September 12-16, 2016; Munich, Germany; Full Report – Obesity – Draft

Executive Highlights

This chapter includes oral presentations from EASD 2016 related to chronic weight management. We heard about the promise of canagliflozin (J&J’s Invokana) and high-dose liraglutide (Novo Nordisk’s Saxenda), and also learned about potential new targets for obesity treatment – namely, the appetite-inhibiting hormones PYY and GLP-1. Before you get to the talks, our theme explains our overall thoughts on the obesity research presented at this year’s meeting. Talk titles highlighted in blue are new full report additions, and were not part of our daily highlights coverage during the conference.

Themes

Obesity

• We saw very little new clinical data on obesity management, with the exception of a phase 3 trial of canagliflozin/phentermine coadministration and a study investigating the appetite-inhibiting hormones PYY and GLP-1 post-gastric bypass surgery. Dr. Priscilla Hollander (Baylor College of Medicine, Houston, TX) presented on the first, sharing data from a 26-week trial of overweight or obese individuals without type 2 diabetes. Canagliflozin (J&J’s Invokana) and phentermine together produced 7.5% weight loss vs. 0.6% weight loss in the placebo arm (p<0.001), and while coadministration also caused greater weight loss than either drug alone, these comparisons were not statistically significant. Dr. Hollander argued that the synergistic effects of these two agents are clinically meaningful for chronic weight management, and advocated for further research into a canagliflozin/phentermine combination based on these promising results. Dr. Maria Svane (University of Copenhagen, Denmark) discussed the critical role of PYY and GLP-1, which act in concert after a patient’s bariatric surgery to suppress appetite and help sustain weight loss. Her small study (n=12 patients post-RYGB) found that food intake increased when GLP-1 and PYY were both inhibited with Exendin 9-39 and sitagliptin, respectively. Dr. Svane thus concluded that the synergistic effects of GLP-1 and PYY might explain successful maintenance of weight loss following a gastric bypass procedure. She expressed enthusiasm for these hormones as new targets for obesity pharmacotherapies. Notably, Novo Nordisk’s Chief Scientific Officer Dr. Mads Thomsen also brought up this PYY research during a company press conference at this year’s EASD. He shared excitement over PYY as a molecule to investigate further in developing better obesity drugs, which are much-needed. We’re particularly intrigued by Novo Nordisk’s phase 1 PYY candidate, which is being investigated both alone and in combination with its once-weekly GLP-1 agonist semaglutide. Still, we can’t ignore the persistent patient/provider reluctance to turn to pharmacotherapies for weight loss – as many have quoted, “drugs don’t work in patients who don’t take them.” Within the exhibit hall or among the industry-sponsored symposia, Novo Nordisk was the only company present with an obesity drug in its portfolio, and there was little mention of Saxenda (liraglutide 3.0 mg) at the company’s exhibit hall booth (at least, relative to the square footage dedicated to Tresiba, Xultophy, and Victoza). We might’ve hoped for more insight on obesity management as a type 2 diabetes prevention tactic, especially as prediabetes prevalence continues to skyrocket.

Detailed Discussion and Commentary

Oral Presentations: Pills or Surgery for Weight Loss?

Coadministration of canagliflozin and Phentermine for Weight Management in Overweight and Obese Adults

Priscilla Hollander, MD (Baylor College of Medicine, Houston, TX)

Dr. Priscilla Hollander presented phase 2 findings that the co-administration of canagliflozin (J&J’s Invokana) and phentermine produced significant weight loss vs. placebo in a 26-week trial enrolling overweight and obese participants without type 2 diabetes, but did not achieve statistically greater weight loss vs. either drug alone. We first saw these results in a poster at ADA 2016. After 26 weeks of treatment, participants randomized to canagliflozin/phentermine coadministration (n=61) experienced 7.5% weight loss on average from a baseline of ~222 lbs (101 kg) vs. 0.6% mean weight loss in the placebo arm (n=57) from a baseline of ~229 lbs (104 kg), p<0.001. Participants receiving 300 mg canagliflozin only (n=53) saw 1.9% weight loss from a baseline of ~227 lbs (103 kg), while those receiving 15 mg phentermine only (n=60) saw 4.1% weight loss from a baseline of ~227 lbs (103 kg). Comparing the canagliflozin/phentermine group with the canagliflozin only and phentermine only groups, there was no statistically significant weight loss benefit to the combination therapy. Dr. Hollander noted that the independent weight loss effects of canagliflozin and phentermine weren’t completely additive in the study, in that average weight loss from coadministration is less than the sum of average weight loss from canagliflozin treatment and phentermine treatment separately. That said, she didn’t rule out the possibility that more substantial weight loss from co-administration could be clinically meaningful – she argued that the two drugs are still working synergistically, and pointed out that individuals taking both agents at once may be compensating for the SGLT-2 inhibitor’s physiological effects by increasing their daily caloric intake. Dr. Hollander concluded that further research into a canagliflozin/phentermine combination will reveal its true potency for chronic weight management.

• Canagliflozin/phentermine coadministration was well-tolerated. Dr. Hollander summarized safety data from the 26-week trial, showing only one serious adverse event (occurring in the combination treatment group), no deaths, and no significant difference in adverse event rates between treatment arms. There were 47 adverse events (57%) in the placebo group, 50 (60%) in the canagliflozin group, 46 (54%) in the phentermine group, and 55 (66%) in the canagliflozin/phentermine group. In total, 22 adverse events led to treatment discontinuation: 5 (6%) in the placebo arm; 9 (11%) in the canagliflozin arm; 5 (6%) in the phentermine arm; and 3 (4%) in the coadministration arm. There were also 69 adverse events that investigators attributed directly to the study drug: 8 (10%) in the placebo group; 21 (25%) in the canagliflozin group; 15 (18%) in the phentermine group; and 25 (30%) in the coadministration group. The most common adverse events were related to osmotic diuresis or were genital mycotic infections occurring in females.

Questions and Answers

Q: Given the current discussion on DKA in response to SGLT-2 inhibitors, I assume that with a combination of canagliflozin/phentermine alongside reduced food intake, the issue came up. Did you see differences in ketone levels between treatment arms?

A: There did not appear to be differences in ketones between treatment groups. But I should emphasize, again, that these were not diabetes patients.

Q: Did you do any analysis that factored in weight loss during the run-in period?

A: Yes, that analysis has been done, but I don’t yet have access to it.

Q: The majority of participants were female, and many would have been around the age of menopause. Do you think menopause symptoms affected any side-effects of patients on phentermine?

A: No, this doesn’t appear to have been the case. Phentermine isn’t used that often in Europe, but the most common side-effects are tachycardia, along with possible increases in blood pressure and insomnia. If we’re thinking about characteristic menopause symptoms – hot flashes, etc. – then there were no increases. We didn’t observe any mood changes or psychological differences.

Q: How does canagliflozin cause weight loss in people who already have normal blood glucose?

A: The molecule lowers the renal threshold, so patients do excrete more glucose, regardless of their baseline glycemia.

Q: I understand that these patients were without diabetes, but perhaps some of them had prediabetes. Have you thought about analyzing the effects of canagliflozin/phentermine on individuals with prediabetes vs. participants with truly normal glycemia?

A: That’s a fantastic question. People with A1c up to 6.5% were enrolled in this study, so there was indeed a small number of patients with prediabetes signed-up, but in our analysis we didn’t see any difference in weight loss for this subset of participants.

Comparable Efficacy and Safety of Liraglutide 3.0 mg for Weight Management Across Baseline BMI Subgroups: Results from the SCALE Obesity and Prediabetes Trial

Sten Madsbad, MD (University of Copenhagen, Denmark)

Dr. Sten Madsbad (University of Copenhagen, Denmark) described glycemic control, body weight, and safety endpoints from a post-hoc analysis of the SCALE trial focused on obesity and prediabetes. While the initial analysis showed that treatment with liraglutide 3.0 mg (Novo Nordisk’s Saxenda) causes significant weight loss and significant delays in the onset of type 2 diabetes, this new post-hoc divided participants from the study into four baseline BMI subgroups: 27-29.9 kg/m² (n=62); 30-34.9 kg/m² (n=624); 35-39.9 kg/m² (n-737); ≥40 kg/m² (n=831). There were no signals for BMI-specific intermediate factors. In other words, the post-hoc analysis found no significant interaction between treatment and baseline BMI in terms of weight loss (p=0.48), change in fasting plasma glucose (p=0.55), or regression to normoglycemia (p=0.92) over three years – the effects of Saxenda on these parameters were sustained across all four BMI sub-groups. The interaction between treatment and baseline BMI did reach near statistical significance for A1c reduction (p=0.06), with a smaller A1c difference between liraglutide vs. placebo in the lowest BMI subgroup (0.01% difference in A1c-lowering). Dr. Madsbad also pointed out the lack of noticeable pattern in treatment-related adverse events (including GI side-effects and hypoglycemia episodes) across baseline BMI subgroups. Overall, he positioned this post-hoc as confirming the efficacy of Saxenda in managing obesity and prediabetes.

Weight Loss (%)

Regression to Normoglycemia (%)

A1c Reduction (%)

Questions and Answers

Q: During three years of this trial, did you find any problems with renal function?

A: No. I haven’t shown this data, but we’ve had a great experience with GLP-1 agonists – and liraglutide, especially – and kidney safety.

Combined Actions of PYY and GLP-1 Contribute to Weight Loss after Roux-en-Y Gastric Bypass

Maria Svane, MD (University of Copenhagen, Denmark)

Sweden’s Dr. Maria Svane (University of Copenhagen, Denmark) showed how the appetite-inhibiting hormones GLP-1 and PYY act in concert to decrease food intake following a patient’s Roux-en-Y gastric bypass (RYGB) surgery. In a small study (n=12 individuals following RYGB, with normal glucose tolerance, mean age 36, mean BMI 33.5 kg/m²), participants were given four treatments on four different days: (i) an injection of GLP-1 antagonist Exendin 9-39 and a sitagliptin (Merck’s Januvia) pill; (ii) an injection of Exendin 9-39 and a placebo pill; (iii) a saline infusion and a sitagliptin pill; and (iv) a saline infusion and a placebo pill. Dr. Svane explained how Exendin 9-39 can differentially suppress action of the GLP-1 hormone, and how the hormone DPP-4 is required to activate PYY – thus, a DPP-4 inhibitor like sitagliptin keeps PYY inactive. This strategic study design allowed researchers to isolate and compare the appetite-inhibiting effects of GLP-1 and PYY post-gastric bypass (on average, participants were enrolled ~5 months after a procedure). Food intake increased 19.5% when patients were treated with both active agents (thus inhibiting both GLP-1 and PYY) vs. placebo (p=0.04). Participants’ food intake also increased slightly when they were given Exendin 9-39 alone (inhibiting just GLP-1) or sitagliptin alone (inhibiting just PYY) vs. placebo, but these increases were not statistically significant. Dr. Svane emphasized the critical takeaway that there are no significant or clinically meaningful effects seen for GLP-1 or PYY alone (as evidenced by the fact that inhibiting one or the other doesn’t markedly increase food intake), but acting in tandem, these hormones effectively suppress a patient’s appetite following gastric bypass surgery. The synergistic effects of GLP-1 and PYY could explain successful maintenance of weight loss post-gastric bypass, and Dr. Svane suggested excitedly that this finding presents new drug targets for obesity pharmacotherapies that can mimic the metabolic effects of bariatric surgery without the same level of risk or invasiveness. Novo Nordisk currently has a PYY candidate (PYY 1562 [NN9747/8]) in its phase 1 pipeline for both diabetes and obesity and the company has shared that the candidate could be co-formulated with once-weekly GLP-1 agonist semaglutide. This study certainly lends support for the efficacy of this approach and we believe it’s likely that Novo Nordisk will pursue this further.

Questions and Answers

Q: I see that you’ve shown how GLP-1 and PYY are important for food intake, but I’m not sure how this shows the mechanism of weight loss via bypass surgery. What would this same study show if healthy people or obese people pre-surgery were enrolled?

A: I’m not aware of a study that has used the same combination of agents in un-operated participants. Studies with infusion of Exendin 9-39 in un-operated subjects have shown that the agent doesn’t increase food intake, but also that PYY secretion is upregulated, so I think ultimately infusion of Exendin-939 would be difficult to interpret.

Q: It seems paradoxical that sitagliptin, which is known to be weight-neutral, has a weight-lowering effect here. Are you aware of any data on the combination of a high-dose GLP-1 agonist with sitagliptin?

A: The weight-neutral effect of DPP-4 inhibitors could be happening through this suppression of PYY. I’m not aware of any studies using the particular combination you mention, but that could be a very interesting approach.

Q: There’s also an increase in GLP-1 hormone when individuals are sleep-deprived, so do you think sleep deprivation has a similar effect?

A: It’s not completely clear whether the same effects occur after sleep. Hormone concentration does consistently increase after sleep, so I think the same mechanisms are operating after bariatric surgery-induced weight loss.

Beneficial Effects of Canagliflozin on Energy Metabolism and Visceral Fat Volume through a Possible Mechanism of Fatty Acid Oxidation in an Animal Model of Obese Type 2 Diabetes

Hiroharu Mifune, MD (Kurume University School of Medicine, Japan)

SGLT-2 inhibitors, and particularly canagliflozin, have shown a pronounced weight loss benefit in individuals without hyperglycemia. Canagliflozin (J&J’s Invokana) can reduce body weight without actually improving glycemic status, and this study in mice was designed to figure out why, according to Dr. Hiroharu Mifune (Kurume University, Japan). Dr. Mifune described how mice with induced obesity and mild hyperglycemia were dosed with either 10 mg/kg of canagliflozin or placebo daily. Canagliflozin treatment resulted in a significant reduction in visceral fat vs. placebo after 14 days (p<0.05). Subcutaneous fat also declined for canagliflozin-treated mice over two weeks, but this reduction was not statistically significant vs. placebo. While this study didn’t yield conclusive evidence on the SGLT-2 inhibitor’s mechanism for weight loss, Dr. Mifune ended his talk with a detailed, multifaceted hypothesis: (i) The drug decreases glucose reabsorption by the kidneys, which in turn leads to greater urinary excretion of glucose and calorie loss – the essential precursor to weight loss; (ii) Excess urinary excretion of glucose makes fat the body’s primary fuel resource, increasing lipolysis and thus lowering adipose tissue volume – another precursor to weight loss; (iii) Increased lipolysis spurs more beta-oxidation and greater synthesis of ketone bodies, which decreases a mouse’s food intake during the light phase – the smaller calorie intake contributes directly to weight loss. On the other hand, the increase in carbohydrate intake among humans on SGLT-2 inhibitors has been well-characterized, so we’re curious how Dr. Mifune’s hypothesis may translate to humans.

-- by Helen Gao, Payal Marathe, Jennifer Zhao, and Kelly Close