The potential role of sodium glucose co-transporter 2 inhibitors in combination therapy for type 2 diabetes mellitus

Abstract

Introduction: Sodium glucose co-transporter 2 (SGLT2) inhibitors are a new class of glucose-lowering agents developed for the treatment of type 2 diabetes mellitus (T2DM). These agents have a mechanism of action that is independent of pancreatic β-cell function or the degree of insulin resistance; consequently, SGLT2 inhibitors have the potential to be used not only as monotherapy but also in combination with any of the existing classes of glucose-lowering agents, including insulin. As part of the extensive clinical development programs for modern T2DM therapies, SGLT2 inhibitors have been studied in combination with the most commonly used classes of glucose-lowering medications.

Areas covered: This report summarizes the key clinical trials data for combination therapies using SGLT2 inhibitors currently approved in the United States and/or the European Union, namely, dapagliflozin, canagliflozin, and empagliflozin.

Expert opinion: When given as add-on combination therapy with other glucose-lowering agents, or as monotherapy, SGLT2 inhibitors produced modest but clinically meaningful reductions in glycated hemoglobin, body weight, and systolic blood pressure. These changes have been sustained over long-term follow-up. SGLT2 inhibitors have a generally favorable safety profile similar to that of placebo, and are well tolerated. The risk of hypoglycemia appears to depend on coadministered glucose-lowering agents: when used as monotherapy, the frequency is comparable to that of placebo, but an increased risk is associated with concomitant use of sulfonylureas or insulin. In addition, an increased risk of genitourinary infections has been reported with SGLT2 inhibitors. However, these infections are usually mild, nonrecurrent, and respond to standard treatment.

Keywords::

• canagliflozin
• combination therapy
• dapagliflozin
• empagliflozin
• sodium glucose co-transporter type 2 inhibitors
• type 2 diabetes mellitus

1. Introduction

Type 2 diabetes mellitus (T2DM) is characterized by hyperglycemia and defective insulin secretion, which is unable to compensate for insulin resistance [1]. Patients with T2DM are at increased risk of macrovascular events (coronary artery disease, peripheral artery disease, and stroke) as well as microvascular complications (diabetic retinopathy, nephropathy, and neuropathy) [1]. It is well established that improved control of hyperglycemia reduces the risk of microvascular complications and improves cardiovascular (CV) outcomes in people with diabetes [2-8]. However, despite numerous treatment options, a large proportion of patients with T2DM do not achieve optimal glycemic control, even when individualized targets are accounted for [9-12].

Although lifestyle interventions such as promoting a healthy diet and physical activity are the basis of T2DM treatment [13], most individuals with T2DM will eventually require pharmacotherapy to combat hyperglycemia caused by the pathophysiological abnormalities of T2DM [14], namely, β-cell failure and insulin resistance. In fact, most patients are likely to be recommended pharmacotherapy at diagnosis or soon after [13]. A range of glucose-lowering agents is available, such as metformin, sulfonylureas (SUs), glinides, thiazolidinediones (TZDs), α-glucosidase inhibitors, dipeptidyl peptidase-4 (DPP-4) inhibitors, glucagon-like peptide 1 receptor agonists, and insulins. Given the progressive nature of T2DM, glycemic targets are often achieved by escalating doses of monotherapy and then advancing to dual or triple combination therapy as required [13]. However, as these agents have side effects (e.g., worsening hypoglycemia or weight gain, gastrointestinal side effects, and/or fluid retention), the use of combination therapies can further diminish overall tolerability [13]. These factors are of chief importance to patients with T2DM when selecting their preference for oral glucose-lowering agents [15]. Given the persistent rise in the global prevalence of diabetes, with projections in excess of 590 million cases of diabetes within the next 25 years [16], coupled with the fact that T2DM accounts for 90 – 95% of all cases of diabetes [1], there is a pressing need for new effective therapies for T2DM with improved safety and tolerability profiles.

Renal glucose handling is deregulated in T2DM, causing increased reabsorption of glucose from the kidneys and its release back into the circulation, even in the presence of hyperglycemia. Sodium glucose co-transporter 2 (SGLT2) inhibitors are a novel class of pharmacologic agents developed for the treatment of T2DM. SGLT2 inhibitors target the kidney to promote urinary glucose excretion (UGE) and reduce hyperglycemia. These agents have a mechanism of action that is independent of pancreatic β-cell function or the degree of insulin resistance; consequently, SGLT2 inhibitors have the potential to be used in combination with any of the existing classes of glucose-lowering agents, including insulin.

This report examines the role of SGLT2 inhibitors in renal glucose transport and the rationale for their use in the treatment of T2DM, and summarizes the key clinical trials data for combination therapies using the main SGLT2 inhibitors currently approved in the United States (US) and/or the European Union (EU).

Three SGLT2 inhibitor compounds, canagliflozin, dapagliflozin, and empagliflozin, have progressed to marketing application and approval in the EU; these agents have also been approved in the US. Other ‘gliflozins’ recently entered into Phase III clinical trials, or soon to do so, include ertugliflozin and sotagliflozin. In Japan, canagliflozin, dapagliflozin, ipragliflozin, tofogliflozin, and luseogliflozin have all gained approval.

2. Renal glucose transport

In healthy adults, approximately 180 g of glucose is filtered by the kidneys each day, virtually all of which is reabsorbed and returned to the circulation [17]. This is achieved by the sodium-coupled active transport proteins SGLT2 and SGLT1, located in the proximal kidney tubule [18,19]. The majority of filtered glucose is reabsorbed by SGLT2 in the first part of the tubule [18,19]. SGLT2 has a low affinity for glucose but a high capacity, so it can transport glucose into the tubule cell quickly. The remaining glucose is reabsorbed by SGLT1, which is situated further along the tubule [18,19]. SGLT1 has a high affinity for glucose but a low capacity, so its actions complement those of SGLT2. The action of SGLT2 and SGLT1 is independent of insulin. Reabsorbed glucose is returned to the blood via passive glucose transporters known as GLUTs [18,19].

As plasma glucose levels rise, the amount of glucose filtered increases to a point where the transport proteins become saturated and glucose reabsorption is at maximum capacity – the renal transport maximum for glucose, or Tm glucose. Any further increase in plasma glucose, for example, as in uncontrolled T2DM, causes the excess glucose in the filtrate to be excreted in the urine (i.e., glucosuria). This threshold for glucosuria is variable, but normally occurs at a plasma glucose concentration of approximately 200 mg/dl [20]. In healthy individuals, the renal Tm glucose is not normally exceeded but individuals with the rare disorder familial renal glucosuria, caused by loss-of-function mutations in the SGLT2 gene, are reported to experience UGE in the order of < 10 to > 200 g/1.73 m²/day [21,22]. Affected individuals are usually otherwise asymptomatic [22], and the condition is not known to be associated with diabetes mellitus.

3. Rationale for SGLT2 inhibition

Contrary to expectation, the renal glucose reabsorption capacity appears to be increased in diabetes [23-25] and the kidneys continue to reabsorb glucose even when plasma glucose concentrations are high, with levels that usually exceed the Tm observed in healthy individuals. This results in a continuous flow of glucose from the kidney filtrate back into the circulation, even in the presence of hyperglycemia, which will exacerbate diabetes-associated complications. Therefore, inhibiting the renal reabsorption pathway provides a target for therapeutic intervention in T2DM; if SGLT2 promotes glucose conservation, it follows that inhibiting SGLT2 can have the opposite effect, namely to stimulate UGE and reduce hyperglycemia [26]. The observation that familial renal glucosuria is a benign disorder suggests that pharmaceutical inhibition of SGLT2 would be safe.

Further expectations of SGLT2 inhibition would include reducing glycated hemoglobin (HbA1c) and fasting plasma glucose (FPG). In addition, as SGLT2 inhibition has no effect on the normal bodily responses to hypoglycemia [27], and does not cause insulin release [28,29], SGLT2 inhibitors should not increase the risk of hypoglycemia. Based on their mechanism of action, SGLT2 inhibitors would not be expected to have direct effects on restoring pancreatic β-cell function or insulin sensitivity, but recent studies suggest that these underlying defects could be improved via reduced glucotoxicity [30-32]. In addition, SGLT2 inhibitors should not be affected by pancreatic β-cell function or the degree of insulin resistance present, thus maintaining efficacy in advanced or refractory T2DM. Therefore, SGLT2 inhibitors could be used at any stage of T2DM from early disease onwards. Given their non-insulin–dependent mechanism of action, SGLT2 inhibitors have the potential to be used in combination with any of the existing classes of glucose-lowering agents, including insulin.

4. Position of SGLT2 inhibitors in the treatment guidelines

The first SGLT2 inhibitor to be marketed in the US, canagliflozin, was launched there in 2013, and later that year the American Association of Clinical Endocrinologists (AACE) algorithm was updated to include SGLT2 inhibitors, stating they could provide a therapeutic alternative in patients with T2DM in whom metformin is not tolerated or otherwise contraindicated [33]. The AACE 2013 algorithm also states that SGLT2 inhibitors could be used as add-on therapy to two or three other agents, including insulin, in patients who would benefit from weight loss [33]. The most recent position statement of the American Diabetes Association and the European Association for the Study of Diabetes (ADA/EASD) predates the availability of SGLT2 inhibitors [13], but the five criteria to be considered when escalating anti-hyperglycemic therapy, namely, efficacy (HbA1c reduction), hypoglycemia, weight, side effects, and cost, would allow the use of SGLT2 inhibitors if the physician determined there would be a benefit to the patient. With the availability of comprehensive Phase III trials data on these agents, SGLT2 inhibitors can be expected to be discussed in future treatment guidelines from expert groups.

5. Pharmacologic factors and dosing

While the various SGLT2 inhibitor agents in current development are structurally similar, they differ in their respective selectivity profiles for SGLT2 over SGLT1: empagliflozin has the highest degree of selectivity (> 2500-fold), followed by tofogliflozin (> 1875-fold), dapagliflozin (> 1200-fold), ipragliflozin (> 550-fold), and canagliflozin (> 250-fold) [34]. This value of selectivity may be important when considering the role of SGLT1 in normal intestinal glucose/galactose absorption, as reducing intestinal absorption of glucose/galactose leads to increased fermentation of these sugars, and the acidic fermentation products decrease intestinal pH, potentially resulting in increased solubility and absorption of dietary calcium [35,36]. However, any potential safety implications (e.g., regarding bone metabolism) are not yet clear.

Canagliflozin, dapagliflozin, and empagliflozin are dosed orally, and taken once daily [37-41]. All are currently available in two doses (canagliflozin 100 and 300 mg, dapagliflozin 5 and 10 mg, and empagliflozin 10 and 25 mg) [37-41]. For canagliflozin and empagliflozin, starting at the lower doses is recommended for all patients, with the option of increasing to the higher dose if additional glycemic control is required [37,39,41]. For dapagliflozin, this approach is also recommended by the US FDA [40], although the European Medicines Agency currently recommends starting with 10 mg except in patients with severe hepatic impairment, for whom a 5 mg starting dose is recommended (if well tolerated, this can be increased to 10 mg) [38]. No dose adjustment is recommended in patients with mild renal impairment, although careful monitoring of renal function is required. For patients with moderate renal impairment, lower doses or discontinuation is recommended, although precise recommendations differ for the different authorities, as discussed further in Section 8. All currently available agents are contraindicated in patients with end-stage renal disease or those on dialysis, since the efficacy of SGLT2 inhibitors depends on some degree of kidney function [37-41].

Metabolism of all three drugs occurs in the liver and kidneys, and elimination occurs via metabolic clearance of the drug and its metabolites, predominantly via the feces but also in the urine [37-39].

6. Safety and efficacy as monotherapy

A summary of efficacy and safety data for SGLT2 inhibitors as monotherapy is shown in Table 1 [42-44]. Dapagliflozin, canagliflozin, and empagliflozin given as monotherapy for approximately 6 months provided statistically significant and clinically relevant improvements in glycemic control in patients with T2DM, compared with placebo. The effect of SGLT2 inhibitor monotherapy on glycemic control was sustained, as confirmed by longer-term data from a canagliflozin trial, which reported dose-related decreases in HbA1c from baseline to week 52 of −0.81 and −1.11% for canagliflozin 100 and 300 mg, respectively [45]. As expected, larger reductions in HbA1c were observed in patients with higher baseline HbA1c levels. Modest decreases in body weight and systolic blood pressure (SBP) were also reported with each of these three agents when used as monotherapy [42-44]. When used as monotherapy, these SGLT2 inhibitors were all well tolerated, with few discontinuations due to adverse events (AEs) [42-44]. Dapagliflozin, canagliflozin, and empagliflozin monotherapy was associated with an increased incidence in symptoms suggestive of genital infection and of urinary tract infection (UTI). Where subgroup analyses by gender were available, these events were more common in women than in men [42-44]. The frequency of hypoglycemia was low and was comparable to placebo groups, and there were no episodes of severe hypoglycemia with SGLT2 inhibitor therapy in these three monotherapy studies [42-44].

Table 1. Efficacy and safety of SGLT2 inhibitors as monotherapy.

Reference and trial identification	Study details	Regimen*	N	Treatment and dose, mg/day	Baseline HbA1c, %	Efficacy parameters (change from baseline)^‡				Safety parameters (percentage of patients with a special interest adverse event)^‡
						HbA1c, %	FPG, mg/dl	Body weight, kg	SBP, mmHg	Hypoglycemia, %	Urinary tract infection, % (males and females, if stated)	Genital infection,% (males and females, if stated)^§
Ferrannini Diabetes Care 2010 [42] NCT00528372 (MB102013)	Phase III, 24 weeks	DAPA monotherapy
			75	Pbo	7.84	-0.23	-4.1	-2.2	-0.9	2.7	4.0	1.3
		Morning dose	64	DAPA 5	7.86	-0.77	-24.1	-2.8	-2.3	0	12.5	7.8
		Morning dose	70	DAPA 10	8.01	-0.89	-28.8	-3.2	-3.6	2.9	5.7	12.9
		Evening dose	68	DAPA 5	7.82	-0.79	-27.3	-3.6	-5.2	0	11.8	4.4
		Evening dose	76	DAPA 10	7.99	-0.79	-29.6	-3.1	-2.3	1.3	6.6	2.6
		High HbA1c (≥ 10.1%)	34	DAPA 5	10.82	-2.88	-77.1	-2.1	-5.7	2.9	8.8	5.9
		High HbA1c (≥ 10.1%)	39	DAPA 10	10.73	-2.66	-84.3	-1.9	-2.5	0	15.4	17.9
Stenlöf Diabetes Obes Metab 2013 [43] NCT01081834 (CANTATA-M)	Phase III, 26 weeks	CANA monotherapy
			192	Pbo	8.00	0.14	9.0	-0.5	0.4	2.6	4.2	2.1 (M0%, F3.8%)
			195	CANA 100	8.10	-0.77	-27.0	-2.5	-3.3	3.6	7.2	6.2 (M2.5%, F8.8%)
			197	CANA 300	8.00	-1.03	-34.2	-3.4	-5.0	3.0	5.1	6.6 (5.6%, F7.4%)
		High HbA1c (> 10.0 ≤ 12.0%)	47	CANA 100	10.60	-2.1	-81.1	-3.0	-4.5	Not reported	6.4	12.8 (M4.3%; F20.8%)
		High HbA1c (> 10.0 ≤ 12.0%)	44	CANA 300	10.60	-2.6	-86.5	-3.8	-5.0	Not reported	4.5	4.5 (M5.3%; F4.0%)
Roden Lancet Diabetes Endocrinol 2013 [44] NCT01177813 (EMPA-REG MONO™)	Phase III, 24 weeks	EMPA monotherapy
			228	Pbo	7.91	0.08	11.7	-0.33	-0.3	< 1	5 (M2%, F9%)	0
			224	EMPA 10	7.87	-0.66	-19.5	-2.26	-2.9	< 1	7 (M2%, F15%)	3 (M3%, F4%)
			224	EMPA 25	7.86	-0.78	-24.5	-2.48	-3.7	< 1	5 (M1%, F13%)	4 (M1%, F9%)
			223	SITA 100	7.85	-0.66	-6.85	0.18	0.5	< 1	5 (M3%, F9%)	1 (M1%, F1%)
		High HbA1c (> 10.0%)	87	EMPA 25	11.50	-3.70	-87.6	-2.43	-4.0	0	3 (M3%; F4%)	1 (M2%; F0%)

The SGLT2 inhibitors listed herein are those that gained marketing approval in the US and/or EU; data are listed only for licensed doses (dapagliflozin 5 and 10 mg; canagliflozin 100 and 300 mg; empagliflozin 10 and 25 mg).

*Data for high glycemic subgroups are presented for monotherapy studies only.

^‡Data are presented as reported in each publication (for randomized double-blind arms unless otherwise stated).

^§Genital mycotic infection was specified in canagliflozin studies.

CANA: Canagliflozin; DAPA: Dapagliflozin; EMPA: Empagliflozin; F: Female; FPG: Fasting plasma glucose; HbA1c: Glycated hemoglobin; M: Male; Pbo: Placebo; SBP: Systolic blood pressure; SGLT2: Sodium glucose co-transporter type 2; SITA: Sitagliptin.

7. Safety and efficacy in combination with other antidiabetes agents

A summary of efficacy and safety data from individual trials of dapagliflozin, canagliflozin, and empagliflozin as add-on combination therapy is shown in Tables 2,3,4 and 5 [46-62]. Add-on combinations include metformin [46-51], SU [52-55], DPP-4 inhibitor [56], TZD [57-59], and insulin [60-62]. In addition, several fixed-dose (i.e., single pill) combination products utilizing SGLT2 inhibitors are currently in clinical development (Table 6), of which dapagliflozin + metformin (in 5 mg/850 mg and 5 mg/1000 mg tablets) [63] and canagliflozin + metformin (in 50 mg/850 mg and 50 mg/1000 mg tablets as well as 150 mg/850 mg and 150 mg/1000 mg tablets) [64] are available in the EU. The canagliflozin/metformin combination is also available in the US (50 mg/500 mg and 50 mg/1000 mg tablets, and 150 mg/500 mg and 150 mg/1000 mg tablets) [65].

Table 2. Efficacy and safety of SGLT2 inhibitors as add-on to metformin.

Reference and trial identification	Study details	Regimen*	N	Treatment and dose (mg/day)	Baseline HbA1c, %	Efficacy parameters (change from baseline)^‡				Safety parameters (percentage of patients with a special interest adverse event)^‡
						HbA1c, %	FPG, mg/dl	Body weight, kg	SBP, mmHg	Hypoglycemia, %	Urinary tract infection, % (males and females, if stated)	Genital infection,% (males and females, if stated)^§
Bailey Lancet 2010 [46] NCT00528879 (MB102014)	Phase III, 24 weeks	DAPA + MET		MET ≥ 1500 mg/day
			137	Pbo	8.11	-0.30	-5.95	-0.9	-0.2	3	8	5
			137	DAPA 5	8.17	-0.70	-21.4	-3.0	-4.3	4	7	13
			135	DAPA 10	7.92	-0.84	-23.42	-2.9	-5.1	4	8	9
Henry Int J Clin Pract 2012 [47]	Phase III, 24 weeks	DAPA + MET		MET XR titrated to 2000 mg/day
NCT00643851 (MB102021)		Study 1: DAPA 5	201	Pbo + MET XR	9.14	-1.35	-33.51	-1.29	-1.8	0	7.5	2
			194	DAPA 5 + MET XR	9.21	-2.05	-61.10	-2.66	-2.9	2.6	7.7	6.7
			203	DAPA 5 + Pbo	9.14	-1.19	-42.00	-2.61	-4.2	0	7.9	6.9
NCT00859898 (MB102034)		Study 2: DAPA 10	208	Pbo + MET XR	9.03	-1.44	-34.78	-1.36	-1.2	2.9	4.3	2.4
			211	DAPA 10 + MET XR	9.10	-1.98	-60.36	-3.33	-3.3	3.3	7.6	8.5
			219	DAPA 10 + Pbo	9.03	-1.45	-46.49	-2.73	-4.0	0.9	11.0	12.8
Nauck Diabetes Care 2011 [48] NCT00660907 (D1690C00004)	Phase III, 52 weeks	DAPA + MET		MET 1500 – 2500 mg/day
			400	DAPA 2.5 – 10	7.70	-0.52	-22.34	-3.22	-4.3	3.4	10.8	12.3
			401	GLIP 5 – 20	7.70	-0.52	-18.74	1.44	0.8	39.7	6.4	2.7
Cefalu Lancet 2013 [49] NCT00968812 (CANTATA-SU)	Phase III, 52 weeks	CANA + MET		MET ≥ 1500 or ≥ 2000 mg/day
			483	CANA 100	7.80	-0.82	-25.3	-3.7	-3.3	6	6	8.9 (M7%, F11%)
			485	CANA 300	7.80	-0.93	-27.4	-4.0	-4.6	5	6	11.1 (M8%, F14%)
			482	GLIM 1 – 6 or 8	7.80	-0.81	-18.4	0.7	0.2	34	5	1.7 (M1%, F2%)
Lavalle González Diabetologia 2013 [50] NCT01106677 (CANTATA-D)	Phase III, 52 weeks	CANA + MET		MET ≥ 1500 or ≥ 2000 mg/day
			183	Pbo/SITA	8.00	Not reported	Not reported	Not reported	Not reported	2.7	6.6	1.1 (M1.1%, F1.1%)
			368	CANA 100	7.90	-0.73	-27.0	-3.3	-3.5	6.8	7.9	8.4 (M5.2%, F11.3%)
			367	CANA 300	7.90	-0.88	-36.0	-3.7	-4.7	6.8	4.9	6.5 (M2.4%, F9.9%)
			366	SITA 100	7.90	-0.73	-18.0	-1.2	-0.7	4.1	6.3	1.9 (M1.2%, F2.6%)
Häring Diabetes Care 2014 [51] NCT01159600 (EMPA-REG MET™)	Phase III, 24 weeks	EMPA + MET		MET ≥ 1500 mg/day
			207	Pbo	7.90	-0.13	6.3	-0.45	-0.4	0.5	4.9 (M2.6%, F7.7%)	0
			217	EMPA 10	7.94	-0.70	-20.0	-2.08	-4.5	1.8	5.1 (M0%, F12.0%)	3.7 (M0.8%, F7.6%)
			213	EMPA 25	7.86	-0.77	-22.3	-2.46	-5.2	1.4	5.6 (M0.8%, F11.8%)	4.7 (M0.8%, F9.7%)

*Data for high glycemic subgroups are presented for monotherapy studies only.

^‡Data are presented as reported in each publication (for randomized double-blind arms unless otherwise stated).

^§Genital mycotic infection specified in canagliflozin studies.

CANA: Canagliflozin; DAPA: Dapagliflozin; EMPA: Empagliflozin; F: Female; FPG: Fasting plasma glucose; GLIM: Glimepiride, GLIP: Glipizide; HbA1c: Glycated hemoglobin; M: Male; MET: Metformin; Pbo: Placebo; SBP: Systolic blood pressure; SGLT2: Sodium glucose co-transporter type 2; XR: Extended-release formulation.

Table 3. Efficacy and safety of SGLT2 inhibitors as add-on to sulfonylurea ± metformin.

Reference and trial identification	Study details	Regimen*	N	Treatment and dose (mg/day)	Baseline HbA1c, %	Efficacy parameters (change from baseline)^‡				Safety parameters (percentage of patients with a special interest adverse event)^‡
						HbA1c, %	FPG, mg/dl	Body weight, kg	SBP, mmHg	Hypoglycemia, %	Urinary tract infection, % (males and females, if stated)	Genital infection,% (males and females, if stated)^§
Strojek Diabetes Obes Metab 2011 [52] NCT00680745 (D1690C00005)	Phase III, 24 weeks	DAPA + SU		GLIM 4 mg/day, O/l
			145	Pbo	8.15	-0.13	-1.98	-0.72	-1.2	4.8	6.2	0.7
			142	DAPA 5	8.12	-0.63	-21.26	-1.56	-4.0	6.9	6.9	6.2
			151	DAPA 10	8.07	-0.82	-28.47	-2.26	-5.0	7.9	5.3	6.6
Schernthaner Diabetes Care 2013 [53] NCT01137812 (CANTATA-D2)	Phase III, 52 weeks	CANA + MET + SU		MET ≥ 1500 or ≥ 2000 mg/day, SU ≥ half-max labeled dose
			377	CANA 300	8.10	-1.03	-28.7	-2.3	-5.1	43.2	4.0	11.9 (M9.2%, F15.3%)
			378	SITA 100	8.10	-0.66	-2.2	0.1	0.9	40.7	5.6	2.1 (M0.5%, F4.3%)
Wilding Int J Clin Pract 2013 [54] NCT01106625 (CANTATA-MSU)	Phase III, 26 weeks^§	CANA + MET + SU		MET ≥ 1500 or ≥ 2000 mg/day, SU ≥ half-max labeled dose
			156	Pbo	8.10	-0.13	3.6	-0.8	-2.7	17.9	7.7	3.2 (M1.3%; F5.0%)
			157	CANA 100	8.10	-0.85	-18.0	-1.9	-4.9	33.8	8.3	13.3 (M7.9%; F18.5%)
			156	CANA 300	8.10	-1.06	-30.6	-2.5	-4.3	36.5	8.3	11.5 (M5.7%; F18.8%)
Häring Diabetes Care 2013 [55] NCT01159600 (EMPA-REG METSU™)	Phase III, 24 weeks	EMPA + MET + SU		MET ≥ 1500 mg/day or MTD, SU ≥ half-max labeled dose or MTD
			225	Pbo	8.15	-0.17	5.6	-0.39	-1.4	8.4	8.0 (M2.7%, F13.3%)	0.9 (M0.9%, F0.9%)
			225	EMPA 10	8.07	-0.82	-23.2	-2.16	-4.1	16.1	10.3 (2.7%, F18.0%)	2.7 (M0.9%, F4.5%)
			216	EMPA 25	8.10	-0.77	-23.2	-2.39	-3.5	11.5	8.3 (M0%, F17.5%)	2.3 (M0.9%, F3.9%)

*Data for high glycemic subgroups are presented for monotherapy studies only.

^‡Data are presented as reported in each publication (for randomized double-blind arms unless otherwise stated).

^§Genital mycotic infection specified in canagliflozin studies.

CANA: Canagliflozin; DAPA: Dapagliflozin; EMPA: Empagliflozin; F: Female; FPG: Fasting plasma glucose; GLIM: Glimepiride; HbA1c: Glycated hemoglobin; M: Male; MET: Metformin; MTD: Maximum tolerated dose; Pbo: Placebo; SBP: Systolic blood pressure; SGLT2: Sodium glucose co-transporter type 2; SITA: Sitagliptin; SU: Sulfonylurea.

Table 4. Efficacy and safety of SGLT2 inhibitors as add-on to a DPP4i ± metformin, and add-on to a thiazolidinedione ± metformin.

Reference and trial identification	Study details	Regimen*	N	Treatment and dose (mg/day)	Baseline HbA1c, %	Efficacy parameters (change from baseline)^‡				Safety parameters (percentage of patients with a special interest adverse event)^‡
						HbA1c, %	FPG, mg/dl	Body weight, kg	SBP, mmHg	Hypoglycemia, %	Urinary tract infection, % (males and females, if stated)	Genital infection,% (males and females, if stated)^§
Add-on to DPP4i ± MET
Jabbour Diabetes Care 2013 [56] NCT00984867 (D1690C00010)	Phase III, 24 weeks	DAPA + DPP4i ± MET		SITA 100 mg/day, MET ≥ 1500 mg/day
		Entire cohort	224	Pbo	8.00	0.0	3.8	-0.3	-5.1^¶	1.8	4.0	0.4
		Entire cohort	223	DAPA 10	7.90	-0.5	-24.1	-2.1	-6.0^¶	2.7	4.9	8.4
		(+ SITA)	111	Pbo + SITA	8.10	0.1	4.6	-0.1	-4.2^¶	Safety data presented for entire cohort only	-	-
		(+ SITA)	110	DAPA 10 + SITA	8.00	-0.5	-22.0	-1.9	-6.6^¶	-	-	-
		(+ SITA+ MET)	113	Pbo + SITA + MET	7.90	-0.0	3.0	-0.5	-5.5^¶	-	-	-
		(+ SITA+ MET)	113	DAPA 10+ SITA + MET	7.80	-0.4	-26.2	-2.4	-5.3^¶	-	-	-
Add-on to TZD ± MET
Rosenstock Diabetes Care 2012 [57] NCT00683878 (MB102030)	Phase III, 48 weeks	DAPA + TZD		PIO ≥ 30 mg/day
			139	Pbo	8.34	-0.54	-13.1	2.99	2.0	0.7	7.9	2.9
			141	DAPA 5	8.40	-0.95	-22.8	1.35	-1.0	2.1	8.5	9.2
			140	DAPA 10	8.37	-1.21	-33.1	0.69	-2.2	0	5.0	8.6
Forst Diabetes Obes Metab 2014 [58] NCT01106690 (CANTATA-MP)	Phase III, 52 weeks	CANA + MET + TZD		MET & PIO given at stable background therapy dose (no further details)
			113	CANA 100	8.00	-0.92	-26.7	-2.5	-3.4	4.4	5.3	8.0 (M3.9%, F16.7%)
			114	CANA 300	7.90	-1.03	-31.5	-3.6	-3.7	6.1	7.9	12.3 (M4.8%, F21.6%)
Kovacs Diabetes Obes Metab 2014 [59] NCT01210001 (EMPA-REG PIO™)	Phase III, 24 weeks	EMPA + TZD ± MET		PIO ≥ 30 mg/day, MET ≥ 1500 mg/day or MTD
			165	Pbo	8.20	-0.11	6.47	0.34	0.7	1.8	16.4 (M8.2%, F22.8%)	2.4 (M1.4%, F3.3%)
			165	EMPA 10	8.10	-0.59	-17.0	-1.62	-3.1	1.2	17.0 (M3.6%, F30.5%)	8.5 (M7.2%, F9.8%)
			168	EMPA 25	8.10	-0.72	-22.0	-1.47	-4.0	2.4	11.9 (M2.4%, F21.7%)	3.6 (M1.2%, F6.0%)

*Data for high glycemic subgroups are presented for monotherapy studies only.

^‡Data are presented as reported in each publication (for randomized double-blind arms unless otherwise stated).

^§Genital mycotic infection specified in canagliflozin studies.

^¶Seated SBP in patients with seated SBP ≥ 130 mmHg at baseline, to week 8.

CANA: Canagliflozin; DAPA: dapagliflozin; DPP4i: Dipeptidyl peptidase-4 inhibitor; EMPA: Empagliflozin; F: Female; FPG: Fasting plasma glucose; HbA1c: Glycated hemoglobin; M: Male; MET: Metformin; MTD: Maximum tolerated dose; Pbo: Placebo; PIO: Pioglitazone; SBP: systolic blood pressure; SGLT2: Sodium glucose co-transporter type 2; SITA: Sitagliptin; TZD: Thiazolidinedione.

Table 5. Efficacy and safety of SGLT2 inhibitors as add-on to insulin.

Reference and trial identification	Study details	Regimen*	N	Treatment and dose (mg/day)	Baseline HbA1c, %	Efficacy parameters (change from baseline)^‡				Safety parameters (percentage of patients with a special interest adverse event)^‡
						HbA1c, %	FPG, mg/dl	Body weight, kg	SBP, mmHg	Hypoglycemia, %	Urinary tract infection, % (males and females, if stated)	Genital infection,% (males and females, if stated)^§
Wilding Ann Intern Med Intern Med 2012 [60] NCT00673231 (D1690C00006)	Phase III, 48 weeks	DAPA + INS high dose ± OADs		INS ≥ 30 units/day
			193	Pbo	8.47	-0.47	Not reported	0.82	-1.49	51.8	5.1	2.5
			211	DAPA 5	8.62	-0.96	-16.22	-1.00	-4.33	55.7	10.8	9.9
			194	DAPA 10	8.57	-1.01	-16.22	-1.61	-4.09	53.6	10.2	10.7
Matthews A764 EASD 2012 [61] NCT01032629 (CANVAS, INS substudy)	Phase III, substudy efficacy duration 18 weeks	CANA + INS ± OADs		INS ≥ 30 units/day
			565	Pbo	8.30 (reported mean across the 3 groups)	Δ versus Pbo	versus Pbo	versus Pbo	versus Pbo	37	2.1	(M0.5%, F2.2%)
			566	CANA 100		-0.65	-22.52	-1.9%	-2.6	49	2.3	(M4.0%, F11.8%)
			587	CANA 300		-0.73	-29.01	-2.4%	-4.4	48	3.4	(M8.3%, F9.9%)
Rosenstock ADA 2013 1102-P [62] NCT01011868 EMPA-REG BASAL™	Phase IIb, 78 weeks	EMPA + INS
			170	Pbo	8.10	-0.02	3	0.7	0.1	35.3	8.8	1.8
			169	EMPA 10	8.26	-0.48	-10	-2.2	-4.1	36.1 (pooled)	14.8	7.7
			155	EMPA 25	8.34	-0.64	-15	-2.0	-2.4	(See row above)	11.6	5.2

The following studies are on ClinicalTrials.gov, but study results have not yet been posted: NCT01809327 (Phase III, 26 weeks, CANA + MET XR in treatment naïve); NCT01606007 (Phase III, 24 weeks, DAPA + DPP4i [SAXA] + MET); NCT01646320 (Phase III, 24 weeks, Triple therapy: DAPA + DPP4i [SAXA] + MET); NCT01734785 (Phase III, 24 weeks, EMPA + DPP4i [LINA]); NCT01422876 (Phase III, 52 weeks, EMPA/DPP4i [LINA] FDC).

*Data for high glycemic subgroups are presented for monotherapy studies only.

^‡Data are presented as reported in each publication (for randomized double-blind arms unless otherwise stated).

^§Genital mycotic infection specified in canagliflozin studies.

Δ: Change; CANA: Canagliflozin; DAPA: Dapagliflozin; DPP4i: Dipeptidyl peptidase-4 inhibitor; EMPA: Empagliflozin; F: Female; FDC: Fixed-dose combination (i.e., single pill); FPG: Fasting plasma glucose; HbA1c: Glycated hemoglobin; INS: Insulin; LINA: Linagliptin; M: Male; MET: Metformin; OAD: Oral antidiabetes drug; Pbo: Placebo; SAXA: Saxagliptin; SBP: Systolic blood pressure; SGLT2: Sodium glucose co-transporter type 2; XR: Extended-release formulation.

Table 6. Fixed-dose (i.e., single pill) combination products, containing SGLT2 inhibitor + another oral antidiabetes agent, currently in clinical development.

Drug combination	Sponsor(s)	Indication	Global status
Dapagliflozin + metformin IR	AstraZeneca, Bristol-Myers Squibb	T2DM	Registered in EU (Jan 2014)
Dapagliflozin + saxagliptin	AstraZeneca, Bristol-Myers Squibb	T1DM, T2DM	Phase III
Canagliflozin + metformin IR	Johnson & Johnson	T2DM	Registered in EU (April 2014) and US (August 2014)
Canagliflozin + metformin ER	Johnson & Johnson, PDL BioPharma	T2DM	Phase III
Empagliflozin + linagliptin	Boehringer Ingelheim, Eli Lilly	T2DM	Submitted US
Empagliflozin + metformin IR	Boehringer Ingelheim, Eli Lilly	T2DM	Phase I
Empagliflozin + metformin ER	Boehringer Ingelheim, Eli Lilly, PDL BioPharma	T2DM	Phase I

ER: Extended-release formulation, EU: European Union; IR: Immediate-release formulation; T1DM: Type 1 diabetes mellitus; T2DM: Type 2 diabetes mellitus; US: United States.

7.1 Safety

SGLT2 inhibitors were generally well tolerated in all combination therapies used to date and trials have reported few serious AEs. As with the monotherapy trials, there was an increased frequency of symptoms suggestive of genital infection and of UTI with add-on combination therapy involving dapagliflozin, canagliflozin, and empagliflozin (Tables 2,3,4 and 5). These events could both be associated with the persistent presence of glucose in the urine resulting from SGLT2 inhibitor therapy. This trend appeared to be consistent across studies, irrespective of the combination regimen, and was verified by analyses of pooled data from Phase III trials of these three SGLT2 inhibitors (predominantly from trials using combination therapy) [66-70]. For dapagliflozin, canagliflozin, and empagliflozin trials, the majority of these events were mild in severity and responded to standard therapies, and very few patients discontinued treatment because of these events [66-70].

7.2 Hypoglycemia

As discussed above, SGLT2 inhibitors used as monotherapy are not associated with an increased risk of hypoglycemia; however, based on their mechanism of action, hypoglycemia risk might be increased in certain combination regimens. SGLT2 inhibitors reduce blood glucose levels independent of insulin, so no increased risk of hypoglycemia is predicted when used in combination with drugs that do not affect insulin levels, such as metformin or TZDs; however, with insulin or insulin secretagogues such as SUs, the reduced blood glucose levels would be predicted to increase hypoglycemia risk unless the dosage of insulin was reduced. This is supported by the clinical trial evidence (Tables 2,3 and 5) and, as predicted, the frequency of hypoglycemia with SGLT2 inhibitor combination therapy is dependent upon the choice of glucose-lowering therapy that is coadministered, with an increased frequency of hypoglycemic events reported when used in combination with SU or insulin. This is reflected in prescribing information for available agents, which advise consideration of a lower dose of insulin or an insulin secretagogue (e.g., SU) to reduce the risk of hypoglycemia [37-41].

7.3 Glycemic control

Dapagliflozin, canagliflozin, and empagliflozin given in add-on combination therapy for approximately 6 months provided statistically significant and clinically relevant improvements in glycemic control in patients with T2DM, compared with placebo (Tables 2,3,4 and 5). Dapagliflozin, at daily doses of 5 and 10 mg, demonstrated statistically significant reductions in HbA1c versus placebo when used as combination therapy with different agents (add-on to metformin, SU, TZD, or insulin [± oral antidiabetes agents]) [35]. For canagliflozin add-on therapy, HbA1c results were generally consistent across placebo-controlled Phase III studies, with reductions relative to placebo ranging from −0.70 to −0.92% with the 300 mg dose and from −0.57 to −0.74% with the 100 mg dose; these changes were statistically significant versus placebo (p < 0.001 for each study) [36]. For empagliflozin, mean HbA1c reductions when given as an add-on to metformin + SU were −0.82 and −0.77%, respectively, for empagliflozin 10 and 25 mg, −0.59 and −0.72%, respectively, as add-on therapy to pioglitazone with/without metformin, and −0.48 and −0.64%, respectively, as add-on therapy to basal insulin (Tables 3,4 and 5); these changes were statistically significant versus placebo (p < 0.001 for each study) [55,59,62]. Longer-term data (≥ 52 weeks) from trials of all three SGLT2 inhibitors indicate that the glucose-lowering effect was maintained [48-50,53,62]. Therefore, while mean HbA1c reductions will depend to an extent on the baseline HbA1c in each trial, as well as other patient characteristics such as the duration of diabetes, consistent results have been seen across the class when added to existing therapies.

7.4 Body weight

In the general population, overweight/obesity is associated with an increased risk of CV disease [71]. In T2DM, however, this association is complicated by the issue of how to adjust for confounding factors such as hyperglycemia, hypertension, and dyslipidemia. Hence, it is more challenging to show conclusively a direct correlation between weight loss and reduction in CV risk [72,73]. Despite this, weight loss in T2DM is associated with other benefits such as improvements in quality of life, insulin resistance, and other CV risk factors. As a result, clinical guidelines recommend weight loss for overweight or obese patients with T2DM [72-74]. SGLT2 inhibitors are the first class of oral agents for the treatment of T2DM associated with weight reduction. As observed with monotherapy, moderate decreases in body weight were also reported with dapagliflozin, canagliflozin, and empagliflozin when given as add-on combination therapy. A study of dapagliflozin 10 mg added to metformin over 102 weeks reported a mean body weight reduction of −4.54 kg in dapagliflozin-treated patients with reduced fat mass of −2.80 kg (vs −2.12 kg and −1.46 kg for placebo, respectively) [75]. Imaging studies confirmed that the weight reduction was due to decreased body fat rather than due to dehydration, and was explained by loss of glucose via the kidney [75], with UGE of 50 – 80 g/day being equivalent to 200 – 320 kcal. A further study of dapagliflozin 10 mg added to pioglitazone over 48 weeks revealed that dapagliflozin offset the weight gain caused by pioglitazone (mean change in body weight: +2.99 kg vs +0.69 kg for placebo and dapagliflozin groups, respectively) [57]. Canagliflozin added to metformin significantly reduced body weight after 52 weeks, whereas there was a slight increase in weight with the SU comparator (−3.7 kg and −4.0 kg for canagliflozin 100 and 300 mg, respectively, vs +0.7 kg for SU; p < 0.0001) [49]. Approximately two-thirds of the reduction in body weight in the canagliflozin groups was from fat mass and one-third from lean body mass [49]. Canagliflozin added to pioglitazone over 52 weeks resulted in mean absolute changes in body weight of −2.5 and −3.6 kg for canagliflozin 100 and 300 mg groups, respectively [58]. For empagliflozin, reductions in body weight from baseline levels were seen with doses of 10 and 25 mg after 24 weeks compared with placebo (Tables 2,3,4 and 5). The mean change in body weight after 24 weeks for empagliflozin added to metformin was −2.08 and −2.46 kg for empagliflozin 10 and 25 mg groups, respectively, versus −0.45 kg for placebo (p < 0.001 for each dose vs placebo) [51]. When empagliflozin was added to pioglitazone (± metformin), there were significant reductions from baseline in body weight in the empagliflozin groups compared with placebo, with adjusted mean changes of +0.34 kg with placebo versus −1.62 kg with empagliflozin 10 mg and −1.47 kg with empagliflozin 25 mg (both p < 0.001) [59].

7.5 Blood pressure

In contrast to the available data for the association between obesity/overweight and CV risk, there is clear evidence that hypertension in patients with diabetes increases their risk of CV disease [72,76]. Furthermore, a number of studies have shown that lowering blood pressure (BP) in patients with concomitant diabetes and hypertension lowers that CV risk [77-79]. There is some discrepancy between international treatment guidelines as to the recommended treatment target for BP lowering, but patients with diabetes should be treated to a target of at least < 140/80 mmHg [72]. SGLT2 inhibitors are not indicated for hypertension, but may help in goal attainment in patients close to goal. For most clinical trials of SGLT2 inhibitors, BP has been measured as an exploratory endpoint and, in general, patients receiving dapagliflozin, canagliflozin, or empagliflozin as add-on therapy showed reductions in SBP of approximately 3 – 5 mmHg versus placebo (Tables 2,3,4 and 5). These reductions appeared to be consistent irrespective of background/combination regimen. A systematic review of 10 dapagliflozin randomized controlled trials, including monotherapy and add-on combination therapy studies, reported that the decrease in seated SBP was greater with dapagliflozin versus placebo (weighted mean difference: −3.57 mmHg; 95% confidence intervals: −4.38, −2.77; p < 0.00001; I² = 0% where I² is the percentage of total variation across studies that is due to heterogeneity rather than chance; a value of 0% indicates no observed heterogeneity [80]) [81]. Canagliflozin data from a pooled analysis of six Phase III trials, including add-on combination therapy studies, reported reductions in SBP of −3.3 and −4.5 mmHg for 100 and 300 mg, respectively, relative to placebo. In subjects with elevated SBP (≥ 140 mmHg), placebo-corrected reductions were observed for both doses of canagliflozin (−3.2 and −5.6 mmHg for 100 and 300 mg, respectively; p < 0.001 for each) [82]. Similar data were reported for empagliflozin when a pooled analysis of four 24-week Phase III trials investigating empagliflozin as monotherapy or add-on therapy (+ metformin, or metformin + SU, or pioglitazone ± metformin) demonstrated reductions in SBP for empagliflozin groups versus placebo (placebo-corrected change from baseline −3.4 and −3.8 mmHg for empagliflozin 10 and 25 mg, respectively) [83]. A study of patients with T2DM and hypertension found that empagliflozin 10 and 25 mg significantly reduced mean 24-h SBP, measured via ambulatory BP monitoring, versus placebo (−2.95 and −3.68 mmHg vs +0.48 mmHg, respectively; p < 0.001 vs placebo for each dose) [84]. It is postulated that the UGE stimulated by SGLT2 inhibition causes a diuretic effect leading to decreased BP; however, a recent study of dapagliflozin proposed that SGLT2 inhibitors may possess an additional diuretic-like capacity to lower BP [85].

7.6 Other CV risk factors

Overall, the focus of trials of SGLT2 inhibitors has been on younger patients with longer life expectancy, where treatment decisions must consider long-term benefits, in particular CV outcomes. In clinical trials, SGLT2 inhibitors have demonstrated changes in a number of CV risk factors in addition to HbA1c. As discussed in the previous section, improvements in body weight and BP suggest the potential for reducing the risk of CV events and, based on simulation modeling, significant reductions in the risk of myocardial infarction (MI), stroke, CV death, and all-cause death could be expected with SGLT2 inhibitor treatment versus standard care [86].

Effects on lipids, including low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C), have also been seen in clinical trials, and have received considerable attention, since both elevated LDL-C levels and decreased HDL-C levels are predictors of CV risk in patients with T2DM [87]. A review of dapagliflozin Phase III trials (as monotherapy and as add-on combination therapy) reported that, across the individual studies, the mean change from baseline in LDL-C ranged from −0.5 to +9.5%, and the corresponding values for HDL-C ranged from +2.1 to +9.3%, in patients receiving dapagliflozin [88]. Dose-related increases in LDL-C were observed with canagliflozin; pooled data from four 26-week placebo-controlled trials revealed that the mean percentage increase from baseline in LDL-C for 100 and 300 mg canagliflozin relative to placebo were 4.5 and 8.0%, respectively [41]. Monitoring of LDL-C and treatment per standard care is recommended after initiating canagliflozin [41]. A pooled analysis of four placebo-controlled Phase III trials of empagliflozin reported small increases in both HDL-C and LDL-C with empagliflozin 10 and 25 mg versus placebo after 24 weeks. For HDL-C, the change from baseline was 0.00 mmol/l (0.00 mg/dl), +0.07 mmol/l (+2.71 mg/dl), and +0.07 mmol/l (+2.71 mg/dl) for placebo, empagliflozin 10 and 25 mg, respectively (p < 0.001 for both empagliflozin doses vs placebo). For LDL-C, the corresponding values were +0.02 mmol/l (+0.77 mg/dl), +0.08 mmol/l (+3.10 mg/dl), and +0.10 mmol/l (+3.87 mg/dl), respectively (p < 0.01 for empagliflozin 25 mg vs placebo) [83].

These data show that small, dose-related increases in LDL-C can occur with SGLT2 inhibitor treatment, although increases in HDL-C were also observed in these trials. In the context of CV risk factor management, primary prevention of CV disease has been demonstrated in patients with T2DM following statin therapy [89], and treatment guidelines recommend these agents for the treatment of dyslipidemia in this patient population [73], and if increases are seen with SGLT2 inhibitor therapy, LDL-C should be treated per standard of care. In terms of the increases in HDL-C seen with SGLT2 inhibitors, given that HDL-C levels are inversely related to CV disease in observational studies, an increase in HDL-C levels would be considered desirable. However, data from clinical studies aimed specifically at reducing CV risk through increasing HDL-C levels have been less conclusive [73], to the extent that lifestyle interventions such as increased physical activity and weight reduction remain the cornerstone of increasing HDL-C [73]. Therefore, the increases with HDL-C seen with SGLT2 inhibitors would appear advantageous, but the extent to which this would have an impact on CV risk reduction remains to be determined.

Raised uric acid levels are associated with ischemic heart disease and stroke [90,91], and possible improvements in uric acid levels with SGLT2 inhibitor suggest the potential for reduced risk of CV events with long-term use. Reductions in mean blood uric acid levels from baseline to week 24 of up to −55.32 μmol/l (−0.93 mg/dl) were observed in an analysis of four dapagliflozin Phase III trials, where patients had normal baseline mean uric acid levels [88]. For canagliflozin in add-on combination therapy, decreases in serum urate were observed after 26 weeks compared with placebo (−8.8 and −9.4% for canagliflozin 100 and 300 mg, respectively, vs +0.7% for placebo) [54], and after 52 weeks compared with active comparator (−9.9 and −10.3% for canagliflozin 100 and 300 mg, respectively, vs +8.0% for glimepiride [49]; and −6.5% for canagliflozin 300 mg vs +6.2% for sitagliptin) [53]. Empagliflozin reduced blood uric acid versus placebo at week 24 in a pooled analysis of four Phase III trials (−28.95 μmol/l [−0.49 mg/dl] and −29.55 μmol/l [−0.50 mg/dl] for empagliflozin 10 and 25 mg, respectively, vs +1.03 μmol/l [+0.02 mg/dl] for placebo; p < 0.001 vs placebo for both dose groups) [83].

7.7 CV outcome studies

Although SGLT2 inhibitors have shown a beneficial effect on CV risk factors such as HbA1c, body weight, and BP [92], currently available information on clinical outcomes such as stroke, MI, and CV death is limited. However, large CV trials are underway for dapagliflozin, canagliflozin, empagliflozin, and ertugliflozin, and it is hoped that the results of these trials will provide greater insights into the effects of SGLT2 inhibitors on CV outcomes.

The Dapagliflozin Effect on Cardiovascular Events (DECLARE-TIMI58: ClinicalTrials.gov identifier: NCT017 30534) is a multicenter, randomized, double-blind, placebo-controlled trial to evaluate the effect of dapagliflozin 10 mg once daily on the incidence of CV death, MI, or ischemic stroke in patients with T2DM. Recruitment began in 2013, and aims to enroll 27,000 patients with T2DM and a high risk of CV events. The primary endpoint is time to first event included in the composite endpoint of CV death, MI, or ischemic stroke. The study is expected to complete in 2019.

The Canagliflozin Cardiovascular Assessment Study (CANVAS; ClinicalTrials.gov identifier: NCT01032629) [93] completed recruitment for its first phase in 2012. CANVAS is a randomized, double-blind, placebo-controlled, parallel-group, multicenter trial designed to evaluate the effects of canagliflozin on the risk of CV disease and to assess safety and tolerability in patients with inadequately controlled T2DM and increased CV risk. The first of two phases randomized 4330 individuals to placebo or to canagliflozin 100 or 300 mg (1:1:1) with planned follow-up of approximately 2 years to substantiate potential CV protection by assessing key biomarkers and to achieve initial safety objectives. The primary outcome specified for the evaluation of the effects of canagliflozin on the risk for CV disease is the composite of CV death, nonfatal MI, and nonfatal stroke [93].

The Empagliflozin Cardiovascular Outcome Event Trial (EMPA-REG OUTCOME™, ClinicalTrials.gov identifier: NCT01131676) [94] has also completed recruitment. The aim of the trial is to determine the long-term CV safety of empagliflozin, as well as investigating potential benefits on microvascular outcomes. Patients with T2DM who were drug-naïve or on background glucose-lowering therapy, and at high risk of CV events, were randomized (1:1:1) to either empagliflozin 10 mg, empagliflozin 25 mg, or placebo (double-blind, double-dummy) superimposed on standard of care. The primary outcome is time to first occurrence of CV death, nonfatal MI, or nonfatal stroke [94].

A further CV outcomes trial, for ertugliflozin, has recently started recruitment (Cardiovascular Outcomes Following Treatment with Ertugliflozin in Participants with Type 2 Diabetes Mellitus and Established Vascular Disease; ClinicalTrials.gov identifier: NCT01986881). The main objective of this randomized, double-blind, placebo-controlled, parallel-group study is to assess the CV safety of ertugliflozin. The study includes a predefined glycemic substudy in participants receiving background insulin with or without metformin and another predefined glycemic substudy in participants receiving background SU monotherapy. The primary outcome is the time to first occurrence of any component of the composite endpoint of CV death, nonfatal MI, or nonfatal stroke. Estimated enrollment is 3900 patients, and the study is due to complete in 2021.

8. Other issues to consider

Patients with T2DM are a heterogeneous group, ranging from relatively young patients with little comorbidity to older patients with difficult-to-manage conditions, such as chronic kidney disease (CKD). As renal impairment is usually associated with more advanced T2DM, this population is more likely to require combination therapy to maintain glycemic control; however, many antidiabetes agents are unsuitable for use in patients with renal impairment. Metformin, for example, is generally considered as first-line therapy for patients with T2DM, but is contraindicated in patients with CKD. SGLT2 inhibitors are thought to have a good safety profile in patients with renal impairment but reduction in efficacy is predicted, based on the need for renal function for the mechanism of action. However, since options are limited in this group, SGLT2 inhibitors may still be considered, since trials with empagliflozin and canagliflozin have shown significant reductions in HbA1c in patients with mild or moderate CKD [95,96]. These HbA1c improvements were accompanied by weight loss and reductions in BP [82,83].

When dapagliflozin was tested in patients with moderate renal impairment, changes in HbA1c were not significantly improved compared with placebo, although reductions in weight and BP were observed [97]. Currently, prescribing information for dapagliflozin states that, in patients with mild renal impairment (estimated glomerular filtration rate [eGFR] ≥ 60 to < 90 ml/min/1.73 m²), no dose adjustment is required. However, dapagliflozin is currently not recommended for use in patients with either moderate to severe (eGFR < 60 ml/min/1.73 m²) or severe (eGFR < 30 ml/min/1.73 m²) renal impairment [38,40]. Similarly, canagliflozin and empagliflozin can be used without dose adjustment in patients with mild renal impairment, but not in patients with moderate to severe renal impairment. In patients tolerating canagliflozin or empagliflozin whose eGFR falls persistently below 60 ml/min/1.73 m², the dose should be adjusted to or maintained at 100 mg once daily (canagliflozin) or 10 mg once daily (empagliflozin). Canagliflozin and empagliflozin should be discontinued when eGFR is persistently below 45 ml/min/1.73 m² [37,39,41]. It is further recommended in the labeling information for these SGLT2 inhibitors that renal function is assessed before commencing SGLT2 inhibitor therapy and that assessment is repeated periodically during treatment [37-41].

The treatment of older patients with T2DM poses a particular challenge for several reasons, including their increased prevalence of comorbid conditions, and increased propensity to experience treatment-related hypoglycemia. Therefore, while guidelines recommend individualizing treatment for older patients based on their overall patient profile, rather than making decisions on age alone, in those with a shorter life expectancy, strategies that minimize the risk of hypoglycemia are often preferred. DPP-4 inhibitors, with their ease of use and low risk of hypoglycemia, are often recommended for this group [98], but where weight loss is desirable or additional glycemic control is required, SGLT2 inhibitors may be a useful add-on choice. While adult patients of a range of ages were included within the various Phase III trial programs for SGLT2 inhibitors, only a small number of studies looked specifically at older age groups. One study that did this evaluated the efficacy and safety of canagliflozin in T2DM patients (n = 716) aged 55 – 80 years (mean 63.6 years) who had inadequate glycemic control (HbA1c ≥ 7.0 to ≤ 10.0%) on their current therapy of blood glucose-lowering agents (defined as any oral or injectable treatment) [99]. At week 26, treatment with canagliflozin 100 and 300 mg significantly reduced HbA1c versus placebo (−0.60, −0.73, and −0.03%, respectively; p < 0.001) [99]. Both canagliflozin doses also significantly reduced body weight, FPG concentration, and SBP (p < 0.001 for each parameter vs placebo) [99]. Documented hypoglycemia was slightly higher in the canagliflozin groups compared with placebo (patients not on baseline glucose-lowering agents associated with hypoglycemia: 6.7 and 4.8% vs 3.2% for 100 and 300 mg vs placebo, respectively; patients on baseline glucose-lowering agents associated with hypoglycemia: 43.1 and 47.4% vs 37.7% for 100 and 300 mg vs placebo, respectively) [99]. Severe hypoglycemia events occurred most frequently in the placebo group in patients on baseline glucose-lowering agents associated with hypoglycemia (4.0 vs 1.1 and 0.6% for canagliflozin 100 and 300 mg, respectively) [99]. Overall, canagliflozin was well tolerated in older patients, and the safety profile was consistent with that observed in other Phase III trials of canagliflozin [99].

In terms of prescribing information, both European and US labels for the SGLT2 inhibitors recommend that factors such as renal function and volume depletion should be taken into account before initiating therapy although, in general, no dosage adjustment is recommended based on age [37-41]. In Europe, the following recommendations are made: for canagliflozin, care should be taken when increasing the dose in patients aged ≥ 75 years [39]; for dapagliflozin and empagliflozin, therapy is not recommended in patients aged ≥ 75 years (dapagliflozin) [38] or ≥ 85 years (empagliflozin) [37] due to limited therapeutic experience.

9. Expert opinion

As metformin remains the classic first-line drug for patients starting glucose-lowering therapy, SGLT2 inhibitors could be considered among the options for add-on therapy in patients who have not achieved glycemic targets with their initial glucose-lowering therapies. Their novel mechanism of action makes SGLT2 inhibitors suitable for use in combination with any background glucose-lowering agent, including insulin. This is supported by data from numerous clinical trials, where SGLT2 inhibitors have been successfully used as monotherapy and as add-on therapy with metformin alone or in combination with SUs, TZDs, DPP-4 inhibitors, and insulin. The optimum combination therapy using SGLT2 inhibitors will always depend on individual patient factors, such as the need for weight control/loss or the risk of hypoglycemia, and so on. For example, an SGLT2 inhibitor + a DPP-4 inhibitor + metformin may offer good glycemic control with a low risk of hypoglycemia, and would be weight-neutral or induce weight loss. Furthermore, as SGLT2 inhibitors are not dependent on the production of insulin, they may be considered for use at any stage of T2DM, from newly diagnosed patients to those with long-standing disease. In addition, they have the potential for use in T2DM patients already receiving insulin to provide an alternative to increasing the insulin dose or frequency.

Clinical trials of the SGLT2 inhibitors dapagliflozin, canagliflozin, and empagliflozin have investigated their efficacy, safety, and tolerability as monotherapy and as add-on combination therapy with other antidiabetes agents. In terms of efficacy, these agents reduced HbA1c, and produced reductions in body weight and SBP. While modest, the improvements seen are likely to be clinically meaningful, and changes have been sustained over long-term follow-up. The potential implications for CV risk factors could be favorable but this remains to be proven. To date, SGLT2 inhibitors have a generally favorable safety profile that is similar to that of placebo, and are well tolerated. Few serious AEs have been reported from clinical trials. The frequency of hypoglycemia is low and episodes are usually mild in severity and their frequency was comparable to that of comparators. The risk of hypoglycemia appears to depend on the choice of coadministered glucose-lowering agent, with an increased hypoglycemic risk associated with concomitant use of an SU or insulin. In addition, an increased risk of genitourinary infections has been reported with SGLT2 inhibitors, probably because the presence of glucose in the urine provides a suitable environment for the growth of micro-organisms; however, these infections are usually mild, nonrecurrent, and respond to standard treatment. Nevertheless, care should be taken in patients with a history of genitourinary infections.

In conclusion, SGLT2 inhibitors have the potential to make an important contribution to the treatment of T2DM. Their role in treatment algorithms is still to be refined, but given their benefits over SUs, long-term studies of combination regimens substituting SGLT2 inhibitors in place of SUs are eagerly awaited.

Article highlights.

• SGLT2 inhibitors are a class of glucose-lowering agents for T2DM with a novel, non-insulin–dependent mechanism of action.

• As monotherapy or combined with other glucose-lowering agents, SGLT2 inhibitors produce modest but clinically meaningful reductions in HbA1c, body weight and systolic blood pressure.

• SGLT2 inhibitors are generally well tolerated with a safety profile similar to that of placebo.

Declaration of interest

The author meets criteria for authorship as recommended by the International Committee of Medical Journal Editors (ICMJE). The author received no direct compensation related to the development of the manuscript. Writing assistance was provided by Debra Brocksmith, MB ChB, PhD, of Envision Scientific Solutions, which was contracted and funded by Boehringer Ingelheim Pharmaceuticals, Inc. (BIPI). BIPI was given the opportunity to review the manuscript for medical and scientific accuracy as well as intellectual property considerations. R Lajara has served as a scientific advisor for Eli Lilly and Dexcom; as a scientific advisor and on the speaker’s bureau for Boehringer Ingelheim, Novo Nordisk, and Sanofi; and on the speaker’s bureau for AstraZeneca, Abbott, Insulet and Valeritas.

Notes

This box summarizes key points contained in the article.