New GLP-1 analogs, mimetics in the pipeline for type 2 diabetes

FDA Advisory Committee’s split vote in April complicates the future for liraglutide.

All signs indicate that several glucagon-like peptide-1 analogs and mimetics may be poised to hit the market and provide clinicians with a variety of new options for the treatment of type 2 diabetes.

Exenatide (Byetta, Amylin) received FDA approval in 2005 and is currently the only GLP-1 incretin mimetic available. The drug made an immediate impact on treatment regimens and continues to perform well in post-approval clinical trials.

There are a number of other GLP-1 class drugs in the pipeline. Once-weekly exenatide, another product of Amylin, is receiving a great deal of attention due to the long-acting mechanism of action. Also from Amylin is a version of exenatide, currently in investigative stages, that can be administered nasally.

Liraglutide (Victoza, Novo Nordisk) is a once-daily injectable GLP-1 analog that has completed phase-3 clinical trials. A recent vote by the FDA Endocrinologic and Metabolic Drugs Committee on the drug’s safety was split, making its future development uncertain.

GlaxoSmithKline has announced the initiation of a phase-3 trial of albiglutide (Syncria), and taspoglutide (Roche) is also in phase-3 trials.

Zachary Bloomgarden, MD, said GLP-1 drugs address the basic goals of diabetes treatment. Photo by: Stephen Fuller

“The GLP-1 receptor agonists are good drugs,” Zachary Bloomgarden, MD, clinical professor in the department of medicine at the Mount Sinai School of Medicine, said in an interview. “They lower the blood sugar in a glucose-sensitive fashion and they induce weight loss. It is hard to argue with a treatment that addresses these two basic goals of diabetes treatment.”

The most recent set of ADA guidelines placed drugs in the GLP-1 class in the second tier of recommended treatment options due to a scarcity of clinical data. However, recent open-label trials involving exenatide have demonstrated positive results and bode well for the future of all GLP-1 treatments.

Exenatide

A retrospective cohort study involving 3,262 patients assigned exenatide and 3,038 assigned insulin glargine found that exenatide yielded lower rates of comorbidities and hypoglycemic events, as well as better adherence rates.

The trial began in 2005, shortly after exenatide received FDA approval, and continued through 2007.

Patients assigned insulin were 1.8 times more likely to have a hypoglycemic event than those assigned exenatide (P<.001). Patients in the exenatide arm had an average score of 1.45 on the Deyo-Charlson comorbidity index compared with an average score of 1.82 for patients assigned insulin glargine.

By the 12-month mark, the medication possession rate among patients who had received more than one prescription was 68.4% for patients assigned exenatide vs. 57.9% for those assigned insulin glargine (P<.001). Nearly 47% of patients in the exenatide arm achieved a medication possession rate of ≥80%, compared with 29.4% in the insulin glargine arm (P<.001).

Bloomgarden said the twice-daily injection regimen with exenatide is a problem for some patients. “The data demonstrate that, in terms of effectiveness, there is no winner among the GLP-1 drugs in development. They are all effective and may be more effective than exenatide taken twice daily. And for many people, taking a shot twice a day is a turnoff,” he said.

“Taking a shot once a day is less of a turnoff, but still a turnoff. Once a week, still less; once a month, still less. Ultimately, the hierarchy that develops among these medicines may simply come down to how often the patient has to self-administer,” Bloomgarden said.

Liraglutide

Alan Garber, MD, PhD, and colleagues conducted a double-blind, double-dummy, active-control parallel group study — the LEAD-3 Mono trial — that included 746 patients with early type 2 diabetes. They randomly assigned 251 patients to 1.2 mg liraglutide once daily, 247 patients to 1.8 mg liraglutide once daily, and 248 to 8 mg glimepiride once daily.

The primary outcome was change in HbA1c; secondary endpoints were changes in body weight, fasting plasma glucose, self-measured eight-point plasma glucose profiles, blood pressure, beta cell function, fasting glucagon and a self-reported quality of life assessment.

Garber, chief medical editor of Endocrine Today, and colleagues reported a 0.51% decrease in HbA1c in patients assigned glimepiride, a 0.84% decrease with 1.2 mg liraglutide and a 1.14% decrease with 1.8 mg liraglutide. The difference was –0.33% (P=.0014) between glimepiride and the lower dose of liraglutide and –0.62% (P<.0001) between glimepiride and the higher dose of liraglutide.

In the glimepiride arm, 16% of participants reached HbA1c levels <6.5% compared with 28% in the liraglutide 1.2-mg arm (P=.0025) and 38% in the liraglutide 1.8-mg arm (P<.0001). Decreases in fasting plasma glucose from baseline were significantly greater in the liraglutide groups than in the glimepiride groups.

In the drug naïve population, 1.8-mg of liraglutide produced stable and durable 52-week reductions of HbA1c levels of 1.6% without any loss of efficacy over the full 52-weeks trial period, Garber said.

In contrast, he said, glimepiride at 8 mg produced an HbA1c reduction of about 0.9% with loss of efficacy during the 52-week observation period of the trial. Further, there was weight loss with liraglutide but weight gain with glimepiride; incidence of hypoglycemia was 10 times higher with glimepiride vs. liraglutide.

Insulin resistance decreased in the liraglutide groups but increased in the glimepiride group. Patients assigned liraglutide lost weight, whereas patients assigned glimepiride gained weight. Further analysis demonstrated that nausea was unrelated to weight loss.

James R. Taylor, PharmD, CDE

James R. Taylor, PharmD, CDE, clinical associate professor, University of Florida College of Pharmacy, offered some perspective on the results that were presented at the 2008 ADA annual meeting. “The preliminary results for liraglutide look good,” he said.

“It seems to have all of the same effects as exenatide. There appears to be a similar drop in HbA1c, and the rates of nausea are similar as well. The once-daily injection is definitely an advantage over twice a day,” Taylor said.

However, the future is uncertain for liraglutide, given the FDA’s Endocrinologic and Metabolic Drugs Advisory Committee split vote in early April over the potential for development of c-cell thyroid disease with the drug. The committee voted 12-1 that the data from animal studies could apply to humans, and then voted 6-6 with one abstention on the question of whether the thyroid cancer risk should keep the drug from winning FDA approval.

The committee voted 12-0 with one abstention that available papillary thyroid cancer data permit marketing of liraglutide, provided that the remainder of the risk:benefit data were found to be acceptable. (See perspective on this action below.)

The committee voted 8-5 that there was adequate evidence that liraglutide would not convey unacceptable cardiovascular risk.

Exenatide once-weekly

A recent head-to-head, noninferiority trial evaluated the efficacy of twice-daily exenatide 10 mcg vs. a once-weekly injection of exenatide 2 mg. Two hundred ninety-five patients with type 2 diabetes were included. Eligible participants were either treatment-naïve or were on one or more oral antidiabetic agents.

The primary endpoint was change in HbA1c at 30 weeks. Secondary endpoints included evaluation of safety and tolerability, analysis of fasting and postprandial glucose concentrations, body weight, fasting glucagon and lipids, pharmacokinetics, blood pressure and paracetamol absorption.

HbA1c decreased by 1.9% with the once-weekly formulation compared with a 1.5% decrease with the twice-daily version (P=.0023). HbA1c levels of 7% or less were reached by 77% of patients assigned the weekly injection and 61% assigned the twice-daily injection (P=.0039).

Patients in the once-weekly injection group lost 3.7 kg; patients in the twice-daily injection group, 3.6 kg (P=0.89). Significant reductions in both fasting and postprandial glucose were observed in both groups at the 30-week mark. Researchers reported an average reduction of 2.3 mmol/L in fasting plasma glucose in the weekly-injection group vs. a reduction of 1.4 mmol/L in the twice-daily group.

Adverse event profiles for both groups were similar, although patients assigned the weekly injection were less likely to report nausea and vomiting.

Ken Wilhelm, MD, senior director of medical affairs at Amylin, discussed these results with Endocrine Today. “A lot of our data suggest that all of the parameters are going to be better with the once-weekly exenatide. We observed better efficacy with improved tolerability,” Wilhelm said.

Albiglutide

A recently published trial of albiglutide, a GLP-1 mimetic generated by genetic fusion of DPP-IV-resistant GLP-1 dimer to human albumin, demonstrated that the half-life of the drug may lend itself to once-weekly dosing.

There were 54 participants with type 2 diabetes in the trial, and 46 patients in a second study.

In the first study, participants were randomly assigned to 9 mg, 16 mg or 32 mg albiglutide or placebo on days one and eight. In the second study, patients were assigned a single dose of 16 mg or 64 mg albiglutide in the arm, leg or abdomen.

As reported in the Journal of Clinical Endocrinology and Metabolism, albiglutide “improved fasting plasma glucose and postprandial glucose with a favorable safety profile.”

After adjusting for placebo, fasting plasma glucose decreased on day two by 26.7 mg/dl (95% CI, –46.3 to –7.06) and on day nine by 50.7 mg/dl (95% CI, –75.4 to –26.0). Albiglutide also reduced postprandial glucose.

The half-life of the drug was six to seven days, and pharmacokinetics or pharmacodynamics was not affected by injection site. Adverse event levels were similar across treatment and placebo groups.

Murray Stewart, DM, FRCP, of GlaxoSmithKline, discussed the design of this phase-2 trial. “With a low dose of 16 mg and high dose of 64 mg, we saw large decreases in glucose after only one injection,” he said. It could be extrapolated that those decreases would continue, which may allow patients to reach and maintain steady state glucose levels, according to Stewart.

“The main reason we chose this study design was because we wanted to see the profile for a weekly injection. It is easier to see the pharmacokinetic profile on one or two injections. From there, we are able to predict whether the drug can be used weekly, bimonthly or monthly,” Stewart said.

Taspoglutide

In a phase-2 trial of taspoglutide, 133 patients with diabetes inadequately controlled with metformin were randomly assigned 20 mg taspoglutide or placebo subcutaneously for four weeks, followed by either maintenance at 20 mg (20/20) or a dose increase to 30 mg or 40 mg weekly with matched placebo for an additional four weeks. Patients were then followed for four weeks after the last administered dose.

Fasting plasma glucose changed by –2.3 mmol/L in the maintenance group, –1.6 mmol/L in the 30-mg titration group, –2.2 mmol/L in the 40-mg titration group and –0.6 mmol/L in the placebo group. HbA1c decreased to ,7% in 72% of patients in the maintenance group, in 53% of patients in the 30-mg titration group, in 70% in the 40-mg titration group and in 19% of patients receiving placebo.

Transient, mild-to-moderate nausea was the most frequently reported adverse event, with nausea resolving spontaneously in most patients.

In another phase-2 trial involving taspoglutide, 306 patients with type 2 diabetes inadequately controlled with metformin were randomized to eight weeks of treatment with placebo or taspoglutide, either 5 mg, 10 mg or 20 mg weekly or 10 mg and 20 mg once every two weeks and followed up for four additional weeks after the last administered dose.

HbA1c reductions ranging from –1.0% to –1.2% were observed in all groups assigned the active drug vs. –0.2% in patients assigned placebo (P<.0001). The percentage of patients that achieved the target of HbA1c levels ≤7% was 79% and 81% in the 10-mg and 20-mg weekly groups vs. 17% with placebo.

Body weight decreased progressively and dose dependently with significant reductions from baseline in the 10-mg weekly group (–2.0 kg, P<.05) and the 20-mg weekly group (–2.8 kg, P<.0001).

Dose-dependent, transient, mild-to-moderate nausea was the most commonly reported adverse event. Six patients experienced serious adverse events considered by the investigator to be unrelated to the study drug, Rajiv Patni, MD, taspoglutide life cycle leader at Roche, told Endocrine Today.

Nausea

Although nausea is often transient for drugs in the GLP-1 class, its consistent presence continues to be a concern.

“Probably the most important thing to remember about the nausea and vomiting is that, in most people, it will go away in a few weeks,” Taylor said. “However, we have put some people on exenatide and they call back a few days later to say that they are unable to tolerate it.”

Other clinicians have said that they prescribe more DPP-IV inhibitors for this reason, even though multiple trials have demonstrated superior efficacy of drugs in the GLP-1 class.

Most researchers believe that clinicians will gain knowledge about the right drug dose for individual patients, helping to curtail nausea. Because two dose levels of exenatide are available — 5 mcg and 10 mcg — many physicians first initiate treatment with the lower dose and then increase it once adjustments have been made.

“Changing the timing of administration can be helpful for some people,” Wilhelm said. “Encouraging patients to pay attention to quality and quantity of food they eat can help manage the nausea. Being particularly careful around the time of dose titration may also be helpful.”

Other researchers say that longer-acting drugs will be key to reducing nausea. “There has been some speculation that if you look at the pharmacokinetics of exenatide, the drug concentration goes up and down quite quickly” Stewart said.

“Some of the nausea and vomiting may be related to the rate of change in pharmacokinetics. The drug concentration of the long-acting formulations rises slowly. Many people believe that a smoother profile will lead to less nausea and vomiting,” he said.

The question of whether longer-acting drugs will have value is debatable. Some clinicians believe that while the convenience of a bi-monthly or monthly regimen could be attractive, a longer-acting mechanism may also be problematic.

Patni said that it “may become a serious concern for clinicians if a person taking a monthly injection develops nausea and vomiting. This mechanism may not be amenable to monthly or less frequent dosing given the known adverse event profile.”

There has been some speculation about the relationship between nausea and weight loss in patients assigned drugs in the GLP-1 class. Even though that relationship may exist, most researchers have indicated that the primary causes of weight loss are the increased feeling of satiety and delay in gastric emptying induced by GLP-1 drugs.

“The effects of these drugs on satiety and gastric motility are being seen in weight loss,” Wilhelm said. “The appreciation of the value of treating weight is growing. Weight reduction is clearly being recognized as key component of management of this disease.”

Convenience

There is both optimism and concern about drugs with mechanisms of action that are longer than a week. “It will definitely be helpful to improve presentation to the patient and make it easier for people to get exposed to the drugs.” Wilhelm said.

“However, we have yet to clearly demonstrate the value of a monthly or quarterly administration. As the interval increases, it is less clear that the incremental benefit on compliance is significant,” he said.

Another factor that may influence patient compliance is the size of the needle used to inject the drug. Some concerns have been raised that the once-weekly exenatide uses a 23-gauge needle, which is considerably larger in diameter than the 29-gauge needle used for taspoglutide and the 29-gauge used for albiglutide. Part of the package for taspoglutide is a delivery system that includes the smaller needle, which is hidden and disposable.

“This market is very device-sensitive,” Patni said. “While differentiation has to start with efficacy and tolerability, our device may provide ancillary differentiation.”

How this class of drugs impact beta cell function in humans is still unclear. Animal models have demonstrated varying results, including that of improved beta cell function and the prevention of beta cell death.

However, no such evidence has yet to be proven in humans. “If it can be proven that there are significant effects on beta cell function in humans, it would probably move these to first-line drugs,” Taylor said. “You would want to get patients on them right away because that would be a big benefit.”

“The problem researchers are having is that it is difficult to look at the beta cell,” Patni said. “Because those cells constitute such a small percentage of the pancreas, there is really no way to directly measure or visualize beta-cell mass or function. The best we can do is look at indirect or surrogate measures.” – by Rob Volansky

Perspective on FDA Committee vote on liraglutide

FDA Advisory Committee’s Decision on Liraglutide

Alan J. Garber

Rats and mice both have spontaneous c-cell tumors with or without liraglutide. Long-acting GLP-1 agents may increase this potential. There is no evidence of this occurring in nonhuman primates or in man. The study data in patients in the liraglutide trials saw no increase in c-cell proliferation and no increase in calcitonin secretion or levels. Thus, the rodent hypersentitivity to GLP-1 agonists seems not to carry over to man.

– Alan J. Garber, MD, PhD

Endocrine Today Chief Medical Editor

Papillary thyroid cancer is common and found in 3% to 5% of autopsies in North America. This coupled with its generally excellent prognosis serves as a sound basis for not screening patients for thyroid nodules or thyroid cancer. Thus, finding a cluster of papillary thyroid cancers in patients being screened for anatomic thyroid abnormalities comes as no surprise and should not generally serve as a basis for major concern.

– Jeffrey Garber, MD

Endocrine Today Editorial Board member

Liraglutide caused c-cell stimulation in mice leading to hyperplasia and neoplasia including medullary thyroid carcinoma almost certainly through a GLP-1 receptor stimulation of the c-cells. The drug caused c-cell neoplasia and medullary thyroid carcinoma in rats, but the FDA did not think this was mediated through simple GLP-1 receptor stimulation, whereas the company thought it was.

In humans there was no compelling evidence of c-cell stimulation based on my review of the data. Any difference seen in calcitonin concentrations between groups (liraglutide vs. placebo vs. active comparator) was in a range that would be considered very low normal if conventional calcitonin assays could measure that low.

Nonetheless, the committee was concerned that they could not prove that the rodent data would not/might not apply to humans.

The c-cell disease found in individuals during the liraglutide development program did not support an effect of liraglutide on c-cell disorders; specifically, one medullary thyroid carcinoma and one medullary thyroid carcinoma in situ found in the comparator group.

Three cell hyperplasia and one nodular hyperplasia (medullary thyroid carcinoma in situ) were found in the liraglutide group. The randomization was 2:1 in terms of numbers liraglutide vs. comparators. Three of the four patients discovered with c-cell disease had abnormal calcitonin concentrations at baseline.

– Gilbert H. Daniels, MD

Co-Director, Thyroid Clinic at Massachusetts General Hospital Cancer Center, Boston

Will longer-acting GLP-1 drugs be accepted by patients?

For more information:

• Balena R. 108-OR. Presented at: the American Diabetes Association 68th Annual Scientific Sessions; June 6-10, 2008; San Francisco.
• Drucker D. The Lancet. 2008; doi:10.1016/S0140-6736(08)61206-4.
• Fabunmi R. Curr Med Res Opin. 2009;25:777-786.
• Garber A. The Lancet. 2009; doi:10.1016/S0140-6736(08)61246-5.
• Matthews J. J Clin Endocrinol Metab. 2008;93:4810-4817.
• Ratner R. 10-OR. Safety and tolerability of high doses of the long acting, human GLP-1 analogue R1583 in diabetic subjects treated with metformin: a double-blind, placebo-controlled phase-2 study. Presented at the American Diabetes Association 68th Annual Scientific Sessions; June 6-10, 2008; San Francisco.