3. Integration of GLP-1RA or SGLT2i in Patients With T2D With or at Risk of atherosclerotic CVD

PICO 4: In patients with T2D and either atherosclerotic CVD (ASCVD) or high CV risk, what is the role of novel anti-hyperglycemic agents compared with placebo for reduction of a composite of CV death, nonfatal MI, or nonfatal stroke?

The initial trial that used empagliflozin published in 2015, through to the most recent trial that used efpeglenatide published in 2021 were evaluated in detail in the accompanying systematic review and meta-analysis.^[7] Table 1 shows that MACE was reduced similarly by both classes with a relative risk reduction of 12%-14%. These classes were also associated with similar relative risk reductions in all-cause (12%-15%) and CV mortality (13%-15%). Reduction in nonfatal MI was noted only with SGLT2i but the effect was modest (10% relative risk reduction). Because this effect was not statistically different from the neutral effects on nonfatal MI associated with the GLP-1RA class, we make no recommendation on the basis of this end point. The SGLT2i class showed significant relative risk reduction for the prevention of the composite kidney outcomes (35%) and for hospitalization for HF (32%) compared with placebo and superior to GLP-1RA. As noted previously, currently we do not have any large clinical trials for the treatment of HF or CKD in patients with or without T2D using GLP-1RA. Finally, the important but less common outcome of nonfatal stroke was reduced with GLP-1RA, particularly in the Semaglutide Unabated Sustainability in Treatment of Type 2 Diabetes (SUSTAIN) 6 trial (semaglutide once weekly injection) and Researching Cardiovascular Events With a Weekly Incretin in Diabetes (REWIND; dulaglu- tide) trials.^{[9],[46],[47]} The systematic review and meta-analysis (Table 1) indicate a relative risk reduction of nonfatal stroke of 16% associated with use of GLP-1RA.

General discussion

The principles of pharmacotherapy for patients with T2D have been thoroughly reviewed by Diabetes Canada.^[3],^[4]Achievement of target glucose levels, especially in the early years after T2D diagnosis, reduces the incidence and progression of microvascular complications and, in the long term (more than 10 years), is associated with reduced CV outcomes.^[48-51] In parallel with achieving the A1c goal, it is also recommended that GLP-1RA or SGLT2i be included for patients with T2D with or at high risk of ASCVD to reduce cardiorenal risk, irrespective of A1c. Thus, substitution of (replacing rather than adding) an agent with cardiorenal benefit might be appropriate if people are at or near A1c target.

The choice of initial pharmacotherapy has emerged as an area of uncertainty in patients with newly diagnosed T2D who have or are at risk of ASCVD. Although most guidelines continue to recommend metformin as first-line anti-hyperglycemic therapy, the European Society of Cardiology^[52] recommends that GLP-1RA or SGLT2i should be first-line therapy in individuals with ASCVD or at high or very high CV risk. Although there have been no specific trials to show cardiorenal benefit for GLP-1RA or SGLT2i when used as first-line therapy or as monotherapy or in newly diagnosed T2D, the benefit seen in the CV outcome trials has not been found to vary with the duration of diabetes, suggesting that similar benefits might be seen early in the course of disease.4 The benefits are also not dependent on the presence of metformin.^[8],[53-55] Therefore, the inclusion of GLP-1RA or SGLT2i at the time of diagnosis of T2D in patients with ASCVD or multiple risk factors is a reasonable option and aligns with the views of Diabetes Canada.^[3],[4] In addition, the traditional role of metformin in the early management of T2D is not always appropriate if not tolerated or contraindicated (eg, eGFR < 30 mL/min/1.73 m²). In HF, including HFrEF and HFpEF, SGLT2i used in addition to other evidence-based HF therapies but without treatment with metformin were shown to improve major HF-related outcomes and quality of life within a short period of time after initiation of therapy. Moreover, benefit was seen in patients with and those without T2D.^[8],^[53-55]

SGLT2i reduce hospitalization for HF^[15],[20] and reduce progression of nephropathy^[15],[19] in persons with T2D and CV risk factors only; benefits to reduce MACE or mortality, at least within the short-term duration of the trials to date, are less certain.^[56],[57] Conversely, in such patients, GLP-1RA seem to reduce MACE,^{[9],[12],[13]} a factor that might help selection between SGLT2i or GLP-1RA for A1c reduction. Our analyses (Table 1) indicate that the reduction of nonfatal stroke is strongest for GLP-1RA, which might also factor into the initial choice of classes.

Opinions vary about whether beneficial effects are general to a class or specific to individual agents. Although network meta-analyses have attempted to provide comparisons of specific SGLT2i or GLP-1RA, no head-to-head trials are currently available that help differentiate between medications within either of these 2 classes.^[56],[58] Consequently, the writing group consensus emphasizes class effects but recognizes that some outcomes have been associated with specific agents (Table 3). Using both classes together to achieve glycemic targets when needed appears to be a reasonable option. However, it is not known whether additional cardiorenal benefit can be expected by combining both classes, although the potential mechanisms might be complementary. The most recent GLP-1RA CV outcome trial showed similar benefit whether the patient was using SGLT2i or not.^[10],[59] Finally, an individualized approach to therapy should also weigh the individual’s preferences, costs and coverage, side effect profile, consideration of kidney function and glucose-lowering efficacy, desire for weight loss, and comorbidities such as frailty. Although diminished kidney function attenuates the glucose-lowering effects of SGLT2i, cardiorenal protection is maintained with an eGFR > 20-25 mL/min/1.73 m².^{[21],[22],[36]}