Summary | Guidance from outcome trials: What are the clinical implications
Professor Zannad set out to share ideas on how to deal with a patient with diabetes and HF, based on data of the latest trials. On the one hand, evidence comes from HF therapy outcome trials, in which a subset of patients had diabetes. Patients with diabetes benefit from various types of HF therapy in the same way as patient without diabetes. On the other hand, there are glucose-lowering diabetes outcome trials, in which it is interesting to look at the effect on HF outcomes. HF as an independent endpoint has long been neglected in diabetes trials15, but this should be improved in future studies.
A meta-analysis of the effect of intensive glucose control as compared with less intense control, based on data of the pivotal ADVANCE, UKPDS, ACCORDS and VADT studies, showed a reduction in MI (by 15%) in patients with T2DM, while the effect on hospitalization or death from HF was neutral (HR: 1.00).16 Although glucose control may improve microvascular and/or macrovascular disease, this does not necessarily translate into protection against HF. In the RECORD trial, rosiglitazone, as compared with metformin, showed better glucose control, but was neutral on CV outcomes. Most strikingly, the risk of HF was increased with rosiglitazone.17
When comparing the newest glucose-lowering drug classes DPP-4 inhibitors, GLP-1 receptor agonists (GLP-1RAs) and SGLT-2 inhibitors, various effects on HF events are seen. In the first category, saxagliptin showed an excess of HF events in the SAVOR-TIMI 53 trial, while in the latter category, both empagliflozin and canagliflozin showed a reduction of HF events. In the middle category, at least so far, the GLP-1RAs seem to be neutral on HF events. This suggests that glucose control, which is a property common to all these medications, is not associated to an effect on HF.
The class of DPP-4 inhibitors shows the broadest range of effects on HF; different members have shown different outcomes. While the results of the SAVOR-TIMI 53 trial surprised the field with an increase in hospitalization for HF in patients receiving saxagliptin (HR: 1.80, 95%CI: 1.29-2.55, P=0.001 at 180 days after randomization, and HR: 1.27, 95%CI: 1.07-1.51, P=0.007 at 720 days)18, alogliptin showed total neutrality with respect to the outcome of CV death and hospitalization for HF (HR: 0.997, 95%CI: 0.820-1.212, at 30 months), as did sitagliptin in TECOS (HR: 1.00, 95%CI: 0.83-1.20). These findings suggest that there is no class effect for the DPP-4 inhibitors.
As Sattar had already touched upon, a big positive surprise came form the EMPA-REG OUTCOME trial, in which the SGLT2 inhibitor empagliflozin improved mortality, CV death and MACE. HF was a secondary outcome. Much improvement was seen very soon after initial exposure to the drug.4 Thus, empagliflozin appeared to be a glucose-lowering agent, which is not only safe, but also improves CV outcomes including HF. Irrespective of the definition of HF that was applied, the finding that empagliflozin treatment was beneficial for HF outcomes was very robust. Also in patients without HF at baseline, empagliflozin reduced HF events, thus new onset symptomatic HF was prevented.19
HF, especially when it co-exists with diabetes, is often accompanied with chronic kidney disease (CKD). But patients with eGFR <60 ml/min/1.73m2 have been excluded in many trials. The proportion of patients with reduced renal function ranged between 5 and 30% in outcome trials evaluating novel glucose-lowering drugs. Thus, current findings cannot be extrapolated to patients with CKD and moderate to severe renal dysfunction. Two dedicated trials (CREDENCE with canagliflozin and CARMELINA with linagliptin) study these subpopulations, to evaluate whether the HF protection extends to patients with the frequent comorbidity of CKD, diabetes and HF.
When comparing empagliflozin and canagliflozin, results seen in the EMPA-REG OUTCOME trial and the CANVAS program are largely consistent. Thus, the protection against HF seems to be a class effect of SGLT2 inhibitors. A hint at potential mechanisms underlying this benefit comes from the increase in hematocrit, seen with empagliflozin. A consistent and persistent increase in hematocrit of 45 % over the course of follow-up was seen, which was not related to the number of red blood cells, suggestive of hemoconcentration.20 Thus, diuretic decongestion effect takes place. Moreover, the sodium and water excretion induced by SGLT2 inhibition can help prevent HF events. A mediation analysis suggested that the effect of empagliflozin on hematocrit and hemoglobin, and also albumin, explained most of the clinical benefit on HF seen in EMPA-REG OUTCOME (37.3, 44.4 and 25.3% respectively).20
GLP-1RAs showed a benefit on CV death and MI in LEADER and SUSTAIN-6, but not on HF. Thus, considering equal glucose control, different CV outcomes are to be expected with different glucose-lowering agents.21 GLP-1RAs are thought to be more effective in prevention of atherosclerotic events. Novel guidelines and treatment recommendations introduce new sub-groups and different phenotypes of HF, as not all diabetes patients should be considered the same. Empagliflozin is now considered the preferred second line or third line therapy in addition to metformin, especially in patients at risk of HF and CVD. Based on the LEADER trial data, liraglutide is considered second or third line therapy in patients with stage three kidney disease, and in patients with atherosclerotic disease.21 Thus, treatment of patients with diabetes and HF is increasingly finetuned.
More trials are ongoing or coming up, which will teach us more about the use of these glucose-lowering agents in various patient populations. For instance, the EMPEROR trial investigates the effect of empagliflozin in patients with HFrEF (LVEF ≤40%, n=2850) or HFpEF (LVEF ≥40%, n=4126), in parallel trials. In each HF arm, patients will be randomized to empagliflozin 10 mg or placebo, on top of standard of care. EMPEROR includes both patients with and without T2DM, thus the results may inform us whether empagliflozin may become HF therapy, independent of diabetes.22
Similarly, DAPA-HF includes patients with HFrEF (LVEF ≤40%), also including patients with eGFR <30 ml/min/1.73m2. The primary outcome is a composite of CV death, hospitalization for HF or an urgent HF visit, and a number of secondary outcome measures focus on aspects of HF, making DAPA-HF a typical HF trial.23
Some of these recent insights are the unintended consequence of the 2008 FDA Guidance on trials of new antidiabetic medications, which mandated that all these new agents should be evaluated for CV safety. Some trials were also designed to test for superiority. As described above, certain trials have demonstrated that a drug that lowers glycated hemoglobin, can also affect major CV events within five years of use, which has nothing to do with glycemic control. Ultimately, these novel glucose-lowering drugs may turn out to become HF therapy. This is an important paradigm shift, and the consequence of serendipity: an accident and sagacity while in pursuit of something else. Zannad concluded by citing Louis Pasteur, who said ‘in the fields of observation, chance favors only the prepared mind’ and stating that fortunately HF specialists were prepared to capture this signal from SGTL2 inhibitors.