According to a biomarker sub-study of patients with heart failure with reduced ejection fraction (HFrEF) from the DAPA-HF trial, presented during the 2021 American Heart Association (AHA) Scientific Sessions, increases in high-sensitivity cardiac troponin T (hsTnT) during a 1-year time interval were highly predictive of subsequent risk of worsening heart failure and cardiovascular death.
Physician’s Weekly spoke with the presenter, David Berg, MD, MPH (Brigham and Women’s Hospital, Harvard Medical School) about the data, which was simultaneously published in Circulation.2 For the DAPA-HF (Dapagliflozin and Prevention of Adverse-outcomes in Heart Failure; NCT03036124) trial, treatment with the sodium-glucose cotransporter 2 inhibitor (SGLT2i) dapagliflozin (10 mg daily) significantly reduced cardiovascular death or worsening heart failure in ambulatory patients with HFrEF (HR, 0.74; 95% CI, 0.65-0.85; P<0.001).3 Dr. Berg and colleagues assessed whether higher levels of circulating high-sensitivity cardiac troponin T (hsTnT), known to reflect myocardial injury, are associated with a higher risk of worsening heart failure, using data from DAPA-HF. Specifically, they studied whether changes in hsTnT over time could carry prognostic significance, if the effects of dapagliflozin on clinical outcomes were related to baseline hsTnT levels, and what the effect might be of dapagliflozin on hsTnT levels.
For this analysis, hsTnT was measured at baseline in 3,112 patients (median baseline hsTnT concentration, 20 ng/L). At 1 year, hsTnT was measured in 2,506 patients; 67.9% of patients had at least a 10% relative increase or decrease in hsTnT concentrations, and 43.5% had at least a 20% relative change. The primary endpoint was adjudicated worsening heart failure or cardiovascular death, and the researchers observed that both relative and absolute increases in hsTnT during 1 year were associated with worse clinical outcomes.
The relative reduction in the primary endpoint with dapagliflozin was consistent across quartiles of baseline hsTnT (P-interaction, 0.55), but patients in the top quartile tended to have the greatest absolute risk reduction (absolute risk difference, 7.5). Although not statistically significant, the authors noted a trend for dapagliflozin to attenuate the increase in hsTnT over time when compared with placebo (relative least squares mean reduction, -3% [-6% to 0%]; P=0.076).
Dr Berg concluded that higher baseline hsTnT and greater increase in hsTnT during 1 year are associated with a higher risk for heart failure worsening and/or cardiovascular death. Baseline hsTnT levels had no impact of the efficacy of dapagliflozin; there was no evidence of attenuated treatment benefit in those with very high hsTnT concentrations.
PW: Is troponin a clinically relevant biomarker for HFrEF?
DB: Broadly speaking, we have known for some time that patients with HF have elevated biomarkers of myocardial injury, and the most widely used biomarker for myocardial injuries is cardiac troponin. It is pretty well established that patients who have HF and elevated troponin have worse subsequent outcomes, such as higher risk of adverse events like cardiovascular death, worsening HF events, worsening HF, or hospitalizations.
What is less well known is what happens over time with these biomarkers and specifically with cardiac troponin and the prognostic implications of that. There is also uncertainty as to what extent the efficacy of dapagliflozin varies as a function of troponin level. A third aspect, that is less well known, is the effect of dapagliflozin on troponin. Those were the questions we tried to address in this study.
DAPA-HF was a study of patients with ambulatory HF. They all had HFrEF, which was defined as an LVEF less than or equal to 40%. Enrolled patients all had elevated natriuretic peptides and were randomized in a one-to-one allocation ratio to treatment with dapagliflozin versus placebo. A total of 4,744 patients were followed for a median of 1.5 years. The primary outcome was a composite of worsening HF, HF hospitalizations, urgent HF visits, or cardiovascular death.
In this biomarker analysis of the DAPA-HF data, we confirmed what has been previously shown, that patients with elevated troponin levels had higher risks of having cardiovascular death and worsening HF. But the efficacy of dapagliflozin was consistent across all troponin quartiles. And we modeled this a couple of ways. We looked at troponin as a continuous variable and also in quartiles and as a categorical variable.
The benefit of dapagliflozin, with respect to reducing cardiovascular death and worsening HF, was consistent irrespective of baseline troponin concentration. That is really important to establish, because some HF therapies work less well in patients who have more advanced disease. Two scenarios can be envisioned in which this is important: when we measure very high biomarkers in certain patients, we can determine that the chosen HF therapy may not be working adequately, or on the other end of the spectrum, low biomarker levels may indicate that patients do not actually have advanced disease. On the contrary, we observed equal efficacy of the therapy, in this case dapagliflozin, across the spectrum of myocardial injury.
Probably one of the most exciting results from our analysis is that a single measurement of troponin was predictive of worse outcomes. In addition, the change of troponin over time was highly predictive of subsequent events. Patients who had an increasing concentration of troponin during a 12-month period had subsequent increase in risk. Our data support the hypothesis that troponin levels can provide insight into myocardial disease progression in addition to underlying disease events.
Because we can leverage the randomization of a clinical trial, we could deduce the effect of dapagliflozin on troponin. We saw a modest effect of dapagliflozin on troponin levels; we observed about a 3% reduction in troponin from baseline to 12 months. It was not a dramatic effect of dapagliflozin on troponin, and it was just shy of statistical significance. I think this suggests that troponin is a very good biomarker of prognosis, and that change in troponin over time is a very good biomarker of myocardial disease progression and subsequent risk. That modest effect, however, was very consistent regardless of where you are on the spectrum of disease. That finding probably does not provide complete insight into why dapagliflozin is so effective as a HF therapy.
What are the next steps to validate this prospectively?
There are a couple of reasons that we might use biomarkers in clinical practice and in the context of clinical research. One is to identify patients who are at higher risk and who may benefit from therapies. In HF trials, we see the use of natriuretic peptide elevation almost universally applied as a criterium for trial eligibility. It is clear is from our data that troponin offers an orthogonal axis of risk. Baseline troponin levels give the medical team another perspective on what is going on with the patient’s heart and insight into risk.
Going forward, we need to consider whether we should stratify risk upfront based on troponin in terms of identifying patients who should be getting into clinical trials. I think that is an important next step. Furthermore, how do we use this in clinical practice to try to help our patients, given that there is compelling evidence from this analysis that measurement of troponin over time gives you a lot of information? How should we change our practice with that information? If a patient’s troponin levels are increasing over time, should we add dapagliflozin to the therapy? After all, we saw that dapagliflozin is effective, irrespective of the what the actual troponin levels were. We have an explosion of new HF therapies beyond the four pillars of pharmacologic management of HFrEF, which is beta blockers, angiotensin receptor neprilysin inhibitors, mineralocorticoid receptor antagonists, and now SGLT2 inhibitors. There are a number of other therapies available as well. If troponin is going up over time, despite keeping the patient on optimal medical management, maybe that is a signal that we should be adding other therapies or thinking about referring for advanced therapies. Serial measurement of cardiac troponin may provide objective assessment of clinical trajectory, which might in turn be used to inform clinical decision-making.