitalization for heart failure. We pooled data using random effects meta-analyses. 1c
RESULTS: Seven trials (1913 patients) were eligible. Compared with GLP-1RA, GLP-1RA/SGLT2i combination therapy was associated with a greater reduction in HbA1c (weighted mean difference -0.61%, 95% CI -1.09% to -0.14%, four studies), body weight (-2.59 kg, -3.68 to -1.51 kg, three studies) and systolic blood pressure (-4.13 mmHg, -7.28 to -0.99 mmHg, four studies). Compared with SGLT2i, GLP-1RA/SGLT2i combination therapy reduced HbA1c (-0.85%, -1.19% to -0.52%, six studies) and systolic blood pressure (-2.66 mmHg, -5.26 to -0.06 mmHg, six studies), but not body weight (-1.46 kg, -2.94 to 0.03 kg, five studies). After excluding data for one trial that had a considerably longer duration than the remaining studies, body weight was also reduced versus SGLT2i (-1.79 kg, -2.99 to -0.59 kg, five studies). Combination therapy did not increase the incidence of severe hypoglycaemia. Data for mortality and cardiovascular outcomes were scarce.
CONCLUSIONS: GLP-1RA/SGLT2i combination therapy seems to reduce HbA, body weight and systolic blood pressure without increasing the risk of severe hypoglycaemia compared with either GLP-1RA or SGLT2i. No conclusions can be made regarding long-term effectiveness or the effect on cardiovascular outcomes. 1c
