Specific SGLT-2 blockers were developed from phlorizin,1 a naturally occurring compound found in apple tree bark, which is a potent glycosuric agent acting as a non-selective SGLT inhibitor. Structural modifications improved metabolic activity and stability as well as selectivity for SGLT2 inhibition, which is desirable in order to reduce the gastrointestinal side effects of SGLT-1 inhibition.

The SGLT-2 blocker dapagliflozin (Forxiga) was introduced into the UK in 2012. Canagliflozin (Invokana) was introduced into the USA in 2013, to be launched in UK late 2013; empagliflozin was submitted to the FDA and EMEA in March 2013 hoping to be introduced summer 2014.

Clinical studies of dapagliflozin in type 2 diabetes

An initial dose-ranging placebo controlled study2 was conducted in diet-treated subjects with a HbA1c range of 7-10% and a mean age of 53 years. It showed that 10mg dapagliflozin over 24 weeks produced a placebo subtracted decrease in HbA1c of 0.66% (P<0.0001) with a greater reduction if given in the morning, or if the initial HbA1c was higher.

There were also placebo subtracted decreases of 3.2kg weight (NS), and 2.9mmHg systolic blood pressure. There was no significant increase in hypoglycaemic and hypotensive rates between dapagliflozin and placebo. There was a non-significant increase in urinary tract infections from 4% to 8% on dapagliflozin, and a significant increase in genital infections from 1% on placebo to up to 13% on dapagliflozin. There was a slight rise in haematocrit, and decrease in uric acid, but no significant changes on the blood tests, ie. importantly no decline in eGFR.

Bailey et al3 randomised 546 subjects with type 2 diabetes on metformin 1500mg/day or more, with HbA1c 7-10%, to dapagliflozin (2.5mg, 5mg or 10mg) or placebo. The subjects had renal function acceptable for the metformin use, and importantly had no osmotic symptoms and no heart failure (NYHA III or IV). Subjects had a mean age of 52.7 to 55.0 years. At 24 weeks reductions in HbA1c were significantly greater on dapagliflozin, (-0.84% on 10mg) than on placebo (-0.30%; p<0.0001). Reductions in fasting glucose were apparent within one week, and most of the maximum reduction was achieved within four weeks. There were 2-4% of subjects in each arm who reported symptoms suggestive of hypoglycaemia (all mild). There was greater weight loss on dapagliflozin (eg 2.9kg on 10mg) than on placebo (0.9kg) and dapagliflozin was associated with a reduction in systolic blood pressure of 2-5mmHg compared to placebo reduction of 0.2mmHg.

Symptoms suggestive of urinary tract infection were similar in all four groups (8% on both placebo and 10mg dapagliflozin), but symptoms of genital infection were more common on dapagliflozin (8-13%) than on placebo (5%); most subjects with genital infections had just one episode which resolved with standard treatment and did not lead to discontinuation of dapagliflozin. Hypotension or syncope was rare in all four groups (0-1%), and there were no adverse events associated with dapagliflozin use. Unfortunately, the number of elderly patients in this trial was low, and not stated.

87% of these patients continued in a double blind extension to a total of 102 weeks,4 but the completion rate for the extension period was lower for the placebo group (63.5%) than for the dapagliflozin groups (68% to 80%). At week 102, all dapagliflozin doses had a significant decrease in HbA1c, with a placebo subtracted HbA1c decrease of 0.8% on 10mg dapagliflozin (p<0.0001). All dapagliflozin groups had sustained reductions in body weight (-1.10 to -1.74kg) at 102 weeks (although this is a slight gain during the extension period), whereas increases were noted in placebo-treated patients. Hypoglycaemia was rare and not severe.

The rates of genital infection and urinary tract infection were the same as in the initial study, with two patients on dapagliflozin withdrawing, one due to urinary tract infection and one due to genital infection. Creatinine levels improved slightly on dapagliflozin, and uric acid levels dropped approximately 50 micro-mols per litre. Thus the efficacy and relative safety seems maintained at two years in up to 80% of subjects.

What to add in after initial metformin monotherapy failure was studied by Nauck et al.5 Dapagliflozin or glipizide were added on to metformin in 801 patients poorly controlled on metformin (1500mg per day or more; mean age 59 years). Exclusion criteria included calculated creatinine clearance <60 mL/min, and cardiac failure. Doses were up-titrated at three weekly intervals with maximum possible doses dapagliflozin 10mg and glipizide 20mg. At 52 weeks, HbA1c reduction on dapagliflozin was non-inferior to that of glipizide (-0.52% versus 0.52%). However, hypoglycaemia was significantly more common on glipizide occurring in 40.8% on glipizide and 3.5% on dapagliflozin; weight decreased on dapagliflozin and increased on glipizide resulting in a mean weight difference of 4.65kg. Again, dapagliflozin was associated with a slight decrease in blood pressure.

Dapagliflozin side effects included symptoms of genital infections in women (21% on dapagliflozin versus 5.4% on glipizide, 1-3 events per year), genital infections in men (12.3% on dapagliflozin versus 2.7% on glipizide) and symptoms of urinary tract infections (10.8% versus 6.4%). Interestingly, dizziness was more common on glipizide (9.1%) than dapagliflozin (3.7%), presumably due to hypoglycaemia rather than hypotension. More subjects on dapagliflozin were likely to experience a decline in calculated renal function than on glipizide (4.2% versus 1.7%).

Adding dapagliflozin 10mg daily to a sulphonylurea, glimepiride, in a 24 week placebo-controlled randomised controlled trial led to a placebo subtracted HbA1c reduction of 0.69%, 2.26kg mean weight loss, but with an increased hypo rate from 4.8% on placebo to 7.1% on dapagliflozin.6

In a placebo-controlled randomised controlled trial adding dapagliflozin to pioglitazone, dapagliflozin improved glycaemic control, had the genital infections as anticipated, but also ameliorated the weight gain and peripheral oedema of the pioglitazone.7

Wilding et al8 randomised 808 subjects with type 2 diabetes on insulin plus OHA to placebo or various doses of dapagliflozin (HbA1c 7.5 to 10.5%; mean age 59 years) for 48 weeks. The glucose levels declined rapidly over the first eight weeks on dapagliflozin; at 48 weeks, 10mg dapagliflozin had a placebo subtracted decreases in HbA1c of 0.54%, and weight loss of 2.43kg. 42% of subjects on placebo needed an increase in insulin dose to achieve target glucose levels, whereas only 15% needed insulin up-titration on the dapagliflozin 10mg od. Disappointingly, the hypoglycaemia rate was unchanged by the use of dapagliflozin (any hypo occurred in 52 to 60% of subjects), but one would expect a study driving glucose levels down to a target with insulin to increase the hypoglycaemia rate.

All the above studied type 2 diabetes, but dapagliflozin has been used in type 1 diabetes in a small, short trial.9 Groups of 14 poorly controlled subjects were randomised to different doses of dapagliflozin or placebo for 14 days. Fasting and mean glucose levels, weight and insulin requirements were reduced by 10mg dapagliflozin, without an increase in the hypo rate; these did not reach statistical significance, but this pilot study is promising for the future.


In a 52 week, randomised, double-blind, active-controlled10 trial, patients aged 18-80 (mean age 56) with type 2 diabetes and HbA1c of 7.0-9.5% on stable metformin were randomised to canagliflozin 100mg (n=478) or 300mg, (n=474) or glimepiride (up-titrated to 6mg or 8mg per day) (n=473). All subjects had to have an eGFR >55mL/min/1.73m2. The subjects were well matched with mean HbA1c of 7.8% initially; HbA1c reductions were 0.82% on canagliflozin 100mg, 0.93% on canagliflozin 300mg, and 0.81% on glimepiride. Whereas subjects on glimepiride gained 0.7kg on average, those on canagliflozin lost 3.7 to 4.0kg; systolic blood pressure rose marginally on glimepiride, (0.2mmHg) but fell by 3.3 to 4.6mmHg on canagliflozin. The drugs were relatively well tolerated with 5-7% withdrawing due to side effects.

As anticipated, more patients had documented hypos on glimepiride (34%) than on canagliflozin (5-6%). 5-6% of subjects had a urinary tract infection with no difference between canagliflozin and glimpiride, but genital thrush was much more common on canagliflozin (7 to 14%) than on glimepiride (1-2%); there was more urinary frequency on canagliflozin (3%) than on glimepiride. The slight changes in eGFR over the trial were interesting; eGFR dropped quickly on canagliflozin and then plateaued out with a one year decline of 1.7 to 3.0mL/min/1.73m2, whereas on glimepiride, the eGFR gently declined over the year by 5.1mL/min/1.73m2. There were very similar low frequencies of postural hypotension in each group.


Patients inadequately controlled on metformin and sulfonylurea (HbA1c ≥7 to ≤10%) were randomised to once-daily empagliflozin 10mg (n = 225), empagliflozin 25mg (n = 216), or placebo (n = 225) for 24 weeks.11 There were significant reductions in placebo subtracted HbA1c (0.6 to 0.65%) weight, and systolic blood pressure using empagliflozin; adverse events were similar in all three groups.


These agents may not work with lower degrees of renal function; dapagliflozin should not be used if eGFR less than 60, because of poor glycaemic effect.12

The canagliflozin DIA 3004 trial studied subjects with eGFR 30 to 50, HbA1c reductions were modest eg. 0.33%.14 Usefully, data from many canagliflozin trials has been analysed by renal function.13

Unfortunately decreased renal function is very common in elderly diabetic folk. For instance, in elderly diabetic Australians (93% Caucasian),14 53% had eGFR 30 to 59 mL/min/1.73m2 and 2% had eGFR <30mL/min/1.73m2 .

Co-prescription with loop diuretics is also advised against,12 since the SGLT2 inhibitors might produce an osmotic diuresis. This is a problem since cardiac failure is common with a prevalence of 22% in one study of diabetic subjects aged 65 or over, with a 12% annual incidence rate.15

Studies in elderly

The studies on these agents have included elderly subjects, but in small numbers. Since the renal threshold for glycosuria increases with age,16 one would wonder if these agents do work so well in the elderly. The DIA3010 canagliflozin study17 examined subjects aged 55 to 80 years (38% were aged 65 or over) on any treatment for type 2 diabetes and with HbA1c 7.0 to 10.0% and eGFR> 50. After 26 weeks, canagliflozin 300mg od had a placebo subtracted HbA1c reduction of 0.70% (0.57% on 100mg od of canagliflozin), a weight reduction of 2.7kg, and a systolic blood pressure reduction of 7.9mmHg. The fasting glucose levels reached their lower steady state after six weeks of canagliflozin. Interestingly, in subjects on agents able to cause hypoglycaemia, the severe hypo rate was non-significantly lower with the canagliflozin than placebo, although, overall, any hypo was more common on canagliflozin (not significant). Although urinary frequency and urinary tract infections were slightly more common on canagliflozin (NS), genital mycotic infections were more common on canagliflozin.

Data from the canagliflozin programme has been analysed by age;13 there is a difference in the PP analysis where there is a high dropout rate on placebo ie. only those doing acceptably on placebo carry on, compared to the LOCF reporting.

But it would appear that this class of drugs is not particularly effective in subjects aged over 75 years, and this may partly be explained by their poorer renal function.13,17

Indeed the manufacturers of dapagliflozin state that "Due to the limited therapeutic experience in patients 75 years and older, initiation of dapagliflozin therapy is not recommended."13

NICE guidance

The NICE advice on dapagliflozin18 is that it is to be used as add on to metformin or to be used with insulin in type 2 diabetes, with the same criteria for efficacy and continuation of dapagliflozin, as the DPP-4 inhibitors,19 ie. a 0.5% reduction in HbA1c at six months.

Use of SGLT-2 inhibitors in the elderly

Use is constrained firstly by the need for adequate renal function for the drug to work eg. an eGFR >50-60mL/min/1.73m2 and the drugs may not work well in subjects over 75 years old (whose renal function may well be worse than that estimated by the standard MDRD formula.20 Secondly for safety, one needs to be cautious about initiating an SGLT-2 inhibitor in subjects either on loop diuretics or already prone to hypotension, and to monitor them carefully during intercurrent illnesses for fluid depletion and hypotension, possibly withholding the agent temporarily. As with many new drugs, they are slightly more expensive than recent branded oral agents, and considerably more expensive than established generic oral agents. But using an SGLT-2 inhibitor would result in modest but significant reductions in weight, blood pressure, and uric acid levels, whereas insulin, sulphonylureas and pioglitazone result in a gain in weight, and insulin therapy with its associated costs is not cheap.

Hence, a small proportion of older diabetic folk will be suitable for an SGLT-2 inhibitor, but if one is suitable for an SGLT-2 inhibitor, one will also be suitable for metformin, a DPP-4 inhibitor, and a GLP-1 injection, although side effects (predominantly gastro-intestinal) might limit use of other agents.

It is possible, that one would follow NICE and use it as add-on to diet or metformin, but other options such as a gliptin would also be possible and marginally cheaper and a sulphonylurea would be considerably cheaper but also more risky due to possible hypoglycaemias.

But for someone poorly controlled on insulin, particularly the brittle diabetic person, it is a very attractive option. Some insulin treated subjects experience wild fluctuations in the blood glucose levels between symptomatic hyperglycaemia and symptomatic hypoglycaemia despite our best efforts. It should be feasible to decrease the insulin dose to decrease the risk of hypos, and add in an SGLT-2 inhibitor to remove the symptomatic hyperglycaemia. Unfortunately, we are not aware of any clinical trials assessing the brittle diabetic person, and in the type 2 diabetes add on to insulin study,8 dapagliflozin non-significantly increased the hypo rate. This use would fit within license and NICE for type 2 diabetes; it should work in type 1 diabetes and diabetes secondary to pancreatic disease, but this would be against current license and NICE.

The SGLT-2 inhibitor could also be used to avoid injectable treatments for patients on maximum tolerated other oral hypoglycaemic agents, with inadequate control. SGLT2 use as part of triple therapy ie. added to two other OHAs21,22 had a placebo subtracted decrease in HbA1c of 0.4 to 0.6%. Add-on to triple therapy has not been studied, and is not approved by NICE; the cost of an SGLT-2 inhibitor and a DPP-4 inhibitor would be similar to the cost of a GLP-1 agent such as daily liraglutide or weekly exenatide (ignoring the costs of administering these injections).

One would warn the subject about the risk of genital thrush, and supply a tube of canesten cream.

The efficacy in subjects aged over 75 is uncertain; thus in clinical practice if one used an SGLT-2 inhibitor in those aged 75 years plus, one would need to monitor for loss of glycaemic control over four to six weeks.3,7,17 and review at six months (as per NICE guidance) that there had been apparent benefit. There is an urgent need for RCT data in subjects aged 75 plus in terms of glycaemic benefit and safety including specific elderly endpoints such as falls, frailty and using more accurate formulae to predict renal function.

Overall, this is an interesting class of drugs with the added benefit of weight and blood pressure reduction, but due to the requirement for reasonable renal function and absence of heart failure treatment, its use may be limited in older subjects.


Conflict of interest: Simon Croxson has received funding for meetings or speakers/consultancy fees from all companies involved in diabetes treatment.


1. Nair S, Wilding JP. Sodium glucose cotransporter 2 inhibitors as a new treatment for diabetes mellitus. J Clin Endocrinol Metab 2010 Jan; 95(1): 34-42

2. Ferrannini E, Ramos S.J, Salsali A, et al. Dapagliflozin monotherapy in type 2 diabetic patients with inadequate glycemic control by diet and exercise: a randomized, double-blind, placebo-controlled, phase III trial. Diabetes Care 2010; 33(10): 2217-24

3. Bailey C, Gross J, Pieters A, Bastien A, List J. Effect of dapagliflozin in patients with type 2 diabetes who have inadequate glycaemic control with metformin: a randomised, double-blind, placebo-controlled trial. Lancet 2010; 375(9733): 2223-33

4. Bailey CJ, Gross JL, Hennicken D, et al. Dapagliflozin add-on to metformin in type 2 diabetes inadequately controlled with metformin:a randomized, double-blind, placebo-controlled 102-week trial BMC Medicine 2013, 11:43 Accessed from:-http://www.biomedcentral.com/content/pdf/1741-7015-11-43.pdf

5. Nauck M, Del Prato S, Meier J, et al. Dapagliflozin versus glipizide as add-on therapy in patients with type 2 diabetes who have inadequate glycemic control with metformin: a randomized, 52-week, double-blind, active-controlled non-inferiority trial. Diabetes Care 2011; 34(9): 2015-2022

6. Strojek K, Yoon K, Hruba V, et al. Effect of dapagliflozin in patients with type 2 diabetes who have inadequate glycaemic control with glimepiride: arandomised, 24-week, doubleblind,placebo-controlled trial. Diabetes Obes Metab 2011; 13: 928-938.

7. Rosenstock J, Vico M, Wei L, et al: Effects of dapagliflozin, an SGLT2 inhibitor, on HbA(1c), body weight, and hypoglycemia risk in patients with type 2 diabetes inadequately controlled on pioglitazone monotherapy. Diabetes Care 2012; 35:1473-78

8. Wilding JP, Woo V, Soler NG, Pahor A, Sugg J, Rohwedder K, Parikh S; Dapagliflozin 006 Study Group. Long-term efficacy of dapagliflozin in patients with type 2 diabetes mellitus receiving high doses of insulin: a randomized trial. Ann Intern Med 2012; 156(6): 405-15

9. Henry RH, Rosenstock J, Chalamandaris A-G, et al. Exploring The potential of Dapagliflozin in Type 1 diabetes; phase 2a pilot study. Abstract ADA 73rd meeting, 2013

10. Cefalu WT, Leiter LA, Yoon KH, et al. Efficacy and safety of canagliflozin versus glimepiride in patients with type 2 diabetes inadequately controlled with metformin (CANTATA-SU): 52 week results from a randomised, double-blind, phase 3 non-inferiority trial. Lancet. 2013 Jul 11. pii: S0140-6736(13)60683-2. doi: 10.1016/S0140-6736(13)60683-2. [Epub ahead of print]

11. Häring HU, Merker L, Seewaldt-Becker E, et al; on behalf of the EMPA-REG METSU™ Trial Investigators. Empagliflozin As Add-on to Metformin Plus Sulfonylurea in Patients With Type 2 Diabetes: A 24-week, randomized, double-blind, placebo-controlled trial. Diabetes Care 2013 Aug 20. [Epub ahead of print]

12: Bristol Myers Squibb. Summary of Product characterisitcs; Forziga http://ec.europa.eu/health/documents/community-register/2012/20121112124487/anx_124487_en.pdf

13. Parks, MH. Division Director Summary Review, application number 204042Orig1a000. Center for Drug Evaluation and Research, 2012. Accessed from http://www.accessdata.fda.gov/drugsatfda_docs/nda/2013/204042Orig1s000PharmR.pdf

14. Chadban SJ, Briganti EM, Kerr PG, et al. Prevalence of Kidney Damage in Australian Adults: The AusDiab Kidney Study. JASN 2003; 14; suppl 2; S131-S138

15. Bertoni AG, Hundley WG, Massing MW, et al. Heart failure prevalence, incidence, and mortality in the elderly with diabetes. Diabetes Care 2004; 27(3): 699-703

16. Butterfield WJH, Keen H, Whichelow MJ. Renal glucose threshold variations with age. BMJ 1967; 4: 505-7

17. Bode B, Stenlöf K, Sullivan D, Fung A, Usiskin K.Efficacy and safety of canagliflozin treatment in older subjects with type 2 diabetes mellitus: a randomized trial. Hosp Pract  2013; 41 (2): 72-84

18. National Institute for Clinical Excellence. NICE technology appraisal guidance 288; Dapagliflozin in combination therapy for treating type 2 diabetes. June 2013 http://www.nice.org.uk/nicemedia/live/14193/64270/64270.pdf

19. National Institute for Clinical Excellence. NICE short clinical guideline 87; Type 2 diabetes: newer agents. May 2009 http://www.nice.org.uk/nicemedia/live/12165/44318/44318.pdf

20. Schaeffner ES, Ebert N, Delanaye P, Frei U, Gaedeke J, Jakob O, Kuhlmann MK, Schuchardt M, Tölle M, Ziebig R, van der Giet M, Martus P. Two novel equations to estimate kidney function in persons aged 70 years or older. Ann Intern Med 2012; 157(7): 471-81

21. Jabbour S, Hardy E, de Bruin TW, Martin P, Gause-Nilsson I, Rohwedder K, Parikh SJ Dapagliflozin as part of triple combination therapy helps reduce HbA1c and weight in patients with type 2 diabetes Abstract ADA 73rd meeting, 2013

22. Clar C, Gill JA, Waugh N. Systematic review of SGLT2 receptor inhibitors in dual or triple therapy in type 2 diabetes BMJ Open 2012;2:e001007 doi:10.1136/bmjopen-2012-001007. Accessed from:- http://bmjopen.bmj.com/content/2/5/