Comparable Cardiorenal Benefits of SGLT2 Inhibitors and GLP-1RAs in Asian and White Populations: An Updated Meta-analysis of Results From Randomized Outcome Trials

Sodium–glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide 1 receptor agonists (GLP-1RAs) are two newer antidiabetes classes of medication that have shown distinct cardiovascular (CV) and renal benefits, primarily in White populations and people of European ancestry. However, over half of the people living with diabetes worldwide are of Asian ethnicity (1). Previous reports have suggested race and ethnicity variations in the pathophysiology and metabolic phenotype of type 2 diabetes (T2D) (2,3). For example, Asian individuals had a higher risk of T2D despite a lower BMI compared with non-Asian populations (4). In addition, Asian individuals have a lower β-cell function and experience a greater decrease in insulin sensitivity compared with White individuals, making Asian populations more susceptible to diabetes than White populations (2,3,5). Therefore, there has been great interest in demonstrating the extent to which the cardiorenal benefits of these newer agents are generalizable to Asian individuals. The findings of one prior meta-analysis with a focus only on Asian patients with T2D suggested a significant reduction in the risk of major adverse CV events (MACE) with GLP-1RAs, compared with placebo, but not with SGLT2 inhibitors (6). Another meta-analysis that included both Asian and White populations with T2D suggested that Asian individuals derived greater CV death and/or hospitalization for heart failure (HHF) benefits from SGLT2 inhibitors and greater MACE benefits from GLP-1RAs compared with their White counterparts (7). Nevertheless, since these two meta-analyses, five clinical trials have been published that address the cardiorenal benefits of these newer therapies, including Evaluation of Cardiovascular Outcomes in Patients With Type 2 Diabetes After Acute Coronary Syndrome During Treatment With AVE0010 (Lixisenatide) (ELIXA) (8), Evaluation of Ertugliflozin Efficacy and Safety Cardiovascular Outcomes Trial (VERTIS CV) (9), the Effect of Sotagliflozin on Cardiovascular and Renal Events in Patients with Type 2 Diabetes and Moderate Renal Impairment Who Are at Cardiovascular Risk (SCORED) trial (10), Dapagliflozin Effect on Cardiovascular Events trial (DECLARE-TIMI 58) (11), and the Effect of Sotagliflozin on Cardiovascular Events in Patients with Type 2 Diabetes Post Worsening Heart Failure (SOLOIST-WHF) trial (12). Therefore, we aimed to conduct an updated meta-analysis by incorporating the most recent evidence to compare cardiorenal benefits of SGLT2 inhibitors and GLP-1RAs between Asian and White populations.

Unlike pairwise meta-analysis in pooling direct evidence (within each pairwise comparison) only, network meta-analysis, in combining both direct and indirect evidence, enables us to compare multiple treatment effects in one analysis (13). Thus, in this updated meta-analysis, we also conducted a network meta-analysis to compare treatment effects between SGLT2 inhibitors and GLP-1RAs among Asian individuals.

We conducted and reported this meta-analysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (14).

We systematically searched PubMed, Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL) from inception to 28 March 2021 using relevant search terms (Supplementary Table 1). We also checked the reference lists of included trials to include additional studies.

We screened retrieved citations and included eligible trials according to the following criteria: 1) randomized, placebo-controlled CV and renal outcome trials designed to assess the cardiorenal safety and efficacy of SGLT2 inhibitors and GLP-1RAs, 2) trials included adults (age ≥18 years) with or without T2D, 3) trials compared SGLT2 inhibitors or GLP-1RAs with each other or with placebo/no treatment), and 4) trials reported the outcome of MACE, CV death/HHF, and composite renal outcomes (including end-stage kidney disease, a sustained decline in the estimated glomerular filtration rate, death from renal or CV causes, or a doubling of the serum creatinine level) with stratification by race/ethnicity (Asian vs. White populations). Details about the above outcome definitions for each trial can be found in Supplementary Table 2.

Two reviewers independently extracted the data using a predesigned form. We extracted the following data for each article: baseline participant characteristics, study drugs and controls, follow-up duration, number of patients included in the study (including Asian and White populations), and end point data (hazard ratio [HR] for Asian and White participants, respectively).

The risk of bias for each study was judged as low, high, or unclear according to Cochrane risk-of-bias tool in the following five domains: sequence generation, allocation concealment, blinding, detection bias, and attrition bias (15).

In this study, we performed a pairwise meta-analysis to compare treatment effects between Asian and White populations and then carried out a network meta-analysis to evaluate the comparative treatment effect between SGLT2 inhibitors and GLP-1RAs on MACE in Asian patients.

We first conducted a pairwise meta-analysis to compare the effect of SGLT2 inhibitors and GLP-1RAs with placebo on MACE, CV death/HHF, and composite renal outcomes. We calculated the pooled HRs and 95% CIs within each drug class using a random-effects model. The magnitude of heterogeneity between studies was assessed with the I² statistic, with values of 25%, 50%, and 75% representing low, moderate, and high levels of heterogeneity, respectively (16). Next, we tested the interaction between treatment effect and race (Asian vs. White) using a meta-regression analysis (17). We assessed publication bias using funnel plots when at least 10 trials were included in each meta-analysis (18,19).

To compare the treatment effect for MACE between SGLT2 inhibitors and GLP-1RAs within Asian individuals, we performed a frequentist network meta-analysis using a random-effects model that accounts for the inconsistency and heterogeneity within a network model (20,21). We estimated HRs and 95% CIs for pairwise comparisons among SGLT2 inhibitors, GLP-1RAs, and placebo. If there was direct and indirect evidence of any particular pairwise comparison (a closed loop in the network), we assessed the transitivity (also called consistency) in the network using the Q statistic under the assumption of a full design-by-treatment interaction random-effects model. If the consistency test was upheld, it implied that the direct and indirect intervention effects are consistent (22).

All statistical analyses were performed with Stata (version 14) and R (netmeta package). A two-tailed P value <0.05 was considered statistically significant.

From a total of 9,648 citations that we identified using our search strategy, we included 17 outcome trials enrolling a total of 127,526 participants, among whom 15,397 (13.4%) were identified as Asian (8–12,23–34) (Fig. 1). Trial and patient characteristics are summarized in Table 1. The median follow-up period was 2.4 years (range 0.75–5.4). Of the 17 trials included, 14 included patients who had T2D and were at high risk for CV disease, established CV disease, or kidney disease (8–12,23–31); 3 were in patients with heart failure (12,32,34); and 3 were in patients with chronic kidney disease (10,31,33). We identified 7 trials (8,23–28) comparing GLP-1RAs with placebo and 10 trials (9–12,29–34) comparing SGLT2 inhibitors with placebo. Overall, the risk of bias for all included trials was judged as low (Table 1).

The results from the pairwise meta-analysis are summarized in Fig. 2, and details can be found in Supplementary Figs. 1–4.

In four outcome trials investigators reported the MACE with SGLT2 inhibitors in patients who had T2D and were at high risk for or established CV disease (9,11,29,30). The HRs for Asian and White participants were 0.85 (95% CI 0.63, 1.14) and 0.91 (95% CI 0.84, 0.98), respectively. The treatment effect was not different between Asian and White populations (P = 0.55). The outcome of CV death/HHF was available in six trials (9–12,32,34). SGLT2 inhibitors were significantly associated with a decrease of risk, by 35% in Asian participants (HR 0.65; 95% CI 0.53, 0.81) and by 20% in White participants (HR 0.80; 95% CI 0.73, 0.88), with a P = 0.13 suggesting no difference between Asian and White groups. With restriction to the patients with heart failure (12,32,34), the HRs for Asian and White populations were 0.61 (95% CI 0.49, 0.75) and 0.77 (95% CI 0.65, 0.91), without observation of a significant difference between Asian and White groups (P = 0.20). For two trials reporting the composite renal outcomes in patients with kidney disease (31,33), SGLT2 inhibitors significantly reduced the risk, by 34% in Asian participants (HR 0.66; 95% CI 0.51, 0.85) and by 34% in White participants (HR 0.66; 95% CI 0.57, 0.78), and there was no significant difference between Asian and White groups (P = 0.97).

Seven trials reported the MACE associated with GLP-1RAs in patients who had T2D and were at high risk for or established CV disease (8,23–28). GLP-1RAs were significantly associated with a decreased risk in Asian participants (HR 0.74; 95% CI 0.61, 0.90) and White participants (HR 0.90; 95% CI 0.83, 0.98). No significant interaction effect of treatment by White/Asian ethnicity was identified (P = 0.10). Low to moderate levels of heterogeneity between studies were observed in the above meta-analyses (Fig. 2). Publication bias was not assessed due to inclusion of <10 trials in our meta-analysis.

The network plot and network meta-analysis results of available interventions for MACE are presented in Fig. 3. The network plot showed that the network in this analysis was star shaped, without observation of any closed loops. Consistent with the results from pairwise meta-analysis, in our network meta-analysis of 10 trials involving Asian participants who had T2D and were at high risk for or established CV disease (8,9,23–30) we found that GLP-1RAs (HR 0.74; 95% CI 0.61, 0.90) and SGLT2 inhibitors (HR 0.83; 95% CI 0.66, 1.06) were associated with a decreased risk of MACE compared with placebo. We did not find a significant difference between GLP-1RAs and SGLT2 inhibitors in MACE (HR 0.89; 95% CI 0.65, 1.21). A consistency test was not allowed due to the absence of a closed loop in the network.

In this study, we found no significant difference between Asian and White populations in the CV and renal benefits of SGLT2 inhibitors and GLP-1RAs. In addition, the results from the network meta-analysis demonstrated comparable benefits of SGLT2 inhibitors and GLP-1RAs for MACE in Asian patients.

Findings of Lee et al. (7) in a prior meta-analysis suggested that Asian individuals derived greater CV death/HHF benefits from SGLT2 inhibitors and greater MACE benefits from GLP-1RAs compared with their White counterparts. In contrast, in our meta-analysis—including data from five additional clinical trials (n = 4,269 additional Asian individuals)—we found that the cardiorenal benefits of SGLT2 inhibitors and GLP-1RAs were comparable between Asian and White populations. The inconsistent results between the two meta-analyses are likely to be explained by the discrepancies in the number of clinical trials included in different analyses. We consider that the inclusion of more trials and especially more Asian individuals in a meta-analysis can result in a higher degree of precision in comparison of the cardiorenal benefits between Asian and White populations. For example, the previous meta-analysis only included two SGLT2 inhibitor trials to conclude that there was a lower risk of CV death/HHF in Asian patients than in White patients (HR 0.60 vs. 0.82, respectively; P = 0.01) (7), while our current data showed no significant difference between the two ethnic groups after inclusion of a third trial (HR 0.61 vs. 0.77; P = 0.20). Similarly, for the meta-analysis of GLP-1RAs trials, there was no significant difference between Asian and White populations for MACE when one more trial was included (P = 0.10 for the current meta-analysis vs. P = 0.03 for the previous meta-analysis by Lee et al. [7]). Indeed, the results from the current meta-analyses are in line with existing studies focusing on the treatment effects on glycemia and other biomarkers. For example, among patients with T2D taking SGLT2 inhibitors, a meta-analysis (included 17 trials with Asian patients and 39 trials with non-Asian patients) showed no significant difference in HbA_1c reduction, weight loss, or blood pressure change between Asian and non-Asian groups (35). Moreover, the risk of all-cause mortality was similar between the two ethnic groups (35). Among T2D patients under treatment with GLP-1RAs, the Asian- and White-dominant groups showed similar glucose-lowering effects (P = 0.90) (36).

One prior meta-analysis focusing on Asian patients with T2D suggested a significant reduction in MACE risk with GLP-1RAs, compared with placebo, but not with SGLT2 inhibitors (6). There has not yet been a study with comparison of SGLT2 inhibitors and GLP1-1RAs in Asian individuals. The results from our network meta-analysis suggested a comparable CV benefit between SGLT2 inhibitors and GLP-1RA among Asian patients with T2D. Nevertheless, it is noteworthy that the strength of inference made in the current network meta-analysis is weakened by the absence of a closed loop in the network. Clinical trials and real-world data for comparing the treatment effects between SGLT2 inhibitors and GLP-1RAs in Asian patients with T2D are needed to support clinical decisions on treatment selection.

We acknowledge that our study is subject to limitations. First, this meta-analysis is based on stratified data from randomized controlled trials, none of which were designed to evaluate the effects of race/ethnicity on treatment. Thus, there may be an imbalance in distributions of risk factors across racial/ethnic subgroups. For example, multiple factors, such as concurrent medications, diet and exercise, socioeconomic factors, and treatment adherence, might not be adequately controlled for in comparisons between Asian and White populations in this study. Second, it is known that there is ethnic heterogeneity in insulin sensitivity and β-cell function between South and East Asian populations (37). However, we were unable to conduct such subgroup analyses because of the lack of information on finer categories for Asian participants. Third, there was no direct comparison across different treatment strategies (all were intervention vs. placebo) in this study, which weakened the strength of inference made in the network meta-analysis between different interventions and made it impossible for us to assess the consistency between direct and indirect evidence.

In conclusion, in CV and renal outcome trials with evaluation of newer glucose-lowering drugs, Asian participants appear to derive cardiorenal benefits from SGLT2 inhibitors and GLP-1RAs similar to those seen in White participants. Furthermore, the two therapeutic classes seem to have comparable CV benefits among Asian participants who were included in these outcome trials. Our data support the wide adoption of either of these two contemporary therapies among Asian individuals who have T2D and are at high risk for or have established CV disease to improve long-term health outcomes. However, given the limitations in this study, future studies with use of real-world data are warranted to further confirm findings from the current meta-analysis.

Duality of Interest. H.T. is a consultant at EvidPro, LLC. No potential conflicts of interest relevant to this article were reported.

Author Contributions. H.T. and J.G. had the idea for the study and led the study design. H.T. and W.S. identified and selected trials and extracted data. H.T. performed all data analyses, checked for statistical consistency, and interpreted results. All authors contributed to data interpretation. H.T. drafted the report, and all other authors critically reviewed the manuscript.