VERTIS-CV Results 'Disappoint,' Raise More Questions

The VERTIS-CV trial results, presented at this year's virtual American Diabetes Association (ADA) Scientific Sessions, raise more questions than they answer.

Before we dive into interpreting the implications of these results and the purported cardiovascular (CV) benefits of SGLT2 inhibitors, let's review some commonsense caveats:

• We still need to wait for the peer-reviewed publication from VERTIS-CV.

• CV outcome trials (CVOTs) are not head-to-head, with numerous study differences, making comparisons difficult.

Nonetheless, the presenters of the VERTIS-CV session asserted that the overall results of this new study are in line with other SGLT2-inhibitor outcome trials, based on a new meta-analysis combining their results with EMPA-REG, CANVAS, CREDENCE, and DECLARE-TIMI 58.

But can clinical decisions be based solely on meta-analyzing such heterogeneous trials? Or should we first consider the overall neutral results (primary and all secondary endpoints) of the CVOT in question? And then, as the next step, compare like-vs-like CVOTs before meta-analyzing a variety of CVOTs and chronic kidney disease outcome trials?

Why did they jump straightaway to comparing and combining apples and oranges?

Comparing Like-vs-Like CVOTs

VERTIS-CV is similar to EMPA-REG in multiple ways. Both trials had similar baseline characteristics and background glucose-lowering and cardioprotective medications, which is expected, given that the enrollment criteria for both studies were limited to secondary prevention populations with type 2 diabetes (ie, those with an event history of coronary artery disease, cerebrovascular disease, or peripheral vascular disease).

Both trials randomly assigned patients 1:1:1 to two doses of the SGLT2 inhibitor in question, combining and comparing CV endpoints for the two doses together against the respective placebo arms. Both trials had similar durations of treatment exposure (mean of 2.9 years for VERTIS-CV and median of 2.6 years for EMPA-REG). Both trials found similar differences in A1c, systolic blood pressure, and weight at the end of the trial periods, favoring the SGLT2 inhibitor arms over the placebo arms. The placebo arms of both trials had similar rates for the primary outcome of 3-point major adverse CV events (MACE; composite of CV death, nonfatal myocardial infarction, and nonfatal stroke) — around 4 per 100 patient-years.

Despite all of these similarities, it is striking that VERTIS-CV does not even show a hint of reduction in MACE and CV death with ertugliflozin, unlike EMPA-REG, which achieved a significant relative risk reduction of 14% for MACE, 38% for CV death, and a number needed to treat of only 39 for all-cause mortality over 3 years with empagliflozin.

So Why the Differences?

A clear explanation for these stark CV outcome differences could not be ascertained from the presentation. Possible reasons that will need to be explored in the main VERTIS-CV publication and later analyses include the following:

• Could the differences in outcomes be explained by study dropouts? EMPA-REG had around 25% of participants discontinuing the study drug, whereas roughly 31% of participants in VERTIS-CV discontinued the study medication. Not only does this affect the statistical power for VERTIS-CV, but we also don't yet know how many of those who discontinued their study medication started taking a prescription SGLT2 inhibitor instead.

• Was there differential "drop-in" of other cardioprotective diabetes medications among trial arms in VERTIS-CV? Because of positive results from the LEADER, SUSTAIN-6, and REWIND trials, GLP-1 receptor agonists have gained widespread acceptance following guideline recommendations for cardioprotection. It is plausible that this change in practice may have influenced the standard of care for the high-risk secondary prevention population in VERTIS-CV because of its timeline (completed in 2019), whereas this was not a factor for EMPA-REG (completed in 2015).

• Was the difference in results due to a protocol amendment in the middle of the trial, which doubled the original sample size of VERTIS-CV to 8000 participants? The proportions of and outcomes for participants randomly assigned before and after the protocol amendment were not presented at the ADA meeting. In addition to any differences in standard of care during the two time periods, another important question is whether VERTIS-CV had less statistical power (compared with EMPA-REG) to detect a difference in CV endpoints because of this amendment.

• Do the differences in reported baseline congestive heart failure (CHF) explain the results? The only apparent difference in baseline characteristics in VERTIS-CV was the 24% history of CHF vs 10% in EMPA-REG. On the basis of the DAPA-HF trial, the higher proportion of CHF should have resulted in a larger reduction in the secondary endpoint of CV death and hospitalization for heart failure in VERTIS-CV. Conversely, it can be postulated that over the past 5 years, the sensitivity of diagnosing and managing CHF has increased among healthcare providers. Could this improved diagnosis and management of CHF have led to a reduced statistical power to detect a difference in VERTIS-CV?

• Did the two global trials enroll participants from different geographical areas? For example, the proportion of trial participants from Asia in EMPA-REG was 19% compared with 6.4% in VERTIS-CV; the proportion from South America was 15% in EMPA-REG vs 8.8% in VERTIS-CV. Could population differences in these geographical regions have a part in explaining the neutral VERTIS-CV results?

If none of the above provide a plausible explanation, we may be scratching our heads about alternative explanations:

• Are ertugliflozin and empagliflozin truly different for CV endpoints despite similarities in molecular targets, SGLT2 selectivity, clinical metabolic effects, etc.?

• Are the differences in trial results a play of chance?

Contrary to the differences in MACE endpoints, hospitalization for heart failure and composite renal outcomes for VERTIS-CV align with results from EMPA-REG, although they are of nominal and borderline significance, respectively, in this latest trial.

The Bottom Line

Even though the VERTIS-CV results disappoint, the neutral MACE results of this latest CVOT reinforce the 2020 ADA standards of care regarding the use of SGLT2 inhibitors and GLP-1 receptor agonists for cardioprotection. An SGLT2 inhibitor with demonstrated CV disease benefit is preferred in patients with established heart failure or diabetic kidney disease, whereas a GLP-1 receptor agonist with demonstrated cardiovascular disease benefit is preferred in those with atherosclerotic cardiovascular disease and those with multiple CV risk factors.

Harpreet S. Bajaj, MD, MPH, is a community endocrinologist in Brampton, Ontario, Canada, and vice chair of the Diabetes Canada Guidelines. His clinical and research interests are the prevention and management of diabetes and its related complications. He is the founder of STOP Diabetes Foundation and volunteers with numerous community public health organizations to raise awareness of diabetes prevention and treatment.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube