Glycaemic control can decrease microvascular complications in patients with type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM). However, the effect of glycaemic control on macrovascular complications using currently available therapeutic options has been disappointing.

Globally, the obesity and diabetes pandemics are driving the increase in macrovascular complications, with ischaemic heart disease and stroke the top causes of death worldwide. Diabetes caused 4.9 million deaths in 2014 (every seven seconds a person dies from diabetes), with cardiovascular disease (CVD) being the principal cause of death. In 2015, the number of deaths globally from diabetes exceeded the number of deaths from HIV/AIDS, malaria and TB combined. The prevalence of diabetes mellitus is expected to increase by 54% globally by 2030. In Africa, the prevalence of diabetes mellitus is expected to increase by 140% by the year 2040. The estimated national prevalence of diabetes in SA (based on HbA1c) in persons older than 15 years was 9.5% in 2012, with about 45% being undiagnosed. Asian and coloured populations have the highest prevalence of diabetes in SA.

Although the Diabetes Control and Complications Trial and UK Prospective Diabetes Study showed a reduction of microvascular complications with improved glucose control in patients with T1DM and T2DM, both studies failed to show a reduction in macrovascular complications within the timeframe of the original study. Since macrovascular complications are the predominant cause of mortality in patients with T1DM or T2DM, it is essential to determine ways to prevent these complications to improve the survival of our patients. A combined analysis of four major trials (27 000 patients) suggested that lowering HbA1c by 1% had a minor effect on heart disease, and no effect on stroke, especially in patients with established CVD. Until recently, there had been no major cardiovascular outcome trial showing that any antihyperglycemic agent could decrease the prevalence of CVD in patients with DM. The EmpaReg Outcome (Empagliflozin) trial was the first randomised-controlled trial to show significant reductions of CVD and mortality in patients with DM using an agent primarily targeting hyperglycaemia.

Subsequently, CANVAS and the LEADER (liraglutide) trial have also demonstrated a significant reduction in CVD and mortality. The addition of empagliflozin, canagliflozin and liraglutide to the treatment options for the management of hyperglycaemia in patients with T2DM may herald the beginning of a new era in which doctors can significantly reduce CVD and improve survival by treating hyperglycaemia with these agents.

These studies have thrust the sodium glucose cotransporter 2 (SGLT2) inhibitors (‘gliflozin’) into the forefront of T2DM guidelines. SGLT2 accounts for 90% of glucose reabsorption in the kidney. The SGLT2 inhibitors increase urinary excretion of glucose and lower plasma glucose levels in an insulin-

independent manner. This class of novel agents can effectively control the blood sugar level without producing weight gain or hypoglycemia, and two of these agents have been shown to reduce CVD and improve mortality.

Three drugs have been accepted by the Food and Drug Administration (FDA) in the US – dapagliflozin, canagliflozin and empagliflozin. Canagliflozin was the first SGLT-2 inhibitor that was approved by the FDA, it was accepted in March 2013. Dapagliflozin and empagliflozin were accepted in 2014.


Originally it was thought that diabetes mellitus was a renal disorder because of the glucose found in the urine. In 1835, French chemists isolated phlorizin from the bark of the apple tree. Due to the bitter taste of phlorizin, it was thought to have antipyretic properties, and was used for fevers and infectious diseases, particularly malaria. In 1886, Von Meiring observed that dogs receiving doses of phlorizin above 1g developed glycosuria.

Phlorizin seemed to have very interesting properties and the results in animal studies were encouraging as it improved insulin sensitivity and in diabetic rat models it seemed to increase glucose levels in urine and also normalised the glucose concentration in plasma without causing hypoglycaemia. Phlorizin has very poor oral bioavailability as  it is broken down in the gastrointestinal tract and therefore has to be given parenterally. Phlorizin was therefore never pursued in humans.

Although intensively investigated at the time, its mode of action remained mysterious until the elucidation of sodium/glucose cotransporter mechanisms in the 1960s.

After insulin was discovered, the focus of diabetes management was on the pancreas. Traditionally, the focus of therapeutic strategies for diabetes have been to enhance endogenous insulin secretion and/or to improve insulin sensitivity. Therefore, most other antihyperglycemic agents available are involved with insulin action. Over the past decade, the role of the kidney in glucose homeostasis has been re-examined.

Experience with phlorizin stimulated the development of a new class of drugs, the SGLT-2 inhibitors.


Normally, the kidneys filter and reabsorb ~180g of glucose daily. Physiologically, the higher the plasma glucose, the more glucose is reabsorbed by the kidneys. Glucose is a polar compound and its solubility and transportability occur through specialised tissue glucose transporters, particularly in the renal tubule, the small intestine, the brain, and peripheral tissues.

Two gene families are involved – the sodium-glucose cotransporters (SGLTs) and facilitated glucose transporters (GLUTs). GLUTs facilitate passive transport along membranes and SGLTs are involved in active transport. After being filtered through the glomeruli, 90% of glucose is is reabsorbed by a high capacity system controlled by SGLT-2 transport proteins in the early convoluted segment of the proximal tubules. The remaining 10% of glucose not re-absorbed by SGLT2 is then reabsorbed by sodium glucose transporter 1 (SGLT-1), therefore all filtered glucose is usually reabsorbed. SGLT2, the most important renal transporter, is minimally expressed elsewhere, while SGLT1 is widely expressed, so inhibition SGLT2 would not be expected to have wide physiological effects. Out of the six families of SGLTs, the SGLT2 cotransporters have been the most successful as therapeutic targets as they are kidney specific.

Inhibition of SGLT-2 reduces the capacity for tubular glucose reabsorption by approximately 30%-50% resulting in loss of filtered glucose in the urine (glucosuria, approximately 70g daily) and also results in better control of the plasma glucose level, lower insulin levels, lower blood pressure (due to osmotic diuresis and intrarenal mechanisms), lower uric acid levels and augments calorie wasting. Some data also supports a direct renoprotective effect of SGLT-2 inhibition. There are multiple hypotheses suggesting why SGLT2 inhibitors are only able to block 30%-50% of tubular glucose reabsorption. These include competitive inhibition progressively raising the local glucose concentration and reducing its effectiveness, upregulation of SGLT1 and other GLUTs, saturation of renal secretion of SGLT2-inhibitors limiting the amount of SGLT2 inhibitor in the proximal tubule.

SGLT2 inhibitors have insulin- independent effects. These compounds induce weight loss and decrease blood pressure through processes that are distinct from those involved in decreasing plasma glucose levels. In experimental models of untreated diabetes, proximal tubular sodium resorption is increased, and there is less sodium delivered to distal portions of the nephron and the juxtaglomerular apparatus. Accordingly, the transduced signal is a diminished plasma volume, which leads to increased intraglomerular pressure and, ultimately, hyperfiltration, along with increases in blood pressure. SGLT2 inhibitors reverse these changes by blocking proximal tubular sodium resorption, which results in negative sodium balance, decreased plasma volume, and reduced blood pressure.

Thus, SGLT2 inhibitors alter intrarenal haemodynamics. It is now recognised that renal abnormalities lead to an increase in cardiovascular risk.


As a class, the SGLT2-inhibitors have been shown to cause a mean reduction in HbA1c of 0.8%-1%. A SGLT2-inhibitor can safely be used as monotherapy, or combined with another antihyperglycemic agent, including metformin, sulphonylureas, thiazolidinediones, DPP4-inhibitors, GLP1-RA or insulin. The HbA1c reduction is in addition to that expected from the background medication, and has been shown to be sustained for two years.


Low risk of hypoglycemia: The mechanism of action for SGLT2- inhibitors is insulin-independent, reducing HbA1c via actions in the kidney, therefore hypoglycemia is extremely rare, unless combined with either a sulphonylurea or insulin. These agents are more effective at higher plasma glucose levels, and will cause less glucosuria during normoglycemia, having a low risk of hypoglycemia.

Weight loss: Due to the glucosuria and approximate daily loss of 280 calories (~ 70g glucose), the SGLT2- inhibitors can cause an average weight loss of up to 3kg. The weight loss is similar to that achieved with intensive lifestyle changes, for example, a 2/3 reduction of fat mass and a 1/3 reduction of lean tissue mass. The weight loss has been shown to be sustained for at least two years. Blood Pressure Reduction: SGLT2 inhibition induces an osmotic diuresis and alters intrarenal haemodynamics (as discussed above). This has a favourable effect on blood pressure (systolic BP reduction 3-4mmHg), heart failure, and cardiovascular outcomes.

Nephroprotection: Kidney disease is a critical determinant of death from cardiovascular causes in persons with diabetes mellitus. Beyond medications that control glycaemia and blood pressure, only medications that inhibit the renin–angiotensin system (RAS) have had a robust renoprotective effect.

A recently published secondary analysis of the EMPA-REG OUTCOME trial confirmed the renoprotective effects of empagliflozin. Furthermore the CANVAS trial also showed a possible benefit of canagliflozin with respect to the progression of albuminuria and the composite outcome of a sustained 40% reduction in the estimated glomerular filtration rate, the need for renal replacement therapy, or death from renal causes.

Macrovascular risk reduction: The cardiovascular outcome trial for empagliflozin and canagliflozin have been reported. Patients enrolled in the EMPA-REG OUTCOME trial were at high risk for vascular disease (close to 50% of the participants had a history of myocardial infarction, 75% had evidence of coronary artery disease, 25% had previous stroke, and 20% had peripheral vascular disease. In patients receiving 10mg or 25mg of empagliflozin daily in addition to standard care, there was a significant 14% risk reduction for threepoint MACE and reduced death from any cause.

Notably, the benefit occurred within weeks of starting the treatment. The analysis showed that 62 patients would need to be treated for three years to provide a benefit in these outcomes.

It seems likely that the observed risk reduction in the study was multifactorial.

The dapagloflozin (DECLARE-TIMI58) cardiovascular outcome trial result is expected in 2019.

The CVD REAL study was an observational, restrospective analysis where treatment with SGLT-2i vs. other glucose-lowering drugs was associated with a lower risk of HHF and death, suggesting that the benefi ts seen with empagliflozin in a randomised trial may be a class effect applicable to a broad population of T2D patients in real-world practice.

The addition of an SGLT2 inhibitor to standard care may ultimately alter vascular reactivity, as well as cardiac and cardiorenal function. A decrease in albuminuria or in uric acid, that is indirectly induced by the drug, may be beneficial as well. Other aspects of the effects of the drug may be at play through mechanisms that remain speculative.


Adverse events associated with SGLT2- inhibitors are mild and (as anticipated with a drug that causes glucosuria) included more genital infections (especially candida infection), polyuria and an increased risk of dehydration.

Infections: The cumulative incidence of urinary tract infections (UTIs) vary between the three available SGLT2 inhibitors, from 2.3% at 24 weeks to 9.4% at 76 weeks. Usually, two UTIs should trigger a change of therapy.

Pyelonephritis is uncommon and similar to the placebo groups. Practically, it is to be expected that a susceptible patient will present with one extra mycotic genital infection per year. The urinary tract and genital mycotic infections

can usually be treated with first-line antibiotics or antifungals. Due to the infection risk, patients with recurrent mycotic genital infections or UTIs may not be ideal candidates for treatment with SGLT2-inhibitors.

Osmotic diuresis: As glucosuria is associated with osmotic diuresis, patients treated with a SGLT2-inhibitor may void an extra 300-400ml over 24 hours. This can be associated with a transient increase in serum creatinine, a drop in eGFR of approximately 5ml/ min/1.73m2, which tends to improve after the first week of treatment and normalise over six months in >95% of patients. Patients at high risk for dehydration and hypotension may therefore not be ideal candidates treatment with SGLT2-inhibitors.

Ketoacidosis: Although this is rare (150 cases described worldwide), fatal ketoacidosis has been described in patients treated with an SGLT2-inhibitor. Ketoacidosis was mainly seen in insulindependent patients, and those patients with latent autoimmune diabetes of adults, who, for various sreasons, stopped their insulin treatment. The absence of insulin triggered ketoacidosis.

Insulin-dependent patients should be educated regarding the risk, and advised to continue insulin whenever using an SGLT2-inhibitor, and to consider omitting both (along with metformin) when ill.

Lipid profile: Slight increase in LDL and HDL with a decrease in LDL: HDL ratio, with no current evidence of any clinical implications on cardiovascular risk.

Other: In the Canagliflozin studies, there has been an observation of an increase in bone fractures and small toe amputations. This has not been observed in studies using other SGLT-2 inhibitors. Dapagliflozin had a signal for increased bladder cancer, although

patients who developed bladder cancer were all found to have had haematuria prior to entering the study, which suggests the possibility of pre-existing bladder pathology.


Currently, the SGLT2-inhibitors are positioned as alternative options in treatment of T2DM to be used as second- or third-line therapy. They have not been recommended as first- line therapy. These agents could be considered as preferred agents in patients with established CVD, where only liraglutide (LEADER trial) and empagloflozin (Empa-Reg outcome trial) and canagliflozin (CANVAS trial) have shown benefit and dapagliflozin (DECLARE-TIMI58) will be reported in 2019. However, SGLT2-inhibitors can be used as monotherapy, or added to any of the other treatment options, as they work independently of insulin and beta-cell function, and have a low risk of hypoglycaemia, unless combined with a SU or insulin. Circumstances where the SGLT2 inhibitors could be a preferred option would include overweight and obese patients where weight gain or hypoglycemia has been, or is likely to be, problematic with other treatment options. Furthermore, it could be considered earlier in the treatment of patients with established CVD. Circumstances where these agents have higher risk of harm would include patients with recurrent UTIs or genital mycotic infections, patients with high risk of hypotension or dehydration, and patients at high risk of stroke, fracture (canagliflozin), amputation (canagliflozin), bladder cancer (dapagliflozin) or ketoacidosis.


SGLT2 inhibitors are a novel therapeutic agent for managing T2DM. These drugs have not yet been registered for use in patients with T1DM. They can be expected to reduce the HbA1c by 0.8%- 1%, induce a 3kg weight loss, slight reduction in blood pressure, and low risk of hypoglycaemia, although this is increased when a SGLT-2 inhibitor is used in combination with sulphonylureas or insulin. These agents could soon be an integral part of our armamentarium for the treatment of patients with T2DM in SA. However, the HbA1c lowering potential is not superior to established antihyperglycemic agents and the macrovascular benefits were seen in studies with patients with established CVD. It is not yet known whether these benefits can be extrapolated to patients without established CVD. However, the added benefits of weight loss and blood pressure reduction, the newly released data on nephroprotection and delay in progression of renal failure and patients needing dialysis, as well as the favourable side-effect profile and low risk of hypoglycaemia, would positively impact our patients. The SGLT2 inhibitors are soon to be the first insulin-independant therapeutic agents available in SA, and given the advantages and favourable side-effect profile, they are likely to become an important component of our management of patients with T2DM.

Three drugs have been accepted by the FDA in the US: Dapagliflozin, canagliflozin and empagliflozin.

Canagliflozin was the first SGLT-2 inhibitor that was approved by the FDA, it was accepted in March 2013. Dapagliflozin and empagliflozin were accepted in 2014. In SA, we are expecting registration of dapagloflozin and empagliflozin within the next few months.

Author: Dr Marius Wasserfall, Physician, Mediclinic Panorama