Association of Sodium-Glucose Cotransporter 2 Inhibitors With Cardiovascular Outcomes in Patients With Type 2 Diabetes and Other Risk Factors for Cardiovascular Disease

Key Points Question What is the updated magnitude of benefit associated with sodium-glucose cotransporter 2 inhibitors (SGLT2-Is) on outcome of cardiovascular death or hospitalization for heart failure (HHF) in different select subgroups of patients? Findings This meta-analysis of 10 high-quality randomized clinical trials (71 553 participants) found that use of SGLT2-Is was associated with lower occurrence of cardiovascular death or HHF by 33% in high-risk patients. For the primary outcome of HHF or cardiovascular death, the study showed equal benefit of SGLT2-Is in both sexes; lower risk in participants who were younger than 65 years compared with those 65 years or older; and greater risk reduction in participants who were Asian, Black, or of other races or ethnicities compared with White participants. Meaning These findings suggest that SGLT2-Is may be associated with an overall cardiovascular benefit for patients with high-risk cardiovascular features compared with placebo in both sexes, different age groups and different races and ethnicities.


Introduction
Sodium-glucose cotransporter 2 inhibitors (SGLT2-Is), also known as a glifozins, constitute a class of medication that was initially approved as an antidiabetic agent because the mechanism of action consisted of lowering blood glucose levels by promoting excretion of glucose through the kidneys via renal tubules. 1 Failure and Reduced Ejection Fraction), 10 DAPA-CKD (Dapagliflozin in Patients with Chronic Kidney Disease), 11 and EMPEROR-Reduced (Cardiovascular and Renal Outcomes with Empagliflozin in Heart Failure) 12 trials enrolled both patients with and without diabetes. Cardiovascular benefits in patients with heart failure regardless of the presence of diabetes in the DAPA-HF, EMPEROR-Reduced, and SOLOIST-WHF trials included only heart failure among patients with reduced ejection fraction. 3,5,6 The VERTIS-CV, 8 EMPA-REG OUTCOME, 6 SCORED, 4 DECLARE TIMI, 7 and CANVAS 5 trials included patients with established cardiovascular disease (CVD) or at least 1 risk factor for CVD. These trials compared cardiovascular outcomes, with each trial focusing on different risk factors in the population. Meanwhile, the new VERTIS-CV 8 trial showed that the outcome of death from cardiovascular causes and hospitalization for heart failure (HHF) did not differ between the SGLT2-I group and the placebo group. A recently published trial, SOLOIST-WHF, 9 studied the cardiovascular outcomes in patients with worsening of existing heart failure requiring hospital treatment. All RCTs found significant reductions in cardiovascular death and HHF in patients receiving SGLT2-Is except CANVAS, 5 which found a reduction in HHF but not in cardiovascular death. All these studies showed a variable degree of benefit in the SGLT2-I group compared with the placebo group for cardiovascular death or HHF.
Our meta-analysis aimed to interpret various outcomes from these studies and compare them with one another. This meta-analysis is taken from large RCTs within the last 7 years. Furthermore, this meta-analysis ensures the generalizability of data from all these trials because patients with and without diabetes, with and without heart failure, and with cardiovascular risk factors were studied and analyzed altogether. In contrast, prior meta-analyses by Butler et al 13 and Docherty et al 14 primarily studied patients with heart failure and their cardiovascular outcomes, whereas our study included patients without heart failure and accounted for the presence of cardiovascular risk factors.
Another meta-analysis by Zelniker et al 15 is similar to our meta-analysis but was performed in 2019 and hence did not include the emerging data from our study. Furthermore, the present meta-analysis is the first large study, to our knowledge, that has also focused on cardiovascular outcomes in both sexes and different age and racial and ethnic groups to understand the magnitude of benefit across the subgroups, because this is essential to initiate SGLT2-I use as standard therapy. reporting guideline providing the search strategy to obtain all eligible studies. The PRISMA flow diagram and the reasons for study exclusion are provided in Figure 1.

Selection of Studies
The following standard criteria were set to select eligible RCTs comparing outcomes of SGLT2-I use vs a placebo control group. The baseline characteristics included established atherosclerotic cardiovascular disease (ASCVD), a high risk of ASCVD, or heart failure with or without kidney disease and diabetes; in addition, the trial reported our primary outcome, which is cardiovascular death or  HHF. After removing the duplicates, a total of 568 studies were screened. A total of 52 studies were selected to screen abstracts with or without a full study review. Ten high-quality studies were eligible for the meta-analysis. One RCT 16 did not match our outcomes of the cardiovascular benefit because its outcome was an improvement in left ventricular systolic function. Multiple observational studies 17,18 were excluded from the meta-analysis. Several trials that have not yet released their final results were not included in the study. Ten placebo-controlled RCTs met the inclusion criteria. [3][4][5][6][7][8][9][10][11][12] Any discrepancies in data extraction or risk-of-bias assessment were resolved with the consensus of all authors.

Quality Assessment
The modified Jadad Score was used to assess the methodological quality of RCTs. A score of 0 to 8 quantified the quality of each trial; high-quality studies (score Ն3) were included, as shown in eTable 1 in the Supplement.

Outcomes
The primary efficacy outcomes were cardiovascular death or HHF. The secondary outcomes were MACE, HHF, cardiovascular death, acute MI, and all-cause mortality. MACE was defined as death due to cardiovascular causes, nonfatal MI, or nonfatal stroke. Both safety and efficacy outcomes were analyzed by performing a meta-analysis. The prespecified subgroup analyses were performed for sex (men vs women), age groups (<65 or Ն65 years), and racial and ethnic groups, the latter of which were defined in the original trials and were categorized as Asian, Black, or other race or ethnicity (in which the category "other" was not specified consistently) or White for the purposes of our analyses.
In the selected population the benefit of SGLT2-I is not clearly understood when SGLT2-Is are emerging as standard therapy in this current era. The data was abstracted by observing event rates in different race and ethnicity groups from each included trial. The exclusion and inclusion criteria of each study, including primary and secondary outcomes, are all presented separately (eTable 2 in the Supplement)

Data Analysis
Cochrane Review Manager (RevMan) software, version 5.3 (The Nordic Cochrane Center) was used to perform pairwise analysis. The data from studied trials were used to calculate odds ratios (ORs) and 95% CIs comparing the intervention vs the placebo groups. The analysis of all outcomes was performed using a Mantel-Haenszel equation and the random-effects model. A 2-sided P < .05 was considered statistically significant for all analyses. Heterogeneity was tested using I 2 and χ 2 tests. I 2 index values of 25% to 50% were considered low heterogeneity; 51% to 75%, moderate heterogeneity; and greater than 75%, high heterogeneity. Publication bias was evaluated by visual inspection of the funnel plot (eFigure 1 in the Supplement).

Results
A total of 10 trials involving 71 553 patients in a high-risk cardiac disease study group were included in the analysis. 3 The primary outcome of cardiovascular death or HHF was reported in all 10 studied trials, with Further subgroups analyses of the primary outcome were performed based on sex and age ( Figure 2) and race and ethnicity (Figure 3). There was no difference in the outcome of the SGLT2-I group in men and women when compared with the placebo group.

Discussion
This meta-analysis presents the comprehensive pooled data from 10 contemporary large, placebocontrolled RCTs. The major findings include the following. First, cardiovascular death or HHF was decreased in the SGLT2-I group compared with the placebo group. Second, all-cause mortality was decreased in the SGLT2-I group compared with the placebo group. Third, MACE outcomes were decreased in the SGLT2-I group compared with the placebo group. Fourth, the SGLT2-I group had a decreased rate of HHF or emergency department visits for heart failure compared with the placebo group. Fifth, there was no difference in outcome for acute MI between the SGLT2-I and placebo groups.
Our study found that SGLT2-I use was associated with decreased risk of cardiovascular death or HHF by 33%, with a number needed to treat of 5.7 at P < .001. This result was significant when HHF admissions and cardiovascular death were analyzed independently as well. There was a 2.44% decreased event rate in the SGLT2-I group compared with the placebo group. These results are consistent with those reported by Butler et al 13 in their meta-analysis that studied SGLT2-I use Test for overall effect: z = 6.22 (P <.001) 9 Test for subgroup differences: χ 2 = 0.10 (P =.75); I 2 = 0% 1 and CREDENCE, 3 contributed to the positive outcomes of this study. Both trials specifically studied patients with diabetes and chronic kidney disease and had strongly positive MACE outcomes. The VERTIS-CV, 8  that the rate of cardiovascular death or HHF in men is slightly higher than that of women (9.01% vs 5.34%). There were more men in the study than women, and the greater rate of cardiovascular death or HHF in men could be attributed to the difference in the sample size.

JAMA Network Open | Cardiology
For another subgroup analysis, participants were divided into those younger than 65 years and 65 years or older. Age is a significant independent risk factor for cardiovascular death. 19 Heart failure is more common in elderly individuals and is more common in older women than in older men. 19 Our findings suggest that SGLT2-Is effectively reduce the outcome of cardiovascular death or HHF in patients in both age groups with similar risk reduction. However, the cardiovascular death or HHF  Test for subgroup differences: χ 2 = 1.87 (P =.17); I 2 = 46.6% 1 outcome was reduced in patients younger than 65 years when compared with those 65 years or older. This could imply a benefit of starting SGLT2-I therapy at an earlier age rather than an older age (>65 years) to prevent further advance of the disease. There is significant potential to reduce morbidity and mortality caused by CVD and provide substantial economic benefit to one of the leading causes of US health care expenditure.
Similarly, both White groups and groups who were Asian, Black, or of other race or ethnicity had a reduction in cardiovascular death or HHF in the SGLT2-I group compared with the placebo group (ORs, 0.82 and 0.66, respectively). However, the sample size for the groups who were Asian, Black, or of other race or ethnicity was smaller in all the studies (total, 6900 vs 26 646 for the White group).
This finding is important because the rates of CVD and comorbidities were greater in the cohort of participants who were Asian, Black, or of other race or ethnicity, and SGLT2-I use confers benefits to both groups equally. There was a lack of heterogeneity in all subgroups in our study. With the findings from a subgroup analysis, use of SGLT2-Is as a novel therapeutic approach is suggested as an important measure to prevent cardiovascular death or HHF in both sexes, age groups, and racial and ethnic groups regardless of the presence of multiple comorbidities.
The study population in this meta-analysis included patients with heart failure and those at a high risk of ASCVD. We emphasize that at present more than 5.8 million patients have congestive heart failure in the US alone and more than 1 million are hospitalized per year as a result of heart failure. It has become the most common hospital admission diagnosis of patients covered by Medicare and is the leading cause of rehospitalization within 30 days. 20,21 Among patients with heart failure 64 years and older, 23.2% are readmitted within 30 days. Each admission for heart failure incurs an estimated cost of $14 631, resulting in a substantial economic burden. 22 The amalgamated pooled data from these trials in our study suggests an overall significant reduction in cardiovascular death or HHF, MACE outcomes, and all-cause mortality in patients treated with SGLT2-Is with every degree of comorbidity from low to high risk. The study population included only patients with congestive heart failure from high-quality trials, including DAPA-HF, 10 VERTIS-CV, 8 and SOLOIST. 9 Five studies (EMPA-REG outcome, 6 DAPA-HF, 10 EMPEROR REDUCED, 12 VERTIS-CV, 8 and SOLOIST 9 ) include only patients with established cardiovascular disease, and 4 studies (CANVAS, 5 CREDENCE, 3 DECLARE-TIMI, 7 and DAPA-CKD 11 ) include patients with prior ASCVD rates of 65.6%, 50.5%, 40.6%, and 37.4%, respectively.
The cardiac benefits of SGLT2-Is have been hypothesized to be due to several mechanisms. The cardioprotective effects of SGLT2-Is are secondary to an improvement in ventricular load through a reduction in preload by eliciting natriuresis and osmotic diuresis and a decrease in afterload by lowering blood pressure and improving vascular function. 23 In addition, SGLT2-Is have been compared in other studies with loop diuretics; both were associated with a similar natriuretic effect and reduced interstitial fluid. These natriuretic effects may explain the findings in reducing HHF. 23 In addition, SGLT2-Is are believed to improve cardiac metabolism by optimizing the use of ketones. 23 Last, SGLT2-Is have been associated with the inhibition of cardiac fibrosis, which is considered an important heart failure pathway. 24 These mechanisms of action will not be relevant in the pathophysiology of acute ischemia to the myocardium. Sodium-glucose cotransporter 2 inhibitors do not have known antianginal properties or vasodilatory effects, do not reduce myocardial oxygen consumption, and do not prevent remodeling of the cardiac muscle. This is most likely the reason why SGT2-I is not beneficial for acute MI. A previous study 25 analyzing the broad spectrum of safety profiles of SGLT2-Is demonstrated a nonsignificant occurrence of major hypoglycemic events, acute kidney injury, fracture, bladder cancer, Fournier gangrene, amputation, and urinary tract infection but a slightly increased association of diabetic ketoacidosis.