Effect of Dapagliflozin in DAPA-HF According to Background Glucose-Lowering Therapy

Although sodium–glucose cotransporter 2 inhibitors (SGLT2is) have been shown to improve cardiovascular outcomes in patients with type 2 diabetes, they are usually prescribed as second-line glucose-lowering therapy (GLT), most often in addition to metformin (1–3). Uncertainty about the place of SGLT2is in the management of patients with type 2 diabetes is reflected in the differing recommendations in recent guidelines (4–8). The placebo-controlled Dapagliflozin And Prevention of Adverse-Outcomes in Heart Failure trial (DAPA-HF), in which the SGLT2i dapagliflozin reduced the risk of worsening HF and cardiovascular mortality in patients with HF and reduced ejection fraction (HFrEF), provides a unique opportunity to examine the efficacy of SGLT2i alone and in combination with other GLTs in patients with type 2 diabetes (9).

DAPA-HF was a prospective, randomized, double-blind, placebo-controlled trial in patients with HFrEF that evaluated the efficacy and safety of 10 mg dapagliflozin once daily, compared with placebo, added to standard care (9,10).

In this post hoc analysis, we included randomized patients with either undiagnosed (defined as central laboratory HbA_1c ≥6.5% [48 mmol/mol] at both screening and randomization visits) or a medical history of type 2 diabetes. We examined the effect of dapagliflozin, compared with placebo, in subgroups (limited to those with >200 individuals to minimize the likelihood of type 1 errors) by the use or not of background GLT and by individual GLT classes: biguanides (hereafter referred to as metformin), sulfonylureas, dipeptidyl peptidase 4 (DPP-4) inhibitors, and insulin. We examined the primary outcome, a composite of an episode of worsening HF (either an unplanned hospitalization or an urgent visit resulting in intravenous therapy for HF) or cardiovascular death, along with the individual components of cardiovascular death and HF hospitalization, and the prespecified secondary end points of all-cause mortality and the composite of total (first and recurrent) HF hospitalizations and cardiovascular death.

The cumulative incidence of the primary end point by treatment group in subgroups of interest was plotted using the Kaplan-Meier method. The effect of dapagliflozin compared with placebo was examined using Cox proportional hazards models with history of hospitalization for HF and treatment-group assignment as fixed-effect factors (history of hospitalization for HF was not included in the models for all-cause mortality). An interaction test using a subgroup–by–randomized treatment interaction term was performed to assess for treatment effect modification within each subgroup. Analyses were performed using Stata, version 16 (StataCorp, College Station, TX). A P value <0.05 was considered statistically significant.

Of the 4,744 randomized patients in DAPA-HF, 1,983 (41.8%) had a documented medical history of type 2 diabetes, and 156 (3.3%) had undiagnosed type 2 diabetes. Therefore, 2,139 (45.1%) patients with type 2 diabetes were included in the analysis. Of these, 1,596 (74.6%) were treated with GLTs: metformin (47.7%), insulin (25.2%), sulfonylurea (20.6%), DPP-4 inhibitor (14.5%), and glucagon-like peptide 1 (GLP-1) receptor agonist (1.0%) (each alone or in combination). The baseline characteristics of patients by use of GLT and type of GLT are summarized in Supplementary Tables 1 and 2.

Supplementary Fig. 1 shows the cumulative incidence of the primary composite end point by randomized treatment in the subgroups of interest. The effect of dapagliflozin on the primary end point was consistent in patients taking GLT (hazard ratio 0.72; 95% CI 0.58–0.88) and in those who were drug-naive (0.86; 0.60–1.23; interaction P = 0.39) (Fig. 1). When considering individual GLT classes (Fig. 1) or combinations (Supplementary Fig. 2), there was no statistically significant interaction between background GLT and the effect of randomized therapy on the primary composite outcome.

In general, the effect of dapagliflozin on cardiovascular death and HF hospitalization was similar for individual GLTs (Supplementary Fig. 3) and combinations of these (Supplementary Fig. 2). Furthermore, no modification of treatment effect by background GLT was observed for the composite end point of total (first and recurrent) HF hospitalizations and cardiovascular death (Supplementary Fig. 4) or all-cause mortality (Supplementary Fig. 5).

In this post hoc analysis of DAPA-HF, we found that the benefit of dapagliflozin compared with placebo in patients with type 2 diabetes and HFrEF was not influenced by background GLT use. The benefit of dapagliflozin was consistent in drug-naive patients and across all classes of commonly used GLTs, including metformin.

Perhaps the most interesting group of participants was the ∼25% of individuals with type 2 diabetes in DAPA-HF who were not prescribed any GLT at baseline, i.e., those in whom randomized dapagliflozin became first-line GLT and pharmacological monotherapy. Despite limited power for subgroup analysis due to a relatively small number of patients and a lower event rate, the benefit of dapagliflozin on the primary end point seemed to be consistent with the effect in type 2 diabetes patients overall.

Metformin was the most commonly used GLT in DAPA-HF, taken by approximately half of patients with type 2 diabetes and HFrEF, despite limited evidence for its cardiovascular safety in this patient group (11). Nevertheless, international HFrEF management guidelines support the use of metformin as the first-line GLT in patients with type 2 diabetes (12). It has been suggested that the benefit of SGLT2i is modified by metformin use based on a subgroup analysis of the Canagliflozin Cardiovascular Assessment Study (CANVAS) trials (13). This is clearly not the case from the present analysis of DAPA-HF or a post hoc analysis of the BI 10773 (Empagliflozin) Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients (EMPA-REG OUTCOME) (14).

Examination of outcomes in patients receiving the other major classes of GLT was also of interest. After metformin, insulin was the most widely used GLT, and dapagliflozin was as effective in these participants as compared with patients not taking insulin. Given the substantially higher event rate experienced by patients receiving insulin compared with those receiving other GLTs, the relative risk reduction in insulin-treated individuals translated into an even larger absolute risk reduction and a number needed to treat of only 16 to prevent one patient having the primary outcome over the median 18.2 months of follow-up. Furthermore, the benefits of dapagliflozin were again consistent whether added to a sulfonylurea or a DDP-4 inhibitor.

We believe our findings are relevant to the discussion that followed recent updated guidance on management of diabetes issued by the European Society of Cardiology (ESC) and jointly by the American Diabetes Association and the European Association for the Study of Diabetes (4–7). Both recommendations emphasized that the cardiovascular benefits of SGLT2i and GLP-1 receptor agonists are obtained independently of starting HbA_1c, an approach supported by the strategy employed in DAPA-HF. More controversially, the ESC guidance supported the use of SGLT2i and GLP-1 receptor agonists as first-line GLT and not necessarily as an adjunct to metformin, which had previously been the recommended initial GLT in most patients with cardiovascular disease (7). Our data also support this recommendation, at least in patients with HFrEF, and provide further evidence, along with the evidence of benefit in HFrEF patients without diabetes, to the view that the mechanisms of action underlying the cardiovascular benefits of dapagliflozin are independent of any glucose-lowering effect (15).

As with all studies of this nature, there are inherent limitations. The analyses were not prespecified and some had limited power, despite only including subgroups with >200 individuals. The small number of patients taking a GLP-1 receptor agonist at baseline prohibited further examination of this subgroup.

In patients with type 2 diabetes and HFrEF, the reductions in the risk of worsening HF and cardiovascular death with dapagliflozin were consistent across a range of background of GLTs and in patients receiving no GLT. Our data provide support for the use of dapagliflozin as first-line monotherapy in type 2 diabetes, at least in patients with HFrEF.

Funding. J.J.V.M. is supported by a British Heart Foundation Centre of Research Excellence Grant RE/18/6/34217.

Duality of Interest. The DAPA-HF trial was funded by AstraZeneca. K.F.D. reports his employer (University of Glasgow) is paid by AstraZeneca for his involvement in the DAPA-HF trial during the conduct of the study; grants from Novartis; and personal fees from Eli Lilly outside the submitted work. P.S.J. reports his employer (University of Glasgow) is paid by AstraZeneca for involvement in the DAPA-HF trial during the conduct of the study; consulting, advisory board, and speaker’s fees from Novartis; advisory board fees from Cytokinetics; and a grant from Boehringer Ingelheim outside the submitted work. O.B. is an employee of AstraZeneca. D.L.D. reports personal fees from Frontier Science, Actelion, Population Health Research Institute, Duke Clinical Research Institute, Bristol-Myers Squibb, Medtronic, Boston Scientific, GSK, Merck, National Institutes of Health (NIH) National Institute of Allergy and Infectious Diseases and National Heart, Lung, and Blood Institute (NHLBI), AstraZeneca, Intercept, Mesoblast, Liva Nova, DalCor, Sanofi; and personal fees and other from D.L. DeMets Consulting outside the submitted work. S.E.I. reports personal fees and nonfinancial support from AstraZeneca during the conduct of the study and personal fees and nonfinancial support from Boehringer Ingelheim, Sanofi/Lexicon, Merck, Zafgen, VTV Therapeutics, Abbott/Alere, and Novo Nordisk outside the submitted work. L.K. reports being an executive committee member for the DAPA-HF study, payment from which will be administered by Rigshospitalet University Hospital, from Astra-Zeneca, during the conduct of the study; personal fees from Novartis as speaker; and personal fees from Bristol-Myers Squibb as speaker outside the submitted work. M.N.K. reports personal fees from AstraZeneca during the conduct of the study; grants, personal fees, and other from AstraZeneca; grants and personal fees from Boehringer Ingelheim; personal fees from Sanofi, Amgen, Novo Nordisk, Merck (Diabetes), Janssen, Bayer, Glytec, Novartis, Applied Therapeutics, Amarin, Eli Lilly, and Vifor Pharma outside the submitted work. A.M.L. is an employee and shareholder of AstraZeneca. F.A.M. reports personal fees from AstraZeneca during the conduct of the study. M.S.S. reports grants and personal fees from AstraZeneca during the conduct of the study; personal fees from Althera, Anthos Therapeutics, Bristol-Myers Squibb, CVS Caremark, Dalcor, Dr Reddy’s Laboratories, Dyrnamix, Esperion, and IFM Therapeutics; grants and personal fees from Amgen, Intarcia, Jansen Research and Development, Medicine Company, Medimmune, Merck, and Novartis; grants from Bayer, Daichii-Sankyo, Eisai, Pfizer, Quark Pharmaceuticals, and Takeda outside the submitted work; and is a member of the TIMI Study Group, which has also received institutional research grant support through Brigham and Women’s Hospital from Abbott, American Heart Association, Aralez, Roche, and Zora Biosciences. M.S. is an employee and shareholder of AstraZeneca. S.D.S. reports grants from AstraZeneca during the conduct of the study; grants and personal fees from Alnylam, Amgen, AstraZeneca, Bristol-Myers Squibb, Gilead, GSK, MyoKardia, Novartis, Theracos, Bayer, and Cytokinetics; grants from Bellerophon, Celladon, Ionis, Lone Star Heart, Mesoblast, NIH/NHLBI, Sanofi Pasteur, and Eidos; and personal fees from Akros, Corvia, Ironwood, Merck, Roche, Takeda, Quantum Genomics, AoBiome, Janssen, Cardiac Dimensions, Tenaya, Daichi-Sankyo, Cardurion, and Eko.Ai outside the submitted work. J.J.V.M. reports his employer (University of Glasgow) being paid by AstraZeneca during the conduct of the study, and his employer (University of Glasgow) being paid by Bayer, Cardiorentis, Amgen, Oxford University/Bayer, Theracos, Abbvie, Dalcor, Pfizer, Merck, Novartis, GSK, Bristol-Myers Squibb, Vifor-Fresenius, and Kidney Research UK (KRUK) outside the submitted work.

Author Contributions. K.F.D., P.S.J., O.B., and J.J.V.M. contributed to the data analysis. All authors were involved in data interpretation and the writing or editing of the report, read and approved the submitted version of the report, and contributed to the study design. K.F.D. and J.J.V.M. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.