Treatment options following metformin in primary prevention populations with type 2 diabetes: which is the right road to take?

“Two roads diverged in a yellow wood,

And sorry I could not travel both”

Robert Frost in the “Road Not Taken”

1. Introduction

According to the most recent ADA/EASD consensus report on the management of hyperglycemia in type 2 diabetes (T2D), agents with proven cardiovascular (CV) or renal benefits should be considered for the management of people with established atherosclerotic vascular disease (ASCVD), heart failure (HF) and/or chronic kidney disease (CKD), irrespectively of their glycated hemoglobin (HbA1c) values [1]. In particular, glucagon-like peptide 1 receptor agonists (GLP-1 RAs) should be preferred when ASCVD predominates. Differently, sodium-glucose cotransporter-2 inhibitors (SGLT-2is) should be the first choice in people with HF and/or CKD, whereas each of the two categories can be added on top of the other, if patients are unable to meet glycemic targets.

While the therapeutic algorithm adopts a straight-forward approach in suggesting specific drug classes in individuals within secondary prevention, the second step after metformin in primary prevention populations with T2D is still less well defined. The use of sulfonylureas (SUs) and meglitinides in the daily clinical setting presents specific advantages, given that these agents are inexpensive, rendering them an important choice in a setting of limited financial resources. They also manifest strong hypoglycemic effects, although it is known that these are not sustainable over time. The main concern over their use is that they might induce severe hypoglycemic reactions due to their mechanism of action, given that SUs stimulate insulin release independently of blood glucose levels. Although recent studies have been reassuring regarding the CV safety of the class, there is some uncertainty about a potential unfavorable effect of hypoglycemia per se on CV outcomes [2]. Thiazolidinediones (TZDs) are also cheap and effective in controlling blood glucose by dramatically improving insulin resistance. While pioglitazone has been proved to reduce the risk of macrovascular events, the CV safety of rosiglitazone has been repeatedly questioned [3].

There is a recommendation for choosing therapies involving a low risk of hypoglycemia, including dipeptidyl peptidase (DPP)–4 inhibitors, TZDs and the new agents when there is a ‘compelling need to minimize hypoglycemia.’ DPP-4 inhibitors present an excellent safety profile and the cost of treatment with these drugs is expected to fall significantly, once they go off patent. However, hypoglycemia is rarely a problem in solely metformin-treated individuals. Therefore, in a setting of adequate financial resources as it is a big part of the modern Western world, SGLT-2is or GLP-1 RAs should be considered as the most appropriate add-on to biguanides, when there is a need to promote weight loss, which is the case with the vast majority of the T2D population. The question then arises: which of the two is the best option for our patients, or in other words, which is the right way to take?

2. SGLT2is versus GLP-1 RAs: metabolic effects and safety profile

We are privileged to practice diabetology in the era of Cardiovascular Outcome Trials (CVOTs), whose findings have shifted diabetes care from a glucose-centric to a complication-centric perspective. Improvement in glycemia has been shown to be a key mediator of CV benefits observed in various CVOTS [4]; therefore, metabolic outcomes remain of paramount importance in the management of T2D and both drug classes have proven significant benefits in this field. Although head-to-head trials are currently unavailable, it could be argued that GLP-1 RAs can promote better reductions in HbA1c and body weight compared with SGLT-2is. Actually, these differences might become even greater in the near future, when GLP-1 and glucose‐dependent insulinotropic polypeptide agonists are introduced in daily practice. These dual agonists have been shown to be more powerful in reducing glucose and weight than the existing diabetes therapies. Regarding blood pressure lowering efficacy, GLP-1RAs present significant variability, whereas the SGLT2i effects are more homogenous. Therefore, it is difficult to claim that one class provides a clear advantage compared to the other [5].

It is also worth noting that the metabolic effects of these drugs might present interindividual variability. For example, glycemic and weight loss response to GLP-1 RAs have been shown to relate to genetic polymorphisms and baseline weight and HbA1c of participants in relevant studies [6]. Similarly, weight loss and glucose reduction promoted by empagliflozin as add-on to metformin therapy were greater among people with greater body mass index and HbA_1c before treatment initiation [7]. Obesity has been demonstrated to be a stronger driver of treatment choice than CV or renal risk in T2D [8], highlighting the importance of individual patient characteristics when prescribers are asked to build the therapeutic regimen.

Other important aspects to consider include the administration route and adverse events of each class. Although an oral preparation of semaglutide is available, GLP-1 RAs are mostly injectable therapies at the moment, and this might be a barrier in adherence and persistence to treatment for a number of patients [9]. Moreover, real-world data reveal high rates of treatment discontinuation (even as high as 47.7% at 12 months) among GLP-1 RAs recipients [10]. This might relate to the administration route or to gastrointestinal symptoms, such as nausea and vomiting, which however, wane over time in the majority of people treated with these agents. On the other hand, diabetic ketoacidosis is a potentially lethal adverse effect of SGLT-2is; however, it is rare in the daily clinical setting. Mycotic genital infections often occur with these drugs, but they are in most cases easily manageable and do not always require treatment discontinuation [11].

3. SGLT2is versus GLP-1 RAs for primary prevention: is there a winner?

Micro- and macrovascular complications continue to represent a huge burden for both persons with T2D and health care systems around the world. Despite the progress made during the past few years, more people with T2D die from CV disorders than any other cause [12], whereas diabetes remains the leading cause of end-stage renal disease (ESRD) in the Western world [13]. DPP-4 inhibitors are still more widely used as a second step after metformin than the new agents, despite a similar cost to SGLT-2is and an abundant body of evidence showing a neutral effect of the former on cardiorenal outcomes [14]. On the other hand, it should be noted that robust data on cardiorenal protection for both SGLT-2is and GLP-1 RAs are mainly available at the secondary prevention level, whereas the situation is less clear regarding primary prevention populations with T2D. The scenario becomes even more puzzling, considering data showing that in a large proportion of people with T2D, CV disease, HF, or renal disease remain subclinical and undiagnosed [15], thus, making it difficult to draw clear lines between primary and secondary prevention in diabetes.

When considering the cardiac and renal protective properties of the new agents, prescribers should take into account the different mechanisms of action mediating these benefits. In the case of GLP-1 RAs, these are mainly driven by anti-oxidative, anti-inflammatory and anti-arteriosclerotic effects. This fact explains why these agents mainly reduce ASCVD outcomes but also why enough time is needed before their clinical benefits become evident [5]. Differently, SGLT2is rapidly improve HF and renal outcomes (within the first weeks of their administration) through primarily hemodynamic effects, mediated by osmotic diuresis, natriuresis, improved erythropoiesis, ketogenesis and reduced intraglomerular pressure, among other proposed mechanisms [11].

A recent meta-analysis by Tsapas et al. demonstrated that among people with T2D at low CV risk, no treatment is superior to placebo for reducing vascular outcomes [16]. Differently, another meta-analysis including 4 large-scale SGLT-2is trials showed that the benefit of the class for major adverse cardiovascular events (MACE), hospitalization for HF, CV death and death from any cause was consistent across patient subgroups and independent of the existence of CV disease or HF at baseline [17]. In a similar way, a recently published post hoc analysis of DECLARE-TIMI focusing on the effects of dapagliflozin on the primary prevention subgroup, revealed significant reductions in HF and renal disease risk, which were in line with those of the entire cohort and of the group with established CV disease [18]. On the other hand, ertugliflozin failed to significantly reduce the risk for MACE and renal events compared with placebo in VERTIS-CV [19]. The latter was predominantly a secondary prevention trial; however, its findings questioned whether the cardiac and renal protective properties of SGLT2is are class effects, although inconsistencies in outcomes could be also attributed to different populations studied in various CVOTs.

However, this does not seem to be exactly the case with GLP-1 RAs. In a meta-analysis including data from 56,004 participants and exploring the effect of GLP-1 RAs on MACE reduction, the hazard ratio for the primary prevention group was 0.95 (95% Confidence Interval 0.83–1.08), mainly driven by data from one trial (REWIND with dulaglutide) [20]. Taking into account the relatively large number needed to treat (NNT = 72) seen in the REWIND for the population without established CV disease and that this was mostly driven by stroke reduction and only observed in non-US populations [21], the evidence regarding the primary prevention potential of the class could not be characterized as totally convincing, at least for the moment. Another point to consider is that although both classes seem to reduce kidney outcomes, only SGLT-2is favorably affect ‘hard’ renal endpoints, such as death from renal causes or progression to ESRD, with these benefits observed across the full spectrum of kidney function and albuminuria. In contrast, renal benefits of GLP-1 RAs are predominantly driven by a reduction in macroalbuminuria [1,5].

Table 1 presents a critical evaluation of the main recommendations made by the 2019 ADA/EASD consensus statement.

Table 1. A critical evaluation of the main recommendations made by the 2019 ADA/EASD consensus statement

Recommendation	Critical evaluation
Metformin as the first-line therapy in every person with T2D	Metformin is safe, cheap and confers multiple metabolic benefits. However, data regarding its potential CV benefits mainly come from observational trials and a nested UKPDS study, including only a small number (n = 342) of overweight individuals
SGLT2is and GLP-1 RAs should be considered in people with T2D and established CV disease, HF and/or CKD	These recommendations are based on the findings of well-designed, large-scale, randomized CVOTs showing reduced risk of cardiorenal events in people with T2D treated with the new agents
SGLT2is and GLP-1 RAs should be considered in high-risk individuals independently of HbA1c levels	Mediation analyses of CVOTs have indicated improvement in glycemia as a significant mediator of CV benefits observed with the new agents
When cost is a major issue, SUs and TZDs should be considered for the management of T2D	SGLT2is and GLP-1 RAs might not be affordable or even available in certain parts of the world. On the other hand, hypoglycemia and complications are the main drivers of excessive diabetes cost for health care systems. Thus, the new agents might prove cost-effective in the long run

4. Perspective

It becomes evident that selection of the right drug for the right patient is a clinical challenge. The greatest challenge, however, is to tackle clinical inertia, which might deprive people with T2D of therapy with great metabolic and cardiorenal benefits [14]. Available data show that the introduction of new therapies into the daily clinical setting remains low. This holds true even for those who would benefit most from such an intervention, as happens in people with HF and T2D, who receive SGLT-2is only in a small proportion and certainly contrary to available evidence [22]. Conversely, it is dubious whether it would be cost-effective to treat all people with T2D with SGLT-2is and/or GLP-1 RAs. Although it is difficult to provide a definite answer at the moment, there is evidence suggesting that complications, hospitalizations, and emergency room (ER) visits due to hypoglycemia are the main drivers of increased costs for diabetes care [23]. Hence, the adoption of new therapies might lead to substantial savings for health care systems, through reductions in hospitalizations and ER visits, thus, being cost-effective in the long run.

In conclusion, there are currently more robust data for primary prevention with SGLT-2is than with GLP-1 RAs in people with T2D. Both classes confer remarkable metabolic and cardiorenal benefits through distinct mechanisms of action, offering the opportunity for a tailored approach to patient care. Considering that life with diabetes is more of a marathon than a sprint, many people with T2D will eventually need both classes to control their blood glucose levels and address cardiorenal complications, taking advantage of additive effects on glycemia, weight and potentially other comorbidities, such as fatty liver. At the end of the day, choosing between two equally good roads is rather a fortune than a trouble.

Declaration of interests

T Koufakis has received honoraria as a speaker from AstraZeneca, Boehringer Ingelheim, Pharmaserve Lilly and Novo Nordisk and has participated in sponsored studies by Eli-Lilly. K Kotsa has received honoraria for lectures/advisory boards and research support from Astra Zeneca, Boehringer Ingelheim, Pharmaserve Lilly, Sanofi-Aventis, ELPEN, MSD, and Novo Nordisk. N Papanas has been an advisory board member of Astra-Zeneca, Boehringer Ingelheim, MSD, Novo Nordisk, Pfizer, Takeda and TrigoCare International; has participated in sponsored studies by Astra-Zeneca, Eli-Lilly, GSK, MSD, Novo Nordisk, Novartis and Sanofi-Aventis; has received honoraria as a speaker for Astra-Zeneca, Boehringer Ingelheim, Eli-Lilly, Elpen, MSD, Mylan, Novo Nordisk, Pfizer, Sanofi-Aventis and Vianex; and attended conferences sponsored by TrigoCare International, EliLilly, Galenica, Novo Nordisk, Pfizer and Sanofi-Aventis. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contribution

TK reviewed the literature and drafted the first version of the manuscript. NP, PZ and KK reviewed the literature and critically evaluated the text. All authors have approved the final version and agree to be accountable for all aspects of this editorial article.

ABBREVIATIONS

T2D: type 2 diabetes; CV; cardiovascular; UKPDS: United Kingdom Prospective Diabetes Study; HF: heart failure; CKD: chronic kidney disease; HbA1c: glycated hemoglobin; CVOTs: cardiovascular outcome trials; TZDs: thiazolidinediones; SUs: sulfonylureas; SGLT2is: sodium-glucose cotransporter-2 inhibitors; GLP-1 RAs: glucagon-like peptide 1 receptor agonists

Additional information

Funding

This paper was not funded