Translating recent results from the Cardiovascular Outcomes Trials into clinical practice: recommendations from the Central and Eastern European Diabetes Expert Group (CEEDEG)

Although monotherapy with metformin is commonly used in patients with no CV comorbidities, the early combining of metformin therapy together with agents that reduce CV risk should be preferred in patients with T2D and established CVD.

Progressing patients from mono- to combination therapy should be considered on an individual basis, depending on the patient’s investigation data, and is recommended to achieve good glycaemic control and better CV and renoprotective outcomes.

Current glucose-lowering therapies comprise six classes, each with differing combinations of attributes over and above decreasing HbA1c, such as effects on CV risk, renal outcomes, blood pressure and weight. Suitability for use in the presence of nephropathy should be considered when choosing agents for multiple combinations (Table 3).

As suggested above, to determine the efficacy of glucose-lowering treatments, and to monitor changes in CV risk factors, regular monitoring of the patient should be conducted at 3-monthly intervals. The parameters that should be evaluated are similar to those for initial assessment, and should remain comprehensive to manage both glucose levels and CV risk.

Clinical examination should include blood pressure and BMI. The laboratory tests should ascertain glucose levels to allow for correction of insulin/anti-hypertensive drugs if the patient has been experiencing hypoglycaemia. Laboratory tests should also measure HbA1c level, lipid profile, and proteinuria, GFR/eGFR and UACR.

The SGLT2 inhibitor empagliflozin should be withdrawn if the eGFR has reduced to below 45 ml/min as indicated; however, data from clinical studies show that benefits may still be seen in patients down to an eGFR of 30 ml/min [30, 33]. The GLP-1 receptor agonist liraglutide should also be withdrawn when a minimum eGFR of 30 ml/min is reached.

A repeat ECG should be performed as part of the follow-up to monitor any changes in CV status.

In addition to glucose-lowering agents, patients should also receive an anti-platelet aggregation agent, beta-blocker, angiotensin-converting enzyme (ACE) inhibitor, angiotensin receptor blocker (ARB) and statin therapies if indicated.

Although basal and other insulins have been somewhat superseded by the new generation of anti-hyperglycaemic drugs, they should not be excluded where needed. When prescribing insulins, however, the latest evidence should be considered, which suggests that insulin is best used alongside more modern drugs. Insulin has good glucose-lowering effects, but can lead to hypoglycaemia if not carefully prescribed according to the patient’s lifestyle; therefore, its therapeutic effect should be carefully evaluated. For patients who are receiving other glucose-lowering agents that may offer other benefits, such as CV and renal protection, but who are not achieving their target HbA1c levels, insulin in combination with more modern drugs may enable them to control both aspects of their condition. CEEDEG felt that the evidence for the use of insulin as a monotherapy does not warrant its use in this way, especially for the older, short-acting forms.

Patients in the upper age ranges (75 years and older) are highly variable in terms of individual fitness and general lifestyles. Although care should be taken to avoid hypoglycaemia and its associated increased risk of falls when treating frail elderly patients with T2D and CVD, it should be taken into account that the maximum CV benefits derived from the SGLT2 inhibitor empagliflozin in the EMPA-REG OUTCOME^® study were seen among the older participants, and this population is also more likely to suffer from CV and renal comorbidities. Preventative measures should therefore be taken to avoid hypovolaemia when prescribing empagliflozin in this patient group. In the LEADER^® study of liraglutide, however, patients aged over 60 years experienced fewer CV events in the placebo arm than in the treatment arm [22, 23].

As with all other therapies, therefore, patient preference should be taken together with the medical history when choosing an appropriate agent or agents.

No strong evidence has yet been found for a class effect regarding positive CV outcomes of anti-diabetic agents (Table 4 and Additional file 4: Table S1), although overall, no increased CV risk has been observed [10]. For example, CVOTs for GLP-1 receptor agonists and DPP-4 inhibitors have shown mixed results depending on the individual drug [23, 25, 35‐38]. In the SGLT2 inhibitor class, although the EMPA–REG OUTCOME^® study and CANVAS program differ (for example by study design and the pharmacological characteristics of the drugs), the results of both trials showed a significant reduction on the composite endpoint (3P-MACE); however, individual components of 3P-MACE and key secondary outcomes differed greatly [33, 39]. In the EMPA-REG OUTCOME^® study, CV death had a relative risk reduction of 38% (95% CI 0.49–0.77), whereas in the CANVAS Program CV death was not significantly reduced (13% RRR, 95% CI 0.72–1.06). Additionally, in the EMPA-REG OUTCOME^® study the key secondary outcomes of all-cause mortality (RRR 32%, 95% CI 0.57–0.82, p < 0.001), HHF (RRR 35%, 95% CI 0.50–0.85, p < 0.001) and the composite renal outcome (RRR 46%, 95% CI 0.31–0.85, p < 0.001) all showed a statistically significant benefit when adding empagliflozin versus placebo to standard of care [33]. By contrast, in the CANVAS Program none of the secondary outcomes showed a statistical benefit [39]. This is in part owing to all-cause mortality not being significantly reduced in the CANVAS Program (13% RRR, 95% CI 0.74–1.01) after which all other secondary outcomes analysed in the sequential hypothesis testing plan can only be interpreted as exploratory, such as HHF (33% RRR, 95% CI 0.52–0.87) and the composite renal outcome (RRR 40%, 95%CI 0.47–0.77). In contrast, EMPA-REG OUTCOME^® was powered sufficiently to achieve significance in both HHF (RRR 35%, 95% CI 0.5–0.85, p < 0.001) and the composite renal outcome (RRR 46%, 5% CI 0.31–0.85, p < 0.001) [30, 33, 39]. An additional difference between SGLT2 inhibitors is the observation that there was an increased signal for lower limb amputations related to canagliflozin use in the CANVAS studies, an outcome that, although observed at a higher rate in patients with diabetes than in the general population, has not been observed in any of the empagliflozin or dapagliflozin studies to date [39‐44].

Currently, some agents such as exenatide and dapagliflozin only have CV data from meta-analyses or observational studies [45, 46]. Recently published results from the real-world evidence observational study CVD-REAL, in which the risk of HHF in individuals with T2D who had been newly initiated on SGLT2 inhibitors (canagliflozin, dapagliflozin or empagliflozin) was compared to other glucose-lowering drugs (oGLD), have shown similar trends for risk reductions in HF and all-cause mortality associated with SGLT2 inhibitor use [47].This held true for all six countries involved in the study, despite differences in the background treatment and the use of different SGLT2 inhibitors in the US and Europe. Furthermore, as both HF and the subclinical forms of HF are amongst the earliest vascular complications seen when following up diabetic patients who do not have CVD at baseline [48], it may be suggested that these patients in CVD-REAL benefited from the early use of SGLT2 inhibitors. Of note, there were clear differences between the outcomes in the propensity-matched CVD-REAL and the double-blind, randomised, placebo-controlled, CANVAS Program cohorts, despite canagliflozin comprising the largest overall component of the SGLT2 inhibitors in the CVD-REAL study [39, 47]. For example, in the CANVAS Program, although the combined primary endpoint of 3P-MACE was reduced by 14%, there was no significant reduction in all-cause mortality, yet in the CVD-REAL US cohort, where 75% of patients were receiving canagliflozin, mortality was reduced in the SGLT2 inhibitor group by 62% (HR 0.38, 95% CI 0.29–0.50) [39, 47, 49]. This is most likely to be as a result of the differing methodologies of the two studies and known limitations of using retrospective data. We therefore feel that although well-performed observational studies showing the same trend can lend support to results that have been obtained from RCTs, only prospective well-performed CVOTs, in which strict study criteria, equalisation of baseline characteristics, and elimination of observer bias, can offer robust evidence of CV benefit. Until the results of the dapagliflozin DECLARE CVOT study become available in 2018, we cannot be confident in a class effect for the SGLT2 inhibitor class. We currently advise that when using the treatment algorithms physicians should continue to evaluate each of the members of any of the classes on the basis of their individual efficacy and safety data [23, 33, 35‐38].

A key aspect of therapy for patients with T2D and comorbidities is the provision of a multidisciplinary team, the members of which contribute in complementary ways to improving patient outcomes. CEEDEG recommended the specialities that should be included in the multidisciplinary team and suggested their key functions and responsibilities within the team (Table 3).