Diabetes and cardiovascular risk
Prof. Richard Hobbs gave an introduction to the presentation of prof. Kamlesh Khunti, by outlining the evidence that diabetes is linked to CVD risk. Several studies have indicated that diabetes doubles the risk of vascular disease, and this increase in risk is not fully explained by conventional CV risk factors. The higher risk associated with diabetes is not instantaneous: the duration of having the disease matters for the risk, especially if the condition is poorly treated. After having had diabetes for 10 years, the risk of CV events is about equivalent to that in those with established coronary heart disease. This leads to a lower life expectancy; on average, a 50-year old with diabetes but no history of vascular disease will die about 6 years before a counterpart without diabetes.
Reducing conventional risk factors is effective to mitigate the overall risk in patients with diabetes; those with diabetes benefit from BP lowering and statin therapy to the same extend as those without diabetes. Moreover, the Steno-2 study has demonstrated that intensive multifactorial control of CV risk factors lowers CV risk in patients with T2DM and microalbuminuria. Thus, it is worth the effort to actively reduce risk factors in patients with diabetes. More strategies to prevent diabetes are needed though, for instance by reducing obesity. Also, early diagnosis, possibly by screening, can also be improved.
Prof Kamlesh Khunti then continued with discussing how the CV risk associated with diabetes can be managed. All guidelines emphasise the importance of a multifactorial approach to reducing CV risk, by addressing all modifiable risk factors: lifestyle modification, glycaemic control, management of dyslipidaemia, platelet inhibition, blood pressure control and possibly anti-inflammatory therapy.
About lifestyle intervention he noted that although the LOOK AHEAD trial was stopped because of futility, now there is good evidence from the DaQing study that a 6-year lifestyle intervention programme reduced CV and all-cause mortality and diabetes in Chinese people with impaired glucose tolerance. It should be noted that the CV benefits are not seen before about 10 years.
He shared trial evidence on the beneficial effects of lowering LDL-c and lowering systolic BP. The ASCEND trial suggested that aspirin should not be given as primary prevention to patients with diabetes, because the CV benefits are outweighed by extra major bleeding. In a meta-analysis of traditional glucose-lowering trials, a modest reduction in non-fatal MI and coronary heart disease was seen, but only after about 20 years. No benefit on stroke events and all-cause mortality was seen. In patients with heart failure (HF), blood glucose control was even associated with a higher rate of mortality (in ACCORD).
Recent CV outcome trials with novel antidiabetic therapies have painted a different picture. The trials were set up to test safety of the agents. The first class is that of the DPP-4 inhibitors: they did not cause harm to the high-risk diabetes patients, nor showed a CV benefit (SAVOR-TIMI 53 - saxagliptin, EXAMINE - alogliptin, TECOS - sitagliptin and CARMELINA - linagliptin). Diverse results were seen among trials for the outcome of hospitalization of HF: a higher rate was observed in SAVOR-TIMI 53, while no significant effects were seen in the other DPP-4 inhibitor trials.
Another new class consists of GLP-1 receptor agonists (GLP-1RA). Populations included in the CV outcome trials varied, as did the specific outcomes. No significant reductions in 4- or 3-point MACE were observed in ELIXA (lixisenatide) and EXSCEL (exenatide), whereas the respective GLP-1RAs in LEADER (liraglutide), SUSTAIN-6 (semaglutide) and HARMONY (albiglutide) showed superiority over placebo in reducing the primary composite CV outcome.
The SGLT2 inhibitors yielded even more excitement in the field than GLP-1RAs, as all three members of this class (EMPA-REG OUTCOME: empagliflozin, CANVAS: canagliflozin, DECLARE: dapagliflozin) showed reduction of the primary endpoint. The DECLARE trial was different from the other two, in that a large proportion (59.4%) of T2DM patients without established CVD were included (as opposed to <1% in EMPA-REG OUTCOME and 34.4% in CANVAS).
Separation of the CV death and hospitalization for HF curves occured after about a year with SGLT2 inhibition, whereas MACE curves always take longer to separate. In the overall DECLARE population, MACE was not significantly reduced. A meta-analysis combining data of the three SGLT2 inhibitor outcome trials revealed that the risk of atherosclerotic CV disease was significantly reduced in T2DM patients with established atherosclerotic CV disease, but not in those with only multiple risk factors, but not atherosclerotic disease. Hospitalisation for HF was reduced in all trials, and in the pooled data, in those with eGFR<90 ml/min per m². The same was true for MACE. These data are relevant for patients with T2DM, as renal dysfunction is common. Kidney disease increases the risk of mortality in T2DM. The meta-analysis of SGLT2 inhibitor trials showed that the therapy had good effects on renal outcomes both in primary and secondary prevention setting, and in all categories of renal function, when stratified based on eGFR.
Khunti also dedicated some time to discussing what real world evidence (RWE) can further teach us about the use of therapies. He summarised the differences between what results from randomised controlled trials (RCTs) and real-world data can tell us as follows: ‘Can it work?’ vs. ‘Does it work?’. RCTs evaluate an intervention under carefully controlled conditions and in a highly selected population. These data need to be complemented with other sources of information to obtain a true picture of how an intervention will perform in the wider patient population encountered in daily practice. A hierarchy of quality of data obtained outside clinical trials is: registration RCTs at the top, followed by long-term phase 3 trials, pragmatic randomized trials and observational studies. While RCTs assess efficacy under ideal circumstances, RWE can inform about the effectiveness under the usual circumstances of healthcare practice. When comparing the results in these settings, sometimes an 'efficacy-to-effectiveness gap' is observed: implementation of interventions do not always perform as well in clinical practice as they do in clinical trials. This can be the consequence of population (comorbidities!), biological or behavioural differences (of physician or patient) and of failure to follow up minor safety signals. In the case of diabetes, it is relevant to note that people with hypoglycaemia in the past year are often excluded from diabetes trials. In general, multimorbid patients do not tend to be included in trials. Khunti compared available RCT evidence on SGLT2 inhibition with RWE. He showed that consistent findings were seen in real world practice for patients with and without established CVD, something which was not seen in the trials. This might be the consequence of dealing with a sicker population in real world practice.
In addition to discussing the therapeutic recommendations in the flowchart in the recent ADA/EASD consensus statement, Khunti also raised the topic of quaternary prevention. This refers to stopping treatment when needed. Many physicians not only fail to advance therapy when needed, but also do not de-intensify therapy when appropriate. It has been demonstrated that delayed intervention has consequences for CV risk; patients with HbA1c >7% who did not receive treatment intensification within one year showed a higher risk of CV events at 5 years. In addition, poor treatment adherence is common, and this is also associated with higher CV risk. The ADA/EASD consensus statement considers metformin still as the first-line glucose-lowering treatment for most people with T2DM, based on its efficacy, safety, tolerability, and the extensive clinical experience. While treatment paradigms are evolving, metformin will likely stay on that first place for a while, as it is ‘cheap and cheerful’, said Khunti. In any case, it is important to avoid inertia and glycaemic variability and to improve adherence.
He ended by sharing the ERICAS RWE Resource, a site that posts bi-weekly reviews of real-world observational studies in cardiometabolic medicine.