Why Choose Between SGLT2 Inhibitors and GLP1-RA When You Can Use Both?

The Time to Act Is Now

Alice Y.Y. Cheng, MD


Circulation. 2021;143(8):780-782. 

If one were to reflect on why people involved in the management of type 2 diabetes (T2D) work as hard as they do, especially patients with T2D, the answer would be to improve quality of life through reducing the burden of T2D complications and prolonging life while minimizing the burden of the treatment itself. To support these goals, guidelines and society recommendations from around the world have promoted a multifactorial approach to the management of T2D, with emphasis on glycemic, blood pressure, and low-density lipoprotein cholesterol control and therapeutic lifestyle modifications.[1–3] Although multifactorial management remains a cornerstone of T2D management, there has been a shift in the past 5 years to establish another cornerstone, which is the preferential selection of therapies proven to improve cardiovascular and kidney clinical outcomes. Some of these therapies with cardiovascular and kidney benefits are well-established, such as statins, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and acetylsalicylic acid, when appropriate. Now there are 2 additional therapies that can be added to the list—sodium-glucose cotransporter-2 inhibitors (SGLT2i) and glucagon-like peptide receptor agonists (GLP-1RA)—on the basis of multiple large randomized cardiovascular outcome trials.[4,5] Benefits have been demonstrated in a variety of patients with T2D, ranging from those with multiple risk factors to those with established atherosclerotic cardiovascular disease (ASCVD) or kidney disease.

Although the cardiovascular and kidney benefits of SGLT2i and certain GLP-1RA are well-established, the question remains whether their combined use has any added risk reduction. Despite the lack of definitive evidence, guidelines and society recommendations have promoted the addition of an SGLT2i after a GLP-1RA or vice versa for patients with or at high risk of ASCVD and for patients with chronic kidney disease.[1,3,6] This has been justified on the basis of the different mechanisms of action for SGLT2i and GLP-1RA and the common-sense approach of preferentially choosing a therapy with at least some evidence of benefit over agents with evidence of no benefit. Although this is logical, comparative clinical trial data in support of this advice would be helpful. The individual cardiovascular outcome trials are not able to adequately answer this question given the small numbers of patients who were on both therapies, largely a function of the timing of the various studies. The few short-term trials of combination therapy only evaluated metabolic measures, demonstrating additional metabolic benefits of the combination without providing any outcome data.[7] In this issue of Circulation, Dave et al[8] provide a glimpse into what the effects of combining SGLT2i and GLP-1RA may be on a composite cardiovascular end point of myocardial infarction, stroke, and all-cause mortality, as well as heart failure hospitalization.

In this observational study using real-world data from 3 insurance claims databases in the United States, people with T2D who were already taking GLP-1RA who had added either SGLT2i or sulfonylurea from April 2013 and up to June 2018 were identified. After 1:1 propensity score matching on >95 variables, 12 584 patients in each group were analyzed for the primary endpoints of composite cardiovascular end point and heart failure hospitalization. Compared with the initiation of sulfonylurea, the addition of SGLT2i to GLP-1RA therapy was associated with lower composite cardiovascular end point incidence (pooled adjusted hazard ratio, 0.76 [95% CI, 0.59–0.98]) and lower incidence of heart failure hospitalization (pooled adjusted hazard ratio, 0.64 [95% CI, 0.50–0.82]), which is strikingly similar to what was observed in the randomized cardiovascular outcome trials of SGLT2i, particularly the Kaplan-Meier curve suggesting that the separation in heart failure hospitalization began within the first 3 months of medication exposure. The lower incidence of composite cardiovascular end point appeared to be driven primarily by numerically lower incidence of myocardial infarction and all-cause mortality and not stroke. To support the findings, multiple sensitivity analyses were conducted, which were consistent with the primary analyses. Analyses of the unmatched, unadjusted cohorts were also provided and demonstrated lower incidence for the primary outcomes of comparable magnitude. The majority of sulfonylurea used was glimepiride, a sulfonylurea that has been shown to have cardiovascular effects not inferior to linagliptin from analyses of a randomized, head-to-head cardiovascular outcome trial.[9] This adds credibility to the current findings.

The results from these observational analyses support the hypothesis that the beneficial cardiovascular effects of SGLT2i and GLP-1RA therapy in T2D may be additive. The mechanisms of cardiovascular benefit from SGLT2i and GLP-1RA remain unclear but the proposed mechanisms are not contradictory and may even be complementary, with SGLT2i focusing on hemodynamics, kidney, vascular, and myocardial energetics, and GLP-1RA focusing on metabolic, anti-inflammatory, and possible direct vascular, cardiac, and kidney effects.[10] There are limitations to this type of analysis and the authors outlined these in detail along with their efforts to limit their effect. Despite these efforts, this is not a randomized controlled trial and the potential for residual confounding attributable to lack of randomization remains. One could imagine a randomized cardiovascular outcome trial comparing the addition of SGLT2i versus other antihyperglycemic therapy on a background of GLP-1RA, or vice versa. The question also remains whether one should use SGLT2i first or GLP-1RA first or both therapies up front in a patient with T2D with or at high risk of ASCVD. Large randomized cardiovascular outcome trials could be designed to answer these questions, but is it realistic to expect these trials to be conducted and, perhaps more importantly, would it be a wise use of limited resources? The largest barrier is not the lack of impressive, well-conducted randomized outcome trials, but rather the inadequate implementation of the existing knowledge.[11–14] A cross-sectional study of people with T2D across 13 countries from 5 continents conducted in 2019 estimated the prevalence of any cardiovascular disease to be 34.8% (95% CI, 32.7–36.8) and ASCVD to be 31.8% (95% CI, 29.7–22.7).[13] The use of either SGLT2i or GLP-1RA among those with any cardiovascular disease was only 21.5% and only 21.4% among those with ASCVD, which was similar to use in patients without any cardiovascular disease (22.2%).[14] This already represents an improvement, as a study of the use of guideline-directed therapy among patients with T2D and ASCVD in a single insurance claims database in the United States showed only 8% use of either SGLT2i or GLP-1RA as of January 2018.[10] It was estimated that if 100% of those patients were to receive either SGLT2i or GLP-1RA therapy, there would be an additional 3.3% absolute reduction in 5-year composite of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death.[11] If all the existing evidence-based recommendations were implemented in 70% of the world's population with T2D (age 30 to 69 years), it is estimated that there would be 800 000 fewer premature deaths in 3 years.[15] Thus, given the limited collective resources, would they not be better spent tackling the issue of implementation rather than more expensive trials of SGLT2i and GLP-1RA therapies?

All practitioners involved in the care of people living with T2D need to heed these data as a call to action. Tremendous resources have been spent proving the clinical benefits of SGLT2i and certain GLP-1RA. The low implementation rates are a concern and it would be a shame to waste the efforts of the many investigators and the thousands of patients who volunteered to participate in the many cardiovascular and renal outcome trials, with the common ambition of ultimately improving the lives of others. There are many factors that contribute to the low implementation rates and detailed strategies have been proposed in a recent Lancet Commission on Diabetes, aptly published on World Diabetes Day.[15] Cost and access challenges persist and with time, as certain therapies become less expensive, these challenges will lessen. As individual practitioners, we should reflect on our own practices and determine what our personal barriers may be, such as discomfort with the data, lack of time, lack of reminders, discomfort with the practical aspects of using these therapies, uncertainty as to which practitioner should initiate, uncertainty as to which practitioner should follow-up after initiation, or discomfort with addressing potential adverse effects. All of these barriers are surmountable. SGLT2i and GLP-1RA may have started their lives as antihyperglycemic therapies, but they have proven to do much more than glucose-lowering. As practitioners who provide care for people with T2D, be it cardiology, endocrinology, primary care, internal medicine, or nephrology, we are responsible for ensuring that these patients receive evidence-based outcome-reducing therapies. Instead of struggling over which one (GLP-1RA or SGLT2i) to start first in a given patient with T2D with or at high risk of cardiovascular disease, we must pick one and start, knowing that many patients will require additional treatment and that the combination (evidence now for adding SGLT2i to GLP-1RA) is highly likely to provide further outcome reduction. The time to act is now.