Sodium–Glucose Cotransporter Inhibitors Diabetes Care Volume 38, December 2015

2344 Diabetes Care Volume 38, December 2015

Sodium–Glucose Cotransporter Sunder Mudaliar,1,2 David Polidori,3 Brian Zambrowicz,4 and Robert R. Henry1,2 Inhibitors: Effects on Renal and Intestinal Glucose Transport From Bench to Bedside

Diabetes Care 2015;38:2344–2353 | DOI: 10.2337/dc15-0642

Type 2 diabetes is a chronic disease with disabling micro- and macrovascular complications that lead to excessive morbidity and premature mortality. It affects hundreds of millions of people and imposes an undue economic burden on populations across the world. Although insulin resistance and insulin secretory defects play a major role in the pathogenesis of hyperglycemia, several other metabolic defects contribute to the initiation/worsening of the diabetic state. Prominent among these is increased renal glucose reabsorption, which is maladaptive in patients with diabetes. Instead of an increase in renal glucose excretion, which could ameliorate hyperglycemia, there is an increase in renal glucose reabsorption, which helps sustain hyperglycemia in patients with diabetes. The sodium–glucose cotransporter (SGLT) 2 inhibitors are novel antidia-

REVIEW betes agents that inhibit renal glucose reabsorption and promote glucosuria, thereby leading to reductions in plasma glucose concentrations. In this article, we review the long journey from the discovery of the glucosuric agent phlorizin in the bark of the apple tree through the animal and human studies that led to the development of the current generation of SGLT2 inhibitors.

It took nearly 200 years from the isolation of phlorizin, a chemical found in apple tree bark that inhibits sodium–glucose cotransporters (SGLTs) (1), to the approval of the first medications inhibiting SGLTs for treatment of type 2 diabetes (T2D). During this time, several SGLTs were discovered and the roles of SGLT1 and SGLT2 in intestinal and renal glucose reabsorption have been elucidated in studies in genet- 1Veterans Affairs Medical Center, San Diego, CA ically manipulated rodents, humans with SGLT gene mutations, healthy humans, 2 and humans with diabetes (Fig. 1). This review provides an overview of the basic and School of Medicine, University of California, San Diego, San Diego, CA clinical research that led to the translation of the initial findings of increased glu- 3JanssenResearch&Development,LLC,San cosuria with phlorizin to the development and approval of SGLT inhibitors and a Diego, CA summary of the clinical trial results obtained to date. 4Regeneron Pharmaceuticals, Inc., Tarrytown, NY Corresponding author: Sunder Mudaliar, Identification, Distribution, and In Vitro Characterization of the SGLT Inhibitors [email protected]. In the 1980s and 1990s, Wright and colleagues cloned SGLT1 (2) and SGLT2 (2,3) and Received 3 April 2015 and accepted 4 September did much of the in vitro characterization, demonstrating that SGLT1 has a higher 2015. This article contains Supplementary Data online affinity for glucose than SGLT2 (Km for glucose ;0.4 mmol/L and 2 mmol/L, respectively), whereas SGLT2 has a higher capacity (4). SGLT1 is expressed at high at http://care.diabetesjournals.org/lookup/ suppl/doi:10.2337/dc15-0642/-/DC1. levels in the intestine and is also expressed in the kidney, heart, and skeletal muscle, © 2015 by the American Diabetes Association. whereas SGLT2 is expressed almost exclusively in the kidney (4). Renal SGLT2 ex- Readers may use this article as long as the work is pression is increased in hyperglycemic rodents (5,6) and in humans with T2D (7). properly cited, the use is educational and not for Intestinal SGLT1 expression is regulated by diet and other factors (8) and is profit, and the work is not altered. care.diabetesjournals.org Mudaliar and Associates 2345

Figure 1—Time line of some key developments in the understanding of SGLT-mediated glucose uptake and the development of SGLT inhibitors. increased in subjects with T2D (9). Fur- (12). Wright and colleagues later dem- SGLT1 mutations only mildly increase UGE ther details on the structure and func- onstrated that these two systems were (17). Similarly, only minimal UGE is ob- tion of the transporters can be found in accounted for by SGLT2 and SGLT1, served in SGLT1 knockout (KO) mice, ref. 4. respectively (4). whereas high UGE is seen in SGLT2 KO The effects of SGLT2 inhibitors on re- mice (17,18). Role of SGLT2 and SGLT1 in Renal nal glucose kinetics were assessed using While it is often stated that SGLT2 Glucose Reabsorption controlled glucose infusion experiments accounts for 90% of glucose reabsorp- In the 1930s, Shannon and Fisher eluci- in rats (13) and humans (14,15). These tion, SGLT2 and SGLT1 appear sequen- dated the renal glucose reabsorption ki- experiments showed that SGLT2 inhibi- tially in the proximal tubule, so it is an netics in dogs (10). Their work showed tion leads to a reduction in TMGandRTG oversimplification to provide a single that 1) there is a maximum capacity for while maintaining a threshold-like rela- value to describe their relative contribu- renal tubular glucose transport (the tu- tionship between PG and the UGE rate tion to glucose reabsorption. For exam- bular maximum glucose reabsorption (Fig. 2). Importantly, in subjects with ple, under normoglycemic conditions, rate) (TmG), 2) nearly all filtered glucose T2D treated with SGLT2 inhibitors, the there is sufficient SGLT2 capacity to re- is reabsorbed when plasma glucose (PG) UGErateishighwhenPGishighbut absorb virtually all filtered glucose, and concentrations remain below a thresh- diminishes as PG approaches hypogly- only minimal UGE is observed in humans old value called the renal threshold for cemic levels, suggesting a low risk of or rodents lacking SGLT1. However, glucose (RTG), and 3) urinary glucose ex- treatment-induced hypoglycemia. The maximally effective SGLT2 inhibitor cretion (UGE) increases nearly linearly SGLT2 inhibitor–induced increases in doses typically prevent ;50% of the fil- with PG when PG is above RTG. UGE are sustained at similar levels with tered glucose from being reabsorbed, In the early 1970s, Vick, Diedrich, and long-term treatment (16). suggesting that SGLT1 has considerably Baumann demonstrated that glucose re- The roles of SGLT2 and SGLT1 in renal greater capacity for glucose reabsorp- absorption occurred in the proximal tu- glucose reabsorption were further con- tion than expected based on the com- bule (11), and Turner and Moran later firmed through human and rodent ge- monly quoted 90% value for SGLT2. This demonstrated that this occurs through netic studies. In humans, familial renal is supported by data from genetic mod- two distinct sodium-dependent glucose glucosuria is a rare, benign condition els, since SGLT2 KO mice exhibit only transport systems, one with relatively arising from SGLT2 mutations that reduce about 30% of the UGE observed in low affinity and high capacity and one renal glucose reabsorption and lead to SGLT1/SGLT2 double KO mice (17). with higher affinity and lower capacity UGE ranging from 1 to 170 g/day, whereas These data, combined with human 2346 Sodium–Glucose Cotransporter Inhibitors Diabetes Care Volume 38, December 2015

absorption, metabolism to phloretin, which inhibits GLUTs, and potential intestinal malabsorption with SGLT1 inhibition. This led to the pursuit of selective SGLT2 inhibitors with improved properties. The first publication demonstrating the potential of a selective SGLT2 inhibitor (T-1095) as a treatment for diabetes in rodent models appeared in 1999 (26). Since then, several SGLT2 inhibitors have been developed and three compounds are currently approved for use in the U.S. and Europe (dapagliflozin, canagliflozin, and empagliflozin) (Table 1). Additionally, luseogliflozin, topogliflozin, and ipragliflozin have been approved in Japan and other Figure 2—Relationship between blood glucose concentrations and UGE rates in rodents and compounds are in late-stage clinical tri- fl 6 humans treated with the SGLT2 inhibitor canagli ozin. Left panel: Results (mean SE) from a als. While these compounds have been graded glucose infusion study in Zucker diabetic fatty (ZDF) rats (13). Right panel: Results (mean 6 SD) from a stepwise hyperglycemic clamp study in human subjects with T2D (15). In primarily designed to be highly selec- both studies, canagliflozin treatment produces a left shift in the relationship between blood tive for SGLT2 compared with SGLT1, glucose and UGE with no apparent change in the slope, leading to a reduction in RTG. there is variability in the selectivity, most notably for sotagliflozin, which is only 20-fold selective for SGLT2 com- data showing that phlorizin can block meal (20), indicating that SGLT1 may pared with SGLT1. A selective SGLT1 in- virtually all renal glucose reabsorption be required for the early GLP-1 response hibitor was tested in a phase 1 study in (19), suggest that dual SGLT1/2 renal and that there is a second more pre- 12 subjects and showed that SGLT1 in- inhibitors might achieve consider- dominant phase of GLP-1 release that hibitors block intestinal glucose ab- ably greater UGE than selective SGLT2 does not require SGLT1 and is enhanced sorption, reduce GIP secretion, and inhibitors. in the absence of SGLT1. The increased enhance GLP-1 and peptide YY (PYY) GLP-1 seen in SGLT1 KO mice may be secretion (27). Role of SGLT1 in Intestinal Glucose due to increased glucose reaching the Absorption distal small intestine and colon where SGLT1 is essential for intestinal glucose/ Effects of SGLT2 Inhibitors on it, or its metabolites, can trigger GLP-1 galactose absorption and represents Glycemia release (21,22). SGLT1 heterozygous SGLT2 inhibitors are effective in lower- the primary mechanism of glucose/ mice thrived normally on a regular diet ing PG when used as monotherapy or in galactose uptake from the lumen into but also exhibited elevated glucose in combination with other oral agents/ enterocytes (4). The essential nature of the distal small intestine and cecum insulin. These effects have been demon- SGLT1 is confirmed by the rare genetic and elevated postmeal GLP-1 levels strated in large multicenter, multina- disease glucose-galactose malabsorp- (17), indicating that partial SGLT1 inhi- tional, placebo- and active-controlled tion (GGM), which arises from missense bition might provide benefits without GI studies. Due to limited space, efficacy mutations in SGLT1 (19). This condition intolerability observed in the absence of studies summarized are limited to those causes severe diarrhea if glucose or ga- SGLT1 activity. at least 24 weeks long with compounds lactose is consumed and can be fatal un- approved for use in the European Union less glucose and galactose are removed Development of Pharmaceutical SGLT (EU) and U.S. from the diet. SGLT1 KO mice develop a Inhibitors as Treatment for T2D similar glucose-galactose malabsorp- Although phlorizin was known to in- Monotherapy tion syndrome when fed glucose but crease UGE and was demonstrated to Although metformin is the first choice thrive normally when fed a glucose- completely inhibit renal glucose reab- pharmaceutical treatment for T2D, it and galactose-free diet (18). sorption in the 1930s (23), the potential causes intolerable gastrointestinal (GI) Studies in KO animals have further for using increased UGE as a means to side effects in occasional patients. In elucidated the role of intestinal SGLT1. regulate PG was not demonstrated until such patients, SGLT2 inhibitors can be SGLT1 KO mice exhibited elevated glu- the 1980s. Experiments by Rossetti and used as monotherapy, and in clinical tri- cose in the distal small intestine and co- colleagues in diabetic rats demon- als, compared with placebo/active com- lon and decreased cecal pH when strated that sustained phlorizin treat- parator, they lowered fasting PG (FPG) challenged with a meal containing glu- ment normalized blood glucose by 20–46 mg/dL and HbA1c by 0.54– cose (17,20). Reductions in serum total concentrations resulting in reduced 1.45% in patients with baseline HbA1c glucagon-like peptide 1 (GLP-1) have “glucotoxicity” and improvements in 7.9–9.1%. Those with higher baseline been reported 5 min after a meal chal- b-cell function and insulin sensitivity HbA1c had greater glycemic benefits, lenge (17,20), but serum GLP-1 was in- (24,25). However, phlorizin was not a as did those on higher doses of SGLT2 creased from 30 min to 6 h after the suitable therapeutic agent due to poor inhibitors (Supplementary Table 1). care.diabetesjournals.org Mudaliar and Associates 2347

Table 1—Summary of the most advanced SGLT2 inhibitor compounds (refs. 28–31)

SGLT2 IC50 SGLT1 IC50 SGLT2/SGLT1 Highest approved Compound (nmol/L) (nmol/L) selectivity dose (mg)a Status Canagliflozin 4.2 663 160 300 Approved in U.S., EU, Japan, other countries Dapagliflozin 1.2 1,400 1,200 10 Approved in U.S., EU, Japan, other countries Empagliflozin 3.1 8,300 2,700 25 Approved in U.S., EU Ipragliflozin 5.3 3,000 570 50 Approved in Japan Luseogliflozin 2.3 3,990 1,770 5 Approved in Japan Tofogliflozin 6.4 12,000 1,875 20 Approved in Japan Ertugliflozin 0.9 1,960 2,200 25 Phase 3 LX-4211 (sotagliflozin) 1.8 36 20 400 Completed phase 2 aHighest approved dose or the highest dose still in development for compounds that have not yet been approved.

Dual Oral Combination Therapy inhibitors/pioglitazone has a low poten- albeit with a potential for diarrhea, SGLT2 Inhibitor and Metformin Combination. tial for hypoglycemia (Supplementary bloating, and GI discomfort due to intes- Over time, due in part to disease pro- Table 3). tinal glucose/galactose malabsorption. gression, there is worsening glycemia in Both sotagliflozin and canagliflozin have Triple Oral Combination metformin-treated patients. These pa- been associated with some intestinal The SGLT2 inhibitors are also effective in tients are often treated with the addition SGLT1 inhibition, but neither is believed improving glycemia in triple combina- of sulfonylureas or dipeptidyl peptidase-4 to have any meaningful renal SGLT1 in- tion with metformin and either sulfonyl- (DPP-4) inhibitors. In metformin-treated hibition. Hence, it is not currently known ureas, DPP-4 inhibitors, or glitazones, patients, adding an SGLT2 inhibitor results what the efficacy and safety profile with placebo-subtracted FPG lowered in additional glycemic benefit with low po- would be for a dual inhibitor that also by 20–38 mg/dL and HbA lowered by tential for hypoglycemia (unlike sulfonyl- 1c provides renal SGLT1 inhibition. 0.4–1.03% in patients with mean HbA ureas) and modest reductions in weight 1c In clinical studies, sotagliflozin has 7.8–8.1%. Of note, higher doses of and blood pressure (not seen with DPP-4 been shown to improve glycemia with- SGLT2 inhibitors resulted in greater inhibitors or sulfonylureas). The addition of out any clinically significant increase in HbA lowering and canagliflozin SGLT2 inhibitors as add-on to metformin 1c GI side effects. As monotherapy in T2D, 300 mg had superior HbA lowering results in FPG lowering by 15–40 mg/dL 1c sotagliflozin 150/300 mg once daily for compared with sitagliptin 100 mg in and HbA1c by 0.54–0.77% compared with 28 days lowered placebo-subtracted patients on metformin and sulfonyl- placebo in patients with mean baseline FPG by 39 and 55 mg/dL and HbA by urea combination (Supplementary 1c HbA1c between 7.9 and 8.2%. Compared 0.66 and 0.76%, respectively, in patients Table 4). with sulfonylureas, placebo-subtracted with baseline HbA1c 8.1% (31). Of note, FPG is lowered by 20–27 mg/dL and SGLT2 Inhibitor and Insulin Combination in another study, sotagliflozin 400 mg HbA1c by 0.52–0.93% in patients with Most patients with T2D eventually re- QD in combination with metformin baseline HbA1c 7.7–7.9%. Compared with quire exogenous insulin therapy to resulted in greater reductions in FPG DPP-4 inhibitors, placebo-subtracted FPG achieve and maintain glycemic goals. and HbA1c compared with sotagliflozin is lowered by 27–36 mg/dL and HbA1c by The addition of insulin is associated 200 mg QD despite similar amounts of 0.73–0.88% in patients with baseline with weight gain and increased hypogly- UGE (32). This suggests that part of the HbA1c 7.9–8%. Of note, while dapagliflo- cemia risk. The addition of SGLT2 inhib- efficacy with 400 QD is through SGLT1 zin 10 mg, empagliflozin 25 mg, and can- itors in patients inadequately controlled inhibition in the GI tract. fl agli ozin 100 mg were associated with with insulin and mean HbA1c 8.3–8.5% is The 300-mg dose of canagliflozin also equivalent glycemic control, canagliflo- associated with improved glycemic has transient intestinal SGLT1 inhibition zin 300 mg achieved superior glycemic control with placebo-subtracted FPG and reduces postprandial PG excur- control compared with glimepiride and lowered by 6–63 mg/dL and HbA1c low- sions in the meal after dosing in both sitagliptin (Supplementary Table 2). ered by 0.39–1.27% in the setting healthy subjects and subjects with SGLT2 Inhibitors and Nonmetformin Oral of modest weight loss (1.31–3.5 kg) T2D (14,33,34) through a non-UGE- Combination. In patients treated with andlowerinsulinrequirements(9–19 associated mechanism. This additional nonmetformin oral agents (sulfonyl- units), without increasing major hy- effect on postprandial glucose is postu- ureas, DPP-4 inhibitors, or glitazones), poglycemic episodes (Supplementary latedtobeduetotransientproximal the addition of SGLT2 inhibitors results Table 5). intestinal SGLT1 inhibition. in improved glycemia with placebo-sub- tractedFPGloweredby10–40 mg/dL Effects of Dual SGLT1/2 Inhibitors on Effects of SGLT2 Inhibitors in Type 1 and HbA1c lowered by 0.4–0.9% in pa- Glycemia Diabetes tients with baseline HbA1c 8.0–8.4%. Combined renal SGLT2 and intestinal Currently, SGLT2 inhibitors are not ap- Higher doses of SGLT2 inhibitors re- SGLT1 inhibition have the potential to proved in type 1 diabetes (T1D). How- sulted in greater HbA1c lowering, and increase renal glucosuria and delay/ ever, given their insulin-independent combining SGLT2 inhibitors with DPP-4 reduce dietary glucose absorption, mechanism of action, there is potential 2348 Sodium–Glucose Cotransporter Inhibitors Diabetes Care Volume 38, December 2015

to use these agents in T1D. In small pilot always reach statistical significance. steady-state weight loss appears to studies, use of SGLT inhibitors in addi- Potential mechanism(s) leading to im- be due to fat loss. In studies with DEXA tion to insulin increased UGE and mod- proved insulin sensitivity include amelio- measurement of body composition, estly improved glycemia and body ration of glucotoxicity and body weight ;70% of weight loss was attributed to weight with lower insulin doses, less glu- reduction. fat and numerically greater reductions cose variability, and no increase in hypo- Endogenous Glucose Production occurred in visceral compared with sub- – glycemia (35 37). Longer-term studies Treatment with SGLT2 inhibitors in- cutaneous adipose tissue (48,49). In- fi fl are in progress to characterize the ef - creases endogenous glucose production terestingly, treatment with empagli ozin fl cacy and importantly the safety of these (EGP). EGP increased after a single dose or dapagli ozin also shifted substrate agents, especially regarding potential de- of canagliflozin in healthy subjects (14) utilization from carbohydrate to lipid velopment of diabetic ketoacidosis (DKA). and increased by ;17–25% after single metabolism (38,39). Metabolic, Renal, Cardiovascular, and and multiple doses of empagliflozin and Renovascular Effects of SGLT GI Effects of SGLT Inhibitors dapagliflozin in patients with T2D Inhibitors In addition to lowering PG by increasing (38,41). Despite this increase in EGP, In addition to increasing UGE, inhibiting UGE, SGLT2 inhibitor treatment is asso- SGLT2 inhibitors still lower fasting and SGLT2-mediated renal glucose and so- ciated with additional metabolic, reno- postprandial glucose and improve dium reabsorption leads to changes in vascular, GI, and cardiovascular effects. glycemia in T2D patients. Although an fluid balance, blood pressure, and renal increase in EGP appears paradoxical, it Insulin Secretion function. is possible that this is a physiological re- Consistent with the improvements in Urine Volume b-cell function observed in rats treated sponse to counter the acute glucosuric Although the magnitude of UGE is gen- with phlorizin, improvements in mea- effect of SGLT2 inhibition. Notably, in the erally sustained with continued treat- sures of b-cell function have been ob- above studies, increases in plasma gluca- ment, the increases in urine volume served in patients with T2D treated gon and decreases in plasma insulin appear to be largely attenuated after with SGLT2 inhibitors. Improvements were observed, leading to an increased multiple dosing. No significant changes in model-based measures of b-cell glu- glucagon-to-insulin ratio that may be re- in urine volume were noted after 2 or cose function obtained from mixed- sponsible for the observed EGP increase. 12 weeks of treatment with canagliflozin meal tolerance tests were observed in Recent work documenting expression of (44) or after 4 weeks of treatment with a subjects treated with empagliflozin SGLT1/2 in human pancreatic -cells sug- empagliflozin (50) in phase 1 studies, (38) and canagliflozin (39). These im- gests that SGLT2 inhibitors may act di- and only modest increases in mean daily provements were observed within the rectly on these cells to increase glucagon urine volume (;100–500 mL/day) were first day of treatment (38) and with secretion (42). Preliminary studies indi- reported with SGLT2 inhibitors in phase treatment of 6–12 months (39). With cate that combining an SGLT2 inhibitor 3 studies (45–47). use of the frequently sampled intrave- with a DPP-4 inhibitor blunts the glucagon increase seen with SGLT2 inhibitor Plasma Volume nous glucose tolerance test method, nu- Given the mechanism of action of SGLT2 merical improvements were observed in monotherapy and further improves glycemic control (43). However, it is not inhibitors to produce osmotic diuresis, the acute insulin response to glucose in it is expected that there would be changes subjects treated with dapagliflozin for known whether the combination of SGLT2 inhibitors with GLP-1 receptor in plasma volume. In a 12-week study 3 months, although the increase relative fl agonists blunts the increase in EGP. with dapagli ozin, plasma volume was to placebo did not reach statistical sig- measured in a subset of subjects using fi 125 ni cance (P = 0.06). Longer studies are Body Weight/Body Composition I-labeled human serum albumin. After needed to assess whether SGLT2 inhib- Increased UGE with SGLT2 inhibition re- 12 weeks’ treatment, median plasma vol- itors slow the progressive decline in sults in caloric loss and osmotic diuresis ume decreased by ;7% with dapagliflozin b -cell function that occurs in diabetes. leading to transient fluid loss that appears compared with an increase of 5% with Peripheral Insulin Sensitivity largely attenuated with sustained treat- placebo. However, these results were Improvements in peripheral insulin sen- ment (44). Both processes can lead to based on a small sample size of 8–10 sitivity have also been observed in pa- weight loss, particularly during the early subjects/group (51). In a 12-week study tients treated with SGLT2 inhibitors. In treatment period. In patients treated with canagliflozin in patients with T2D, two hyperinsulinemic-euglycemic clamp with SGLT2 inhibitors, a progressive reduc- plasma volume measured using indocyanine studies ranging from 2 weeks to 3 months tion in body weight is typically observed green dilution decreased ;10% com- after treatment with dapagliflozin, in- over the first 12–26 weeks, followed by pared with placebo after 1 week of creases in glucose disposal rate of maintenance of the reduced body weight treatment, and this effect was largely ;15–20% occurred relative to placebo with minimal further reduction after 26 attenuated with sustained treatment (40,41). With use of mixed-meal toler- weeks. In the phase 3 clinical studies, (44). In phase 3 clinical studies, volume- ance test–based measurements of insu- SGLT2 inhibition typically provided mean related adverse events were generally lin sensitivity in patients treated with placebo-subtracted weight loss of ;2–5% higher in the SGLT2 inhibitor groups, par- empagliflozin and canagliflozin, numeri- (;1.5–6kg)(45–47). ticularly in elderly subjects, those with cal increases in insulin sensitivity mea- While fluid loss may contribute to low estimated glomerular filtration rate sures were observed in multiple studies the initial weight loss with SGLT2 in- (eGFR), or those on diuretics (especially (38,39), although the changes did not hibitor treatment, the majority of the loop diuretics). care.diabetesjournals.org Mudaliar and Associates 2349

Glomerular Filtration Rate an HbA1c decrease of 0.32%). Of note, in Studies suggest that in hypertensive sub- Since SGLT2 inhibitors cause osmotic di- the empagliflozin study, significant low- jects with T2D with normo- or microal- uresis and small reductions in plasma ering was only seen in patients with buminuria, persistent hyperfiltration is volume and blood pressure, it is im- CKD2 and CKD3 but not in those with an independent risk factor for accelera- portant to document their effects on CKD4. Current clinical guidelines for ted renal function loss and develop- renal function (Table 2). In the study SGLT2 inhibitor use in renal impairment ment or progression of nephropathy, with dapagliflozin in patients with mod- are listed in Table 3. whereas amelioration of hyperfiltration erate renal impairment (52), mean Hemoglobin and Hematocrit is renoprotective (56). In a recent study eGFR and creatinine clearance fell by Small increases in hemoglobin and he- in patients with T1D and no macroalbu- ; – 2 fl 3 5 mL/min/1.73 m after 1 week of matocrit are consistently seen in phase minuria, treatment with empagli ozin treatment but stabilized thereafter 3 studies with SGLT2 inhibitors. While for 8 weeks improved HbA1c by 0.5% through 104 weeks of therapy, whereas these increases are consistent with with lower insulin requirements and these parameters slowly declined in the fi small reductions in fluid volume, small was associated with a signi cant atten- placebo group. Similar changes have fi increases in reticulocytes, erythropoie- uation of renal hyper ltration (55). The been seen with canagliflozin where the tin, and red cell mass were reported in a authors concluded that although sev- reductions in eGFR were largest at week 12-week study with dapagliflozin (51), eral factors may have contributed to 3 (the first postbaseline measurement) fi suggesting that changes in hematopoie- the decrease in glomerular ltration and trended back toward baseline over sis may contribute to changes in hemo- rate to near-normal levels, they postu- the 26-week treatment period (53). Sim- globin and hematocrit. lated that activation of tubulo-glomerular ilarly, with empagliflozin treatment in feedback by empagliflozin made a sub- patients with stage 2, 3, or 4 chronic Electrolytes/Uric Acid stantial contribution. They speculated kidney disease (CKD), initial small de- In clinical studies, changes in mean se- that long-term SGLT2 inhibitor use could creasesineGFRreturnedtobaseline rum electrolytes were infrequent. With be renoprotective by reducing intraglo- fl by the end of the 3-week follow-up after dapagli ozin, there were no changes merular pressure, thereby reducing the treatment completion at 52 weeks (54). from baseline levels of mean serum so- risk of developing overt diabetic ne- The initial eGFR reduction with SGLT2 dium, potassium, bicarbonate, calcium, phropathy. Existing data from phase 3 inhibition may be related not only to or chloride at week 24 and up to studies in patients with T2D and CKD antihypertensive and diuretic effects 102 weeks. There were small increases show modest improvements in albumin- but also to increased tubulo-glomerular in mean serum inorganic phosphorus lev- uria progression with SGLT2 inhibitor feedback (55). els from baseline (46). SGLT2 inhibitor use treatment compared with placebo is associated with decreases in serum uric (52–54). Long-term studies are being Effects on Glycemia in Patients With CKD acid (56). Hyperuricemia is known to be With decreasing eGFR, there is lower conducted to determine whether the associated with an increased risk of gout, glycemic efficacy of SGLT2 inhibitors SGLT2 inhibitors retard/prevent the de- kidney stones, and cardiovascular disease. due to a lesser filtered load of glucose. velopment and progression of diabetic Whether lowering uric acid has beneficial In studies in patients with T2D and mod- nephropathy. effects on renal or cardiovascular compli- erate renal impairment, SGLT2 inhibi- cations will require evaluation in longer- tion was associated with approximately a term studies. Cardiovascular Effects of SGLT 0.3–0.45% fall in HbA1c compared with Inhibitors baseline (52–54). Though modest compared Renal Hyperfiltration and Diabetic Blood Pressure with placebo, these reductions reached Nephropathy SGLT2 inhibitor treatment is associated statistical significance for canagliflozin Glomerular hyperfiltration is an early re- with reductions in blood pressure that are and empagliflozin but not for dapagliflozin nal hemodynamic abnormality reflect- likely attributable to both an osmotic (where the placebo group experienced ing increased intraglomerular pressure. diuretic effect and weight loss. In a

Table 2—Changes in renal function with SGLT2 inhibitors (refs. 45–47) Canagliflozin (mg) Dapagliflozin (mg) Empagliflozin (mg) Placebo 100 300Placebo 5 10Placebo 10 25 n 90 90 89 84 83 85 87 90 89 eGFR (mL/min/1.73 m2) Baseline 40.0 39.8 38.8 45.6 44.2 43.9 CKD2 71.8 CKD2 70.8 CKD2 72.3 CKD3 44.3 NA CKD3 45.4 CKD4 22.0 NA CKD4 24.4 Change from baseline to 6 months 21.4 23.6 23.9 20.25 22.38 24.80 NA NA NA Change from baseline to 1 year NA NA NA 22.58 22.08 24.46 CKD2 20.71 CKD2 22.04 CKD2 22.47 CKD3 20.3 CKD3 22.8 CKD4 21.1 CKD4 21.4 Change from baseline to 2 years NA NA NA 22.38 21.71 23.50 NA NA NA 2350 Sodium–Glucose Cotransporter Inhibitors Diabetes Care Volume 38, December 2015

Table 3—Dose adjustment with renal function when using SGLT2 inhibitors (refs. 45–47) eGFR (mL/min/1.73 m2) Drug ,30 30–45 45–60 .60 Dapagliflozin - Do not start drug - Do not start - Do not start drug - Start at 5 mg QD - D/C if on drug - D/C if on drug - D/C if on drug and eGFR - Increase to 10 mg QD as needed persistently ,60 to decrease glucose Canagliflozin - Do not start drug - Do not start drug - Start at 100 mg QD - Start at 100 mg QD - D/C if on drug - D/C if on drug and eGFR persistently - Do not increase dose - Increase to 300 mg QD as needed ,45 - If on 300 mg, decrease dose to decrease glucose to 100 mg Empagliflozin - Do not start drug - Do not start - Start at 10 mg and increase to - Start at 10 mg and increase to - D/C if on drug - D/C if on drug and eGFR ,45 25 mg QD as needed to 25 mg QD as needed to persistently decrease glucose decrease glucose D/C, discontinued.

prespecified pooled analysis of 12 placebo- composite measure of cardiovascular mechanisms of action. In a small pilot controlled studies (46), treatment with da- death, nonfatal myocardial infarction, study, combination sotagliflozin and sita- pagliflozin 10 mg for 24 weeks resulted in a and nonfatal stroke by 16% compared gliptin treatment elevated active GLP-1 systolic blood pressure (SBP)/diastolic with placebo, with much of the benefit levels after meals above levels achieved blood pressure change from baseline of driven by a 38% reduction in cardiovascu- with sitagliptin alone (62). Whether this 24.4/22.1 mmHg vs. 20.9/20.5 mmHg lar death (58). Notably, reductions in the combination results in greater glycemic with placebo. Similar placebo-corrected risks of death from cardiovascular causes benefit remains to be determined. changes from baseline in SBP have been and from any cause occurred early in the seen with canagliflozin 100/300 mg trial, and these benefits continued Long-term Efficacy of SGLT2 of 23.7/25.4 mmHg and empagliflozin throughout the study. Further research Inhibitors 10/25 mg of 23.35/23.93 (45,47). In a is needed to understand the mecha- At 208 weeks, in metformin-treated pa- recent meta-analysis of .50 studies, nisms responsible for the reduction in tients, dapagliflozin compared with gli- compared with other glucose-lowering cardiovascular events. Cardiovascular pizide produced sustained reductions in fl 2 2 agents, SGLT2 inhibitors reduced mean outcome studies with canagli ozin HbA1c ( 0.30%), body weight ( 4.38 kg), fl SBP by 24.45 mmHg (57). Although these and dapagli ozin are still ongoing and SBP (23.67) with lower hypoglycemia changes in blood pressure are favorable, (59,60) and will provide further infor- rates (5.4 vs. 51.5%). Of note, glycemic longer-term studies are needed to deter- mation on the effects of SGLT2 inhibi- control gradually deteriorated over time mine whether these changes are sustained tors on cardiovascular outcomes. in both study arms but was slower with fl and, importantly, lead to lower cardiovas- dapagli ozin. At 52 weeks, mean HbA1c GI Effects of Dual SGLT1/2 Inhibitors cular morbidity and mortality. reduction was 0.5% and similar in both Intestinal Glucose Absorption and Incretin/ arms (baseline mean 7.7%). At 208 weeks, Lipids PYY Secretion the HbA1c decrease from baseline was LDL cholesterol (LDL-C) is a major car- Gut hormones (GLP-1, GIP) play an impor- 0.10 with dapagliflozin versus a 0.20% in- diovascular disease risk factor, and tant role in glucose homeostasis. Growing crease with glipizide (16). reduction of LDL-C is a primary compo- evidence suggests that SGLT1 transport nent of cardiovascular disease risk re- plays a role in entero-endocrine hormone Side Effects and Safety Profile duction strategies. SGLT2 inhibitor release. In human studies, treatment Urinary Tract and Genital Infections fl treatment is associated with small in- with sotagli ozin increased GLP-1 and SGLT2 inhibitor use is associated with creasesinLDL-CandHDLcholesterol. PYY levels after meals and reduced blood increased incidence of both urinary In long-term data over 2 years, the pla- GIP levels after breakfast in patients with tract infections (UTIs) and genital tract cebo-subtracted increases in LDL-C with T2D (31) and increased GLP-1 and PYY in infections (GTIs) (45–47). UTIs occurred dapagliflozin, canagliflozin, and empa- healthy subjects (61). Similar gut hormone more frequently in female patients, and gliflozin were ;5, 3, and 6 mg/dL in pa- changes have been seen in healthy most diagnosed infections were mild/ tients with a baseline LDL-C of ;103, 92, subjects after single 300-mg doses of moderate and responded to standard an- and 93 mg/dL, respectively. For HDL canagliflozin (14). Transient effects on timicrobial treatment. There was no in- cholesterol, the placebo-subtracted in- intestinal glucose absorption with crease in serious or upper UTIs. In the crease from a baseline of ;47 mg/dL canagliflozin are believed to be due to phase 3 studies, the incidence of UTIs was ;1, 0.6, and 3.5 mg/dL with dapa- locally high intestinal drug concentra- was 4.0, 5.9, and 4.3% with placebo, can- gliflozin, canagliflozin, and empagliflo- tions occurring shortly after dosing. agliflozin 100 mg, and canagliflozin 300 mg, zin, respectively (45–47). However, virtually all of the ingested respectively; 3.7, 5.7, and 4.3% with pla- Data from the first cardiovascular out- glucose is absorbed over a 6-h period. cebo, dapagliflozin 5 mg, and dapagliflo- comestudywithanSGLT2inhibitorwere There is a potential opportunity to com- zin 10 mg, respectively; and 7.6, 9.3, and recently reported and demonstrated bine dual SGLT1/2 inhibitors with DPP-4 7.7% with placebo, empagliflozin 10 mg, that empagliflozin treatment reduced a inhibitors based on their complementary and empagliflozin 25 mg, respectively. care.diabetesjournals.org Mudaliar and Associates 2351

Most GTIs in the clinical studies were been minimal changes in serum cal- mechanism of action of these agents mild to moderate and resolved spontane- cium, phosphorus, magnesium, 25-OH- means that these drugs could also be ously or responded to standard antifungal vitamin D, and parathyroid hormone of glycemic benefit in T1D. However, re- therapy. Infections rarely led to treat- that appear to be clinically insignificant. cent reports of DKA with SGLT2 inhibi- ment discontinuation. A small minority In a randomized, double-blind, placebo- tors in T2D and in T1D with off-label of patients experience recurrent events. controlled study, dapagliflozin treat- use mandate further detailed study, es- In some studies, male genital mycotic in- ment over 2 years did not affect markers pecially in T1D. Other side effects include fections occurred more commonly in un- of bone turnover or bone mineral den- an increase in the incidence of GTIs and circumcised males and those with a prior sity in patients with T2D inadequately in some studies a numerical excess of history of balanitis/balanoposthitis. controlled on metformin (48). In studies UTIs and bone fractures. Dual SGLT1/2 These patients were more likely to expe- with canagliflozin, there have been small inhibition is also emerging as a viable rience recurrent infections. In the phase 3 changes in bone markers and bone min- therapeutic option without an increase studies, the incidence of female GTIs was eral density that appear to be clinically in GI symptoms associated with more 3.2, 10.4, and 11.4% with placebo, cana- insignificant (49). Of note, there is a nu- extensive SGLT1 inhibition. The potential gliflozin 100 mg, and canagliflozin 300 mg, merical excess of bone fractures in for benefit due to the effects of SGLT1 on respectively; 0.9, 5.7, and 4.8% with pla- some studies with canagliflozin and gut hormones remains to be deter- cebo, dapagliflozin 5 mg, and dapagliflo- dapagliflozin (45,46). A more definitive mined. The long-term implications of zin 10 mg, respectively; and 1.5, 5.4, and answer to the deleterious effects of increased glucosuria in patients with 6.4% with placebo, empagliflozin 10 mg, SGLT2 inhibitors on bone should become diabetes are not known. Early data and empagliflozin 25 mg, respectively. available from the results of the large suggest that SGLT2 inhibition leads to The incidence of male GTIs was 0.6, cardiovascular outcome trials currently increased glucosuria and increased de- 4.2, and 3.7% with placebo, canagliflozin in progress (58–60). livery of sodium to the distal tubule, 100 mg, and canagliflozin 300 mg, respec- which may modulate tubulo-glomerular Hypoglycemia tively; 0.3, 2.8, and 2.7% with placebo, feedback and reduce glomerular Due to the insulin-independent mecha- dapagliflozin 5 mg, and dapagliflozin hyperfiltration. The benefits of such ef- nism of action of the SGLT2 inhibitors, 10 mg, respectively; and 0.4, 3.1, and fects require long-term studies that are hypoglycemia incidence rates are low 1.0% with placebo, empagliflozin 10 mg, in progress to evaluate the consequences with SGLT2 inhibitor use, except when and empagliflozin 25 mg, respectively. of SGLT2 inhibitors on cardiovascular they are used in combination with sul- disease, a major contributor to disease fonylureas and insulin, the doses of burden in patients with diabetes. DKA which may need to be lowered to avoid Some cases of DKA have recently been hypoglycemia (45–47). reported with SGLT2 inhibitor use in clinical practice (63–65). Many of these oc- Malignancies Funding. This work was supported by a grant curred in patients with T1D in whom the In the preapproval clinical studies, there from the Medical Research Service, Department of Veterans Affairs, to R.R.H. drug was used off-label, although some was an excess of bladder cancers with fl Duality of Interest. S.M. is onthe advisory board cases have occurred in patients with T2D. dapagli ozin treatment. However, there and speaker’s bureau and serves as a consultant In most cases, there were other contribut- were too few cases to determine rela- for AstraZeneca and has also received research ing factors including acute illness, infec- tionship to the drug. Hence, until addi- support paid to the Veterans Medical Research tions, reduced carbohydrate intake, tional data become available dapagliflozin Foundation from Janssen Pharmaceuticals. D.P. should not be used in patients with known is an employee of Janssen Research & Develop- missed insulin doses/pump failures, recent ment, LLC, and B.Z. is an employee of Regeneron surgery, and alcohol use. In several pa- bladder cancer (46). Pharmaceuticals, Inc. R.R.H. is on the advisory tients the blood glucose level was lower board for AstraZeneca and is on the advisory than commonly seen in DKA and in rare Conclusions board and serves as a consultant for Boehringer cases in the euglycemic range. All patients T2D is a chronic disease with significant Ingelheim and Janssen Pharmaceuticals. morbidity and mortality. The introduc- Author Contributions. S.M. and D.P. re- recovered with intravenous fluids along searched data and wrote the manuscript. B.Z. with insulin and glucose infusions and dis- tion of SGLT2 inhibitors has provided a and R.R.H. researched data, contributed to continuation of the SGLT2 inhibitor. Poten- paradigm shift in diabetes management. discussion, and reviewed and edited the man- tial mechanisms that may make patients Glucosuria, once considered a manifes- uscript. All authors contributed to the compo- tation of poor glycemic control, is now sition of the present version of the article and taking SGLT2 inhibitors more susceptible approved the final version. to developing DKA include an increase in being used to lower blood glucose lev- the glucagon-to-insulin ratio, increased els. Increased glucosuria with SGLT2 in- References free fatty acids, a shift in substrate oxida- hibition improves glycemia and leads to caloric loss and modest weight reduc- 1. Petersen C. Analyse des phloridzins. Ann tion from carbohydrate to fat, and possibly Acad Sci Fr 1835;15:178 reductions in ketone body clearance (66). tion, small decreases in blood pressure 2. Hediger MA, Coady MJ, Ikeda TS, Wright EM. (mainly SBP), and a low incidence of hy- Expression cloning and cDNA sequencing of the Skeletal Effects poglycemia. These properties have led Na+/glucose co-transporter. Nature 1987;330: The SGLT2 inhibitors may potentially to the increasing use of these agents in 379–381 3. Wells RG, Pajor AM, Kanai Y, Turk E, Wright affect calcium and phosphorus homeo- clinical practice in combination with EM, Hediger MA. Cloning of a human kidney stasis leading to adverse skeletal ef- metformin and other agents includ- cDNA with similarity to the sodium-glucose co- fects (67). In clinical studies, there have ing insulin. The insulin-independent transporter. Am J Physiol 1992;263:F459–F465 2352 Sodium–Glucose Cotransporter Inhibitors Diabetes Care Volume 38, December 2015

4. Wright EM, Loo DD, Hirayama BA. Biology of 19. Wright EM, Turk E, Martin MG. Molecular blind, crossover study. Diabetes Obes Metab human sodium glucose transporters. Physiol basis for glucose-galactose malabsorption. Cell 2015;17:188–197 Rev 2011;91:733–794 Biochem Biophys 2002;36:115–121 34. Stein P, Berg JK, Morrow L, et al. 5. Vestri S, Okamoto MM, de Freitas HS, et al. 20. Powell DR, Smith M, Greer J, et al. LX4211 Canagliflozin, a sodium glucose co-transporter Changes in sodium or glucose filtration rate increases serum glucagon-like peptide 1 and 2 inhibitor, reduces post-meal glucose excur- modulate expression of glucose transporters peptide YY levels by reducing sodium/glucose sion in patients with type 2 diabetes by a non- in renal proximal tubular cells of rat. J Membr cotransporter 1 (SGLT1)-mediated absorption renal mechanism: results of a randomized trial. Biol 2001;182:105–112 of intestinal glucose. J Pharmacol Exp Ther Metabolism 2014;63:1296–1303 6. Freitas HS, Anheˆ GF, Melo KF, et al. Na(+) - 2013;345:250–259 35. Pieber TR, Famulla S, Eilbracht J, et al. glucose transporter-2 messenger ribonucleic 21. Parker HE, Adriaenssens A, Rogers G, et al. Empagliflozin as adjunct to insulin in patients acid expression in kidney of diabetic rats corre- Predominant role of active versus facilitative with type 1 diabetes: a 4-week, randomized, lates with glycemic levels: involvement of hepa- glucose transport for glucagon-like peptide-1 placebo-controlled trial (EASE-1). Diabetes tocyte nuclear factor-1alpha expression and secretion. Diabetologia 2012;55:2445–2455 Obes Metab 2015;17:928–935 activity. Endocrinology 2008;149:717–724 22. Ezcurra M, Reimann F, Gribble FM, Emery E. 36. Henry RR, Rosenstock J, Edelman S, et al. 7. Rahmoune H, Thompson PW, Ward JM, Smith Molecular mechanisms of incretin hormone se- Exploring the potential of the SGLT2 inhibitor CD, Hong G, Brown J. Glucose transporters in hu- cretion. Curr Opin Pharmacol 2013;13:922–927 dapagliflozin in type 1 diabetes: a randomized, man renal proximal tubular cells isolated from 23. Chasis H, Jolliffe N, Smith HW. The action of double-blind, placebo-controlled pilot study. the urine of patients with non-insulin-dependent Phlorizinontheexcretionofglucose,xylose, Diabetes Care 2015;38:412–419 diabetes. Diabetes 2005;54:3427–3434 sucrose, creatinine and urea by man. J Clin In- 37. Sands AT, Zambrowicz BP, Rosenstock J, 8. Stearns AT, Balakrishnan A, Rhoads DB, vest 1933;12:1083–1090 et al. Sotagliflozin, a dual SGLT1 and SGLT2 in- Tavakkolizadeh A. Rapid upregulation of sodium- 24. Rossetti L, Shulman GI, Zawalich W, hibitor, as adjunct therapy to insulin in type 1 – glucose transporter SGLT1 in response to intesti- DeFronzo RA. Effect of chronic hyperglycemia diabetes. Diabetes Care 2015;38:1181 1188 nal sweet taste stimulation. Ann Surg 2010; on in vivo insulin secretion in partially pancrea- 38. Ferrannini E, Muscelli E, Frascerra S, et al. 251:865–871 tectomized rats. J Clin Invest 1987;80:1037– Metabolic response to sodium-glucose cotrans- 9. Dyer J, Wood IS, Palejwala A, Ellis A, Shirazi- 1044 porter 2 inhibition in type 2 diabetic patients. J – Beechey SP. Expression of monosaccharide 25. Rossetti L, Smith D, Shulman GI, Papachristou Clin Invest 2014;124:499 508 fl transporters in intestine of diabetic humans. D, DeFronzo RA. Correction of hyperglycemia 39. Polidori D, Mari A, Ferrannini E. Canagli ozin, Am J Physiol Gastrointest Liver Physiol 2002; with phlorizin normalizes tissue sensitivity to a sodium glucose co-transporter 2 inhibitor, im- 282:G241–G248 insulinindiabeticrats.JClinInvest1987;79: proves model-based indices of beta cell function 10. Shannon JA, Fisher S. The renal tubular re- 1510–1515 in patients with type 2 diabetes. Diabetologia 2014;57:891–901 absorption of glucose in the normal dog. Am J 26. Oku A, Ueta K, Arakawa K, et al. T-1095, an 40. Mudaliar S, Henry RR, Boden G, et al. Physiol 1938;122:765–774 inhibitor of renal Na+-glucose cotransporters, Changes in insulin sensitivity and insulin secre- 11. Vick H, Diedrich DF, Baumann K. Reevalua- may provide a novel approach to treating dia- tion with the sodium glucose cotransporter 2 tion of renal tubular glucose transport inhibition betes. Diabetes 1999;48:1794–1800 inhibitor dapagliflozin. Diabetes Technol Ther by phlorizin analogs. Am J Physiol 1973;224: 27. Dobbins RL, Greenway FL, Chen L, et al. Se- 2014;16:137–144 552–557 lective sodium-dependent glucose transporter 41. Merovci A, Solis-Herrera C, Daniele G, et al. 12. Turner RJ, Moran A. Heterogeneity of sodium- 1 inhibitors block glucose absorption and impair Dapagliflozin improves muscle insulin sensitiv- dependent D-glucose transport sites along the glucose-dependent insulinotropic peptide re- ity but enhances endogenous glucose produc- proximal tubule: evidence from vesicle studies. lease. Am J Physiol Gastrointest Liver Physiol tion. J Clin Invest 2014;124:509–514 Am J Physiol 1982;242:F406–F414 2015;308:G946–G954 42. Bonner C, Kerr-Conte J, Gmyr V, et al. In- 13. Liang Y, Arakawa K, Ueta K, et al. Effect of 28. Grempler R, Thomas L, Eckhardt M, et al. hibition of the glucose transporter SGLT2 with canagliflozin on renal threshold for glucose, gly- Empagliflozin, a novel selective sodium dapagliflozin in pancreatic alpha cells triggers cemia, and body weight in normal and diabetic glucose cotransporter-2 (SGLT-2) inhibitor: char- glucagon secretion. Nat Med 2015;21:512– animal models. PLoS One 2012;7:e30555 acterisation and comparison with other SGLT-2 517 – 14. Polidori D, Sha S, Mudaliar S, et al. inhibitors. Diabetes Obes Metab 2012;14:83 90 43. Hansen L, Iqbal N, Ekholm E, Cook W, fl Canagli ozin lowers postprandial glucose and in- 29. Mascitti V, Maurer TS, Robinson RP, et al. Hirshberg B. Postprandial dynamics of plasma sulin by delaying intestinal glucose absorption in Discovery of a clinical candidate from the struc- glucose, insulin, and glucagon in patients with addition to increasing urinary glucose excretion: turally unique dioxa-bicyclo[3.2.1]octane class type 2 diabetes treated with saxagliptin plus results of a randomized, placebo-controlled of sodium-dependent glucose cotransporter 2 dapagliflozin add-on to metformin therapy. – – study. Diabetes Care 2013;36:2154 2161 inhibitors. J Med Chem 2011;54:2952 2960 Endocr Pract 2014;20:1187–1197 15. Polidori D, Sha S, Ghosh A, Plum-Morschel¨ 30. Kakinuma H, Oi T, Hashimoto-Tsuchiya Y, 44. Sha S, Polidori D, Heise T, et al. Effect of the L, Heise T, Rothenberg P. Validation of a novel et al. (1S)-1,5-anhydro-1-[5-(4-ethoxybenzyl)- sodium glucose co-transporter 2 inhibitor cana- method for determining the renal threshold for 2-methoxy-4-methylphenyl]-1-thio-D-glucitol gliflozin on plasma volume in patients with glucose excretion in untreated and canagliflozin- (TS-071) is a potent, selective sodium-depen- type 2 diabetes mellitus. Diabetes Obes Metab treated subjects with type 2 diabetes mellitus. J dent glucose cotransporter 2 (SGLT2) inhibitor 2014;16:1087–1095 Clin Endocrinol Metab 2013;98:E867–E871 for type 2 diabetes treatment. J Med Chem 45. Canagliflozin [package insert]. Titusville, NJ, 16. Del Prato S, Nauck M, Duran-Garcia´ S, et al. 2010;53:3247–3261 Janssen Pharmaceuticals, Inc., 2014 Long-term glycaemic response and tolerability 31. Zambrowicz B, Freiman J, Brown PM, et al. 46. Dapagliflozin [package insert]. Wilmington, of dapagliflozin versus a sulphonylurea as add- LX4211, a dual SGLT1/SGLT2 inhibitor, improved DE, AstraZeneca Pharmaceuticals LP, 2014 on therapy to metformin in patients with type 2 glycemic control in patients with type 2 diabetes 47. Empagliflozin [package insert]. Ridgefield, diabetes: 4-year data. Diabetes Obes Metab in a randomized, placebo-controlled trial. Clin CT, Boehringer Ingelheim Pharmaceuticals, Inc., 2015;17:581–590 Pharmacol Ther 2012;92:158–169 and Indianapolis, IN, Eli Lilly and Company, 2014 17. Powell DR, DaCosta CM, Gay J, et al. Im- 32. Rosenstock J, Cefalu WT, Lapuerta P, et al. 48. Bolinder J, Ljunggren O,¨ Johansson L, et al. proved glycemic control in mice lacking Sglt1 Greater dose-ranging effects on A1C levels than Dapagliflozin maintains glycaemic control while and Sglt2. Am J Physiol Endocrinol Metab on glucosuria with LX4211, a dual inhibitor of reducing weight and body fat mass over 2 years 2013;304:E117–E130 SGLT1 and SGLT2, in patients with type 2 diabe- in patients with type 2 diabetes mellitus inade- 18. Gorboulev V, Schurmann¨ A, Vallon V, et al. tes on metformin monotherapy. Diabetes Care quately controlled on metformin. Diabetes Na(+)-D-glucose cotransporter SGLT1 is pivotal 2015;38:431–438 Obes Metab 2014;16:159–169 for intestinal glucose absorption and glucose- 33. Sha S, Polidori D, Farrell K, et al. Pharmaco- 49. Cefalu WT, Leiter LA, Yoon KH, et al. Efficacy dependent incretin secretion. Diabetes 2012; dynamic differences between canagliflozin and and safety of canagliflozin versus glimepiride in 61:187–196 dapagliflozin: results of a randomized, double- patients with type 2 diabetes inadequately care.diabetesjournals.org Mudaliar and Associates 2353

controlled with metformin (CANTATA-SU): 52 cotransporter 2 inhibition in patients with postprandial glucose, insulin, glucagon-like pep- week results from a randomised, double-blind, type 1 diabetes mellitus. Circulation 2014;129: tide 1, and peptide tyrosine tyrosine in a dose- phase 3 non-inferiority trial. Lancet 2013;382: 587–597 timing study in healthy subjects. Clin Ther 2013; 941–950 56. Gilbert RE. Sodium-glucose linked trans- 35:1162–1173.e8 50. Heise T, Seewaldt-Becker E, Macha S, et al. porter-2 inhibitors: potential for renoprotection 62. Zambrowicz B, Ding ZM, Ogbaa I, et al. Ef- Safety, tolerability, pharmacokinetics and pharma- beyond blood glucose lowering? Kidney Int fects of LX4211, a dual SGLT1/SGLT2 inhibitor, codynamics following 4 weeks’ treatment with 2014;86:693–700 plus sitagliptin on postprandial active GLP-1 and empagliflozin once daily in patients with type 2 57. Vasilakou D, Karagiannis T, Athanasiadou E, glycemic control in type 2 diabetes. Clin Ther diabetes. Diabetes Obes Metab 2013;15:613–621 et al. Sodium-glucose cotransporter 2 inhibitors 2013;35:273–285.e7 51. Lambers Heerspink HJ, de Zeeuw D, Wie L, for type 2 diabetes: a systematic review and meta- 63. U.S. Food and Drug Administration. Drug – Leslie B, List J. Dapagliflozin a glucose-regulating analysis. Ann Intern Med 2013;159:262 274 Safety Communication: FDA warns that SGLT2 in- drug with diuretic properties in subjects with 58. Boehringer Ingelheim. BI 10773 (Empagli- hibitors for diabetes may result in a serious condi- fl type 2 diabetes. Diabetes Obes Metab 2013; ozin) Cardiovascular Outcome Event Trial in tion of too much acid in the blood [article online], Type 2 Diabetes Mellitus Patients (EMPA-REG 15:853–862 2015. Available from http://www.fda.gov/Drugs/ OUTCOME). In: ClinicalTrials.gov [Internet]. 52. Kohan DE, Fioretto P, Tang W, List JF. Long- DrugSafety/ucm446845.htm. Accessed 18 July 2015 Bethesda, MD, National Library of Medicine, term study of patients with type 2 diabetes 64. European Medicines Agency. Review of diabe- 2015. Available from https://clinicaltrials.gov/ct2/ and moderate renal impairment shows that tes medicines called SGLT2 inhibitors started [arti- show/NCT01131676?term=empagliflozin+and+ dapagliflozin reduces weight and blood pres- cle online], 2015. Available from http://www.ema cardiovascular&rank=1 sure but does not improve glycemic control. .europa.eu/ema/index.jsp?curl=pages/medicines/ 59. Neal B, Perkovic V, de Zeeuw D, et al. Ratio- – Kidney Int 2014;85:962 971 nale, design, and baseline characteristics of the human/referrals/SGLT2_inhibitors/human_referral_ fi 53. Yale JF, Bakris G, Cariou B, et al. Ef cacy and Canagliflozin Cardiovascular Assessment Study prac_000052.jsp&mid=WC0b01ac05805c516f. fl safety of canagli ozin in subjects with type 2 (CANVAS)–a randomized placebo-controlled Accessed 18 July 2015 diabetes and chronic kidney disease. Diabetes trial. Am Heart J 2013;166:217–223.e11 65. Peters AL, Buschur EO, Buse JB, Cohan P, Obes Metab 2013;15:463–473 60. AstraZeneca. Multicenter Trial to Evaluate Diner JC, Hirsch IB. Euglycemic diabetic ketoacido- 54. Barnett AH, Mithal A, Manassie J, et al.; the Effect of Dapagliflozin on the Incidence sis: a potential complication of treatment with EMPA-REG RENAL Trial Investigators. Efficacy of Cardiovascular Events (DECLARE-TIMI58). sodium-glucose cotransporter 2 inhibition. Diabe- and safety of empagliflozin added to existing In: ClinicalTrials.gov [Internet]. Bethesda, tes Care 2015;38:1687–1693 antidiabetes treatment in patients with type 2 MD, National Library of Medicine, 2015. 66. Taylor SI, Blau JE, Rother KI. SGLT2 inhibi- diabetes and chronic kidney disease: a rando- Available from http://clinicaltrials.gov/show/ tors may predispose to ketoacidosis. J Clin En- mised, double-blind, placebo-controlled trial. NCT01730534 docrinol Metab 2015;100:2849–2852 Lancet Diabetes Endocrinol 2014;2:369–384 61. Zambrowicz B, Ogbaa I, Frazier K, et al. Ef- 67. Taylor SI, Blau JE, Rother KI. Possible ad- 55. Cherney DZ, Perkins BA, Soleymanlou N, et al. fects of LX4211, a dual sodium-dependent glu- verse effects of SGLT2 inhibitors on bone. Lan- Renal hemodynamic effect of sodium-glucose cose cotransporters 1 and 2 inhibitor, on cet Diabetes Endocrinol 2015;3:8–10