American Diabetes Association 74th Scientific Sessions

June 13-17, 2014; San Francisco, CA – SGLT-2 and SGLT-1 Inhibitors and Other Oral Antidiabetic Agents (Non-Incretin) - Draft

Executive Highlights

As with last year’s ADA, this year’s scientific sessions featured a great deal of new data on SGLT-2 inhibitors, although the focus was less on pivotal phase 3 data and more on understanding the nuances of the class’ mechanism and exploring new applications. Leading up to the meeting, we heard a great deal of optimism about SGLT-2 inhibitor/DPP-4 inhibitor fixed-dose combination (FDCs), which bring together insulin-dependent and insulin-dependent mechanisms for (theoretically) additive or even synergistic efficacy. The results presented by AZ and Lilly/BI (on empa/lina [empagliflozin/linagliptin] and saxa/dapa [saxagliptin/ dapagliflozin], respectively) were certainly very positive: saxa/dapa demonstrated a nearly 1.5% A1c reduction from baseline (albeit from a relatively high mean baseline of >9%) and empa/lina yielded solidly >1% A1c reductions from baseline. However, the results fell short of the truly additive efficacy many were hoping for, and in most cases it appeared that the FDCs’ efficacy was largely driven by the SGLT-2 inhibitor component. We would imagine that these FDCs will best distinguish themselves in real-world situations, when the advantages of consolidated dosing and co-pays will have a greater positive impact on adherence.

In a packed oral presentation session dedicated to SGLT-2 inhibitors, we saw some favorable long-term efficacy and safety data on BI/Lilly’s Jardiance (empagliflozin), which was recently approved in Europe and resubmitted to the FDA. The EMPA-REG H2H-SU found that empagliflozin provided statistically superior A1c reductions after a full two years than the SFU glimepiride (-0.66% vs. -0.55% from a relatively low baseline of 7.9%) along with weight and hypoglycemia benefits. We also saw empagliflozin come out on top of Merck’s DPP-4 inhibitor Januvia (sitagliptin) with regards to A1c and weight in a 52-week extension study of the 24-week EMPA-REG MONO trial. The same session also featured long-term data on J&J’s Invokana (canagliflozin) and AZ’s Farxiga (dapagliflozin). Out in the poster hall, we saw impressive phase 2 results on one of the newest kids on the SGLT-2 inhibitor block, Islet Science’s remogliflozin etabonate. The 12-week dose ranging study found a mean placebo-adjusted A1c reduction of 1.1% from baseline, quite striking for a 12-week study (albeit a small one). The current formulation requires twice-daily dosing, but the company has developed an intriguing biphasic formulation that should allow for once-daily dosing; this is theorized to preserve efficacy while reducing side effects.

There was not much this year on selective SGLT-1 inhibition, but there was some new data on SGLT-1/2 dual inhibition, namely Lexicon’s LX4211. A poster presented the positive results of a study of LX4211 in type 2 diabetes patients with renal impairment – we covered topline results from that trial earlier this year. The poster suggests that glycemic improvement in the patients with the most severe renal impairment is driven by the SGLT-1 inhibition component. Our full ADA 2014 SGLT-2 Inhibitor & SGLT-2 Inhibitor Category Report also features coverage of two high-profile symposia on SGLT inhibition, a well as a product theater on AZ’s Farxiga.

Regarding other classes of oral antidiabetic agents, ADA 2014 reprised results of the phase 3 AleCardio trial for Roche/Genentech’s PPAR alpha/gamma dual agonist, which was terminated in 2013 for futility. Presentations that touched upon sulfonylurea use emphasized the heterogeneity of the class and the need to mitigate SFUs’ negative effect by selecting the right agent – there seem to be an increasing number of negative references to hypoglycemia and weight gain, which makes sense as the bar is raised for diabetes therapies. To be sure, an agent with both weight gain and hypoglcyemia associated with the SFU class (to say nothing of other negatives) would almost certainly never get through FDA (or be put forward for development by industry). Dr. Scot Simpson (University of Alberta, Edmonton, Canada) conducted a meta-analysis that found that glimepiride and gliclazide had lower mortality risk than glyburide and glipizide.

Below you’ll find detailed reports on all of these topics and more. Presentation titles highlighted in blue were not previously published in our daily highlight reports during ADA. We’ve highlighted in yellow presentations we found particularly notable.


Table of Contents 

SGLT-2 Inhibitors and SGLT-1 Inhibitors

Oral Presentations: SGLT-2 Inhibitors

Empagliflozin (EMPA) Monotherapy for ≥76 Weeks in Drug-Naïve Patients with Type 2 Diabetes (T2DM) (264-OR)

Michael Roden, MD (Leibniz Center for Diabetes Research, Dusseldorf, Germany)

Dr. Michael Roden presented the results of a 52-week extension study of a 24-week registrational trial (EMPA-REG MONO) comparing Lilly/BI’s SGLT-2 inhibitor Jardiance (empagliflozin) at both 25 mg and 10 mg doses against Merck’s Januvia (sitagliptin 100 mg) and placebo in drug-naïve type 2 diabetes patients. The extension study enrolled 68% of patients enrolled in the initial study. The 24-week data demonstrated that empagliflozin provided comparable A1c reductions to sitagliptin, but the curves diverged slightly during the extension study. At week 76, empagliflozin 25 mg led to a significantly greater placebo-adjusted drop in A1c (-0.89%) relative to sitagliptin (-0.66%) (p = 0.005). The empagliflozin 10 mg arm showed a non-significant trend towards increased efficacy, with a placebo-adjusted reduction of -0.78%. As at 24 weeks, both empagliflozin doses led to significantly greater weight loss (~2.5 kg, ~6 lbs) than sitagliptin, which was weight-neutral. There were slightly more drug-related adverse events with empagliflozin than sitagliptin, which were (unsurprisingly) driven by a ~5% increase in genital infections. The sustained efficacy over 76 weeks and superior A1c reduction over sitagliptin should provide Lilly/BI with some good talking points as it begins empagliflozin’s launch in Europe and in new discussions with the US FDA after the recent re-submission.

  • Results from the original 24-week study (EMPA-REG MONO) were published in the journal Lancet in September of last year.
    • J&J’s Invokana (canagliflozin) previously demonstrated superiority to sitagliptin in a 52-week trial in patients on metformin +/- sulfonylurea – we covered the results at ADA 2012 (read our report). AZ’s Forxiga/Farxiga (dapagliflozin) has been studied as an add-on to sitagliptin, but not versus sitagliptin in phase 3.
  • Of the 899 patients treated in the initial 24-week study, 68% patients continued in the extension study. Characteristics of the patients that continued into the extension study were comparable to those of patients that did not continue. There were slightly more male patients in the study, with a median age of approximately 55 years. Notably, nearly two-thirds of the patients in the study were Asian, with 50% of that group from China, 30% from Japan, and the remaining 20% from India. It is unclear how this might impact the interpretation of the results – some very early studies have suggested that DPP-4 inhibitors have greater efficacy in East Asian populations – but Dr. Roden noted during Q&A that no interaction was found between efficacy and ethnicity (Asian vs. Caucasian). Diabetes duration was on the low side, with most patients having a disease duration of 1 – 5 years. The four arms of the study were placebo, sitagliptin 100 mg, empagliflozin 10 mg, and empagliflozin 25 mg. The placebo group seemed to lose patients to follow-up at a more rapid rate than the other study arms during both the initial study and the extension study, and at week 76 only 65 patients remained in the placebo arm (relative to 132 in each of the empagliflozin arms). Perhaps as a result, the A1c results were displayed as absolute reductions and not placebo-adjusted reductions, although we calculated placebo-adjusted values below.
  • After 76 weeks from the beginning of the initial study, the empagliflozin 25 mg arm achieved a significantly greater placebo-adjusted A1c reduction (-0.89%) than the sitagliptin arm (-0.66%). The difference was statistically significant (p = 0.005); baseline A1c was ~7.9%. The empagliflozin 10 mg arm achieved a numerically greater placebo-adjusted A1c reduction than sitagliptin (-0.78% vs -0.66%), but the difference was not statistically significant (p = 0.131). For background, neither empagliflozin dose achieved a statistically greater A1c reduction than empagliflozin at 24 weeks; the difference emerged gradually during the extension study, as the sitagliptin arm’s A1c crept slowly upwards while the empagliflozin groups’ A1cs held slightly more level.
  • Empagliflozin led to significantly weight loss and reductions in systolic blood pressure than sitagliptin. Both empagliflozin groups achieved weight loss of approximately 2.5 kg (~6 lbs), while sitagliptin and placebo were weight neutral (p<0.001 for both empagliflozin groups vs sitagliptin). Systolic blood pressure fell by approximately 4 mmHg in the empagliflozin groups, and here to there was no appreciable change in the placebo or sitagliptin groups (p=0.001). 
  • As expected, there were more genital infections in the empagliflozin arms (6%) than the sitagliptin (1%) or placebo (2%) arms. This difference drove a difference in overall treatment-related adverse events between empagliflozin and sitagliptin. Other than this difference, there were no imbalances in safety or tolerability between empagliflozin and sitagliptin. Hypoglycemia incidence was identical between the four arms.

Questions and Answers

Q: You said that the enrollment criterion for “drug-naïve” was no diabetes drug use for 12 weeks prior to the study. Were there patients in the trial who had previously been on diabetes drugs and were taken off? If so, do you know the proportion of patients who were previously treated versus those who were truly drug naïve?

A: Subjects could have been on previous metformin treatment, as long as it was discontinued more than 12 weeks before the trial. We don’t have data on the distribution, but there was no imbalance in that proportion of previously-treated patients between the groups.

Q: You mentioned that around two-thirds of the study population was Asian. Could this have introduced any genetic discrimination?

A: Yes, a specific feature of this study is that it included a high percentage of people from Asia. Out of that group, around 50% was from China, 30% from Japan, and 20% from India. There have been subgroup analyses to compare the efficacy between Asians and Caucasians, and there was no real effect – the P-value for the interaction was higher than 0.1.

Q: Could you provide more detail on the temporality of the 4 mmHg reduction in systolic blood pressure you saw at week 76? Was the difference more dramatic at the beginning of the trial, and was subsequently attenuated?

A: If you look at the time curves, the reduction was more or less maintained throughout the study. There were no marked changes at the very end, or at points in between.

Empagliflozin (EMPA) Compared with Glimepiride (GLIM) as Add-on to Metformin (MET) for 2 Years in Patients with Type 2 Diabetes (T2DM) (266-OR)

Martin Ridderstråle, MD, PhD (Steno Diabetes Center, Gentofte, Denmark)

Dr. Martin Ridderstråle presented the results of the two-year EMPA-REG H2H-SU trial, which compared Lilly/BI’s Jardiance (empagliflozin 25 mg) to the sulfonylurea (SFU) glimepiride (1-4 mg), both in addition to metformin. At week 104, from a baseline of 7.9%, empagliflozin yielded a greater reduction in A1c (-0.66%) than glimepiride (-0.55%); the difference, although modest, was enough to achieve the statistical margins for both non-inferiority and superiority. Notably, in the glimepiride arm, the average daily dose was only 2.7 mg per day – Dr. Ridderstråle suggested during Q&A that the glimepiride dosing does impact the interpretation of the efficacy results, but that the more notable findings were in areas beyond A1c. Empagliflozin was associated with a far lower incidence of hypoglycemia (2.5%) than glimepiride (24.2%). This high rate of hypoglycemia suggests to us that pushing the dose of glimepiride further would likely have been unsafe. Empagliflozin came out ahead with regards to body weight (4.5 kg [~10 lbs] difference) and blood pressure, but slightly increased cholesterol and the incidence of genital infections. These results show that empagliflozin is strongly differentiated from SFUs when it comes to factors such as hypoglycemia and weight that are very important for patients, and that the drug might even be able to deliver better A1c-lowering efficacy in the long-term. 

  • This study enrolled 1545 type 2 diabetes patients who were on stable metformin therapy. Approximately one-third of enrolled patients were Asian, and approximately 17% had diabetes duration of over 10 years. The study randomized patients 1:1 to empagliflozin 25 mg or glimepiride, with a dose range of 1-4 mg.
  • Empagliflozin led to a statistically superior reduction in A1c from baseline (-0.66%) at week 102 compared to the -0.55% reduction seen with glimepiride (p = 0.015). Dr. Ridderstråle noted that the mean maximum glimepiride dose in the glimepiride arm was 2.7 mg, with only 40% reaching the maximum dose. When asked about that fact during Q&A, Dr. Ridderstråle suggested that the efficacy should be viewed as more or less comparable, with the more meaningful differentiation coming in the form of benefits beyond A1c. The time-course of A1c reductions was in line with what we have seen in previous trials: glimepiride caused a sharp initial decline that began to creep upward, while empagliflozin led to a more gradual initial reduction that held fairly steady. The curves crossed at around week 40.
  • There was markedly less hypoglycemia with empagliflozin than glimepiride. Approximately 24% of patients on glimepiride experienced a confirmed hypoglycemic adverse event, while the incidence was under 3% in the empagliflozin arm (p<0.001). The only events requiring assistance (n = 2) were in the glimepiride group. Beyond providing reassurance on empagliflozin, these results indicate that the glimepiride dose was probably titrated well, as raising the dose would likely have caused an unacceptable level of hypoglycemia.
  • The empagliflozin group lost 3 kg (~7 lbs) on average, while the glimepiride group lost 1kg (~3 lbs) on average. The empagliflozin group saw a slight 3 mmHg reduction in systolic blood pressure, while the glimepiride group saw a rise of 3 mmHg (p<0.001).
  • There were more drug-related adverse events in the glimepiride group, driven by the increased hypoglycemia. There were more serious adverse events, however in the empagliflozin arm, although Dr. Ridderstråle noted that there was no specific signal or pattern among individual serious adverse events that drove this imbalance. Genital infections were higher with empagliflozin (~12%) than glimepiride (~2%).
  • As has been seen in other SGLT-2 inhibitor trials, empagliflozin led to slight increases in total cholesterol, LDL-C, and HDL-C.

Questions and Answers

Q: Given that the mean daily dose of glimepiride in the study was only 2.7 mg, which is a ways away from the maximum licensed dose, is the superiority result questionable?

A: I do not think that it is questionable, but I do think that you have to view the data in terms of that difference in dose. I’m inclined not to emphasize the superiority finding, as I do not think that is the important finding. A better way to look at the efficacy is that the results were more or less comparable, and then interpret the other effects in that light.

Q: Why was there an elevation in genital infections but not urinary tract infections?

A: I’m not an expert on infectious disease, but from the data I have seen, apparently bacteria in the urinary tract like glucose but do not need it, while mycotic agents absolutely love glucose.

Q: Did you measure if there was a correlation between infections, volume depletion, and baseline blood sugar levels?

A: There was no difference in the level of volume depletion or infections with respect to blood sugar levels. They are seen in all levels of glycemia.

Q: I noticed that LDL cholesterol and triglycerides went up despite a reduction in weight loss and hyperglycemia. Did you measure ApoB?

A: The lipid findings were interesting, but we haven’t evaluated those details yet. Concerning triglycerides, we actually saw a decrease at 52 weeks, then a slight increase. Regarding LDL, this slight increase appears to be consistent but small, and we don’t know if it is clinically significant. There are some hypotheses about why you see that increase in LDL. One is that there is a slight increase in hematocrit, and that if you adjust for that, you see less of an increase. Another interesting theory is that if you lose sugar through urine, you need to compensate with increased energy intake and your dietary intake and composition may change. Supporting that theory is the fact that you see increases in both LDL and HDL.

Improvement in Glycemic Control and Reduction in Body Weight over 52 Weeks with Dapagliflozin as Add-on Therapy to Metformin plus Sulfonylurea (267-OR)

Stephan Matthaei, MD (Diabetes-Center Quackenbruck, Quackenbruck, Germany)

Discussing dapagliflozin 10 mg/day as add-on to metformin and sulfonylurea, Dr. Stephan Matthaei presented 52-week data from a randomized controlled trial (24-week phase 3 trial plus follow-on period). The study included adults with type 2 diabetes inadequately controlled on metformin and sulfonylurea (n=216). Baseline mean A1c was 8.1%-8.2%, age 61 years, BMI 32 kg/m2, diabetes duration 9.3-9.6 years, and systolic blood pressure 135-136 mm Hg. Patients were randomized 1:1 to receive either placebo or dapagliflozin plus their current medications. At 52 weeks dapagliflozin led to statistically significant reductions in A1c (1.10% vs. 0.39% for placebo) and body weight (-2.66 kg vs. -0.48 kg for placebo [5.9 lbs vs. 1.1 lbs]). Systolic blood pressure was reduced at 24 weeks but then returned to baseline by 52 weeks, albeit remaining lower than placebo (decrease of 0.9 mm Hg vs. increase of 1.3 mm Hg for placebo). Minor hypoglycemic events – blood glucose >63 mg/dl with symptoms, or <63 mg/dl without symptoms – were more common with dapagliflozin than placebo (15.7% vs. 8.3%). Dr. Matthaei attributed the increase to sulfonylurea therapy and recommended that when starting on dapagliflozin, dosage of sulfonylurea (or insulin) should be reduced. Genital tract infections were also more common with dapagliflozin (10.2% vs. 0.9%), especially in women (14.3% vs. 2.0%). Overall adverse events were balanced between dapagliflozin and placebo (70.1% vs. 74.1%), serious adverse events were also balanced (6.5% vs. 7.4%).

Questions and Answers

Q: The blood pressure went down at 24 weeks but then trended back, despite the persistence of weight loss. Do you have any idea why?

A: Based on our four-year comparison to glipizide, when we had stable reduction in weight and blood pressure, this was kind of unexpected. Mean absolute systolic blood pressure at 24 weeks was 129 mm Hg; perhaps doctors reduced other blood pressure meds. We need to look at this.

Q: What was the standard deviation of weight loss?

A: (Shows slide – we estimate the SD at roughly 1 kg [2.2 lbs]). This is good news for us as doctors; it encourages compliance. In Germany, where we have the largest experience, since we’ve had access for 18 months, compliance is very high. These patients have tried to lose with for decades, and suddenly they are able to do it with this new medication

Q: Were dapagliflozin-related adverse events mainly the effect of hypoglycemia due to glycosuria, or were they having other drug-related comorbidities?

A: This was mainly due to more minor episodes of hypoglycemia, which is due to sulfonylurea co-medication. Whenever we try to reach these ambitious glycemic goals with this co-medication, we should adapt the dose or even discontinue sulfonylureas.

Q: Many times in the presentation, themes emerged. The hypoglycemia was striking as you add dapagliflozin, also the increase in genital infections. Should we give any instructions to patients up front?

A: The best instruction is triple combination-therapy: metformin, DPP-4 inhibitor, and SGLT-2 inhibitor. But sulfonylureas are still widely used. The next question is, why not halve the dose of sulfonylurea, and then increase it again only if glycemic goal is not met? It’s the same question with insulin. My recommendation would be to reduce insulin right away by about 20% to have no hypoglycemic episodes.

Dapagliflozin is Safe and Well Tolerated in Older Patients with T2DM (269-OR)

Traci Mansfield, PhD (Bristol-Myers Squibb, Princeton, NJ)

Dr. Traci Mansfield compared the safety and tolerability of Farxiga (AstraZeneca’s dapagliflozin) 10 mg vs. placebo in three age ranges (<65, ≥65, and ≥75 years old), from a pooled analysis of nine phase 3 trials. At baseline, mean A1c was roughly 8.0-8.1% in all groups. Contrasting the younger, older, and oldest patients (n=2707 vs. 1275 vs. 174), the main baseline differences were mean age (55 vs. 70 vs. 77 years old), diabetes duration (8 vs. 13 vs. 15 years), and percentage of patients with moderate renal impairment (8.3% vs. 22% vs. ~25%). Frequency of total adverse events were more common with dapagliflozin than placebo in all three age groups, with adverse events most common in the oldest patients (74.4% vs. 71.5%; 73.1% vs. 70.7%; 80.4% vs. 75.3%). Severe adverse events were balanced between dapagliflozin and placebo (13.7% vs. 14.6%) and were similar in all three age groups. Genital infections and urinary tract infections were generally more common with dapagliflozin than placebo, and rates were similar across age groups. Renal adverse events were more common with dapagliflozin, and they were more common in older patients (3.5% vs. 2.3%; 14.0% vs. 7.9%; 29.9% vs. 20.8%). Dr. Mansfield explained that most renal events involved increased serum creatinine and/or decreased creatinine clearance and that many events occurred in patients with moderate renal impairment, who would be contraindicated today. Volume-related events were more common with dapagliflozin and were seen most in the oldest patients (1.7% vs. 1.2%, 2.3% vs. 1.7%; 3.1% vs. 2.6%); Dr. Mansfield said that the volume-related events were not related to renal function. 

Questions and Answers

Q: I am curious about adding more AstraZeneca products to the mix – what would happen if you added Bydureon and Kombiglyze?

A: My expertise is really in dapagliflozin and SGLT-2 inhibition, and I would probably apply my comments to that.

Questioner: In our clinic we stack often with SGLT-2, Bydureon, metformin of course – I’m not sure about adding a DPP-4 inhibitor in people already taking a GLP-1 receptor agonist – then add basal insulin, and then short-acting insulin for glucose above 180 mg/dl. We get tremendous benefits, including in phenomena of wellness; it takes more than the one bullet of dapagliflozin.

Q: About 20% of older patients had moderate renal impairment, who would be contraindicated for Farxiga. How’d these patients do with renal and volume-related adverse events?

A: These studies were conducted before the label was created. When we exclude those patients, you see a lot fewer renal events. Of the renal events in older subgroup, a lot were driven by patients with eGFR<60. Volume-depletion events were still more common with dapagliflozin than placebo.

Q: Did you perform any subgroup analyses of patients taking loop diuretics or thiazide diuretics?

A: Patients taking loop diuretics were more prone to renal and volume-related events, but we don’t see this with other antihypertensive agents.

Long-term Efficacy and Safety of Canagliflozin (CANA) in Older Patients with Type 2 Diabetes Mellitus (T2DM) Over 104 Weeks (268-OR)

Bruce Bode, MD (Atlanta Diabetes Associates, Atlanta, GA)

Dr. Bruce Bode presented the results of a 78-week extension of a 26-week primary study on the effects of J&J’s Invokana (canagliflozin) in elderly type 2 diabetes patients. The double-blind placebo-controlled study randomized 716 type 2 diabetes patients (age 55 – 80) with inadequately controlled diabetes to placebo, canagliflozin 100 mg, or canagliflozin 300 mg after a two week run-in period; 521 patients completed the follow-up. Canagliflozin 300 mg led to a placebo-adjusted A1c reduction of -0.60%, and canagliflozin 100 mg led to a reduction of -0.49%, from a relatively low baseline of 7.7%. This represented a very slight (and unsurprising) rise in A1c from the week 26 results (-0.70% and -0.57% for canagliflozin 300 mg and 100 mg, respectively). Canagliflozin also caused sustained reductions in fasting plasma glucose, body weight, and blood pressure. The incidence of serious adverse events, which are of particular concern in older patients, were similar for treatment and control groups. Although adverse events such as genitourinary infections were more prevalent in the treatment groups, this was consistent with the effect of canagliflozin on the broader type 2 diabetes patient population.

  • Study participants had a mean baseline A1c of 7.7±0.8%, mean fasting plasma glucose (FPG) of 157 mg/dl, and had lived with diabetes for an average of 12 years. Approximately 40% of participants were over 65 years of age; 74% took sulfonylureas or insulin, while 23% were on metformin monotherapy or other agents like DPP-4 inhibitors and GLP-1 analogs that are not associated with hypoglycemia (this balance had implications on the hypoglycemia results). Slightly under 3% were on diet and lifestyle change only at baseline.
  • Over 104 weeks, the canagliflozin groups showed reductions in A1c, FPG, body weight, and blood pressure as seen in the primary study. A1c and FPG levels rose slightly from week 26 levels, but weight continued to gradually decline, with a final weight loss of 2.3% and 3.2% for the canagliflozin 100 mg and 300 mg groups, respectively. This was somewhat surprising, as weight loss seen with SGLT-2 inhibitors generally plateaus in the space of six to eight months. During Q&A, Dr. Bode suggested that reduced eating in elderly patients might be the cause.
    • In a modified intent-to-treat analysis (last observation carried forward) analysis, canagliflozin 300 mg led to a placebo-adjusted A1c reduction from baseline of -0.60%, and a reduction in FPG of 23.3 mg/dl. Canagliflozin 100 mg led to a placebo-adjusted A1c reduction from baseline of -0.49%, and a reduction in FPG of 21.3 mg/dl. Approximately 42% of patients on canagliflozin 300 mg and 36% of patients on canagliflozin 100 mg achieved an A1c of below 7.0%.
  • Consistent with other studies, there were slight rises in both LDL and HDL cholesterol seen with canagliflozin. At the end of the study, patients on canagliflozin 100 mg or 300 mg experienced a 2.5% or 2.8% placebo-adjusted increase in LDL cholesterol. HDL cholesterol rose as well, and remained higher than placebo (3.6% and 4.8% for canagliflozin 100 mg and 300 mg, respectively). Also consistent with expectations, there was a slight reduction in blood pressure seen with canagliflozin.
  • The incidence of serious adverse events was similar across groups, although patients on canagliflozin had higher incidences of hypoglycemia, genital infections, and urinary tract infections. Discontinuation rates due to adverse events were similar for canagliflozin 100 and 300 mg compared to placebo (5% and 10% vs. 7%), and severe hypoglycemia only occurred once in the entire trial. Patients on canagliflozin had higher rates of UTIs (15% with canagliflozin 100 mg and 17% with canagliflozin 300 mg vs. 10% with placebo) as well as adverse events related to osmotic dieresis and volume depletion – these did not lead to many discontinuations, but given the risk of falls with elderly patients, providers should be careful to warn patients about this effect. As expected, the rates of genital mycotic infections were elevated with canagliflozin, especially in women.

Questions and Answers

Q: Why were people older than eighty excluded in a study of older people?

A: With people older than eighty, you get too many adverse events that cloud a study, so it is wise to cut the age range off at eighty.

Q: There seems to be an attenuation of the initial A1c improvement as time went on. You said that reflects the progressive nature of diabetes, but the SGLT-2 mechanism has nothing to do with beta cell failure.

A: The difference in A1c from 26 to 104 wks was only 0.1%. Also, the rise in A1c was the same in all groups. In the elderly, beta cell function declines, so you need to improve beta cell function or make up for it with medications that improve insulin resistance or replace insulin secretion.

Q: Regarding weight loss, can you identify patients which patients lost weight and which did not? Are there markers, like higher BMI?

A: I have no specific percentage data, but about 25% of people don’t lose weight. The cause of weight loss is calorie spillage in the urine, and if you compensate by eating, you won’t lose weight. But most people do lose weight.

Q: We saw some extension data in the past where A1c stayed same, and weight loss did not continue. This presentation was different from other presentations, since here A1c increased but weight continued to decrease. Why is that?

A: Beta cell continues to decline with age, which accounts for the increase in A1c, and older people often eat less, which may account for the continued weight loss.

Time Course of Changes in Glycemic Parameters and Body Weight in Patients Receiving Dapagliflozin as Add-on or as Initial Combination Therapy with Metformin (265-OR)

Arie Katz, MD (AstraZeneca, Wilmington, Delaware)

Dr. Arie Katz presented a post-hoc analysis of the time-course of glycemic and body weight changes during the first 24 weeks of treatment with AZ’s Forxiga (dapagliflozin) as an add-on or an initial combination therapy with metformin. Data was drawn from sub-populations of two large phase 3 trials for dapagliflozin. Treatment with dapagliflozin 10 mg as an add-on or initial combination therapy with metformin reduced A1c noticeably even after four weeks of treatment (-0.35% as add-on therapy [baseline = 8%] and -0.39% as initial combination therapy [baseline = 9%]), and the treatment difference continued to expand thereafter. Curves for fasting plasma glucose diverged by week one, and leveled off by week four at a ~15 – 25 mg/dl treatment difference. Surprisingly, body weight curves also appeared to diverge from week one – we imagine the initial loss could be due to diuresis. The curves continued to diverge, and were beginning to stabilize at an ~2.5 kg (~6 lb) treatment difference between dapagliflozin and placebo. Data on the time-course of SGLT-2 inhibitors’ effects will be particularly useful for insulin-treated patients, who walk a fine line between hyperglycemia and hypoglycemia when  starting a new agent. 

  • This 24-week post-hoc analysis drew data from two primary trials. The first trial compared dapagliflozin 10 mg vs. placebo as an add-on to metformin, while the second tested initial combination therapy with metformin XR and dapagliflozin 10 mg vs treatment with metformin XR alone. Baseline A1c was ~8% for the add-on trial and ~9% for the initial combination therapy trial.
  • There were no instances of severe hypoglycemia events in any of the studies, although genital infections and urinary tract infections were more frequent in patients on dapagliflozin.

Questions and Answers

Q: How many people achieved target A1c levels within 24 weeks?

A: We have no data for this analysis, but consistently more people achieved goals than in comparator groups.

Q: Comparing add-on therapy and the initial combination therapy, do you have any idea whether a similar percentage reached target A1c levels?

A: I’m not sure.

Temporal Changes in Urinary Glucose Excretion (UGE), Urine Volume (UV), and Plasma Volume (PV) in Subjects with Type 2 Diabetes Mellitus (T2DM) Treated with Canagliflozin (CANA) (263-OR)

Sue Sha, MD, PhD (Janssen Research & Development, Raritan, NJ)

Dr. Sue Sha presented results of a randomized, double-blind, placebo-controlled study (n=36 people with type 2 diabetes) characterizing changes in plasma volume and other measures of fluid/electrolyte imbalance on J&J’s Invokana (canagliflozin) after one week and after 12 weeks of treatment with high dose (300 mg) Invokana. As a reminder, a safety concern associated with Invokana is that its diuretic mechanism cause dangerous plasma volume depletion. This study demonstrated that Invokana decreased plasma volume by about 10% after one week of treatment, but that plasma volume returned to baseline after 12 weeks of treatment. However, some markers of plasma volume depletion did not return to baseline after 12 weeks (eGFR, BUN/Cr, and hematocrit), which may suggest that lasting effects of the initial plasma volume decrease may persist longer than the volume depletion itself.

  • Invokana lowered A1c (-0.6% placebo-adjusted) and fasting glucose (-29 mg/dl placebo-adjusted) over 12 weeks from a baseline A1c of 7.6-7.7% and baseline FPG of ~150 mg/dl. As the drug is designed to do, Invokana increased urinary glucose excretion by ~90 g/day at 1 week, and this persisted to 12 weeks.
  • The small increase in urine volume (~150 ml placebo-adjusted) observed at one week returned to baseline by week 12. Correspondingly, placebo-adjusted mean plasma volume decreased by 9.7% at week one (a 300 ml placebo-adjusted decrease), and essentially returned to baseline by week 12.
  • Although the plasma volume effect attenuated at week 12, a small reduction in eGFR, small increase in BUN/Cr, and small increase in hematocrit persisted through week 12, suggesting that some small reduction in plasma volume functionally remained despite the return of plasma volume to baseline.
  • Invokana lowered systolic blood pressure pretty significantly in this trial (placebo-adjusted reductions of 10 mmHg sitting and 8 mmHg standing at week one) that were maintained through week 12 (placebo-adjusted reductions or 13 mmHg sitting and 14 mmHg standing). Dr. Sha noted during Q&A that the mechanism for continued blood pressure lowering despite the attenuation of urine volume is not known. The results of this study will be published in Diabetes, Obesity, and Metabolism.

Questions and Answers

Q: You showed quite a dramatic drop in systolic blood pressure, yet diminishing of the urine volume and diuresis at week 12. Any idea how the blood pressure reduction can be sustained despite the reduction of diuresis? Do you have any sodium excretion data?

A: The mechanism is not well understood. It’s very similar to results from other diuretics such as thiazides. The patterns are very similar. There are several mechanisms hypothesized, but still not well understood.

Oral Presentations: ADA Presidents Oral Session

The Glucose Transporter SGLT 2 Is Expressed in Human Pancreatic Alpha Cells and Is Required for Proper Control of Glucagon Secretion in Type 2 Diabetes (386-OR)

Caroline Bonner, PhD (European Genomic Institute for Diabetes, Lille, France)

Dr. Caroline Bonner’s presentation revealed some striking new findings about SGLT-1 and -2 expression and activity in the pancreas. Her group found that SGLT-1/2 mRNA is found to an appreciable extent in pancreatic alpha cells, where it could be serving as a glucose sensor or playing some other role in endocrine signaling. The expression of SGLT-2 mRNA in particular is increased in obese individuals, but paradoxically falls in individuals with type 2 diabetes, apparently due to glucotoxicity. When Dr. Bonner’s team inhibited SGLT2 in human islets using siRNA or dapagliflozin, it led to a substantial stimulation of glucagon expression. Additionally, administering dapagliflozin to an insulin-resistant mouse model caused a significant three-fold increase in plasma glucagon. Until now, the prevailing opinion has been that SGLT-2 is only expressed in the kidney, and these findings (if confirmed in further study) have significant implications on the best way use the drug class. During Q&A, Dr. Bonner suggested (as other KOLs have recently) that GLP-1 agonists might be a complementary partner for SGLT-2 inhibitors due to their glucagon-suppressing activity. An attendee cast the findings in a positive light, suggesting that SGLT-2 inhibitors’ combined reduction in hyperglycemia and increase in glucagon secretion might be a great match for type 1 diabetes.

  • Two high-profile studies published earlier this year (Merovci et al. & Ferrannini et al., JCI 2014) demonstrated that SGLT-2 inhibitors led to significant increases in plasma glucagon and hepatic glucose production. We heard Dr. Ralph DeFronzo (University of Texas Health Science Center, San Antonio, TX) discuss this point at CODHy Latin America earlier this year (read our report). During that talk, Dr. DeFronzo suggested that this effect could mean that SGLT-2 inhibitors should be paired with a GLP-1 agonist, which could blunt the increase in glucagon – Dr. Bonner echoed this point during Q&A.
  • The current view of SGLT-2 inhibitors largely depends on the view that SGLT-2 transporters are only expressed to an appreciable extent in the kidney. As a result, Dr. Bonner noted, the expression and activity of the transporter elsewhere in the body is poorly understood.
  • Through immunofluorescence staining, Dr. Bonner’s group showed that SGLT-1 and SGLT-2 co-localize with glucagon/alpha cells in pancreatic islets. Further studies using cell purification showed that SGLT-1 and SGLT-2 mRNA is enriched in alpha cells in particular – while expression in beta cells is relatively low or nonexistent.
  • The expression of SGLT genes is deregulated over the course of the progression into type 2 diabetes. Using cross-sectioned samples from human islets, Dr. Bonner’s group showed that SGLT-1 and SGLT-2 mRNA expression increases as individuals progress into obesity and glucose intolerance. However, interestingly, SGLT-2 mRNA expression drops off dramatically as patients progress from glucose intolerance to diagnosed type 2 diabetes. Glucagon mRNA expression, in contrast, spikes substantially as patients progress from glucose intolerance to type 2 diabetes. The drop-off in SGLT-2 expression is remarkable, as SGLT-2 expression in the kidney is significantly up-regulated as patients progress to type 2 diabetes. When Dr. Bonner’s group exposed human islets in culture to varying levels of glucose, it appeared that the drop-off in SGLT-2 expression could be linked with glucotoxicity.
  • Both a knockdown in SGLT-2 using siRNA and the SGLT-2 inhibitor dapagliflozin (AZ’s Farxiga/Forxiga) stimulated glucagon gene expression. Dr. Bonner stated that the two means of blocking SGLT-2 activity also caused a increase in SGLT-1 expression, but in our view it was harder to conclude that from the data. The mechanism by which an SGLT-2 inhibitor might reduce SGLT-2 mRNA is not immediately clear; generally, counterregulatory mechanisms would cause increased expression of a gene if the protein it codes for is being inhibited. During Q&A, Dr. Bonner somewhat cryptically suggested that the action of SGLT-2 inhibition in the pancreas could be at the transcriptional level.
  • In a preclinical model of insulin resistance, Dr. Bonner’s team found that dapagliflozin administration led to an increase in plasma glucagon. The mouse model used was a C57blk lineage treated with S961 peptide. The increase in plasma glucagon (ostensibly from pancreatic alpha cells) was roughly three-fold, and was statistically significant.

Questions and Answers

Q: These results raise the questions of whether SGLT-2 inhibitors might be useful in type 1 diabetes, especially brittle type 1 diabetes. Aside from using them to blunt hyperglycemia due to the glycosuric effect, perhaps it could blunt hypoglycemia by increasing glucagon expression.

A: We are doing studies now in STZ models – that’s something we’re looking into.

Q: Regarding therapeutics, if in theory this is a prominent mechanism, you would expect to see more hypoglycemia with SGLT-2 inhibitors, but if anything you see a reduction in glucose. Does that suggest that this intriguing interaction with islets has less therapeutic impact than the class’ effect in the kidneys?

A: We have to wait and see, but I do think that this drug class is very good. Perhaps it might not be used best as a standalone therapy. It might be better used in combination with a GLP-1 analog to combat the oversecretion of glucagon.

Q: Why was it that you saw relatively low expression of SGLT-1 and SGLT-2 in lean individuals, but then the levels were much higher in obese patients, as were levels of glucagon? In healthy individuals there is no connection between SGLT and glucagon expression.

A: In the kidney, SGLTs are very well characterized as glucose transporters. We are not sure what the function is in the pancreas. Perhaps it is a glucose sensor, or is involved in endocrine signaling. We do know that expression is reduced in type 2 diabetes due to an overload of glucose.

Q: Dapagliflozin is meant to be an inhibitor of the SGLT-2 protein, but in vitro you saw reduced SGLT-2 mRNA expression. How can you explain that?

A: There is some evidence in our lab suggesting that SGLT-2 is inhibited at the transcriptional level.

Q: Generally, one of the strongest determinants of glucagon secretion is insulin secretion. The results we saw in the JCI articles were unexpected, as there was an increase in glucagon but no increase in insulin. Did you study the effect of changing insulin levels in the growth media?

A: We haven’t studied the effect of decreasing insulin concentration in the cell media.

Q: Is there anything known about glucagon levels in humans that have loss-of-function mutations for SGLT-2?

A: I do not know that off the top of my head.


Dual Add-On Therapy in Poorly Controlled Type 2 Diabetes on Metformin: Randomized, Double-Blind Trial of Saxagliptin + Dapagliflozin vs. Saxagliptin and Dapagliflozin Alone (127-LB)

J Rosenstock, L Hansen, P Zee, Y Li, W Cook, B Hirshberg, N Iqbal

This poster featured the first detailed phase 3 results of AZ’s SGLT-2 inhibitor/DPP-4 inhibitor fixed dose combination (FDC), saxagliptin/dapagliflozin (n=534). This 24-week study compared the FDC to each of its individual components: the SGLT-2 inhibitor Forxiga (known as Farxiga in the US; dapagliflozin) and the DPP-4 inhibitor Onglyza (saxagliptin). All patients were on background metformin. AZ released topline results from this trial in May, announcing A1c reductions of 1.5% on the saxa/dapa arm compared to 1.2% in the dapagliflozin arm and 0.9% in the saxagliptin arm (not quite additive, and definitely not synergistic, as had been hoped). From the poster, we learned that the baseline A1c in the trial was quite high: 8.9% in the saxa/dapa and dapagliflozin arms and 9.0% in the saxagliptin arm. See the table below for A1c reductions stratified by baseline A1c – across these subgroups, saxa/dapa consistently had greater A1c reductions than the other two groups, but as would be expected, the A1c-lowering effect was smaller in people with lower starting A1cs. Changes in body weight, as would be expected, seemed driven by the dapagliflozin component. The saxa/dapa group lost 2.1 kg (4.6 lb), the saxagliptin group had zero weight change, and the dapagliflozin group lost 2.4 kg (5.3 lb) from a baseline BMI of 32 kg/m2 (baseline weight not specified). No adverse events of interest (including major or minor hypoglycemia, urinary tract infections, or genital infections) were any higher on saxa/dapa compared to dapagliflozin or saxagliptin. Notably, urinary tract infections on saxa/dapa (1%) were actually numerically lower than on either saxagliptin (5%) or dapagliflozin (5%) (p-value not specified), and genital infections in the saxa/dapa arm (0.6%) were the same as the saxagliptin arm (0.6%) and lower than the dapagliflozin arm (6%) (p-value not specified). Overall, it appears that the saxa/dapa combination has an excellent glycemic, weight, and safety profile compared to other oral agents (e.g., metformin, DPP-4 inhibitors, SFUs, TZDs), but is perhaps not quite as different from SGLT-2 inhibitors as had been hoped at this stage – we look forward to seeing longer term data to assess duration, etc.

  • As reported in the topline release, both fasting plasma glucose (FPG) and two-hour postprandial glucose (PPG) were significantly better on saxa/dapa vs. saxagliptin, but not significantly different from dapagliflozin. Mean change in FPG was -38 mg/dl on saxa/dapa from a baseline of 180 mg/dl, -14 mg/dl on saxagliptin from a baseline of 192 mg/dl, and -32 mg/dl on dapagliflozin from a baseline of 185 mg/dl. Mean change in two-hour PPG was -80 mg/dl from a baseline of 242 mg/dl on saxa/dapa compared -36 mg/dl from a baseline of 256 mg/dl on saxagliptin and -70 mg/dl from a baseline of 246 mg/dl on dapagliflozin.
  • As reported in the topline release, more people on saxa/dapa achieved an A1c goal of <7% compared to saxagliptin or dapagliflozin. On saxa/dapa, 41% of patients got to goal, compared to 18% on saxagliptin and 22% on dapagliflozin. Given that patients had a relatively high starting baseline A1c, this is a pretty strong finding.


Saxa/dapa + met

Saxagliptin + met

Dapagliflozin + met

Baseline A1c <8% subgroup

Mean baseline A1c








Adjusted mean change from baseline




Baseline A1c ≥8% to <9% subgroup

Mean baseline A1c








Adjusted mean change from baseline




Baseline A1c ≥9% subgroup

Mean baseline A1c








Adjusted mean change from baseline




Fixed Dose Combinations of Empagliflozin/Linagliptin for 24 Weeks in Drug-Naïve Patients with Type 2 Diabetes (T2DM) (129-LB)

A Lewin, R DeFronzo, S Patel, D Liu, R Kaste, HJ Woerle, UC Broedl

The first of two phase 3 posters on Lilly/BI’s Jardiance/Trajenta (empagliflozin/linagliptin; “empa/lina”) presented the results of a study in 667 drug-naïve type 2 diabetes patients. There has been a great deal of excitement about the combination of DPP-4 inhibitors and SGLT-2 inhibitors, as the combination of an insulin-dependent and insulin-independent mechanism of action were thought to potentially yield additive or synergistic efficacy. While the efficacy seen with the high-dose FDC (empagliflozin 25 mg/linagliptin 5mg) was certainly strong for an oral compound, it fell well short of additive efficacy – in fact, it did not achieve a statistically significantly greater A1c reduction than empagliflozin monotherapy. Empa/lina 25 mg/5 mg yielded a mean A1c reduction of 1.08% after 24 weeks, compared to -0.67% with linagliptin 5mg (p<0.001) and -0.95% with empagliflozin 25 mg (p = 0.179), from a baseline of ~8%. From what we could tell, the results may have been dampened by a weaker-than-expected performance from the high-dose FDC group, as the lower-dose FDC (empa 10 mg/lina 5 mg) had a greater mean A1c reduction (-1.24%) and achieved statistically significantly better efficacy than its component monotherapy doses. A sub-analysis in patients with a baseline A1c at or above 8.5% yielded slightly more logical results: although the high-dose FDC once again did not achieve significantly greater A1c reduction than empagliflozin 25 mg monotherapy, it was not less effective than the lower-dose FDC. Although the results were not altogether negative (A1c reductions of over 1% for an oral are impressive), it was somewhat disappointing to not see truly additive efficacy with the combination.

  • The phase 3 study randomized 677 drug-naïve type 2 diabetes patients – 667 completed the trial. Patients were randomized to one of five treatments: empagliflozin 25 mg/linagliptin 5 mg, empagliflozin 10 mg/linagliptin 5 mg, empagliflozin 25 mg, empagliflozin 10 mg, and linagliptin 5 mg. The poster presented 24-week data, but the study will go on for a total of 52 weeks.
  • Empa/lina demonstrated solid efficacy and non-glycemic effects, but the high-dose combination performance was weaker than we might have expected. The high-dose combination fell short of the low-dose combination across categories, from A1c reduction (1.08% vs. 1.24%, respectively), the number of patients reaching an A1c goal of below 7.0% (55% and 62%, respectively), and weight loss (-2.0 kg [~4 lbs] and -2.7 kg [6 lbs]). While the A1c reductions seen with the low-dose combination were statistically significantly greater than those seen with its component monotherapies, the high-dose combination did not achieve statistical superiority over high-dose empagliflozin monotherapy, although it did over linagliptin monotherapy. The percentage of patients achieving a final A1c below 7.0% provided a more positive framing of the data than did the raw mean A1c reductions. 
  • As opposed to the metformin add-on trial (see 130-LB below), empa/lina did not demonstrate significant reductions in fasting plasma glucose relative to empagliflozin monotherapy. The combinations did achieve reductions in FPG over linagliptin monotherapy, on the order of ~23 mg/dl.
  • Interestingly, in this study, there was no clear increase in genital infections with empagliflozin, either as monotherapy or in combination with linagliptin. However, the number of overall events was quite small. Other adverse events were more or less balanced between groups.

Fixed-Dose Combinations of Empagliflozin/Linagliptin for 24 Weeks as Add-On to Metformin in Patients with Type 2 Diabetes (T2DM) (130-LB)

R DeFronzo, A Lewin, S Patel, D Liu, R Kaste, HJ Woerle, UC Broedl

The second of two phase 3 posters on Lilly/BI’s Jardiance/Trajenta (empagliflozin/linagliptin; “empa/lina”) presented the results of a study in 674 type 2 diabetes patients on background metformin. As opposed to the other empa/lina poster, the efficacy results here were more logical and consistently statistically significant. From a baseline of ~8%, the high-dose combination (empa 25 mg/lina 5 mg) arm achieved a mean A1c reduction of 1.19%, which beat out the 0.62% reduction with empagliflozin 25 mg and 0.70% reduction with linagliptin 5 mg – we found it interesting that linagliptin had numerically greater efficacy than empagliflozin in this trial. The lower-dose combination (empa 10 mg/lina 5 mg) achieved a mean A1c reduction of 1.08%, while the empagliflozin 10 mg arm achieved a mean reduction of 0.66%. Over 60% of patients on the high-dose combination achieved a final A1c below 7%, while only 33% of empagliflozin 25 mg patients and 36% of linagliptin patients achieved that goal. Weight loss appeared to be tied to the empagliflozin dose – empagliflozin 25 mg (with or without linagliptin) led to a mean weight reduction of ~3 kg (~7 lbs), while empagliflozin 10 mg (with or without linagliptin) led to about a pound less weight loss. Genital infections were more common with empagliflozin, but the relationship did not appear to be dose-dependent. Overall, the combination (and each of the component monotherapies) was well tolerated. Although the A1c reduction for the combination was not additive, the percentage of patients who achieved a goal of < 7.0% was fairly close to additive.

  • The study enrolled 686 type 2 diabetes patients on background metformin – 674 patients completed the study. Patients were randomized to one of five treatments: empagliflozin 25 mg/linagliptin 5 mg, empagliflozin 10 mg/linagliptin 5 mg, empagliflozin 25 mg, empagliflozin 10 mg, and linagliptin 5 mg. The poster presented 24-week data, but the study will go on for a total of 52 weeks.
  • Both empa/lina arms achieved a mean A1c reduction over 1%. From a mean baseline of ~8%, the empa 25 mg/lina 5 mg arm achieved a mean A1c reduction of 1.19%, which was significantly greater than the 0.62% reduction in the empagliflozin 25 mg arm and 0.70% reduction in the linagliptin arm. We found it interesting that linagliptin beat out empagliflozin – it seemed like the high-dose empagliflozin arm performed worse than might have been expected. The empa 10 mg/lina 5 mg arm achieved a mean A1c reduction of 1.08%, relative to a 0.66% reduction in the empagliflozin 10 mg arm. All comparisons between the combination arms and component monotherapies were highly statistically significant.
    • The poster also broke out mean A1c reductions for patients with a baseline A1c at or above 8.5%. From a mean baseline of ~9.1%-9.3%, the high-dose combination group experienced a mean reduction of 1.84%, the low-dose combination group experienced a mean reduction of 1.61%, the high-dose empagliflozin group experienced a mean reduction of 1.22%, the low-dose empagliflozin group experienced a mean reduction of 1.29%, and the linagliptin arm achieved a mean reduction of 0.99%. All comparisons between the combination arms and the component monotherapies were highly statistically significant.
    • Empa/lina helped more patients achieve an A1c goal of less than 7%. Approximately 62% of the high-dose combination group and 58% of the low-dose combination group achieved that goal, compared to 33% of the high-dose empagliflozin group, 28% of the low-dose empagliflozin group, and 36% of the linagliptin group.
    • Both empa/lina combinations achieved significantly greater fasting plasma glucose reductions than the component monotherapies. The difference between the high-dose combination and high-dose empagliflozin was 16 mg/dl, while the difference between the high-dose combination and linagliptin was 22 mg/dl.
  • The reduction in weight from baseline appeared to largely be a function of the empagliflozin dose, independent of combination with linagliptin. The high-dose combination arm and high-dose empagliflozin arm lost 3 kg (~7 lbs), while the low-dose combination and low-dose empagliflozin groups lost 2.5 kg (~6 lbs). The linagliptin group lost less than 1 kg (~2 lbs).
  • Adverse events were generally balanced between groups.

Sodium Glucose Cotransporter 2 (SGLT2) Inhibition with Emplagliflozin Reduces Microalbuminuria in Patients with Type 2 Diabetes (1125-P)

D Cherney, M von Eynatten, S Lund, S Kaspers, S Crowe, H Woerle, T Hach

This study pooled data from four phase 3 randomized, controlled trials to examine the effect of empagliflozin on urine albumin to creatinine ratio (UACR) in type 2 diabetes patients with microalbuminuria. Of the 2,477 patients who were randomized to placebo or empagliflozin in the 24 week trials, 458 patients on placebo (n=157), empagliflozin 10 mg (n=146), and empagliflozin 25 mg (n=155) started with microalbuminuria (UACR 30-300 mg/g). At week 24, patients on empagliflozin had significantly lower UACR for both the 10 mg (30% reduction; p <0.001) and 25 mg (25% reduction; p=0.004) dose. Finally, the treatment group did not have more adverse events than the control. These findings suggest that SGLT-2 inhibitors like empagliflozin could be renal-protective. Only one SGLT-2 inhibitor is being studied in a renal outcomes trial, J&J’s Invokana in CREDENCE.

  • The three groups of patients with microalbuminuria had similar baseline characteristics in terms of mean A1c, blood pressure, BMI, UACR, eGFR, age, and time since diagnosis of diabetes. Mean A1c levels for the placebo, empagliflozin 10 mg, and empagliflozin 25 mg groups were 8.13%, 8.26%, and 8.18%. Average blood pressure was 132.0, 132.8, and 133.6 mmHg, and average UACR was 61.9, 57.5, and 60.1 mg/g, respectively. Mean BMI was ~28-29 kg/m2 and mean age was ~55-57 years.
  • Patients with microalbuminuria treated with empagliflozin displayed a significant reduction in A1c and blood pressure. At week 24, patients on empagliflozin 10 mg and 25 mg showed significant placebo-adjusted reductions in A1c of 0.56% from a baseline of 8.26% (p <0.001) and 0.62% from a baseline of 8.18% (p <0.001). After accounting for placebo effects, blood pressure for these groups also decreased by 3.6 mmHg (p=0.011) and 3.5 mmHg (p=0.012), respectively.
  • Of all the patients in the four clinical trials, patients with microalbuminuria experienced a greater relative reduction in UACR. For the overall pooled population (n=2349), the percentage decrease in geometric mean UACR at week 24 for the empagliflozin 10 mg and 25 mg groups were only 10%, compared to 30% and 25% for the patients with microalbuminuria.

LX4211, a Dual Inhibitor of SGLT1/SGLT2, Reduces Postprandial Glucose in Patients with Type 2 Diabetes Mellitus and Moderate to Severe Renal Impairment (132-LB)

P Lapuerta, A Sands, I Ogbaa, P Strumph, D Powell, P Banks, B Zambrowicz

Dr. Pablo Lapuerta and colleagues present the results of a double-blind randomized, seven-day trial of Lexicon Pharmaceuticals’ LX4211 in 31 type 2 patients with moderate to severe renal impairment (mean baseline eGFR of 43 ml/min/1.73m2; other baseline characteristics detailed below). The participants were randomized to LX4211 400 mg once daily (n=16) or to placebo (n=15) in additional to their insulin therapy or oral anti-diabetic medication, with a treatment period of seven days. A standard breakfast meal was administered on days -1, 1, and 7, and data on glucose and GLP-1 were measured 15 minutes before the breakfast, as well as 1, 2, 2.5, 3, and 4 hours post-breakfast. LX4211 treatment resulted in statistically significant reductions in post-prandial glucose vs. placebo (which were evidence in patients with eGFR <45 ml/min/1.73m2), as well as reductions in fasting plasma glucose (average of -20 mg/dl; p=0.056). Participants on LX4211 also experienced statistically significant increases in post-meal total and active GLP-1 levels vs. those on placebo, which reflected the drug’s inhibition of gastrointestinal SGLT-1. The authors highlight that urinary glucose excretion was only slightly elevated in the LX4211 group (37 g/24 hours) compared a minor decrease in those on placebo (-1.4 g/24 hours; p<0.001). They also note that the PK results support the use of LX4211 400 in renally impaired patients, as there was no increase in LX4211 exposure for patients with eGFR <45 ml/min/1.73m2 relative to those with eGFR ≥45 ml/min/1.73m2. Based on these results, the authors conclude that LX4211 improves glycemic control in type 2 patients with renal impairment and call for longer-term clinical studies in this patient population.

  • At baseline, the participants had a mean age of 66 years, BMI of 34 kg/m2, duration of diabetes of 17 years, and eGFR of 43 ml/min/1.73m2. Seventeen percent of the patients were male, and 21% were Caucasian. The participants reported recent or concomitant use of insulin (61%), SFU (39%), metformin (29%), TZD (10%), and DPP-4 inhibitors (10%). As expected, the rates of common co-morbidities were high: hypertension (90%), hyperlipidemia (90%), neuropathy (42%), and cardiovascular disease (39%).
  • The tables below detail the change in fasting plasma glucose and urinary glucose excretion, stratified by eGFR level:

Table 1: Fasting Plasma Glucose


LX4211 vs. Placebo       


eGFR 45-59 ml/min/1.73m2



eGFR <45 ml/min/1.73m2



Mean for all patients



Table 2: Urinary Glucose Excretion





eGFR ≥45 ml/min/1.73m2




eGFR <45 ml/min/1.73m2




Mean for all patients




  • All adverse events were of mild to moderate intensity, and the frequency of adverse events was comparable between the LX4211 and placebo group:


LX4211         (n=16)

Placebo    (n=15)

Number of patients (%) with 1 treatment-emergent adverse event (TEAE)

7 (44%)

5 (33%)

Number of patients (%) with 1 drug-related TEAE

1 (6%)

3 (20%)

Efficacy and Safety of Twice-Daily Remogliflozin for the Treatment of Type 2 Diabetes Mellitus (1103-P)

WO Wilkison, AP Sykes, L Kler, J Lorimer, R O’Connor-Semmes, R Dobbins, S Walker

This poster was one of three that Islet Sciences presented on their novel SGLT-2 inhibitor remogliflozin etabonate. This poster featured results from a 12-week dose-ranging phase 2 trial testing twice-daily remogliflozin etabonate (n=336) at five doses (50 mg, 100 mg, 250 mg, 500 mg, and 1000 mg, each twice daily). The top dose tested, 1000 mg twice-daily, achieved an impressive 1.1% placebo-adjusted A1c reduction from baseline (~8.1%) – while it is challenging to compare trials, this is the highest A1c reduction achieved by an SGLT-2 inhibitor in a 12-week phase 2 dose ranging trial, and an Islet press release referred to the results as “best-in-class efficacy”. Weight loss was approximately 3 kg (~7 lbs) for both groups, and there were slight increases in LDL cholesterol and genital mycotic infections (11% in the 1000 mg dose group). Another poster (1102-P) presented efficacy and safety data on a once-daily formulation of remogliflozin etabonate – it found A1c and weight reductions that were statistically significant but less pronounced than those seen with twice-daily dosing, but a relatively low incidence of genital infections and no apparent increase in LDL (perhaps as a result of reduced nocturnal exposure). Based on these phase 2b results, Islet has developed a “biphasic” combination of immediate-release and delayed-release remogliflozin that it theorizes should preserve the efficacy of twice-daily dosing with the safety profile of once-daily dosing. The company also presented results at this ADA from a PK/PD study (1101-P) showing that the biphasic formulation achieved relatively high daytime exposure and low nighttime exposure, but only further testing will show if this characteristic results in the “best-in-class” safety and efficacy profile the company believes is possible.    

  • In the phase 2b study investigating twice-daily dosing of the original formulation: weight loss was somewhat dose dependent, with both the 500 mg and 1000 mg doses providing about 3 kg [6.6 lb] weight loss from baseline (baseline weights for these two groups were 87 kg [191 lbs] and 89 kg [196 lbs], respectively).
    • As with other SGLT-2 inhibitors, remogliflozin etabonate was associated in this trial with LDL cholesterol increases, although the LDL changes did not appear to have a dose-dependent relationship with the drug (patients experienced up to a mean 13.4 mg/dl increase on the middle [250 mg] dose, but only a 9.4 mg/dl increase on the highest [1000 mg] dose).
    • Rates of genital mycotic infections were somewhat dose dependent, with the highest rate (11%) seen in the highest, 1000 mg, dose.
  • Read our report on the topline phase 2b results for thoughts from Islet CEO James Green on the biphasic formulation.

Sodium Glucose Co-Transporter-2 (SGLT2) Inhibitor Empagliflozin (EMPA) in Type 1 Diabetes (T1D): Impact on Diurnal Glycemic Patterns (1051-P)

B Perkins, D Cherney, H Partridge, N Soleymanlou, H Tschirhart, B Zinman, R Mazze, N Fagan, S Kaspers, H Woerle, U Broedl, O Johansen

To further analyze the effects of Lilly/BI’s SGLT-2 inhibitor empagliflozin in type 1 diabetes patients, this research group used CGM to explore diurnal glycemic patterns in 40 type 1 diabetes patients on empagliflozin in a single-arm open-label pilot study lasting eight weeks. It was exciting to see measures like glucose variability as major endpoints in a trial, although the study’s findings on time in zone were surprising and we’ve asked for follow up on this. There were some reductions in overall glucose exposure (AUC) and improvements in glucose variability, but these were not statistically significant in many cases, which was surprising. Moreover, the initiation of empagliflozin therapy caused a reduction in basal insulin dose, but a slight and not statistically significant reduction in bolus insulin dose. We were surprised that the improvements in variability and stability were as modest as they appeared to be in this trial, although the small sample size (n = 40) and lack of a comparator group likely played a role. When empagliflozin therapy was withdrawn, patients saw a rebound in glucose AUC and variability to levels that were actually above baseline, demonstrating proof-of-effect. 

  • The study compared AGP profiles at baseline, “mid-treatment,” at the end of treatment, and following cessation of treatment. The AGP profiles combined the two weeks of data into a 24-hour profile.
  • For all CGM parameters, the study revealed a general trend of improvement from baseline to mid-treatment and end-of-treatment. However, most of these were not statistically significant. Parameters measured included glucose exposure measured as glucose AUC, glucose variability through blood glucose inter-quartile range, and glucose stability expressed as the mean hourly rate of change of glucose levels. In addition, the parameters also included time spent in hyperglycemia (>180 or >140 mg/dl), normoglycemia (70-140 mg/dl), and hypoglycemia (<70 mg/dl or <60 mg/dl).
  • Interestingly, after beginning treatment with empagliflozin, patients saw a statistically significant reduction in basal insulin (25.7 to 19.5 units, p<0.0001) but not bolus insulin (29.0 to 27.0 units, p = 0.19). SGLT-2 inhibitors are known primarily for their effect on postprandial glucose, so one would expect to see a reduction in bolus insulin. It is possible that the modest reductions in glucose variability were enough to allow patients to better titrate their basal insulin.
  • The positive trends in glucose control were seen both during the day and during the night. This finding does not support the theory that SGLT-2 inhibition during the day plays a disproportionate role in improving glycemia.
  • Demonstrating proof-of-effect, following cessation of treatment, glucose AUC rebounded to levels that were significantly higher than those seen at baseline despite a rebound in daily insulin dose. Glucose stability and glucose variability demonstrated patterns following the cessation of treatment. A rebound would be expected, and is a positive sign that empagliflozin treatment had some effect while it was being administered, but the fact that final glucose exposure was higher than at baseline could indicate the lasting presence of a counterregulatory response (possibly via the glucagon axis), although there is room for it to have been a chance finding.
  • This trial suggests avenues for future study. If empagliflozin acts in a more balanced fashion on both fasting and postprandial glucose in type 1 diabetes patients, then it would affect the titration of insulin when patients begin empagliflozin therapy. The relatively small size of the study of course limits the interpretability of its findings, but interest continues to grow in the use of SGLT-2 inhibitors in type 1 diabetes, and we expect to receive data from larger trials in coming years – AZ recently announced plans to begin phase 3 testing of dapagliflozin in type 1 diabetes patients in 2014. 

Symposium: ADA Diabetes Care Symposium – New Drug Therapies, Innovative Management Strategies, and Novel Drug Targets

Improved Glucose Control with Weight Loss, Lower Insulin Doses, and No Increased Hypoglycemia with Empagliflozin Added-On to Titrated Multiple Daily Injections of Insulin in Obese Inadequately Controlled Patients with Type 2 Diabetes

Julio Rosenstock, MD (Dallas Diabetes and Endocrine Center, Dallas, TX)

Dr. Julio Rosenstock presented results of a 52-week trial comparing placebo, Boehringer Ingelheim/Lilly’s empagliflozin 10 mg, and empagliflozin 25 mg in obese patients with inadequately controlled type 2 diabetes on high-dose basal-bolus insulin therapy, with or without metformin (n=566). Compared to placebo, empagliflozin led to lower A1c at 18 weeks (with minimal insulin titration in any group), lower A1c at 52 weeks (after insulin titration was allowed), lower daily insulin dose at 52 weeks, slight weight loss, and equivalent rates of hypoglycemia. The main side effect with empagliflozin was an increased rate of events consistent with genital infection, consistent with the SGLT-2 inhibitor class.

  • The EMPA-REG MDI trial enrolled adults with type 2 diabetes on multiple-daily-injection (MDI) insulin therapy with more than 60 units of insulin per day, A1c between 7.5% and 10.0%, BMI between 30 and 45 kg/m2, and eGFR at least 60 ml/min/1.73 m2 (i.e., no moderate or severe chronic kidney disease). At baseline, mean A1c was 8.3%, mean age was 57 years, mean weight was 96 kg, mean BMI was 35 kg/m2, mean insulin dose was 92 U/day, and mean number of mealtime insulin injections was 2.6-2.7/day. Besides insulin, patients were either treated with no oral antihyperglycemic therapy (29% of patients) or with at least 1,500 mg/day of metformin (71%). Patients were treated with placebo (n=188), empagliflozin 10 mg (n=186), or empagliflozin 25 mg (n=189)
  • Compared to placebo, empagliflozin 10 mg and 25 mg each led to lower A1c despite lower insulin doses (all comparisons statistically significant). For the first 18 weeks of the trial, clinicians were allowed to adjust insulin doses only by 10% or less, so that empagliflozin’s effect on blood glucose would be apparent. At 18 weeks, A1c decreased by 0.5%, 0.9%, and 1.0% in the placebo, 10 mg, and 25 mg groups, respectively. In weeks 19 through 40, insulin could be freely titrated to meet blood glucose targets of 100 mg/dl before meals and 140 mg/dl after meals. Then from weeks 41 through 52, insulin dose was held within 10% of week-40 levels. At 52 weeks insulin dose had been increased from baseline by 10.2 U/day in the placebo group, increased by only 1.3 U/day in the 10 mg group, and decreased by 1.1 U/day in the 25 mg group. The 52-week A1c decreases in the three groups were 0.8%, 1.2%, and 1.3%.
  • Patients using placebo had slight weight gain at one year (0.4 kg [0.88 lbs]), whereas empagliflozin led to slight weight loss (2.0 kg [4.4 lbs] in both the 10 mg and 25 mg groups. These weight changes occurred largely in the first 18 weeks of the trial, when the placebo group gained 0.34 kg (0.75 lbs), the empagliflozin 10 mg group lost 1.0 kg (2.2 lbs), and the empagliflozin 25 mg group lost 1.5 kg (3.3 lbs). The placebo-vs.-empagliflozin differences were statistically significant at both 18 and 52 weeks.
  • Rates of systolic blood pressure (SBP) declined slightly in all three groups; the change was statistically non-significantly greater with empagliflozin than placebo. From baseline SBP of 133-134 mm Hg, SBP at 52 weeks was 3-4 mm Hg lower in all three groups. Compared to placebo at 52 weeks, empagliflozin was associated with increased HDL cholesterol (statistically non-significant trend) and increased hematocrit (p-values not provided). Diastolic pressure at 52 weeks was statistically significantly lower with empagliflozin 25 mg than placebo, and triglyceride levels were statistically non-significantly higher with empagliflozin 10 mg than placebo. 
  • Patients treated with empagliflozin had higher rates of events consistent with genital infections (in both genders – consistent with the SGLT-2 inhibitor class), events consistent with urinary tract infections (in men), and dizziness; rates of hypoglycemia were similar in all three treatment groups. The percentages of patients who experienced one or more drug-related adverse events in the three groups  – placebo, empagliflozin 10 mg, and empagliflozin 25 mg – were 34%, 30%, and 40%, respectively. The percentages of patients with one or more adverse events of any kind were 90%, 86%, and 85%. The three groups were balanced in their rates of hypoglycemia ≤70 mg/dl and/or requiring assistance (58%, 51%, 58%) as well as hypoglycemia ≤70 mg/dl and requiring assistance (1.6%, 1.6%, 0.5%). Events consistent with genital tract infection were seen more in the empagliflozin groups overall (2%, 4%, 10%), in women (2%, 8%, 11%) and in men (1%, 1%, 8%). Events consistent with urinary tract infections were balanced overall (15%, 16%, 15%) and in women (29%, 24%, 26%). However, in men UTIs were more common with empagliflozin than with placebo (0%, 5%, 4%). Rates of dizziness were 1%, 3%, and 7%. No upper-urinary-tract infections (pyelonephritis) or urosepsis was recorded in the trial.

Questions and Answers

Dr. Robert Ratner (Chief Scientific and Medical Officer, American Diabetes Association): The investigators up-titrated the placebo group’s dose of insulin. What limited the ability to attain an equivalent A1c to the empagliflozin group? What is wrong with high insulin doses?

A: Nothing is wrong with high insulin doses; this population is just hard to treat. In most studies of this population you don’t get to A1c 7.0%. The thing is Dr. Ratner, you haven’t seen patients in quite some time, and that’s why you’re asking this question.

Q: Why do you think triglyceride levels increased in the empagliflozin 10 mg group?

A: I have no explanation for that.

Symposium: The Role of SGLT2 Inhibitors in the Treatment of Type 2 Diabetes

Physiology – The Role of Renal Sodium-Glucose Transport in Diabetes

Edward Chao, DO (UCSD, San Diego, CA)

Dr. Edward Chao presented on SGLT-2 inhibitors’ mechanism of action, and discussed a series of questions that remain unanswered about the new drug class. Some of these questions included: (i) why is the reduction in glucose seen with SGLT-2 inhibition not linearly dose dependent? (i.e.: why are there diminishing returns?); (ii) What role could SGLT-1 inhibition play in the treatment of type 2 diabetes? (iii) Why does SGLT-2 inhibition appear to be correlated with a rise in endogenous glucose production? (iv) What effect might SGLT-2 inhibition have on glucotoxicity? Read on below for Dr. Chao’s answers

  • In phase 3 trials, SGLT-2 inhibitors’ glucose-lowering efficacy seems to taper off at high doses. Although SGLT-2 is responsible for 80-90% of renal glucose reuptake, studies estimate that far less than 80-90% of all filtered glucose ends up in the urine in patients on an SGLT-2 inhibitor. To explain this, Dr. Chao suggested that blocking SGLT-2 inhibition could lead to upregulation in the expression of SGLT-1. If true, this phenomenon could increase the relevance of agents with more of an action on SGLT-1 in addition to SGLT-2.
  • SGLT-1 inhibition in the gut appears to results in delayed glucose reabsorption, which in turn yields multiple possible metabolic benefits. Inhibition of glucose uptake in the proximal small intestine would increase the delivery of glucose to L cells in the distal gut, where GLP-1 and PYY are produced. Therefore, SGLT-1 inhibition in the gut could act through the incretin axis. One study published last year (Polidori et al., Diabetes Care 2013) demonstrated that J&J’s SGLT-2 inhibitor Invokana (canagliflozin) increased GLP-1 and PYY levels, which Dr. Chao hypothesized was a results of residual SGLT-1 inhibition. 
  • Two papers published in JCI earlier this year appeared to show that SGLT-2 inhibition led to an increase in glucagon secretion and endogenous glucose production, which blunted the drugs’ glucose-lowering efficacy. Glucagon appears to drive at least part of this effect. Later at ADA, in the Presidents Oral Presentation Session, we saw preclinical data strongly suggesting that SGLT-2 inhibition did indeed lead to an increase in glucagon secretion. This finding, if true, would suggest that SGLT-2 inhibitors’ efficacy could be greatly increased when paired with an agent like a GLP-1 agonist or glucagon receptor antagonist that could blunt the counterregulatory glucagon response.
  • One very positive finding on SGLT-2 inhibitors in the recent literature was that AZ’s Forxiga (dapagliflozin) appears to improve muscle insulin sensitivity. The study (Merovci et al., JCI 2014) found that the reductions in plasma glucose caused by dapagliflozin treatment were associated with an increase in muscle glucose uptake. In Dr. Chao’s view, this finding constitutes early evidence of the glucotoxicity hypothesis, as it shows that a reduction in glucose through glycosuria was able to improve one of the core deficits of type 2 diabetes.
  • After discussing answers to the aforementioned unanswered questions, Dr. Chao proposed a set of additional forward-looking questions about SGLT inhibition:
    • What is the therapeutic role of SGLT-1/SGLT-2 dual inhibition?
    • Do SGLT-2 inhibitors increase caloric intake?
    • What is the mechanism of the increased endogenous glucose production seen with SGLT-2 inhibitors?
    • Could adding incretin mimetics to SGLT-2 inhibitors have a synergistic effect?

Efficacy – A Review of Current Evidence

Clifford Bailey, PhD (Aston University, Birmingham, United Kingdom)

Dr. Bailey reviewed clinical data with the SGLT-2 inhibitors, providing an overview of trends within the class. Following a review of the history of the class, Dr. Bailey summarized data from phase 2 and phase 3 trials of dapagliflozin (AZ’s Farxiga), canagliflozin (J&J’s Invokana), empagliflozin (Lilly/BI’s Jardiance), and LX-4211 (Lexicon), noting their durable A1c declines, weight loss, and blood pressure reductions. Referencing a plot of A1c decline and weight loss across the trials, he emphasized the tight confidence intervals within the class, suggesting that virtually everyone was responding to the drugs. He reinforced this via FDA Advisory Committee data from canagliflozin, highlighting that amongst subgroup analyses baseline A1c was the only significant associative factor with A1c decline. While he lauded the A1c declines, he suggested “what keeps people particularly interested” in the class is the weight loss, with all candidates demonstrating similar 2-3 kg (4-7 lbs) reductions in body weight. He concluded by noting their compatibility with other diabetes therapies, indicating that provided adequate renal function effectively the drugs could be used anywhere along the natural history of type 2 diabetes.

Questions and Answers

Q: Why does the weight loss bottom out? You’re continuing to lose glucose and everything stables out.

A: I wish I knew the answer to that. It’s probably a combination of improved metabolic efficiency and as your glucose comes down you may be susceptible to feeling hungrier. Although we’ve not been able to get accurate data on this I think folks might be eating more.

Q: Any cases of overdoses in people?

A: I’m not aware of anyone, and I’m not too sure there would be any detriment as there have been studies exploring fairly large doses.

SGLT-2 Inhibitors: Safety and Adverse Effects

Lawrence Leiter, MD (University of Toronto, Toronto, Canada)

Dr. Lawrence Leiter provided a category-by-category overview of the possible safety risks associated with SGLT-2 inhibitor therapy. He ended the presentation by noting that there is not yet enough clinical evidence on the relatively new drug class to fully characterize each of the potential safety risks he outlined, and that the results of the cardiovascular outcomes trials for SGLT-2 inhibitors will be important to see. He pointed out that EMPA-REG OUTCOME (the CVOT for Lilly/BI’s empagliflozin) is progressing faster than expected – earlier this year the company moved the estimated primary outcome date up from early 2018 to early 2015.  

  • Hypoglycemia: Dr. Leiter noted that mechanistically, one would not expect SGLT-2 inhibitors to cause a large increase in hypoglycemia. The slight increase in hypoglycemia seen in some clinical trials has generally been due to the concomitant use of sulfonylureas or insulin.
  • Genital infections: This is a relatively well-characterized side effect of SGLT-2 inhibitors. Detailed data on AZ’s Forxiga (dapagliflozin) seemed to show that the incidence of genital infections did not increase in a dose-dependent manner, suggesting that it may be more of a threshold effect. Other studies (Johnsson et al., J Diabetes Complications 2013) show that the risk of infections is higher in younger individuals, individuals with a history of recurrent infections, and in obese individuals.
  • Urinary tract infections: The signal hear is far less clear than for genital infections. Where increases have been seen, Dr. Leiter noted that there do not appear to be increases in more dangerous upper tract infections.
  • Blood pressure and volume depletion: This is another family of effects that is directly linked to SGLT-2 inhibitors’ mechanism of action, and that is generally not a decision-maker (or breaker) for providers or patients. Risk factors for hypotension or volume depletion are concomitant use of loop diuretics and patients with low baseline eGFR.
  • Bone health: Most studies investigating bone health with SGLT-2 inhibitors have not shown changes in bone biomarkers beyond what would be expected with slight weight loss. The one slight exception is canagliflozin, which appears to be associated with a slight increase in fractures. Dr. Leiter noted that the mechanism behind this signal is unclear, especially in the absence of changes in biomarkers. He creatively postulated that the cause might be more falls, which (in our view) could be due to postural hypotension, especially since most falls happened relatively soon after initiation of therapy.
  • Malignancies: A slight imbalance in bladder cancer was one of the major points of discussion in the most recent FDA Advisory Committee meeting for AZ’s Forxiga (dapagliflozin). Dr. Leiter stated that it is hard to imagine a mechanism that would explain the signal, especially since the majority of the cases were early after the initiation of treatment (most tumors take a long time to develop). He concluded that malignancy is probably not a real safety issue, although data from long-term outcomes studies will be more conclusive.
  • Cardiovascular safety: The entire SGLT-2 inhibitor class appears to be associated with a slight increase in LDL, with perhaps a slightly more prominent signal seen with canagliflozin 300 mg. However, Dr. Leiter pointed out that the risk must be interpreted in the context of simultaneous reductions in blood pressure and weight.

Questions and Answers

Q: There is some evidence suggesting that with chronic use of SGLT-2 inhibitors, the substrate balance for energy usage shifts from glucose to lipids. Could an SGLT-2 inhibitor prompt ketosis? Do we need more data on lipid metabolism?

A: I think it is possible that these agents may alter the balance of fuel utilization. We need to understand the mechanism behind the small increases in LDL.

Q: On lipids, could the mechanism behind the increase be an increase in hematocrit?

A: One possibility may be hematocrit, yes. Another mechanism put forth, though not currently supported by much data, is that these agents may cause patients to eat more and alter their pattern of macronutrient consumption. That is a more speculative theory.

SGLT2 Inhibitors in Diabetes – Unanswered Questions

Zachary T. Bloomgarden, MD (Mt. Sinai, New York, NY)

Dr. Zachary Bloomgarden concluded the session by addressing remaining questions surrounding the SGLT-2 inhibitor class, beginning with “why aren’t the SGLT-2 inhibitors more potent?” He noted the possibility of renal glucose uptake via other SGLT/GLUT transporters or a compensatory rise in endogenous glucose production. With glucagon known to play a role in the kidney, Dr. Bloomgarden posited the latter effect could be mediated by an increase in renal gluconeogenesis following SGLT-2 inhibitor use. He then moved to ask, “Are the SGLT-2 inhibitors nephrotoxic or nephroprotective?” highlighting animal studies supporting a reduction in glomerular mesangial expansion and albuminuria after SGLT-2 inhibitor administration. Clinically, he highlighted the recently-initiated CREDENCE study, which is investigating the effects of canagliflozin in patients with stage 2-3 chronic kidney disease – given the level of unmet need in the population of patients with diabetes and renal impairment, evidence of even a modest protective effect could have sweeping effects for the class.

  • Dr. Bloomgarden began by asking “Why aren’t the SGLT-2 inhibitors more potent?” While SGLT-2 is responsible for 80-90% of the 160-180 g glucose transported daily, SGLT-2 inhibition only induces a maximum of 50-80 g of glycosuria. Dr. Bloomgarden posited a number of theories to explain this, including incomplete exposure of the drug to the SGLT-2 receptor (due to high binding of the agents to plasma proteins or secretion of the drug distal to the site of reabsorption) and possible compensation of SGLT-1 or other SGLT/GLUT transporters for the increased distal tubular glucose delivery.
  • Interestingly, there may be a compensatory response to SGLT-2 inhibition as well, leading to increased endogenous glucose production. Highlighting recent literature (Ferrannini et al., JCI 2014), Dr. Bloomgarden noted increased endogenous glucose production and reduced glucose utilization in type 2 diabetes patients after single dose and four-week administration of empagliflozin, blunting the decline in fasting and prandial glycemia. This was further supported by another recent paper (Merovci et al., JCI 2014), which demonstrated a paradoxical increase in endogenous glucose production in type 2 diabetes patients following two weeks of treatment with dapagliflozin, with an associated increase in glucagon levels. With glucagon known to play a role in the kidney, Dr. Bloomgarden posited this could be mediated by an increase in renal gluconeogenesis following SGLT-2 inhibitor use.
  • Dr. Bloomgarden next considered whether SGLT-2 inhibitors are nephrotoxic or nephroprotective. Looking again to the literature (Arakawa et al., Br J Pharmacol 2001), he noted a reduction in glomerular mesangial expansion and albuminuria in mouse models of type 2 diabetes and obesity following administration of the experimental SGLT-2 inhibitor T-1095, suggesting possible nephroprotective effects. He went on to note numerous studies presented at ADA this year supporting possible benefit to renal function, including 532-P, which indicated suppression of glomerular mesangial expansion, interstitial fibrosis, and diabetes-induced oxidative stress in diabetic mouse models following dapagliflozin treatment. Clinically, he highlighted the CREDENCE study, which is investigating the effects of canagliflozin in patients with stage 2-3 chronic kidney disease.
  • Dr. Bloomgarden concluded with a number of additional remaining questions he was unable to address in his talk. These included 1) “What is the effect of SGLT-2 inhibitors on LDL cholesterol?” 2) “Do the drugs reduce bone mineral density and through what mechanism?” and 3) “Do the drugs alter glucose sensing in the central nervous system, causing increased appetite?”

Corporate Symposium: The Kidneys in Type 2 Diabetes – Disturbed Glucose Homeostasis to Mechanism-Based Therapy (Supported by an educational grant from AstraZeneca)

Case Presentation 1

Lawrence Blonde, MD (Ochsner Medical Center, New Orleans, Louisiana)

Following an introduction by Dr. Daniel Einhorn (UCSD, La Jolla, CA), session moderator Dr. Lawrence Blonde  presented a case study of a 59-year-old Caucasian patient with a current A1c of 8.7%. Dr. Blonde argued that dual therapy, in addition to lifestyle change, would be more effective for this patient. He then provided a broad overview of the two approved SGLT-2 inhibitors in the US, AZ’s Forxiga (dapagliflozin) and J&J’s Invokana (canagliflozin), noting that both drugs promote reductions in A1c, body weight, and blood pressure. He explained that patient selection for SGLT-2 inhibitors requires consideration of renal function and susceptibility to hypotension to avoid undue risk. Dr. Blonde concluded with the reminder that “diabetes is not just glucose.” He emphasized the importance of targeting multiple metabolic goals in each patient such as blood pressure, lipid levels, weight, and A1c (all of which SGLT-2 inhibitors can help patients achieve). 

Case Presentation 2

Richard Pratley, MD (Florida Hospital Diabetes Institute, Orlando, FL)

“These are really exciting times in diabetes,” Dr. Richard Pratley asserted to begin his case study presentation. His patient was a 70-year-old retired African American woman who lived alone, had a family history of cardiovascular disease, and had been diagnosed with type 2 diabetes for thirteen years. She took metformin and insulin glargine at bedtime, had an A1c of 7.9%, and approached her doctor because she had been waking up in the middle of the night sweating and with a headache, possible signs of nocturnal hypoglycemia. Dr. Pratley suggested that a combination of oral agents may be best for this patient, and presented several studies on such combinations. He noted that fixed-dose combinations of SGLT-2 inhibitors and DPP-4 inhibitors are currently in development. 

Panel Discussion

Dr. Pratley: Can we say anything about relative efficacy and safety among SGLT-2 inhibitors?

Dr. Blonde: We will see head-to-head studies in the future. My prediction would be that the drugs will have pretty similar efficacy and we would expect that from the mechanisms.

Q: We talked about how there is an improvement in weight, but what I didn’t show was the time course of that improvement. As with most therapies associated with weight reduction, weight loss plateaus. Given that the mechanism of SGLT-2 inhibitors is the loss of calories through glucose excretion in the urine, should there be a plateau?

Dr. Richard Pratley (Florida Hospital Diabetes Institute, Orlando, Florida): We don’t know why there is a plateau. While we see weight going down, there isn’t much further weight loss after 24 weeks. There must be some sort of compensation connected to appetite, because it’s unlikely to be that the patient is exercising less.

Dr. Lawrence Blonde (Ochsner Medical Center, New Orleans, Louisiana): Is it possible that there is some change or compensation in terms of energy utilization in the body?

Dr. Pratley: In uncontrolled diabetes, there is increased energy expenditure. When you improve glucose levels, you decrease that metabolic rate, which is one possible explanation for the effect.

Dr. Blonde: I think many people suspect that there is some compensation happening involving greater caloric consumption. Can you elaborate on the increased endogenous glucose production and potential for and increase in glucagon?

Dr. Pratley: As you know, the regulation of glucagon is complicated. It’s related to glucose level and it’s also related to insulin levels through paracrine action. We see lower insulin levels with SGLT-2 inhibitors but we still don’t know what drives the higher glucagon.

Dr. Blonde: If that glucagon effect persists, it would make the combination of a DPP-4 inhibitor and an SGLT-2 inhibitor more attractive.

Dr. Pratley: It would, if you operated according to the pathophysiology. If you add an SGLT-2 inhibitor onto a DPP-4 inhibitor, the reductions in A1c are about what you see in other clinical scenarios. You saw a better reduction with the two drugs than with either drug alone, but it wasn’t additive, which is typical of combination therapies. It’s a good combination but it’s not necessarily better than other combinations.

Dr. Blonde: Another combination that we have less data about is the combination of GLP-1 analogs and SGLT-2 inhibitors, which might cause larger A1c, weight and blood pressure effects. Only small pieces of data have become available.

Dr. Pratley: It is a logical combination that is very attractive, and may work, but we have no clinical data that speaks to its efficacy yet.

Q: What is the potential effect of SGLT-2 inhibitors on chronic kidney disease in type 2 diabetes?

Dr. Pratley: Things could go either way. It’s not unusual to see a small increase in creatinine at the beginning, but it’s transient. The real question is whether the effect is good or bad on the kidneys. There’s some data now about renal protective effects. This is an area we’ll hear more about in the future.

Q: Any idea as to why SGLT-2 inhibitors could be associated with a modest decrease in LDL-C?

Dr. Pratley: I don’t think we know the explanation.

Q: What is the mechanism that connects the data about hypoglycemia and adverse events, particularly the cardiovascular events? The ACCORD researchers couldn’t correlate hypoglycemia with the adverse events that occurred.

Dr. Pratley: There are two ways to look at it. One is that hypoglycemia causes an increased risk of adverse events. The other way to look at it is that these are reciprocations and are affected by the same factor. We don’t know yet.

Q: You pointed out that this was an insulin-independent mechanism and we saw that there was good efficacy across the spectrum of type 2 diabetes patients, from those who took it as monotherapy to those who took it as an add-on to insulin. What about giving an SGLT-2 inhibitor to someone with type 1 diabetes? Do you think it would be beneficial?

Dr. Pratley: A lot of people have wondered about that. There is no reason that it wouldn’t work in the context of type 1 diabetes. Those trials are ongoing. The rationale for that is similar to insulin-treated type 2 patients. It comes down to the ability to get by with lower doses of insulin, produce weight loss in patients who often need it, and achieve better post-prandial control.

Q: You showed some data about improved beta cell function with canagliflozin. Has any other method besides HOMA been used to show improvements in beta cell function with SGLT-2 inhibitors? What about the other effects of SGLT-2 inhibitors, like lowering glucotoxicity and improving insulin sensitivity?

Dr. Pratley: A number of published studies, including those by the Ferrannini and DeFronzo groups, have shown that effects on insulin sensitivity and insulin secretion are consistent across SGLT-2 inhibitors.

Q: Over time, will the excess glucose going through the tubules possibly cause nephropathy?

Dr. Pratley: Remember the rare patients that have a genetic condition with a defect in SGLT-2? They don’t appear to progress to nephropathy.

Q: Are there any medication interactions besides with diuretics?

Dr. Blonde: I’m not aware of any drug-drug interactions, but you raise an important issue about loop diuretics that were associated with more volume-related issues. For those with reasonable renal function, loop diuretics aren’t even a good solution for hypertension. But what about people with heart failure?

Dr. Pratley: That’s a good question. Less hypoglycemia is good for these patients, and some diuresis is probably good. We don’t know if lowering glucose levels in patients with heart failure leads to positive outcomes. However, there are very few patients with heart failure that are not on loop diuretics, so we don’t usually have to worry about it.

Product Theaters

Glucose Removal: An SGLT-2 Inhibitor Treatment Option for Adults with Type 2 Diabetes (Sponsored by AstraZeneca)

Ralph DeFronzo, MD (University of Texas Health Science Center, San Antonio, TX)

Dr. Ralph DeFronzo attracted a standing-room-only crowd for his presentation on the role of SGLT-2 inhibition and specifically on AZ’s Farxiga (dapagliflozin). After a review of SGLT-2 inhibitors’ mechanism of action, Dr. DeFronzo presented data from several clinical trials of dapagliflozin as monotherapy and as an add-on to multiple other drug classes. Dapagliflozin reduced A1c, body weight, and blood pressure in all the trials Dr. DeFronzo referenced, with little incidence of hypoglycemia. He concluded by reviewing the safety warnings associated with the drug, noting increased rates of genital mycotic infections in men and women, UTIs, increases in cholesterol, and the small bladder cancer imbalance seen with dapagliflozin’s pooled phase 3 data.

  • Dr. DeFronzo reviewed what he considered to be the three most important clinical studies of dapagliflozin, noting the significant A1c reductions compared to placebo as both monotherapy and add-on therapy. As initial combination therapy, dapagliflozin 5 mg added to metformin XR provided an A1c reduction of 2.1% in 24 weeks from a relatively high baseline of 9.2% (n = 146) with concurrent reductions in body weight of six pounds. As a monotherapy, dapagliflozin demonstrated less extreme but still significant reductions in both A1c among patients with a baseline A1c of roughly 9.1%. As an add-on treatment to metformin, dapagliflozin 5 mg provided a reduction in A1c of 0.7% (n=137; baseline:8.2%) and 0.8% for the 10 mg dose (n=135; baseline: 7.9%) in addition to a reduction in body weight and significant reduction in systolic blood pressure. Finally, Dr. DeFronzo noted that while dapagliflozin plus metformin had comparable reductions in A1c to glipizide plus metformin, the incidence of major and minor hypoglycemia was less for patients treated with dapagliflozin plus metformin compared to patients treated with glipizide plus metformin.
  • Dr. DeFronzo addressed the adverse events associated with dapagliflozin, noting increased rates of genital mycotic infections, UTIs, and cases of bladder cancer. Dr. DeFronzo emphasized that there is no dose response effect in genital mycotic infections or UTIs, but that a history of previous infections increases a patient’s chance of experiencing these outcomes. He spent a significant amount of time explaining the prevalence of cases of bladder cancer among the pooled clinical trial results. Ten cases of bladder cancer included nine cases in the dapagliflozin group and one case in the placebo group. If people with preexisting hematuria and risk factors for bladder cancer were excluded, this narrows down the number of cases to four in the dapagliflozin group, which Dr. DeFronzo emphasized was insufficient data to understand the meaning of these numbers. He reiterated that patients with moderate to severe renal impairment (eGFR <60 mL/min/1.73m2) should not use the drug.

Other Oral, Non-Incretin, Oral Anti-Diabetic Agents

Oral Presentations: Diabetes Complications—From Head to Toe

Risk of All-Cause Mortality Varies amongst Sulfonylureas: A Network Meta-analysis (335-OR)

Scot H. Simpson, PharmD (University of Alberta, Edmonton, AB, Canada)

Dr. Scot Simpson presented findings from a network meta-analysis of 16 different studies looking at mortality rates among users of different types of sulfonylureas. The specific sulfonylureas considered were gliclazide, glimepiride, glipizide, and glyburide. Glyburide was the most commonly used sulfonylurea and served as the comparator for the other compounds. Dr. Simpson and his team found that glipizide’s all-cause mortality risk was identical to that of glyburide (0.98 risk ratio), while glimepiride (0.82 risk ratio) and gliclazide (0.65 risk ratio) mortality risks were lower than that of glyburide. Dr. Simpson concluded that glimepiride and gliclazide were significantly different from glyburide and glipizide, but acknowledged the many limitations of this study and the need for a clinical trial to further explore these results. In the meantime, Dr. Simpson suggested that clinicians use these differences when selecting a sulfonylurea.

  • Current guidelines in North America and Europe recommend sulfonylureas as second-line therapy for type 2 diabetes patients when metformin is insufficient. While they have side effects of hypoglycemia and weight gain, in the last 40 years, concerns of elevated risk of cardiovascular events and mortality associated with sulfonylureas have come up. Dr. Simpson’s research sought to discover whether different types of sulfonylureas had different mortality risks.
  • Dr. Simpson’s team identified 16 directly relevant studies (those containing head-to-head comparisons of sulfonylureas) using Medline and EMBASE. Five of them were randomized controlled trials (n=2,729) and the rest were cohort studies (n=151,011). Only studies that looked at type 2 diabetes patients, used at least two different sulfonylureas, and reported all-cause or cardiovascular mortality (or myocardial infarction rates) were included. Dr. Simpson chose to analyze data comparing the most popular sulfonylureas: gliclazide, glipizide, glyburide (all second generation), and glimepiride (third generation).
  • Using a network meta-analysis design, Dr. Simpson was able to estimate pairwise comparisons of all-cause mortality. Glipizide and glyburide had equal risks of mortality, while gliclazide and glimepiride showed significantly lower risks compared to glyburide. All-cause mortality risk ratios used glyburide as a comparator.  Gliclazide had the lowest risk of mortality among the sulfonylureas. Values are given in the table below:


Pairwise Meta-Analyses Risk Ratio

Network Meta-Analyses Risk Ratio










  • Dr. Simpson recognized limitations of this study (selection bias, various confounds, observational data, unreported data in potential studies) and the need for a randomized clinical trial to test causation. Nevertheless, he recommended clinicians to “consider these possible risk differences when selecting a sulfonylurea.”

Questions and Answers

Q: I noticed that the relatively newer agents were associated with lower mortality. My concern is that even with RCTs and cohort studies that looked at these agents, there is a whole trend with cardiovascular risk protection, more use of statins, more use of blood control. How much of that is confounded with your study? 

A: There are many uncontrolled confounding factors. We did do a sensitivity analysis where we had some of the adjusted correlational data and associations. But there’s certainly some selection bias.

Symposium: Cardiovascular Outcomes in Recent Diabetes Trials

Aleglitazar (AleCardio) Study

Jean-Claude Tardif, MD (University of Montreal, Montreal, Canada)

Dr. Tardif recapped the results of the AleCardio trial, the phase 3 cardiovascular outcomes trial for Roche/Genentech’s dual PPAR-alpha/gamma agonist, aleglitazar. As a reminder, AleCardio was terminated early when an unplanned futility analysis at the accrual of 55% of total events had a HR of 1.01 with a 1% chance of achieving superiority by trial completion (results initially presented at ACC 2014). Dr. Tardif reviewed results of data prior to the database lock on December 17, 2013 (also published in JAMA), concluding that aleglitazar reduced A1c and improved triglyceride and HDL-C levels but without benefit to cardiovascular outcomes. Noting the difficult track record of the PPAR class, he suggested that AleCardio overall reflected the challenges in the development of PPAR drugs given the unique pattern of gene modulation, complex effects on the metabolic pathways, and unpredictable therapeutic profiles – Dr. A. Michael Lincoff (Cleveland Clinic, Cleveland, OH) sadly had similar sentiments at ACC 2014 in his presentation of the data, suggesting in the Q&A that AleCardio may mean the end of new PPAR agonist trials.

Questions and Answers

Dr. Vivian Fonseca (Tulane University, New Orleans, LA): You had similar side effects in SYNCHRONY and your approach was somewhat unusual picking for cardiovascular versus glycemic effects. Was there some rationale to this decision?

A: With aleglitazar if you look at biomarker profile there was a robust benefit to A1c and HDL and triglyceride levels. Muraglitazar did have a profile in the wrong direction for CV events; tesaglitazar stopped due to renal dysfunction. We believed the renal profile for aleglitazar was more favorable than tesaglitazar. So the decision was largely based on the mentioned robust effects – we thought this drug was likely to help patients with ACS and diabetes.

Q: Do you have information on the blood pressure? Do you have corroborated evidence that the rise in creatinine was a result of eGFR as opposed to increased creatinine production?

A: There was a rise in blood pressure, as well as a demonstrated reversible effect on eGFR.

Symposium: China Medical Tribune Symposium – Progress in Diabetes in China – From Molecular Medicine to Clinical Trials

Search for Genetic and Clinical Predictors of Sulfonylurea Treatment Failure in Type 2 Diabetes

Linong Ji, MD (Peking University People’s Hospital, Beijing, China)

Dr. Linong Ji presented a sub-analysis of a prospective cohort study examining the predictive ability of genetic and clinical markers for sulfonylurea treatment failure in type 2 diabetes patients (n=747). A combination of baseline disposition index (DI) and initial treatment response was found to be a more effective predictor of sulfonylurea treatment failure than genetic markers or any single clinical marker. The combination of DI <6.9 and poor initial treatment response was found to predict treatment failure 82% of the time, compared to possessing multiple genetic factors associated with sulfonylurea failure (46%) or good initial treatment response combined with DI <6.9 (49%). Given that sulfonylureas have several dangerous side effects, using these clinical predictors may allow providers to avoid prescribing sulfonylureas to patients who have a high risk of sulfonylurea treatment failure or frequent hypoglycemia.

  • The prospective study followed 747 type 2 diabetes patients from a double-blind, randomized, controlled clinical trial with glibenclamide therapy against traditional Chinese medicine. No difference was found between the two arms of the study, so the subjects were pooled to look at sulfonylurea failure over the course of a year. Treatment failure was defined as a FPG >126 mg/dl measured twice, four weeks after the maximal dose or maximal tolerated dose of glibenclamide was reached.
  • The study examined both genetic and clinical variables to search for predictors of sulfonylurea treatment failure. Genotyping was conducted on patients to search for genetic factors that affect type 2 diabetes susceptibility, MODY, beta cell growth and function, obesity and insulin resistance, and sulfonylurea metabolism. Clinical variables included duration of diabetes, demographics, initial FPG response in the first four weeks, and biological factors such as A1c, disposition index, and HOMA. Poor initial response was defined as a FPG reduction <10% from baseline after four weeks of treatment.
  • As would be expected, patients with more alleles associated with sulfonylurea failure were more likely to experience sulfonylurea treatment failure. Dr. Ji examined eight genes that have been shown to impact sulfonylureas and divided patients into three groups based on the number of alleles they possessed: low risk (fewer than three), middle risk (three to seven), and high risk (over seven). High risk patients had more treatment failure (p=0.043).
  • Of the clinical variables, combining initial response and baseline disposition index (DI) was the best predictor of sulfonylurea success or failure. Dr. Ji did not define how he measured DI during the presentation. Dividing patients into six subgroups based on poor or good initial response as well as baseline DI <6.9, 6.9≤DI<14.0, or DI ≥14.0 revealed that 82% of patients with baseline DI <6.9 and poor initial response experienced sulfonylurea treatment failure, whereas the next best predictor combination was good initial response and baseline DI <6.9 (49%), and the worst predictor of failure was the combination of good initial response and baseline DI ≥14.0 (12%).

Questions and Answers

Q: For patients with good initial response to sulfonylurea treatment, did that increase their chance of hypoglycemia?

A: We did not do this analysis, but it is a good suggestion.

-- by Eric Chang, Jessica Dong, Stephanie Kahn, Joseph Shivers, Manu Venkat, Michelle Xie, Rebecca Xu, and Kelly Close