Benefits and harms of sodium‐glucose co‐transporter‐2 inhibitors (SGLT2‐I) and renin–angiotensin–aldosterone system inhibitors (RAAS‐I) versus SGLT2‐Is alone in patients with type 2 diabetes: A systematic review and meta‐analysis of randomized controlled trials

Abstract Introduction It is uncertain if the combination of sodium‐glucose co‐transporter 2 inhibitors (SGLT2‐Is) and renin‐angiotensin‐aldosterone system inhibitors (RAAS‐Is) provides better cardio‐renal clinical outcomes in people with type 2 diabetes mellitus (T2DM) compared with SGLT2‐Is alone. Using a systematic review and meta‐analysis of randomized controlled trials (RCTs), we evaluated the efficacy and safety with respect to cardio‐renal outcomes of the combination of SGLT2 and RAAS inhibitors vs SGLT2‐Is in patients with T2DM. Methods Studies were identified from MEDLINE, Embase, the Cochrane Library and search of bibliographies to May 2021. The Cochrane risk of bias tool was used to assess the risk of bias of each study. Study‐specific risk ratios (RRs) with 95% confidence intervals (CIs) were pooled. Quality of the evidence was assessed using GRADE. Results Nine articles comprising 8 RCT evaluations (n = 34,551 participants) that compared SGLT2‐Is with placebo in patients with T2DM against a background of standard care and reported subgroup results for those treated with or without RAAS‐Is at baseline were included. No RCT specifically investigated the combination of SGLT2 and RAAS inhibitors compared with SGLT2‐Is alone. The RRs (95% CIs) for composite cardiovascular outcome and composite CVD death/heart failure hospitalization comparing SGLT2‐Is vs placebo in patients on RAAS‐Is were 0.93 (0.85–1.01) and 0.88 (0.76–1.02), respectively. The corresponding estimates for patients not on RAAS‐Is were 0.78 (0.65–0.93) and 0.73 (0.65–0.82), respectively. There was no evidence of interactions between RAAS‐I status and the effects of SGLT2‐Is for both outcomes. Single study results showed that SGLT2‐Is vs placebo reduced the risk of composite kidney outcome and cardiovascular death in patients with RAAS inhibition. The effect of SGLT2 inhibition vs placebo on kidney parameters, genital infections, volume depletion, hyperkalaemia, hypokalaemia, hypoglycaemia and other adverse events was similar in patients with or without RAAS inhibition. The quality of the evidence ranged from very low to moderate. Conclusions Aggregate published data suggest that the combination of SGLT2 and RAAS inhibitors in the treatment of patients with T2DM may be similar in efficacy and safety if not superior to SGLT2‐Is alone. Head‐to‐head comparisons of the two interventions are warranted to inform T2DM management. The use of SGLT2 inhibition as a first‐line therapy in T2DM or its early use in the prevention of renal deterioration and cardiovascular complications in addition to its glycaemic control deserves further study.


| INTRODUC TI ON
Diabetes is a global public burden-it is a leading cause of morbidity, mortality and places substantial socioeconomic and financial pressures on individuals, health systems and global economies. 1,2 In 2015, diabetes (with type 2 diabetes being the most common type) was the sixth leading cause of disability. 3 Chronic kidney disease (CKD), due to diabetic nephropathy, is a common complication in people with type 2 diabetes, 4 with cardiovascular disease (CVD) being the leading cause of morbidity and mortality associated with type 2 diabetes.
Currently, about 422 million people worldwide have diabetes 5 and it has been projected that 592 and 642 million will have diabetes by 2035 and 2040, respectively. 6,7 People with type 2 diabetes need intensive management of glucose and risk factors such as lipids and blood pressure to reduce the risk of disease progression and complications. 8 With the rising global tide of established risk factors such as obesity, physical inactivity and high energy diets, complications and deaths attributable to diabetes will proportionately increase if there is no concomitant improvement in its management. 6 Lifestyle and metformin are the first-line treatment of choice for patients with type 2 diabetes, unless contraindicated in specific situations such as those with advanced renal impairment. 1 Because the kidneys are involved in the pharmacokinetic processing of many antidiabetic drugs [9][10][11][12] or their mechanisms of action, 13 prescribing antidiabetic drugs in patients with type 2 diabetes and renal impairment can be very challenging. There are limited treatment options for glycaemic control in these patients.
Sodium-glucose co-transporter 2 inhibitors (SGLT2-Is) (dapagliflozin, canagliflozin, empagliflozin and ertugliflozin) are the latest therapeutic agents for the treatment of type 2 diabetes. They increase excretion of glucose in the urine by inhibiting glucose reabsorption. 14 Their use is associated with reductions in glycated haemoglobin (HbA1c) levels, systolic blood pressure (SBP), albuminuria and weight loss. 13 There is substantial evidence that SGLT2-Is reduce the risk of cardiovascular outcomes in those at high risk, the need for heart failure hospitalization and the progression of kidney impairment. [15][16][17] SGLT2-Is alone do not cause hypoglycaemia and exert beneficial effects without having significant adverse effects.
Their main common side effect is genital mycotic infections. 18 SLGT2-Is are less effective for glucose control in patients with moderate-to-severe renal impairment (estimated glomerular filtration rate, GFR 30-60 ml/min/1.73 m 2 ) 18 and are not recommended in many guidelines for glycaemic control in people with estimated GFR less than 30 ml/min/1.73 m 2 . For several decades, renin-angiotensinaldosterone system inhibitors (RAAS-Is) (angiotensin-converting enzyme inhibitors (ACE-Is), angiotensin-II type 1 receptor blockers (ARBs) and more recently direct renin inhibitors (DRIs)) have been employed to reduce the rate of progression of diabetes nephropathy in people with type 2 diabetes. 19 Substantial evidence also suggests that RAAS-Is reduce the risk of cardiovascular events. 19,20 SGLT2 and RAAS inhibitors each independently reduce the risk of cardiovascular and kidney complications associated with type 2 diabetes and they appear to have synergistic effects when used as combination therapy. 21,22 Hence, it will be clinically relevant to know the effectiveness of combining SGLT2 and RAAS inhibitors versus SGLT2-Is alone. Though a number of landmark trials comparing SGLT2-Is with placebo have reported outcomes among subgroups of patients with or without RAAS inhibition, no previous systematic review has synthesized the existing evidence. In this context, using a systematic review and meta-analysis of randomized controlled trials (RCTs), we aimed to evaluate whether the combination of SGLT2 and RAAS inhibitors has a superior efficacy and safety profile than SGLT2-Is alone in patients with type 2 diabetes.

| Data sources and search strategy
We registered this systematic review and meta-analysis in the PROSPERO prospective register of systematic reviews (CRD42021251601). It was conducted using a predefined protocol and in accordance with PRISMA guidelines (Appendix 1). MEDLINE, Embase and the Cochrane Library electronic databases were searched from 2012 (being the year of approval of the first SGLT2 similar in patients with or without RAAS inhibition. The quality of the evidence ranged from very low to moderate.

Conclusions:
Aggregate published data suggest that the combination of SGLT2 and RAAS inhibitors in the treatment of patients with T2DM may be similar in efficacy and safety if not superior to SGLT2-Is alone. Head-to-head comparisons of the two interventions are warranted to inform T2DM management. The use of SGLT2 inhibition as a first-line therapy in T2DM or its early use in the prevention of renal deterioration and cardiovascular complications in addition to its glycaemic control deserves further study.

K E Y W O R D S
RAAS inhibitor, SGLT2 inhibitor, Type 2 diabetes inhibitor (dapagliflozin) in the European Union) to 08 May 2021 with no restriction on language. The computer-based searches combined terms related to the intervention (eg SGLT2 inhibitor, dapagliflozin, canagliflozin, empagliflozin and ertugliflozin), comparator (eg RAAS inhibitor, ACE-I), ARB, DRI) and population (eg type 2 diabetes) in humans. A RCT design search filter was employed. Details on the search strategy are provided in Appendix 2. Titles and abstracts of all initially identified citations were initially screened by one author (SS) to assess their suitability for potential inclusion, followed by the acquisition of full texts for detailed evaluation. Full-text evaluation was independently conducted by two authors (SS and SKK).
The reference lists of key studies and review articles were manually scanned for additional studies.

| Study selection and eligibility criteria
Randomized controlled, open or blinded trials that assessed the effects of the combination of SGLT2 and RAAS inhibitors compared with SGLT2 inhibitors in adults with type 2 diabetes and reported on renal or cardiovascular outcomes or adverse events were eligible. Randomized controlled trials that had also compared SGLT2-I treatment with a placebo or standard care and reported outcomes according to whether patients were receiving RAAS-Is or not at baseline were considered. We excluded the following: (i) studies that specifically enrolled only patients with known renal insufficiency or established renal parenchymal disease without diabetes mellitus and (ii) studies that recruited patients with a history of diabetic ketoacidosis, type 1 diabetes mellitus, history of hereditary glucosegalactose malabsorption, primary renal glucosuria or renal disease that required treatment with immunosuppressive agents.

| Data extraction
One author (SKK) initially extracted data from eligible studies using a predesigned data collection form and a second author (SS) independently checked the data with that in original articles. A consensus was reached in case of any inconsistency with involvement of a third (KK). Data were extracted on the following: first author, publication year, study year, specific study design, baseline population including duration of years with type 2 diabetes, proportion of men, geographical location, average age at baseline, numbers enrolled and randomized, allocation concealment, blinding, type of SGLT2-I and dosage; duration of treatment or follow-up; treatment comparisons; and nature of outcome events and their numbers. We extracted risk estimates when reported.

| Outcomes
The primary outcomes were defined as (i) the 3-point major adverse cardiovascular events (MACE), composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke (composite cardiovascular outcome) and (ii) serum creatinine doubling, initiation of renal replacement therapy (RRT) or death from renal disease (composite renal outcome). Secondary outcomes were (i) cardiovascular death, (ii) heart failure (HF) hospitalization, (iii) composite outcome of cardiovascular death or HF hospitalization, (iv) decline in estimated GFR, (v) RRT, (vi) doubling of serum creatinine level, (vii) other renal and cardiovascular outcomes, (viii) glycaemic measures and haemodynamic and metabolic parameters, and (ix) adverse events.

| Risk of bias
The risk of bias of each of the included trials was assessed using the Cochrane Collaboration's risk of bias tool. 23

| Quality of evidence
We assessed the quality of the body of evidence on each outcome using the Grading of Recommendations Assessment, Development and Evaluation (GRADEpro) tool (https://gdt.grade pro.org), based on study limitations, inconsistency of effect, imprecision, indirectness and publication bias. 24 We rated the quality as four levels: high, moderate, low and very low.

| Statistical analysis
Summary measures of association were reported as risk ratios (RRs) with 95% confidence intervals (CIs). Risk ratios were pooled using a fixed effects model given the few studies available for pooling and the absence of substantial heterogeneity across studies. Standard chi-square tests and the I 2 statistic were used to quantify the extent of statistical heterogeneity across studies. 25,26 We employed random effects meta-regression to assess for interactions between RAAS inhibition status and the effect of SGLT2-Is. 27 Only two metaanalysis could be carried out due to limited data. Given the variety of measures reported for some outcomes and inconsistent reporting by some of the trials, a formal meta-analysis could not be performed for some of the outcomes. A narrative synthesis was performed for studies that could not be pooled. The findings of such studies were summarized in tables that included the main characteristics of the study and the results in natural units as reported by the investigators. All tests were two-tailed, and p-values of 0.05 or less were considered significant. All analyses were conducted using Stata version MP 16 (Stata Corp). Following detailed evaluation, 10 articles were excluded because (i) population was not relevant (n = 7); (ii) duplicate studies (n = 2); and (iii) treatment comparison not relevant (n = 1). The remaining nine articles met the inclusion criteria and were included in the review. 28-36

| Study characteristics and risk of bias
The nine articles comprised eight studies, of which one was based on a pooled individual patient data (IPD) analysis of 13 trials (Table 1).
No RCT specifically investigated the combination of SGLT2 and RAAS inhibitors compared with SGLT2-Is alone. All eligible studies were based on trials that had investigated the effects of SGLT2 in-

| Composite kidney outcome
The composite kidney outcome was reported by only one study.

| Composite outcome of cardiovascular death or HF hospitalization
In pooled analysis of four trials, the RR (95% CIs) for the compos-

| Cardiovascular death
The outcome of cardiovascular death was reported by one study.

| Estimated GFR
Two studies reported the effect of SGLT2 inhibition vs placebo on estimated GFR changes across the subgroup of RAAS-I users. 29,35 In the EMPA-REG OUTCOME trial, the initial change in estimated GFR in patients taking the combination of empagliflozin and RAAS-Is was higher than for those taking empagliflozin alone ( Table 2). The  long-and post-treatment changes were also similar for both groups.
In the study that pooled data across 13 trials, the effect of dapagliflozin on estimated GFR was similar in patients with or without RAAS inhibition 35 (Table 2).

| Change in albuminuria
The EMPA-REG OUTCOME trial reported changes in albuminuria status for the effect of empagliflozin vs placebo across the subgroup of RAAS-I users. 29 Though there appeared to be an improvement in albuminuria status in those with baseline RAAS inhibition than those without (Appendix 4), the report noted that there was no significant evidence of interaction across the subgroup. The effect of dapagliflozin on urinary albumin-to-creatinine ratio (UACR) was similar in patients with or without RAAS inhibition in the pooled analysis of 13 trials 35 (Table 2).

| Other kidney outcomes
The EMPA-REG OUTCOME trial reported outcomes for incident or worsening nephropathy, oedema and acute renal failure for the effect of empagliflozin vs placebo across the subgroup of RAAS-I users. 29 Empagliflozin compared with placebo reduced the risk of the incident or worsening nephropathy, oedema and acute renal failure in those on RAAS inhibition at baseline; the risk was only reduced for oedema in those who were not on RAAS inhibition treatment (Appendix 5).

| Metabolic and haemodynamic parameters
In the pooled analysis of 13 trials, the effect of dapagliflozin on HbA1c and haematocrit was similar in patients with or without RAAS inhibition; however, mean reductions in body weight, serum uric acid, SBP and diastolic blood pressure (DBP) were more distinct in patients without RAAS inhibition treatment compared with those with RAAS inhibition treatment at baseline 35 ( Table 2). The effects of SGLT2 inhibition compared with placebo on the risk of hyperkalaemia, hypokalaemia and hypoglycaemia were similar in both groups (Appendix 6).

| Volume depletion
Comparing SGLT2-Is with placebo in those on RAAS-I treatment at baseline, the RR (95% CIs) for volume depletion in pooled analysis of F I G U R E 3 Risk for the composite outcome of cardiovascular death or heart failure hospitalization comparing SGLT2 inhibition with placebo in patients with or without RAAS inhibition treatment at baseline. CI, confidence interval (bars); NR, not reported; RAAS-I, reninangiotensin-aldosterone system inhibitor; RR, risk ratio; SGLT2-I, sodium-glucose co-transporter 2 inhibitor Overall, adverse effects were more common in those with RAAS inhibition treatment at baseline compared with those without -Abbreviations: DBP, diastolic blood pressure; GFR, glomerular filtration rate; HbA1c, glycated haemoglobin; IPD, individual patient data; SBP, systolic blood pressure; SGLT2-I, sodium-glucose cotransporter 2 inhibitor; RAAS-I, renin-angiotensin-aldosterone system inhibitor; T2DM, type 2 diabetes mellitus; UACR, urinary albumin-to-creatinine ratio.

| Genital and urinary tract infections
In pooled analysis of two trials, the effect of SGLT2 inhibition vs placebo on genital infections was similar in patients with or without RAAS inhibition (Appendix 8). For urinary tract infection (UTI), the effect of SGLT2 inhibition vs placebo on UTI was also similar in patients with or without RAAS inhibition (Appendix 9).

| Key findings
In this systematic review and meta-analysis from available RCTs, we have evaluated the efficacy and safety outcomes in patients with

| Comparison with previous studies
A previous pooled analysis of individual level data from 13 placebocontrolled trials investigating the effects of dapagliflozin on cardiorenal risk factors in patients with type 2 diabetes with increased albuminuria treated with or without RAAS-Is at baseline reported similar clinically relevant improvements in metabolic and haemodynamic parameters. 35 In a meta-analysis of 8 RCTs which compared combined therapy of SGLT2-Is and ACEIs/ARBs with placebo plus ACEIs/ARBs in patients with type 2 diabetes, the combination therapy showed significant reduction in glycaemic parameters, body weight, blood pressure and lower risk of adverse events. 21 Another recent meta-analysis demonstrated that combination therapy with SGLT2-Is and ACEIs/ARBs compared with ACEIs/ARBs was welltolerated and achieved better control of blood pressure, improvement of renal outcomes, alleviation of long-term renal function and a decrease in blood glucose and body weight, but an increased risk of hypoglycaemia. 22 To our knowledge, this is the first aggregate metaanalysis to attempt to evaluate whether the combination of SGLT2 and RAAS inhibitors provides better cardio-renal clinical outcomes in patients with type 2 diabetes compared with SGLT2-Is alone. Our overall results suggest that treatment with SGLT2-Is provides similar clinical effectiveness and safety in patients with type 2 diabetes treated with or without RAAS inhibition. The combination of SGLT2 and RAAS inhibition may improve some renal outcomes and parameters such as body weight and blood pressure compared to SGLT2 inhibition alone, but further evaluation is needed.

| Potential explanation of findings
For the past two decades, landmark trials 37,38 have demonstrated that renin-angiotensin-aldosterone system blockade is an efficacious method for the protection of both cardiovascular and renal systems. Despite this, there is some residual risk for both cardiovascular and renal outcomes, 39 thus necessitating the requirement for further additive treatment options. In our analysis, SGLT2 inhibition compared with placebo reduced the risk of major cardiovascular outcome, but this reduction did not reach statistical significance, as this achievement was expected in a well-treated population on renin-angiotensin-aldosterone system blockade.

| Strengths and limitations
The strengths of the current evaluation deserve consideration.

| CON CLUS IONS
In conclusion, emerging data suggest that the combination of SGLT2  Writing-original draft (supporting); Writing-review & editing (equal).

DATA AVA I L A B I L I T Y S TAT E M E N T
The corresponding author had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. This study is based in data from published articles.