Should Baseline Hemoglobin A1c or Dose of SGLT-2i Guide Treatment With SGLT-2i Versus DPP-4i in People With Type 2 Diabetes? A Meta-Analysis and Systematic Review

Our aim was to explore whether the baseline hemoglobin A(1c) or the dose of sodium glucose cotransporter-2 inhibitor (SGLT-2i) chosen better predicted the efficacy of SGLT-2i versus dipeptidyl peptidase-4 inhibitor (DPP-4i) in type 2 diabetes. We searched for randomized trials that compared SGLT-2i with DPP-4i in type 2 diabetes and reported a change in hemoglobin A(1c) over time. We created 2 separate analyses (one based on baseline hemoglobin A(1c) and the other according to US Food and Drug Administration [FDA]-approved SGLT-2i dose). Thirteen trials were included. In the analysis according to baseline hemoglobin A(1c), there was a significantly greater reduction in hemoglobin A(1c) when baseline hemoglobin A(1c) was >= 8.5%, favoring SGLT-2i over DPP-4i but not when baseline hemoglobin A(1c) was <8.5% (mean difference [95%CI], -0.36% [-0.53% to -0.18%] and 0.04% [-0.09% to 0.17%], respectively). On restricting the analysis to trials stratifying hemoglobin A(1c) to <8.0% or >= 8.0%, results did not change. In the analysis based on FDA-approved SGLT-2i doses, higher SGLT-2i doses caused a significantly greater hemoglobin A(1c) reduction at <= 26 and >= 52 weeks compared with the highest DPP-4i doses (mean difference [95%CI], -0.11% [-0.18% to -0.04%] and -0.24% [-0.34% to -0.15%], respectively). Lower SGLT-2i doses caused a significantly greater hemoglobin A(1c) reduction at >= 52 weeks but not at <= 26 weeks compared with the highest DPP-4i doses (mean difference [95%CI], -0.12% [-0.23% to -0.02%] and 0.01% [-0.05% to 0.07%], respectively). In people with type 2 diabetes and a baseline hemoglobin A(1c) >= 8.0%, SGLT-2i produced significantly greater reductions in hemoglobin A(1c) compared with DPP-4i and may be preferred. SGLT-2i dose titration to a higher FDA-approved dose is recommended in suitable patients.