Can biomarkers help find the ‘sweet spot’ for treating patients with diabetes?

This article refers to ‘Relationship between baseline cardiac biomarkers and cardiovascular death or hospitalization for heart failure with and without sodium–glucose co-transporter 2 inhibitor therapy in DECLARE-TIMI 58’ by T.A. Zelniker et al., published in this issue on pages 1026–1036. ‘Sweet spot’, noun; an optimum point or combination of factors or qualities.'1 Cardiac biomarkers are used widely in the diagnosis of heart failure [N-terminal pro-B-type natriuretic peptide ((NT-proBNP), a marker of ventricular stress] and acute coronary syndromes (troponins, markers of myocardial injury). Yet, much interest exists regarding other circumstances when their measurements might be informative for clinical care. For example, both biomarkers are predictive of incident cardiovascular (CV) events across a wide spectrum of patients without obvious heart disease, including patients with type 2 diabetes (T2D).2 In such patients, prior studies suggest elevated concentrations of NT-proBNP and high-sensitivity cardiac troponin (hs-cTn) independently prognosticate incident heart failure hospitalization (HFH) and CV death. To this end, in this issue of the Journal, Zelniker et al.3 have measured NT-proBNP and high-sensitivity cardiac troponin T (hs-cTnT) in the DECLARE-TIMI 58 clinical trial which tested the clinical value of dapagliflozin in patients with T2D at high risk of, or with existing CV disease. Their analysis is the largest so far to address whether baseline levels of these markers help determine which patients achieve greater relative or absolute benefits from a sodium–glucose co-transporter 2 inhibitor (SGLT2i) in the prevention of HFH or CV death. Before discussing their findings, and what they might mean for clinical care, a short reminder of the history of this drug class would be useful. In 2015, the EMPA-REG OUTCOME trial surprised the diabetes and CV communities by demonstrating the SGLT2i, empagliflozin, not only reduced major adverse CV events but also incident HFH, more strikingly so, along with reducing CV deaths.4 If we speed ahead to the present day, a total of five such trials have convincingly shown that this class of drugs lessens incident HFH in people with T2D by around 30–40%, whether they have prevalent CV disease or not, as recently reported in a well conducted meta-analysis.5 These drugs also lessen renal outcomes by a broadly similar extent, and lessen major adverse CV events, albeit more modestly and only in people with existing CV disease. These results are important as heart failure is rising in prevalence, despite reductions over time in rates of myocardial infarction, and people with T2D are at higher heart failure risks.6 As such, heart failure is emerging as an outcome of progressively greater importance in the management of people with T2D. Therefore, that a relatively new class of diabetes drugs may meaningfully lessen development of heart failure at the same time as slowing renal decline in people with T2D is a major boost to the diabetes armoury. We also know now that two such drugs (dapagliflozin, empagliflozin) lessen incident HFH or CV death in people with prevalent heart failure whether they have T2D or not.7, 8 Returning to the results of this new study, Zelniker et al.3 show that whereas the relative benefits of dapagliflozin on risk for HFH or CV death appear near identical across all four quartiles of NT-proBNP or hs-cTnT, the absolute benefits were far greater in those patients with biomarker levels above the median, with the same patterns seen in those with or without heart failure. In fact, the absolute benefits were trivial in those with NT-proBNP or indeed hs-cTnT concentrations below the median, not unexpected as risks can only be reduced meaningfully when they are high to begin with. Whilst such patterns were clear for NT-proBNP, perhaps as expected, they were very similar for hs-cTnT, which is a surprisingly good predictor of incident heart failure, albeit modestly weaker than the ‘gold-standard’ NT-proBNP. That noted, it might have been useful to have also measured high-sensitivity cardiac troponin I in this study, given prior data suggesting this troponin might provide unique information compared to hs-cTnT9; one might have then compared these two troponins in their prediction of both incident HFH and incident acute coronary syndromes. The absolute risk reduction for NT-proBNP levels above the median, in this case being >75 pg/mL, was near 2%. Of course, to what extent one can expand NT-proBNP into the care of patients with T2D is debatable as this test is currently more expensive than many other biochemical tests used in routine practice. Given the prognostic value of NT-proBNP and hs-cTnT in the present study and the heightened benefit of SGLT2i treatment in those with higher values of these biomarkers, measuring NT-proBNP in patients with T2D and new-onset or existing CV disease could be helpful if the clinician is uncertain about starting such a patient on an SGLT2i and may be further useful to decide between an SGLT2i or a glucagon-like peptide-1 receptor agonist (GLP-1RA) (Figure 1). The use of cardiac biomarkers in the secondary prevention diabetes setting is not so controversial, given (i) additional testing volume would be modest at best, and (ii) all patients with T2D and CV disease are now recommended for either SGLT1i or GLP-1RA whether they are at glycaemic target or not.10, 11 What is far less clear is which patients in primary prevention should have NT-proBNP or hs-cTn measured. Certainly, not all patients can have such tests done. As T2D prevalence is upwards of 5% of the population in many European and other high-income countries, universal measurement of NT-proBNP or hs-cTn would not be financially feasible for most health systems. Whether it is possible to develop simple risk algorithm to direct clinicians to patients at higher risk for incident HF who would beneficially have NT-proBNP measured requires further study (Figure 1). Such a two-step approach is in operation for diabetes testing, with a simple risk score determining which patients should have glycated haemoglobin (a more expensive test than fasting glucose) measured.12 At the same time, should costs of NT-proBNP and hs-cTn decline appreciably, this may facilitate their wider use, and, from this, a more rapid understanding of when their measurements may meaningfully and cost-effectively improve clinical care outside their traditional areas of use may emerge. Finally, it is important to reiterate that cardiac biomarkers remain amongst the most powerful predictors of future CV outcomes, being substantially superior to high-sensitivity C-reactive protein in different settings,13, 14 some traditional lipids,15 as well as superior to many other proposed biomarkers. Given the advent of many new drugs in the CV arena, some very expensive, reports like that by Zelniker et al.3 suggest wider but judicious use of cardiac biomarkers have the potential to help ‘find the sweet spot’ for treating the right patients with the right drug at the right time (Figure 1). Further research defining and testing new cardiac biomarker measurement algorithms, and determining the cost-effectiveness of any such approach, would be welcome. Conflict of interest: N.S. is supported by A British Heart Foundation Centre of Research Excellence grant RE/18/6/34217; has received consulting honoraria from AstraZeneca, Boehringer Ingelheim, Eli-Lilly, Merck, Novartis, Novo Nordisk, Pfizer, Roche Diagnostics and Sanofi. He has also received grant support from Boehringer Ingelheim, Roche Diagnostics and Novartis. J.L.J. is supported by the Hutter Family Professorship; is a Trustee of the American College of Cardiology; is a board member of Imbria Pharmaceuticals; has received grant support from Novartis Pharmaceuticals, Roche Diagnostics, Abbott, Singulex, and Prevencio; has received consulting income from Abbott, Janssen, Novartis, Pfizer, Merck, and Roche Diagnostics; and participates in clinical endpoint committees/data safety monitoring boards for Abbott, AbbVie, Amgen, Boehringer Ingelheim, Janssen, and Takeda.