Portuguese-Brazilian evidence-based guideline on the management of hyperglycemia in type 2 diabetes mellitus

Background In current management of type 2 diabetes (T2DM), cardiovascular and renal prevention have become important targets to be achieved. In this context, a joint panel of four endocrinology societies from Brazil and Portugal was established to develop an evidence-based guideline for treatment of hyperglycemia in T2DM. Methods MEDLINE (via PubMed) was searched for randomized clinical trials, meta-analyses, and observational studies related to diabetes treatment. When there was insufficient high-quality evidence, expert opinion was sought. Updated positions on treatment of T2DM patients with heart failure (HF), atherosclerotic CV disease (ASCVD), chronic kidney disease (CKD), and patients with no vascular complications were developed. The degree of recommendation and the level of evidence were determined using predefined criteria. Results and conclusions In non-pregnant adults, the recommended HbA1c target is below 7%. Higher levels are recommended in frail older adults and patients at higher risk of hypoglycemia. Lifestyle modification is recommended at all phases of treatment. Metformin is the first choice when HbA1c is 6.5–7.5%. When HbA1c is 7.5–9.0%, dual therapy with metformin plus an SGLT2i and/or GLP-1RA (first-line antidiabetic agents, AD1) is recommended due to cardiovascular and renal benefits. If an AD1 is unaffordable, other antidiabetic drugs (AD) may be used. Triple or quadruple therapy should be considered when HbA1c remains above target. In patients with clinical or subclinical atherosclerosis, the combination of one AD1 plus metformin is the recommended first-line therapy to reduce cardiovascular events and improve blood glucose control. In stable heart failure with low ejection fraction (< 40%) and glomerular filtration rate (eGFR) > 30 mL/min/1.73 m2, metformin plus an SGLT-2i is recommended to reduce cardiovascular mortality and heart failure hospitalizations and improve blood glucose control. In patients with diabetes-associated chronic kidney disease (CKD) (eGFR 30–60 mL/min/1.73 m2 or eGFR 30–90 mL/min/1.73 m2 with albuminuria > 30 mg/g), the combination of metformin and an SGLT2i is recommended to attenuate loss of renal function, reduce albuminuria and improve blood glucose control. In patients with severe renal failure, insulin-based therapy is recommended to improve blood glucose control. Alternatively, GLP-1RA, DPP4i, gliclazide MR and pioglitazone may be considered to reduce albuminuria. In conclusion, the current evidence supports individualizing anti-hyperglycemic treatment for T2DM.

To develop this guideline, the best evidence available was reviewed and the expert opinions of a Portuguese-Brazilian panel of diabetes specialists were obtained. A list of statements was carefully created and scored. When high-quality evidence was not available from the literature, the panel gave opinions on a variety of clinical scenarios. These opinions were captured and analyzed by an international voting system, which allowed consensus to be reached after multiple rounds of discussion. The main objective of this guideline is to support the decision-making process in clinical practice, taking into account patients' best interests and clinicians' personal preferences.

Methods
The scientific societies appointed 33 specialists with extensive expertise in diabetes to compose the panel. The main clinical topics requiring updated positions in patients with T2DM were heart failure (HF), atherosclerotic CV disease (ASCVD), chronic kidney disease (CKD), and management of T2DM in patients without vascular complications. The panel compiled a narrative review by searching MEDLINE (via PubMed) for randomized clinical trials, meta-analyses, and high-quality observational studies related to type 2 diabetes treatment, using the MeSH terms [ When the results of the search did not yield enough high-quality evidence to answer a specific question or scenario, an expert opinion was sought: a query was sent to all panelists, and responses were recorded. The frequency of responses was analyzed and a consensus opinion was drawn up.
The degree of recommendation depended on the query, following the criteria that are shown in Table 1A. The level of evidence was determined using the same criteria in Table 1B. Specific criteria for atherosclerotic cardiovascular disease (ASCVD) are shown in Table 2 [1]. The panel chose to classify therapeutic options into two groups of glucoselowering agents: antidiabetics with proven CV or renal benefit (AD1) and general glucose-lowering agents (AD). These are specified in Table 3.
The guideline-building process was conducted as described elsewhere [1]. In brief, a preliminary manuscript outlining grades of recommendation and levels of evidence was drafted. Several rounds of discussion were held among the panel members, who reviewed the findings and made suggestions. The manuscript was then returned to the lead author for inclusion of changes. The same procedure was repeated for each of the four modules (ASCVD, HF, CKD, and patients without vascular complications). Subsequently, many other rounds of revision were done by request of subcommittee members. Then, the manuscript was presented for public consultation and discussed with the audience; minor adjustments were suggested. Finally, the consensus version of the document was submitted to the editorial board for final editing and proofreading.

Recommendations and summary of evidence
HbA 1C target 1. In non-pregnant adult patients with T2DM, an HbA 1c target of < 7% is recommended to reduce the incidence of microvascular complications. rate (eGFR) > 30 mL/min/1.73 m 2 , metformin plus an SGLT-2i is recommended to reduce cardiovascular mortality and heart failure hospitalizations and improve blood glucose control. In patients with diabetes-associated chronic kidney disease (CKD) (eGFR 30-60 mL/min/1.73 m 2 or eGFR 30-90 mL/min/1.73 m 2 with albuminuria > 30 mg/g), the combination of metformin and an SGLT2i is recommended to attenuate loss of renal function, reduce albuminuria and improve blood glucose control. In patients with severe renal failure, insulin-based therapy is recommended to improve blood glucose control. Alternatively, GLP-1 RA, DPP4i, gliclazide MR and pioglitazone may be considered to reduce albuminuria. In conclusion, the current evidence supports individualizing anti-hyperglycemic treatment for T2DM.

Summary of evidence
• The UKPDS 33 trial [2] showed that improving glucose control by reducing HbA 1c to a target of below 7% is clearly associated with reduced microvascu-  lar complications. A total of 3867 newly diagnosed patients with T2DM were randomly assigned to intensive treatment (sulfonylurea or insulin) or conventional treatment (diet alone). The aim in the intensive group was to attain a fasting plasma glucose (FPG) of less than 108 mg/dL, versus the best achievable FPG with diet alone in the conventional group. Three aggregate endpoints were considered: (1) any diabetes-related endpoint (sudden death, death from hyperglycemia or hypoglycemia, fatal or non-fatal myocardial infarction, angina, heart failure, stroke, renal failure, any amputation, vitreous hemorrhage, retinopathy requiring photocoagulation, blindness, or cataract extraction); (2) diabetes-related death (death from myocardial infarction, stroke, peripheral vascular disease, renal disease, hyperglycemia or hypoglycemia, and sudden death); and (3) all-cause mortality. After 10 years, the median HbA 1c was 7.0% (interquartile range [IQR], 6.2-8.2%) in the intensive group versus 7.9% (6.9-8.8%) in the conventional group. For any diabetes-related endpoint, risk was 12% lower in the intensive group (95% CI 1-21, p = 0.029) than in the conventional group. The risk reduction in the any diabetes-related composite end- point was largely attributable to a 25% risk reduction (95% CI 7-40, p = 0.0099) in microvascular outcome events. • Similar results were seen in the ADVANCE [3] study, which randomized 11,140 patients with T2DM to undergo either standard or intensive glucose control. After a median of 5 years of follow-up, the mean HbA 1c level in the intensive group was 6.5% compared to 7.3% in the standard-control group. Intensive blood glucose control also reduced the incidence of major microvascular events (new or worsening nephropathy or retinopathy) (HR, 0.86; 95% CI 0.77 to 0.97; p = 0.01). • Although the landmark DCCT trial [4] was conducted in type 1 diabetes (T1DM), the panel considered that the effects of lowering blood glucose and HbA 1c down to 7% by an intensive glucose-lowering regimen with insulin reinforced the target level of < 7% in T2DM. In DCCT, 1441 patients with T1DM were randomized into intensive or conventional treatment with insulin-based therapy, targeting an HbA 1c of less than 6.05%. The HbA 1c attained in the intensive group was around 7%. In relation to conventional therapy, intensive therapy reduced retinopathy by 76% (primary prevention group) and 54% (secondary prevention group). Albuminuria was reduced by 39% (micro) and 54% (macro), whereas peripheral neuropathy was reduced by 60%.

In non-pregnant adult patients with T2DM, an
HbA 1c target < 7% should be considered to reduce the long-term incidence of macrovascular complications.

Summary of evidence
• Long-term post-trial observational follow-up studies [5] have shown that intensive blood glucose control can also decrease macrovascular complications. In the post-trial observational phase of UKPDS, 3277 patients with T2DM were followed for 5 years with no attempts to maintain previously assigned therapies. After the first year, no betweengroup differences in HbA 1c levels remained. In the sulfonylurea/insulin group, relative risk reductions persisted at 10 years for any diabetes-related endpoint (9%, p = 0.04) and for microvascular disease (24%, p = 0.001). However, there were also reduc-tions in risk of myocardial infarction (15%, p = 0.01) and all-cause mortality (13%, p = 0.007). In the metformin subgroup, significant risk reductions persisted for any diabetes-related endpoint (21%, p = 0.01), myocardial infarction (33%, p = 0.005), and all-cause mortality (27%, p = 0.002). • The Epidemiology of Diabetes Interventions and Complications (EDIC) study was a post-trial phase of the DCCT trial [6] in which 93% of DCCT survivors were followed. CV disease (defined as nonfatal myocardial infarction, stroke, death from CV disease, confirmed angina, or need for coronary-artery revascularization) was assessed with standardized measures. During a mean follow-up of 17 years, 46 CV disease events occurred in 31 patients who had received intensive treatment, versus 98 events in 52 patients who had received conventional treatment.
Intensive treatment reduced the risk of any CV disease event by 42% (95% CI 9 to 63%; p = 0.02). The risk of nonfatal myocardial infarction, stroke, or CV death (3P-MACE) decreased by 57% (95% CI 12 to 79%; p = 0.02). The decrease in HbA 1c during the DCCT study was significantly associated with the positive effects of intensive treatment on the risk of CV disease. Intensive glucose control has longterm beneficial effects against CV risk in patients with T1DM as well.

A higher individualized
HbA 1c target level is recommended in frail older adults, in the presence of comorbidities limiting life expectancy or when hypoglycemia is strongly to be avoided.

Summary of evidence
• This postulation was based on expert opinion. There was consensus among the panelists that HbA 1c targets should be higher in special clinical situations, to lower the risk of hypoglycemia.
4. HbA 1c measurements should be obtained at least once every 12 weeks during treatment.

Summary of evidence
• This suggestion was based on the opinion and experience of the great majority of experts in the panel; no specific evidence was found. The objective is to monitor treatment effectiveness and improve adherence.
Initial glucose lowering therapy in treatment-naïve patients with type 2 diabetes mellitus (T2DM) (Fig. 1) 5. Lifestyle modification is recommended during all phases of treatment in T2DM to improve blood glucose control.

Summary of evidence
• Lifestyle measures should be recommended universally as the basis for diabetes treatment, as sustained remission of T2DM is related to the degree of weight loss. The DIRECT study [7] was an openlabel, cluster-randomized, controlled trial conducted at primary health care units in the UK which randomized overweight/obese patients recently diagnosed with T2DM to an integrated structured weight management program (intervention) (n = 149) or the standard of care in accordance with UK guidelines (n = 149 6. In treatment-naïve, non-pregnant adults recently diagnosed with T2DM, without cardiovascular or chronic renal disease, in whom HbA 1c is 6.5-7.5%, METFORMIN monotherapy is recommended to improve blood glucose control and prevent diabetesrelated outcomes.

Summary of evidence
• Metformin is the first-line agent of choice for treatment of T2DM, given its established efficacy, safety profile, low incidence of hypoglycemia and low cost. The efficacy of metformin in reducing diabetesrelated outcomes was demonstrated in overweight and obese patients in the UKDPS 34 trial [9]. The objective of UKPDS 34 was to investigate whether intensive blood-glucose control (treating to a fasting blood glucose of 108 mg/dL) with metformin, a sulfonylurea or insulin had any specific advantage or disadvantage. In a randomized controlled clinical trial, of 4075 patients recruited to UKPDS, 1704 overweight patients with newly diagnosed T2DM (with baseline fasting blood glucose 110-270 mg/ dL) were assigned to either conventional treatment with diet alone (n = 411), intensive control with metformin (n = 342) or intensive control with a sulfonylurea or insulin (n = 951). The median duration was 10.7 years. The primary outcome measures were: any diabetes-related clinical endpoint, diabetes-related death, and all-cause mortality. Overall mean HbA 1c at baseline was 7.2 ± 1.5%. Compared with the conventional group, patients in the metformin group had risk reductions of 32% (95% CI 13-47, p = 0.002) for any diabetes-related endpoint, 42% for diabetesrelated death (9-63, p = 0.017), and 36% for all-cause mortality (9-55, p = 0.011). Among patients allocated to intensive blood glucose control, metformin showed a greater effect than chlorpropamide, glibenclamide, or insulin for any diabetes-related endpoint (p = 0.0034), all-cause mortality (p = 0.021), and stroke (p = 0.032). Intensive glucose control with metformin decreased the risk of diabetes-related endpoints in overweight diabetic patients, and was associated with less weight gain and fewer hypoglycemic attacks than are insulin and sulfonylureas. It is thus the first-line pharmacological therapy of choice in these patients. • Important note: This panel strongly recommends that, before initiating any treatment with antidiabetic agents, the eGFR should be estimated in every patient and the drug initiated only if in accordance with the product label. In the case of metformin, it should not be initiated when eGFR is below 45 mL/ min/1.73 m 2 and should be discontinued whenever eGFR is below 30 mL/min/1.73 m 2 , due to the risk of metabolic acidosis.

7.
In treatment-naïve, non-pregnant adults recently diagnosed with T2DM, without cardiovascular or chronic renal disease, in whom HbA 1c is 6.5-7.5%, DUAL THERAPY, including metformin plus a DPP4i, may be considered to delay hyperglycemia treatment failure.

Summary of evidence
• The VERIFY study [10] was a randomized, multicenter, double-blind, parallel-group trial of newly diagnosed patients aged 18-70 years, diagnosed with T2DM within 2 years of enrollment, with HbA 1c levels between 6.5 and 7.5%. Patients were randomized to "early combination treatment" (metformin and vildagliptin) or "standard care" (metformin monotherapy plus placebo). If HbA 1c did not remain below 7.0% with initial treatment, patients in the metformin group were crossed over to vildagliptin instead of placebo, entering phase 2 of the study, during which all patients were given combination therapy. The primary efficacy endpoint was "time to initial treatment failure", defined as HbA 1c ≥ 7.0% at two consecutive scheduled visits 13 weeks apart from randomization. A total of 2001 participants were randomized. The relative risk for time to failure was significantly lower in the early combination treatment group over the 5-year study period (HR 0.51; 95% CI 0.45 to 0.58; p < 0.0001). Both strategies were safe and well-tolerated. In patients with newly diagnosed T2DM, early intervention with combination therapy (metformin plus a DPP4i) appears to provides greater long-term glycemic control than metformin monotherapy. It is important to highlight that this benefit is restricted to delay treatment failure of glycemic control.
8. In treatment-naïve, non-pregnant, asymptomatic adults recently diagnosed with T2DM in whom HbA 1c is 7.5% to 9.0%, DUAL THERAPY is recommended to improve blood glucose control.

Summary of evidence
• The efficacy and safety of multiple dual therapies were compared with those of monotherapies in a metaanalysis of drug-naïve T2DM patients [11]. A total of 36 clinical trials in T2DM, longer than 12 weeks, in which initial therapy with two antidiabetic agents were compared to one agent were included. The primary endpoint was the change in HbA 1c from baseline. Compared with metformin monotherapy, an initial combination of DPP4i and metformin was associated with a significant decrease in HbA 1c by weighted mean difference (WMD − 0.44%, 95% CI − 0.57 to − 0.31, p < 0.001), without any increase in hypoglycemia nor in serious adverse effects, but with a small increase in body weight (WMD 0.38 kg, p < 0.001). Compared with metformin monotherapy, initial treatment combination of a sulfonylurea plus metformin resulted in significant decreases in HbA 1c (WMD − 0.68%, 95% CI − 0.86 to − 0.50, p < 0.001); however, it significantly increased the risk of hypoglycemia (RR 8.91, p = 0.02). Compared with metformin alone, initial combinations of a thiazolidinedione (TZD) plus metformin led to significant decreases in HbA 1c (WMD − 0.44%, 95% CI − 0.68 to − 0.19, p < 0.001) but also significantly increased the risk of hypoglycemia (RR 1.60, p = 0.03). Compared with metformin monotherapy, initial combinations of SGLT2i plus metformin led to significant decreases in HbA 1c (WMD − 0.47%, 95% CI − 0.58 to − 0.37, p < 0.001), but increased the risk of drug-related AEs (RR 1.45; p = 0.004). Compared with monotherapy, all initial combination therapies resulted in significantly reduced HbA 1c levels in treatment-naïve T2DM. Compared with metformin monotherapy, the initial combination of DPP-4i and metformin or SGLT2i plus metformin was associated with similar risks of hypoglycemia, but the initial combination therapies of sulfonylurea plus metformin and TZD plus metformin increased the risk of hypoglycemia. • This panel recommends clinical judgment for choosing the appropriate drug, considering the level of HbA 1c , the risk of hypoglycemia, tolerability and availability.
9. In treatment-naïve, non-pregnant, asymptomatic adults recently diagnosed with T2DM without overt CVD or renal disease, DUAL THERAPY with metformin plus an AD1 is recommended for renal protection.

Summary of evidence
• Concerning renal protection with SGLT2i, the best evidence in T2DM with preserved renal function comes from the meta-analysis conducted by Neuen et al. [12]. This meta-analysis assessed the effects of SGLT2i on major kidney outcomes in patients with T2DM at different levels of GFR and determined the consistency of effect size across randomized clinical trials that reported effects on major kidney outcomes. The primary outcome was the composite of dialysis, transplantation, or death due to kidney disease. The authors used random-effects models to obtain summary relative risks (RRs) with 95% CIs and random-effects meta-regression to explore effect modification by subgroups of baseline GFR, albuminuria. Four studies met the inclusion criteria: EMPA-REG OUTCOME, CANVAS Program, CREDENCE and DECLARE-TIMI 58. From a total of 38,723 participants, 252 required dialysis or trans- SGLT2i or a GLP 1 -RA should be used along with metformin in DUAL THERAPY. This is due to the finding that the majority of patients were on metformin in all trials: DECLARE (dapagliflozin), 81% [16]; EMPA-REG-OUTCOME (empagliflozin), 73.8% [17]; CANVAS (canagliflozin), 77% [18]; LEADER (liraglutide), 76% [14]; REWIND (dulaglutide), 81.3% [19]; and SUSTAIN-6 (semaglutide), 73% [20]. Thus, the panel considered that the effect of these agents in the aforementioned trials cannot be dissociated from the effects of metformin.
10. In treatment-naïve, non-pregnant, asymptomatic adults recently diagnosed with T2DM without overt CV disease or renal disease, and in whom HbA 1c is 7.5-9.0%, DUAL THERAPY with metformin plus an 11. Whenever an AD1 is not available, in treatmentnaïve, non-pregnant, asymptomatic adults recently diagnosed with T2DM without known CV disease or renal disease, and in whom HbA 1c is 7.5-9.0%, DUAL THERAPY including metformin plus any AD is recommended to improve blood glucose control.

Summary of evidence
Adding DPP4i: • Dual-therapy with DPP4i and metformin is efficacious and safe. A meta-analysis [25] assessing the long-term efficacy and safety of DPP4i combined with metformin compared to metformin alone in patients with T2DM included seven randomized clinical trials lasting at least 24 weeks. The decline in HbA 1c was greater with dual therapy. The difference was − 0.54% (95% CI − 0.63 to − 0.45), with no increase in hypoglycemia (HR 0.79; 95% CI 0.48 to 1.30). • DPP4i have proven CV safety in the noninferiority CV outcome trials (CVOTs): TECOS (sitagliptin) [26], EXAMINE (alogliptin) [27], and CARMELINA (linagliptin) [28]. One exception is vildagliptin, which was not tested for safety in large CVOTs. Although the recent VERIFY [10] study indicated no signal of harm, it was not powered to detect CV safety. In SAVOR TIMI 53 (saxagliptin) [29], however, the frequency of HF hospitalization was higher in those receiving saxagliptin than in the placebo group.
Adding pioglitazone: • Pioglitazone efficacy and safety was studied in patients with CV disease in the PROactive trial [30]. A prospective, randomized controlled trial including 5238 patients with T2DM who had evidence of macrovascular disease assigned to oral pioglitazone (15 to 45 mg) (n = 2605) or matching placebo (n = 2633), taken in addition to their glucose-lowering drugs. The primary endpoint was a composite of all-cause mortality, non-fatal myocardial infarction (including silent myocardial infarction), stroke, acute coronary syndrome, endovascular or surgical intervention in the coronary or leg arteries, and amputation above the ankle.
Overall, safety and tolerability were good, with no change in the safety profile of pioglitazone identified (6% vs. 4% in the pioglitazone and placebo groups, respectively, were admitted to hospital due to heart failure; mortality rates from heart failure did not differ between groups).
Adding sulfonylureas: • The safety of sulfonylureas in relation to CV outcomes was recently demonstrated in the CAROLINA [31] head-to-head randomized clinical trial (glimepiride versus linagliptin), in the TOSCA.IT [32] head-to-head trial (glimepiride versus pioglitazone), and in the ADVANCE [3] trial (gliclazide MR). • In a meta-analysis of randomized clinical trials [33], CV safety was extended to glibenclamide as well. The panel considered that sulfonylureas are safe in relation to CV risk; however, they are associated with increased incidence of episodes of hypoglycemia. Prescription must thus be individualized for each patient. Among the sulfonylureas, gliclazide MR may be associated with a lower risk of hypoglycemia [34].
12. In treatment-naïve, asymptomatic adults with T2DM and no overt CV or renal disease in whom HbA 1c is > 9.0%, DUAL THERAPY including metformin and insulin-based therapy should be considered to improve blood glucose control.

Summary of evidence
• A meta-analysis comparing CV and metabolic outcomes in insulin-based versus non-insulin-based glucose-lowering therapy included 19,300 adult patients across 18 RCTs. In 16 trials, insulin had superior efficacy in attaining blood glucose control (HR 0.20; 95% CI 0.28 to 0.11) and was associated with superior reductions in HbA 1c . There was no significant between-group difference in risk of death from any cause or CV events. Baseline HbA 1c among all included studies ranged from 7.4 to 9.7%. The risk of hypoglycemia was higher among patients receiving insulin (RR 1.90; 95% CI 1.44 to 2.51). Non-insulin treatment was associated with a higher proportion of adverse drug reactions [35] (54.7% versus 45.3%, p = 0.044). • Compared with oral ADs, early intensive insulin therapy in patients with newly diagnosed T2DM is associated with favorable impact on recovery and maintenance of beta-cell function, as well as protracted glycemic remission. A multicenter, randomized clinical trial compared the effects of transient intensive insulin therapy-continuous subcutaneous insulin infusion (CSII) or multiple daily injections (MDI)-versus oral antidiabetic agents on beta-cell function and diabetes remission. A total of 382 treatment-naïve patients with recently diagnosed T2DM were randomized to receive insulin or oral hypoglycemic agents for rapid initial correction of hyperglycemia. The mean HbA 1c at baseline was 9.5-9.8%. Treatment was stopped once normoglycemia had been achieved and remained stable for 2 weeks; patients were then followed on diet and exercise alone. Intravenous glucose tolerance tests were performed and glucose, insulin, and proinsulin levels were measured. The primary endpoint was duration of glycemic remission and remission rate at 1 year. Overall, more patients achieved target blood glucose control in the insulin groups than among those treated with oral ADs. The 1-year remission rate was significantly higher in the insulin groups (51.1% and 44.9% versus 26.7% with oral ADs; p = 0.0012). Beta-cell function, assessed by the HOMA-B and acute insulin response, also improved significantly after intensive therapy. The increase in acute insulin response was sustained in the insulin groups, but had declined significantly in the oral ADs group at 1 year in all patients who achieved remission [36]. that short-term intensive insulin therapy can improve underlying pathophysiology in early T2DM [37].
13. In adult patients with T2DM who are symptomatic (polyuria, polydipsia, weight loss) and present with HbA 1c > 9%, insulin-based therapy is recommended to improve blood glucose control.

Summary of evidence
• The panel recommended the use of insulin-based therapy in T2DM patients with symptoms of hyperglycemia. There is general agreement that insulinbased therapy is needed when symptoms of insulin deficiency are present. This statement is based primarily on the pathophysiology of T2DM, plausibility, and clinical experience.
14. In patients with T2DM without cardiovascular or renal complications, whose HbA 1c remains above target despite dual therapy, TRIPLE THERAPY with metformin plus two AD1 is recommended to improve blood glucose control, renal protection and cardiovascular risk reduction.

Summary of evidence
• The panel considered that, in general, triple therapy is effective and safe for improving blood glucose control. The majority of the cited studies indicate superior HbA 1c -lowering efficacy with 3 than with 2 antidiabetic drugs. • We found no trials directly comparing additive cardiovascular risk reduction or renal protection with a triple combination of SGLT2i and GLP-1 RA plus metformin. However, SGLT2i have demonstrated reduction of renal outcomes in patients with preserved renal function [12], and GLP -1 RA have demonstrated cardiovascular risk reduction in patients in primary prevention who had subclinical atherosclerosis [21].  [41], 106 patients recently diagnosed with T2DM were randomized to metformin/pioglitazone/exenatide (triple therapy) and 115 to metformin followed by sulfonylurea and insulin glargine (conventional therapy) with an HbA 1c target of < 6.5% for 2 years. Participants who received triple therapy had a greater reduction in HbA 1c level than those who received conventional therapy (5.95% versus 6.50%; p < 0.001). Despite lower HbA 1c , participants on triple therapy experienced a 7.5-fold lower rate of hypoglycemia than patients on conventional therapy. Triple therapy was also associated with weight loss versus weight gain in those receiving conventional therapy (− 1.2 kg versus + 4.1 kg respectively; p < 0.01). • A post hoc analysis of three randomized trials of sequential or concomitant add-on of dapagliflozin and saxagliptin to metformin [42] compared the safety of triple therapy (dapagliflozin + saxagliptin + metformin) versus dual therapy (dapagliflozin or saxagliptin + metformin). At 24 weeks, the incidence of any adverse events and serious adverse events was similar between the triple and dual therapy groups, as well as between the concomitant and sequential add-on groups. Urinary tract infections were more common in the sequential groups than in the concomitant groups; genital infections were reported only with sequential add-on of dapagliflozin to saxagliptin/metformin. Hypoglycemia occurred in < 2.0% of patients across all groups. • A network meta-analysis [43]  15. In patients with T2DM without cardiovascular or renal complications, whose HbA 1c remains above target despite triple therapy, QUADRUPLE THER-APY with metformin, two AD1 and one AD is recommended to improve blood glucose control.

Summary of evidence
• Although this panel did not find evidence for using insulin exclusively as a fourth drug in quadruple therapy, there was consensus in the expert opinions for its use due to the efficacy and safety of insulin-based therapy.
16. In patients with T2DM whose HbA 1c remains above target despite triple therapy, QUADRUPLE THERAPY including combinations of metformin plus one AD1 and two AD or even metformin plus 3 AD or insulin based-therapy should be considered to improve blood glucose control.

Summary of evidence
• Quadruple therapy was evaluated in an open-label observational trial [44] conducted in patients with T2DM not controlled (HbA 1c 7.5-12%) despite three different antidiabetic agents. The objective was to address the effectiveness and safety of adding empagliflozin or insulin glargine as a fourth agent in patients already on metformin, glimepiride and a DPP4i. A total of 268 patients were included: 142 on empagliflozin (25 mg/day) and 126 on insulin glargine. After 24 weeks, HbA 1c was significantly reduced from baseline by 1.5 ± 1.2% (p < 0.001) in  [46] was conducted in T2DM to compare the effectiveness and safety of adding empagliflozin 25 mg od or dapagliflozin 10 mg od as part of a quadruple therapy regimen for patients already on metformin, glimepiride and DPP4i and still inadequately controlled (HbA 1c 7.5-12.0%). The outcome measure was change in HbA 1c . In total, 350 patients were enrolled with empagliflozin (n = 176) and dapagliflozin (n = 174), respectively. After 52 weeks, both groups showed significant reductions in HbA 1c , but the reduction was greater in the empagliflozin group (p < 0.001). Safety profiles were similar in the two groups, demonstrating that quadruple therapy can be used effectively in patients with T2DM. • In a 26-week open-label trial [45], patients receiving GLP-1 RA therapy (liraglutide once daily or exenatide twice daily) plus metformin alone or metformin plus pioglitazone and/or a sulfonylurea were randomly assigned to receive insulin degludec plus liraglutide once daily (n = 292) or to continue GLP-1 RA therapy and oral ADs at the pre-trial dose (n = 146). At 26 weeks, superior HbA 1c reductions had been achieved with the insulin degludec/liraglutide combination (estimated treatment difference − 0.94%; p < 0.001).

Treatment of choice:
17. In patients with T2DM and clinical atherosclerosis (ASCVD), the combination of metformin with either an SGLT2 inhibitor or a GLP1-RA (AD1) is recommended to reduce cardiovascular events and to improve blood glucose control.

Summary of evidence
• In a meta-analysis [24]

Summary of evidence
• Studies specifically designed to test whether triple therapy can reduce MACE in T2DM patients with ASCVD were not found in the literature. However, in the EMPA-REG OUTCOME trial [17], 49% of patients were on dual therapy before being rand-

METFORMIN + [SGLT2i or GLP1-RA]
TRIPLE THERAPY omized to empagliflozin. Thus, almost half of patients in whom cardiovascular events were significantly reduced received triple therapy. Considering the robust data in reducing 3P-MACE with both GLP-1 RA and SGLT2i, as described above, this panel considered that the combination of both SGLT2i and GLP1-RA should be preferred among other antidiabetic agents, as they are also safe and effective for reducing blood glucose.

MET+ SGLT2i + GLP1-RA + AD
19. In T2DM patients with ASCVD and HbA 1c above target despite triple therapy, QUADRUPLE THER-APY in a combination of metformin, two AD1 and one AD is recommended to improve blood glucose control.

Summary of evidence
• Evidence from trials using exclusively quadruple therapy in T2DM patients with atherosclerotic cardiovascular disease is lacking. The best evidence available is described in statement 15 of this guideline, referring to quadruple therapy in the general patient with T2DM. This panel considered that the aforementioned evidence does overlap with patients with ASCVD, as high-risk patients were tested in individual trials for safety. This panel agrees that quadruple therapy is recommended whenever HbA 1c targets are not reached despite triple therapy, even in patients with ASCVD.

Alternative treatment in patients with ASCVD:
20. Whenever AD1 is unavailable and HbA 1c is 6.5-7.5%, metformin in MONOTHERAPY is recommended as the initial therapy to improve blood glucose control and reduce cardiovascular events in T2DM patients with clinical atherosclerosis (ASCVD).

Summary of evidence
• A meta-analysis of clinical trials and observational studies [48] assessed the impact of metformin versus placebo and active comparators on mortality and cardiovascular events among T2DM patients, including sub-groups with coronary artery disease (CAD), to evaluate death from all causes, CV death, and incidence of CV events. Alternatives to AD1 in patients with ASCVD:

Alternatively, if only one AD1 is available and
HbA 1c is above 7.5% despite dual therapy in T2DM patients with ASCVD, TRIPLE THERAPY with metformin plus an AD1 and any other AD is recommended to improve blood glucose control.

Summary of evidence
• A network meta-analysis of 176,310 participants across 236 trials [49] found that SGLT2i and GLP-1 RA were associated with significantly lower rates of death from any cause as compared to control. SGLT2i (absolute risk reduction − 0.9%; HR 0.78; 95% CI 0.68 to 0.90) and GLP-1 agonists (absolute risk reduction, − 0.5%; HR 0.86; 95% CI 0.77 to 0.96) were associated with lower mortality, while DPP4i were not associated with significant reductions in death from any cause (absolute risk reduction, 0.1%; HR, 1.02; 95% CI 0.94 to 1.11). Mortality was lower in patients receiving SGLT2i or GLP-1 RA than in those receiving DPP4i, placebo, or no treatment.

Treatment of choice:
23. In patients with T2DM and heart failure (HF) with reduced ejection fraction (< 40%), combined therapy including metformin and an SGLT-2i is recommended to reduce cardiovascular mortality, HF hospitalizations, and to improve blood glucose control.

Summary of evidence
• In a systematic review and meta-analysis [24]

In patients with T2DM and HF with reduced ejection fraction (< 40%) whose HbA 1c is above target despite dual therapy with metformin and an SGLT2i, TRIPLE THERAPY by adding a GLP-1 RA
should be considered to reduce the risk of HF-related hospitalization.

Summary of evidence
• A meta-analysis of seven randomized placebo-controlled trials [21]

Summary of evidence
• Evidence referring exclusively to use of quadruple therapy in patients with T2DM and HF was not found in the literature. The best evidence available is described in statement 15 of this guideline, referring to quadruple therapy for the general patient with T2DM. This panel agrees that quadruple therapy is recommended whenever HbA 1c targets are not reached despite triple therapy, even in patients with ASCVD. However, based on expert opinion, a combination of agents with proven CV safety (i.e., which do not increase risk of HF) is reasonable.

Summary of evidence
• In the SAVOR-TIMI 53 noninferiority trial [29], T2DM patients at risk of CV events were randomly assigned to receive saxagliptin or placebo, and followed for a median of 2.  [30] was a prospective randomized clinical trial of 5238 patients with T2DM and macrovascular disease. Patients were randomized to receive pioglitazone or placebo. The primary endpoint was an expanded MACE composite including death from any cause, nonfatal myocardial infarction, stroke, acute coronary syndrome, endovascular or surgical intervention in the coronary circulation or lower-limb arteries, and above-ankle amputation. The mean observation time was 34.5 months.
The primary endpoint occurred similarly in patients in the pioglitazone group and patients in the placebo group (HR 0.90, 95% CI 0.80 to 1.02, p = 0.095).
Patients receiving pioglitazone experienced more HF episodes than on placebo (11% versus 8%, p < 0.0001). There was also a higher number of HF episodes not needing hospital admission (5% vs. 3%; p = 0.003) and HF episodes requiring hospital admission (p = 0.007) in pioglitazone-treated patients versus placebo. However, there was no difference in the rate of fatal HF.
Alternative treatment:

Summary of evidence
Adding a sulfonylurea: • In the UKPDS trial [53], HF rates were not increased among patients who received sulfonylureas as compared with the conventional treatment group (3.0% vs. 3.3%, HR 0.91, 95% CI 0.54 to 1.52). • An observational study [54] investigated all-cause mortality associated to sulfonylureas (SU) in patients with HF. Patients who were hospitalized for the first time due to HF in 1997-2006, alive 30 days after discharge, and on monotherapy with a specific type of SU were followed for a mean of 744 days. There were 1097 patients on glimepiride; 1031 on glibenclamide; 557 on glipizide; 251 on gliclazide; and 541 on tolbutamide. During the observation period, 2242 patients (64%) died. Compared to gliclazide, which was considered the reference, the risk of death was similar among all types of SU: glimepiride (HR 1.10, 95% CI 0.92 to 1.33); glibenclamide (HR 1.12, 95% CI 0.93 to 1.34), glipizide (HR 1.14, 95% CI 0.93 to 1.38) and tolbutamide (HR 1.04 (0.85-1.26). Significant differences in mortality risk among SU in patients with HF were deemed unlikely.
Adding insulin: • Insulin has a dose-dependent anti-natriuretic effect, and causes weight gain and mild edema at physiologic concentrations. We found no controlled trials addressing safety of insulin in patients with clinically established HF or at high risk of HF. In UKPDS 33 [2], there was no difference in HF rates between patients receiving insulin and those on sulfonylureas.
• Insulin glargine, a long-acting insulin analogue, was studied in the ORIGIN trial [55]. A sub-analysis showed that insulin glargine has a neutral effect on both initial and recurrent hospitalizations for HF. The trial randomized 12,537 patients with prediabetes or diabetes to either insulin glargine or placebo. All were at high cardiovascular risk. However, people with more severe HF-New York Heart Association (NYHA) class 3 or 4-were excluded. There were no differences between groups in hospitalization for HF (HR 0.90, 95% CI 0.77 to 1.05) over the 2.5 years of follow-up [56]. The position of this panel is that insulin can be used as a safe option to control blood glucose in patients with HF. However, close monitoring is advisable in patients with unstable HF.
Adding a DPP4i: • The TECOS [26] noninferiority trial was designed to assess the efficacy and safety of sitagliptin in 14,671 subjects with CV disease. The study found that sitagliptin did not increase hospitalization for HF as compared to placebo (HR 1.00, 95% CI 0.83 to 1.20) during a 3-year follow-up period. CV mortality was similar between sitagliptin and placebo (22.4% versus 23.1%), as well as all-cause mortality, after hospitalization (29.8% versus 28.8%). • CARMELINA [28] was a randomized, placebo-controlled, multicenter noninferiority trial conducted among adults with type 2 diabetes to test linagliptin (n = 3494) against placebo (n = 3485) as add-on over usual care. Hospitalization for HF (an exploratory cardiovascular outcome) occurred in 209 of 3494 patients in the linagliptin group (6.0%) and in 226 of 3485 patients in the placebo group (6.5%). The absolute difference in incidence rate was − 0.27 (95% CI − 0.82 to 0.28), which was nonsignificant (HR 0.90; 95% CI 0.74 to 1.08; p = 0.26). • We did not find CVOTs conducted to assess the CV safety of vildagliptin. However, a retrospective metaanalysis [57] did not find any significant increase in risk of HF in vildagliptin-treated patients.
Chronic kidney disease (CKD) (Fig. 4) Mild to moderate CKD (eGFR 30-60 mL/min/1.73 m 2 or eGFR 30-90 mL/min/1.73 m 2 with albuminuria) 28. In T2DM patients with mild to moderate CKD, DUAL THERAPY with metformin and an SGLT2i is recommended to attenuate loss of renal function, prevent end-stage renal disease, reduce mortality due to renal causes, and to improve blood glucose control.

Summary of evidence
• The CREDENCE study [58] randomly assigned patients with T2DM with HbA 1c 6.5-12% and CKD (calculated eGFR 30-90 mL/min/1.73 m 2 and albuminuria > 300-5000 mg/g) to receive either canagliflozin (100 mg/day) or placebo. Metformin was used by 57.8% of patients. A total of 4401 patients underwent randomization (mean age 63 years, 33.9% female). The mean HbA 1c was 8.3% and the mean eGFR was 56.2 mL/min/1.73 m 2 . CV disease was present in 50% of patients. The median urinary albumin was 927 mg/g. The primary outcome was a composite of end-stage kidney disease (dialysis for at least 30 days, kidney transplantation, or eGFR < 15 mL/ min/1.73 m 2 for at least 30 days, doubling of serum creatinine from baseline or death from renal of cardiovascular disease). Over a median follow-up of 2.62 years, the primary outcome occurred in significantly fewer patients in the canagliflozin group than in the placebo group (43.2 versus 61.2/1000 patientsyear, respectively; HR 0.70, 95% CI 0.59 to 0.82, p = 0.00001). • In a sub-analysis of EMPA-REG OUTCOME [17], 4124 T2DM patients with GFR > 30 mL/min/1.73 m 2 were assigned to either empagliflozin or placebo once daily. The secondary renal outcomes (all prespecified) included incident or worsening nephropathy (progression to macroalbuminuria, doubling of serum creatinine, initiation of renal replacement therapy, or renal death) and incident albuminuria. Incident or worsening nephropathy occurred in 12.7% of patients in the empagliflozin group versus 18.8% in the placebo group (HR 0.61; 95% CI 0.53 to 0.70; p < 0.001).
Doubling of serum creatinine occurred in 1.5% of patients receiving empagliflozin and in 2.6% of those given placebo (a significant relative risk reduction of 44%). Renal replacement therapy was initiated in 0.3% in the empagliflozin group and in 0.6% in the placebo group (a 55% lower relative risk). The rate of incident albuminuria was similar in the two groups. CI 0.20 to 0.82; p = 0.012). Both the cardiorenal and renal-specific composite outcomes were improved by dapagliflozin versus placebo in several prespecified subgroups, including those defined by baseline eGFR and presence or absence of established ASCVD. The mean decrease in eGFR was greater in the dapagliflozin group than in the placebo group 6 months after randomization; however, this decline had equalized by 2 years, and at 3 and 4 years the mean decrease in eGFR was less in the dapagliflozin group than in the placebo group. Metformin is recommended in T2DM and CKD, as it was used by 74% of patients in EMPA-REG OUTCOME trial [17], 58% of patients in the CREDENCE study [58], 76% of participants in CANVAS [18], and 81% of participants in DECLARE-TIMI 58 [16]. The panel 29. In patients with T2DM and mild to moderate CKD whose HbA 1c remains above the target despite dual therapy, TRIPLE THERAPY with metformin, SGLT2i and a GLP-1 RA is recommended to reduce renal outcomes and improve glycemic control.

Summary of evidence
Adding a GLP-1 RA: • The LEADER RENAL [15] sub-study was a prespecified sub-analysis of secondary renal outcomes of the original LEADER trial, in which patients were randomized to receive liraglutide or placebo. In this subanalysis, the secondary outcome was a composite of new-onset persistent macroalbuminuria, persistent doubling of serum creatinine, ESRD, or renal death. Patients who were on insulin or insulin plus an oral AD were randomized to receive dulaglutide (1.5 mg or 0.75 mg) or insulin glargine. Insulin lispro was also added and titrated. The trial lasted 52 weeks. The primary outcome was HbA 1c at 26 weeks. Secondary outcomes included eGFR and UACr. A total of 577 patients were randomized to dulaglutide and insulin glargine. Dulaglutide produced blood glucose control similar to that achieved with insulin glargine and slowed the decline in GFR. This study demonstrates that dulaglutide is safe and effective in patients with moderate-to-severe CKD.
Alternative treatment: 30. In T2DM patients with mild to moderate CKD and HbA 1c above target despite dual therapy, TRIPLE THERAPY with metformin, SGLT2 and an alternative AD (replacing GLP1-RA) is recommended to improve blood glucose control.

Summary of evidence
• The efficacy and safety of triple therapy with an AD in T2DM with chronic kidney disease was addressed in studies using DPP-4, pioglitazone and sulfonylureas.
Adding DPP4i: • This study indicated that low-dose pioglitazone has similar efficacy while promoting less weight gain than standard-dose pioglitazone in patients with CKD.
Adding sulfonylureas: • The safety of sulfonylureas was evaluated in the CAR-OLINA trial [31], a head-to-head, active-controlled, randomized trial that assessed the impact of linagliptin versus glimepiride on CV outcomes in high-risk patients (many with chronic kidney disease

31.
In T2DM patients with mild to moderate CKD and HbA 1c above target despite triple therapy, QUAD-RUPLE THERAPY including metformin, SGLT2i, GLP 1 -RA and a fourth antidiabetic agent (AD) or insulin-based therapy is recommended to improve blood glucose control.

Summary of evidence
• Although we did not find significant efficacy evidence for QUADRUPLE therapy in T2DM patients with mild to moderate renal failure, the panel considered that this strategy is necessary to lower blood glucose in some patients and is reasonably safe when eGFR is between 30 and 90 mL/min/1.73 m 2 , a stage of CKD in which most agents can be used, provided that their dosages are adjusted when appropriate. Special attention is warranted with metformin, which should be replaced when the eGFR falls below 30 mL/min/1.73 m 2 . Sulfonylureas also demand caution due to an increased risk of hypoglycemia in this population.
Adding a DPP4i: • The CARMELINA trial [28]  Adding pioglitazone: • The effect of thiazolidinediones on albuminuria in T2DM was evaluated in a meta-analysis [66] of 15 randomized controlled trials. A total of 2860 T2DM patients with baseline normo-or microalbuminuria, using both rosiglitazone and pioglitazone compared with placebo or other ADs, were evaluated. Overall, in participants with normo-and microalbuminuria, thiazolidinedione therapy was associated with significant reductions in urinary albumin excretion. In studies of pioglitazone, the weighted mean difference of proportional change between the pioglitazone and control groups was 16.2% (95% CI 20.8 to 11.6). The overall mean difference of the change in urine UACR between the thiazolidinedione and control groups was 24.8% (95% CI 39.6 to 10.0]. Thiazolidinediones, especially pioglitazone, reduce urinary albumin and protein excretion significantly in patients with T2DM. Severe CKD (eGFR < 30 mL/min/1.73 m 2 and hemodialysis) 32. In T2DM patients with severe renal failure and HbA 1c above target, insulin-based therapy is the recommended choice to improve blood glucose control.

Summary of evidence
• Insulin glargine is safe and effective in T2DM patients with severe renal failure [67], yielding rapid HbA 1c reductions with a stable half-life and longer duration of action. In a small non-randomized study, 89 patients with T2DM and CKD (mean eGFR 34.1 ± 11.5 mL/min/1.73 m 2 ), who were poorly controlled or experienced frequent hypoglycemia on oral ADs or NPH insulin, were prescribed insulin glargine at bedtime. The dose was started at 0.1 unit/ kg and titrated to the desired target. At 4 months of follow-up, HbA 1c had declined from 8.4% ± 1.6 to 7.7% ± 1.2 (p < 0.001). Body mass index was unaffected (p = 0.96). Mild symptomatic hypoglycemia was experienced by 12.5% of patients. No other adverse events were reported.
33. In T2DM patients with severe renal failure and HbA 1c above target, either a DPP4 inhibitor or a GLP-1 RA (if eGFR 15-30 mL/min/1.73 m 2 ) may be considered to improve blood glucose control.

Summary of evidence
• The DPP4i class (sitagliptin, vildagliptin, alogliptin, saxagliptin and linagliptin) was also tested in small studies in T2DM patients undergoing hemodialysis, and safety should be confirmed in larger studies. • In a small trial [68], 64 patients with T2DM were randomized to sitagliptin (in the reduced dosage of 25 mg/daily) and 65 to glipizide 2.5 mg/daily. There were 28 patients (43%) with eGFR < 30 mL/ min/1.73 m 2 . After 54 weeks, the mean reduction in HbA 1c level from baseline was 0.72% (95% CI 0.95% to 0.48%) in the sitagliptin group and 0.87% (95% CI 1.11% to 0.63%) in the glipizide group. The incidence of symptomatic hypoglycemia was 6.3% in the sitagliptin group vs. 10.8% in the glipizide group (difference, 4.8%; 95% CI 15.7% to 5.6%). Severe hypoglycemia did not occur in the sitagliptin group vs. 7.7% in glipizide group (difference, 7.8%; 95% CI 17.1% to 1.9%). Sitagliptin monotherapy was effective and well tolerated in patients undergoing hemodialysis. No adverse drug reactions related to treatment with vildagliptin were seen. • In a small non-randomized safety trial [70], 16 patients with T2DM undergoing hemodialysis received alogliptin 6.25 mg for 2 years. Baseline serum creatinine was 10.6 ± 1.0 mg/dL. Mean HbA 1c dropped from 7.1 to 5.8% after treatment. None of the patients exhibited significant adverse effects, such as hypoglycemia. One patient experienced a drugrelated rash. Four patients withdrew from this study during the treatment period. • The effects of monotherapy with linagliptin 5 mg in 21 patients with T2DM undergoing hemodialysis was examined in a 6-month non-randomized trial [71]. Linagliptin was administered daily. GA dropped from 21.3% ± 0.6% to 18.0% ± 0.6% over the 6-month treatment period, and body weight did not change. None of the patients experienced hypoglycemia. • Saxagliptin was studied in a sub-analysis of the SAVOR-TIMI 53 trial [72] according to baseline renal function. Patients with T2DM at risk of cardiovascular events were stratified by renal function. There were 336 patients with severe renal impairment (eGFR < 30 mL/min/1.73 m 2 ) who were randomized to receive either saxagliptin or placebo. The primary endpoint was the time to first event of a composite of CV death, myocardial infarction (MI), or ischemic stroke. The major secondary endpoint included the primary composite plus hospitalization for heart failure, coronary revascularization, or unstable angina. After a median duration of 2 years, saxagliptin did not change the risk of the primary and secondary composite endpoints compared with placebo, irrespective of renal function (p = 0.19 for interactions  [15], 2.5% in SUSTAIN-6 (injectable semaglutide) [20], and 1% in REWIND RENAL (dulaglutide) [14]. Thus, data on the safety of GLP-1 RA in this population is limited.

Conclusions
In non-pregnant adults with type 2 diabetes, the recommended HbA 1c target is below 7%. Higher levels are recommended in frail older adults and patients at higher risk of hypoglycemia. Lifestyle modification is recommended at all phases of treatment. In recent diagnosed patients without cardiovascular or renal complications, metformin in monotherapy is the first choice of treatment when HbA 1c is 6.5-7.5%. Optionally, metformin along with a DPP4 inhibitor may be considered to reduce failure in controlling blood glucose. When HbA1c is 7.5-9.0%, dual therapy, including metformin a first line antidiabetic drug AD1 (SGLT2i or GLP-1RA) is recommended, due to their cardiovascular and renal benefits. If an AD1 is unaffordable, other antidiabetic drugs (AD) may be used. Triple or quadruple therapy should be considered when HbA1c remains above target despite dual therapy. In patients with clinical atherosclerosis, the combination of metformin plus one AD1 independently of HbA1c level is also recommended to reduce cardiovascular events. In the stable patient with low ejection fraction heart failure (< 40%) and glomerular filtration rate (eGFR) > 30 mL/min/1.73 m 2 , metformin plus an SGLT2i is recommended to reduce cardiovascular mortality, heart failure hospitalizations and to improve blood glucose control. In patients with mild to moderate diabetes-associated chronic kidney disease (CKD) (eGFR 30-60 mL/min/1.73 m 2 or eGFR 30-90 mL/ min/1.73 m 2 with albuminuria > 30 mg/g), the combination of metformin and a SGLT2i is recommended to attenuate loss of renal function, reduce albuminuria and improve blood glucose control. In patients with severe renal failure (eGFR < 30 mL/min/1.73 m 2 ), insulin-based therapy is recommended to improve blood glucose control. Alternatively, GLP-1RA, DPP4i, gliclazide MR and pioglitazone may also be considered to reduce albuminuria. In conclusion, the current evidences support individualizing anti-hyperglycemic treatment for T2DM according to their cardiovascular and renal status.