Comparator clinical trials of surrogate endpoints with albiglutide are in HARMONY

Abstract

Evaluation of: Ahrén B, Johnson SL, Stewart M et al. HARMONY 3: 104-Week randomized, double-blind, placebo- and active-controlled trial assessing the efficacy and safety of albiglutide compared with placebo, sitagliptin, and glimepiride in patients with type 3 diabetes taking metformin. Diabetes Care 2014;37:2141–8 and Rosenstock J, Fonseca VA, Grass JL et al. Advancing basal insulin replacement in type 2 diabetes inadequately controlled with insulin glargine plus oral agents: a comparison of adding albiglutide, a weekly GLP-1 receptor agonist, versus thrice-daily prandial insulin lispro. Diabetes Care 2014;37:2317–25.

Agonists of glucagon-like peptide-1 (GLP-1) receptors are used in the treatment of Type 2 diabetes. Albiglutide is a new long-acting GLP-1 receptor agonist being developed for once weekly use. This is an evaluation of two clinical trials in the HARMONY clinical trials series. HARMONY 3 compares albiglutide with sitagliptin and glimepiride in subjects with Type 2 diabetes poorly controlled with metformin, and HARMONY 6 compares albiglutide with insulin lispro in subjects poorly controlled with slow/medium release preparations of insulin. Both studies showed that albiglutide lowered HbA1c and had advantages over its comparator drugs. However, questions remain about the safety of albiglutide. Albiglutide is not being used in subjects with a history of thyroid cancer because it is not known whether this is a rare adverse effect with albiglutide. Also, the safety of albiglutide in subjects with Type 2 diabetes and high cardiovascular risk is unknown.

Keywords:

• albiglutide
• cardiovascular safety
• GLP-1 receptor agonists
• HbA1c
• thyroid cancer
• Type 2 diabetes

In the USA, the prevalence of diabetes is about 8% (24 million people), and 90% of it, is Type 2 diabetes, which has both lifestyle and genetic components [1]. Despite the treatments available for Type 2 diabetes, about two-thirds of the subjects die from heart disease or stroke. Diabetes is also a leading cause of blindness, end-stage kidney failure and lower limb amputations [1]. Clearly, there is a need for better treatments for Type 2 diabetes.

In the past 10–15 years, physiological discoveries have opened up new paradigms in the treatment of diabetes. In Type 2 diabetes, the hyperglycemia is caused by a relative lack of insulin production and reduction in insulin sensitivity [1]. Classically, glucose levels in the pancreatic cells, which relate to blood glucose levels, were considered to be the main controller of insulin levels. However, it is now known that some gut peptides influence insulin secretion, and these peptides and their receptors are the target for a range of new drugs introduced for the treatment of Type 2 diabetes in the past 10–15 years.

In response to a meal, glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) are released. Both of these gut hormones augment glucose-mediated insulin secretion, but only GLP-1 suppresses glucagon secretion, inhibits gastric emptying and reduces appetite [2]. Initially, it was reported that in subjects with Type 2 diabetes, the production of GLP-1 and GIP was impaired in response to an oral glucose tolerance test or a meal test, but subsequent studies have disputed this, and recent systematic review and meta-analysis have concluded that there is no impairment of this response (GLP-1 [3], GIP [4]).

Exendin-4 is a 39-amino acid peptide, which acts as an agonist at the GLP-1 receptors, and is resistant to proteolytic breakdown [5]. Synthetic exendin-4 (exenatide) was the first GLP-1 receptor agonist introduced for use in Type 2 diabetes [5]. Subsequently, another short-acting GLP-1 receptor agonists (lixisenatide) is under development [6]. When the short-acting GLP-1 receptor agonists are administered once or twice a day, they are not effective at reducing plasma glucose levels during fasting [6]. Also, the short-acting agents have been shown to delay gastric emptying and this may reduce insulin levels postprandially [6]. To prolong the beneficial effects of stimulating the GLP-1 receptor in subjects with Type 2 diabetes, long-acting GLP-1 receptor agonists (e.g., albiglutide, dulaglutide, liraglutide) are also being developed [6]. These long-acting GLP-1 receptors reduce glucose throughout the day, but with long-term use do not reduce gastric emptying [6].

Albiglutide is being developed in the HARMONY trials, which are being sponsored by GlaxoSmithKline, the manufactures of albiglutide. This evaluation is of HARMONY 3 and 6.

HARMONY 3

The methods and results of HARMONY 3, a Phase III trial comparing albiglutide to placebo, sitagliptin and glimepiride in 1013 subjects with Type 2 diabetes poorly controlled with metformin, are combined and summarized in this section [7]. To be enrolled, subjects had to have good kidney function. Among the subjects excluded were those with symptomatic biliary disease or history of pancreatitis, recent clinically significant cardiovascular and/or cerebrovascular disease, a variety of gastrointestinal disease, personal or family history of medullary thyroid carcinoma and high blood pressure.

The subjects enrolled had a mean age of about 55 years, HbA1c of about 8.15%, were predominantly white, and about half were women. Subjects were randomized to albiglutide 30 mg weekly, sitagliptin 100 mg daily, glimepiride 2 mg daily or placebo. Blinded up-titration was allowed for albiglutide to 50 mg and of glimepiride to 4 mg, based initially on fasting blood glucose levels and subsequently on HbA1c levels, and just over half of the subjects underwent this. There was no active up-titration for sitagliptin or placebo although this was required in 59% of subjects with sitagliptin and 69% with placebo.

The primary endpoint was the change in HbA1c after 104 weeks, which increased with placebo (0.27%) while decreasing with albiglutide, sitagliptin and glimepiride by 0.63, 0.28 and 0.36%, respectively, and this reduction was greater with albiglutide than with sitagliptin or glimepiride. Fasting plasma glucose levels were consistent with HbA1c levels, with the levels at 104 weeks being significantly lower with albiglutide than with placebo, sitagliptin or glimepiride.

Secondary endpoints included body weight, where there was a weight loss with placebo (1 kg), albiglutide (1.2 kg) and sitagliptin (0.9 kg) but a weight gain with glimepiride (1.2 kg). Changes in blood pressure and heart rate were very small in all four groups, and not significant different between the comparator drugs.

Gastrointestinal adverse events were similar in all groups. Symptomatic hypoglycemia occurred to a lesser extent in the placebo (3.0%), albiglutide (4.0%) and sitagliptin (4.3%) groups than with glimepiride (17.9%). Two cases of pancreatitis were possibly related to albiglutide.

HARMONY 6

The methods and results of HARMONY 6, a Phase III, randomized, open-label, multicenter study comparing albiglutide with insulin lispro in subjects with Type 2 diabetes taking insulin glargine, who were badly controlled, despite taking oral anti-diabetic medication and insulin glargine, insulin detemir or NPH insulin, are précised in this section [8]. Exclusion criteria were similar to those for HARMONY 3 (given above).

There were 586 subjected enrolled in HARMONY 6. They had a mean age of about 55 years, HbA1c of 8.45%, were predominantly white, although other races/ethnicity were represented, and about half were women. About 8% of subjects had had a prior myocardial infarction.

In the initial part of the trial, the use of sulphonylureas, gliptins, insulin detemir or NPH insulin was discontinued. Subjects were stabilized on insulin glargine with titration, and continued their other oral anti-diabetic drugs. For most subjects this was metformin (∼69%). Subjects then continued to take insulin glargine titrated to a fasting plasma glucose of <5.6 mmol/l and were randomized to albiglutide 30 mg, which could be uptitrated to 50 mg once weekly or insulin lispro titrated to give a preprandial glucose level of 4.4–7.2 mmol/l and peak postprandial of <10 mmol/l. Over the 26 weeks, the insulin glargine dose was increased by a similar amount in both groups, ∼6 IU. About half the subjects in the albiglutide group were titrated to 50 mg, and (over the 26 weeks) the dose of lispro was nearly doubled (from 15.5 to 30.6 IU).

The primary outcome was the change in HbA1c after 26 weeks, and this was from 8.5 to 7.7% with albiglutide and from 8.4 to 7.8% with the lispro, and these changes were not significantly different. Although fasting blood glucose levels were lower with albiglutide than with lispro, the values were also not significantly different. Subjects in the albiglutide group lost weight (0.7 kg), whereas those in the lispro group gained weight (0.8 kg), and the difference was significantly different.

Severe hypoglycemia only occurred in two subjects, both in the lispro group. Symptomatic hypoglycemia was also higher with lispro (30%) than with albiglutide group (15%). Nausea and diarrhea were more common in the albiglutide group than in the lispro group. One subject treated with albiglutide developed medullary thyroid cancer, but there were no cases of pancreatitis.

Expert commentary

Trials to other groups of antidiabetic drugs

This evaluation has shown that albiglutide lowers HbA1c and has advantages over sitagliptin and glimepiride (HARMONY 3) in subjects with Type 2 diabetes poorly controlled with metformin, and over insulin lispro in subjects who are poorly controlled despite taking insulin glargine (HARMONY 6).

HARMONY 1 showed that albiglutide could reduce HbA1c in the presence of other antidiabetic drugs [9]. In HARMONY 1, 310 subjects with Type 2 diabetes, not controlled with pioglitazone, with or without metformin, were randomized to albiglutide (30 mg sc) or placebo over 52 weeks [9]. At the end of the trial, albiglutide had reduced HbA1c by 0.8%, compared with 0.1% in the placebo group [9]. However, there was no difference in body weight changes in the two groups [9].

HARMONY 4 showed that albiglutide caused weight loss compared to the weight gain with insulin glargine. In HARMONY 4, 779 subjects with Type 2 diabetes being treated with metformin with or without a sulfonylurea were randomized to albiglutide or insulin glargine (01 U once a day) for 52 weeks [10]. The primary endpoint was HbA1c, and this was reduced from 8.3 to 7.6% with albiglutide, and from 8.4 to 7.6% in the insulin glargine group, which indicated that albiglutide was noninferior to insulin glargine in reducing HbA1c [10]. Body weight declined with albiglutide by 1.1 kg, but increased with insulin glargine (1.6 kg), and the difference was significant.

Comparator trials to other GLP-1 receptor agonists

Other GLP-1 receptor agonists available for use in the treatment of Type 2 diabetes include exenatide and liraglutide, and others are in development, for example, lixisenatide, dulaglutide. However, there is only one comparison of albiglutide with another GLP-1 receptor agonist, and this did not clearly establish an overall advantage for albiglutide over liraglutide. In this comparison (HARMONY 7), once weekly albiglutide (titrated to 50 mg) was compared with liraglutide (titrated to 1.8 mg once daily) in 841 subjects with poorly controlled Type 2 diabetes, despite the use of metformin with or without a sulfonylurea [11]. After 32 weeks, HbA1c had been reduced by 0.98 and 0.71% by liraglutide and albiglutide, respectively, which indicated superiority of liraglutide, compared with albiglutide [11]. Body weight loss was greater with liraglutide (2.19 kg) than with albiglutide (0.64 kg) [11]. Nausea was more common with liraglutide (29.2%) than with albiglutide (9.9%), as were injection-site reactions (liraglutide 12.9 vs 5.4%) [11].

Recently, studies in rats, has suggested that liraglutide can cross the blood–brain barrier because of internalization of it receptor and acts at a site distinct from the GLP-1 receptor in the central nervous system to induce weight loss [12]. It is not known whether other GLP-1 receptor agonists share this non-GLP-1 receptor property with liraglutide. Thus, it is possible that liraglutide has a greater effect on weight loss than albiglutide, because of this non-GLP-1 receptor property.

Overall, no clear-cut advantage of albiglutide over liraglutide has been established. Indeed, liraglutide may be preferred for its greater effects on HbA1c and body weight by prescribers. On the other hand, albiglutide may be preferred, by some subjects with Type 2 diabetes, for its once weekly use. However, this has not been established to date and needs to be tested. Also, albiglutide needs to be tested against the other GLP-1 receptor agonists to determine whether it has advantages over these.

Conditions of approval relating to thyroid cancer

Albiglutide (Tanzeum) was approved by the US FDA on 15 April 2014, with a Boxed Warning that tumors of the thyroid gland have occurred in rodents with GLP-1 receptor agonists, and it is not known whether albiglutide causes these tumors in humans. Thus, albiglutide should not be used in subjects with thyroid cancers or a family history of these. In addition, the FDA required that there be a registry over 15 years to determine whether albiglutide increased the risk of this cancer [13].

This evaluation shows that in both HARMONY 3 and 6, subjects with a personal or family history of medullary thyroid carcinoma were excluded from the study, and that there was one case of medullary thyroid cancer in HARMONY 6 in an albiglutide-treated subject [8]. Subjects with a family history of medullary carcinoma were also excluded from the other peer-reviewed published HARMONY studies (1 and 7), and no subject developed thyroid cancer in these studies (1, [9]; 7, [11]). However, these HARMONY studies, or any other use of albiglutide, which excludes subjects with a history of thyroid cancer, will not be useful in determining whether there is a relationship between a history of medullary thyroid carcinoma and albiglutide causing these tumors. However, the registry may be useful in determining whether albiglutide causes thyroid cancers in subjects without a history.

For GLP-1 receptor agonism to be responsible for thyroid cancers, the receptors will have to be found in the normal thyroid gland, and the evidence for this is controversial. Thus, one study reported GLP-1 receptor immunoreactivity in a third of normal human thyroid tissue [13] and another study reported no GLP-1 receptor immunoreactivity or autoradiography in normal human thyroid [14]. Thus, this controversy needs to be resolved, as part of the process of determining whether albiglutide can cause thyroid cancers.

Cardiovascular clinical outcomes

Another condition of approval by the FDA for albiglutide was that a cardiovascular outcomes trial to evaluate the cardiovascular risk of albiglutide be undertaken in subjects with high baseline risk of cardiovascular disease [15]. The HARMONY trials have excluded subjects with recent clinically significant cardiovascular and/or cerebrovascular disease. Thus, it seems unlikely that these trials are going to be useful in determining the effects of albiglutide in subjects with high baseline risk of cardiovascular disease. This has recently been shown to be correct in a meta-analysis, predominated by the HARMONY trials, which showed no cardiovascular safety concerns for albiglutide [16]. The major limitation of this meta-analysis was that the composite of cardiovascular death, nonfatal myocardial infarction and non-fatal stroke was low (∼1%/year) [16].

Thus, it seems to the author that it will be necessary to undertake a new study of the cardiovascular risk of albiglutide in subjects with Type 2 diabetes and high baseline risk of cardiovascular disease. To the author’s knowledge, no such clinical trial has been registered to date [17]. As such studies take up to 10 years to be completed, it is going to be a long time until we have the answer to this very important question.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.