Dr James L Januzzi
Cardiologist, Massachusetts General Hospital. Prof of Medicine, Harvard Medical School

CANVAS study: NT-proBNP and CVD risk reduction


  • In the CANVAS study the distribution of NT-proBNP values at baseline was compared between type 2 diabetics with and without a known history of heart failure. It was observed that these distributions were almost identical, indicating that a significant percentage of patients with T2D without a prior recorded history are at an increased risk of heart failure and its sequelae
  • NT-proBNP concentrations greater than 125pg/mL at baseline are strongly prognostic of various cardiovascular, renal, and neurological complications in this cohort
  • Treatment with the SGLT2 inhibitor used within the CANVAS study may reduce the risk of heart failure in patients with diabetes by reducing NT-proBNP concentrations. This applies to patients with both low or high NT-proBNP, however, absolute risk reduction is greatest in those with an elevated NT-proBNP value

Can you give us a short description of the CANVAS program?

The CANVAS program was a set of two trials, whose goal was to evaluate the role of an SGLT2 inhibitor (Sodium-glucose Co-transporter-2 inhibitor) for the treatment of diabetes in patients with moderate to high cardiovascular risk. The reason for the CANVAS program was because prior therapies for diabetes, specifically thiazolidinediones, had been shown to actually increase cardiovascular risk after they had been released. And so, as a consequence, regulatory agencies recommended that new diabetes drugs be evaluated in what we refer to as Cardiovascular Outcomes Trials or CVOTs. These CVOTs were largely based on a primary endpoint of cardiovascular death, non-fatal MI, and non-fatal stroke, and evaluated therapies such as the SGLT2 inhibitors in this case over a long period of time.

The CANVAS program included patients either with established cardiovascular disease or enhanced cardiovascular risk – they had multiple risk factors – and evaluated the impact of a SGLT2 inhibitor, at 100 or 300 milligram doses versus placebo versus usual care, with the goal to evaluate for safety. Now, as has been the case with other SGLT2 inhibitors, in the CANVAS program it was found to reduce non-fatal myocardial infarction, stroke, and cardiovascular death as a composite endpoint. But notably had a very profound impact on heart failure events. And so, as a consequence, there was a lot of interest to study whether biomarkers might be able to identify risk of patients, in this case with type 2 diabetes and cardiovascular risk, and perhaps predict the benefits of SGLT2 inhibitors in these patients.

Get the latest updates in your inbox

Stay Informed

What were the objectives of the analysis?

So, the objective of our studies was to take advantage of the resources that were available to us from the CANVAS study. So, one of the two trials in the CANVAS programs, specifically CANVAS itself – there’s CANVAS and CANVAS-R – so CANVAS actually had a biorepository. We were able, therefore, to evaluate the role of cardiac biomarker testing, to prognosticate outcomes in the CANVAS study, as well as to evaluate the effects of therapy on NT-proBNP and other biomarkers over the long-term follow-up in this trial – up to six years of follow-up were available.

To evaluate N-terminal-proBNP, we had measurements at baseline, at one year, and then again at six years. And we evaluated the effects of the SGLT2 inhibitor on NT-proBNP concentrations, and conversely evaluated whether NT-proBNP would be predictive of response to the SGLT2 inhibitor in terms of outcomes.

The first observation, very importantly, was that when we looked at the distribution of NT-proBNP values at baseline, we saw something that was quite striking. Specifically, even though only 13% of the subjects in the CANVAS study had a reported history of heart failure at baseline, we found that the distribution of NT-proBNP in those with and without a prior history of heart failure were nearly identical. What this tells us, is that in the CANVAS study, regardless of a known diagnosis of heart failure, a large percentage of patients with type 2 diabetes, without a known diagnosis of heart failure, were at heightened risk for the onset of heart failure, and subsequent complications from heart failure. That’s important. That’s critically important because it tells us that many patients with diabetes have a high risk that might not otherwise be recognized. 

How do you hope that these results will influence clinical practice?

Okay, so we now have the first observation, which is that a large percentage of patients in CANVAS had a risk for heart failure, even though they didn’t realize it. So then, we looked at the treatment effect of the SGLT2 inhibitor on NT-proBNP values, and what we found was that treatment with the SGLT2 inhibitor lowered NT-proBNP concentrations at subsequent measurements. Which was related to both a reduction in NT-proBNP in the on-treatment arm, but also a rise in NT-proBNP in the usual care arm. 

So, this implies two things; first, that patients with diabetes have a rising risk for heart failure over time – hence the placebo arm had a rise in their NT-proBNP. As well, it also shows that treatment with the SGLT2 inhibitor in this population of patients with diabetes, reduced the concentrations of NT-proBNP. This might imply that treatment with the SGLT2 inhibitor reduces risk for heart failure, mediated through NT-proBNP reduction, and we found in mediation analyses that the reduction in NT-proBNP by one year was one of the predictors of a lower risk for heart failure in the future, mediated through treatment with the SGLT2 inhibitor .

The final observation, which is exceptionally important, was that an elevated NT-proBNP at baseline, in this case a concentration greater than 125 picograms per milliliter, strongly prognosticated a wide range of cardiovascular, renal, and neurologic complications, with statistically significant prediction across each endpoint that we examined in adjusted analyses.

Naturally, an open question therefore was, if NT-proBNP at baseline predicted risk for onset of heart failure, stroke, myocardial infarction, death, and other complications, including progression of kidney disease, interestingly, the question that was asked was could an elevated NT-proBNP at baseline therefore predict benefit from the SGLT2 inhibitor? We know that specific SGLT2 inhibitors reduces many of these events, so maybe it’s the patients with an elevated value at baseline for NT-proBNP. It turns out that even in the setting of a low NT-proBNP, the SGLT2 inhibitor reduced these events, however, those patients with an elevated NT-proBNP value had the highest absolute risk for these events, and treatment with the SGLT2 inhibitor, because the relative risk reductions were similar in high and low risk patients, the absolute reductions in risk associated with the SGLT2 inhibitor treatment were greatest in those patients with an elevated NT-proBNP value.

Taken together, these results indicate this interesting interplay between SGLT2 inhibitors, NT-proBNP, and the benefits of SGLT2 inhibitors to reduce cardiovascular, neurologic, and kidney events.


CarDiaLogue (2021). NT-proBNP and Cardiovascular Disease risk reduction in CANVAS Study [Accessed July 01, 2021]

Cardiovascular Diabetology. (2019). https://cardiab.biomedcentral.com/articles/10.1186/s12933-019-0869-2 [Accessed July 01, 2021]

Updates from Leading Experts

STRONG-HF: Post-discharge Heart Failure Management and Implementation of GDMT Heart Failure Therapy

Prof Alexandre Mebazaa shares the key results and highlights from the STRONG-HF study that assessed the safety, tolerability of optimization of guideline directed medical therapy. [...]


CANVAS study: NT-proBNP and CVD risk reduction

Can you give us a short description of the CANVAS program? The CANVAS program was a set of two trials, whose goal was to evaluate the role of an SGLT2...

1 July 2021
Dr James L Januzzi