Meeting the Multi-Hormonal Challenge of Diabetes

At an industry-sponsored symposium held in conjunction with the American Association of Clinical Endocrinologists’ 2003 Annual Meeting and Clinical Congress, three leaders in diabetes research and treatment presented new information on recent advances in the understanding of the pathophysiology of diabetes with emphasis on how newly-discovered pancreatic and gut hormones interact with insulin and glucagon and how it might lead to new improved therapies for patients with type 1 or type 2 diabetes.

This program was supported by an unrestricted educational grant from AMYLIN Pharmaceuticals, Inc.

Intensive Glycemic Control in Type 2 Diabetes: How Important? What are the Barriers?

Endocrinologists are “glucose-centric” when approaching diabetes management, according to Harold E. Lebovitz, MD, Professor of Medicine, the State University of New York Health Sciences Center in Brooklyn.

“As data evolves, it becomes clear that although glucose is very important in the regulation and treatment of diabetes, it is not the sole issue,” said Dr. Lebovitz. “Other metabolic factors play a role in the long-term outcomes.”

The glycemic goal in treating diabetes is near normalization of hemoglobin A1C. Data from the Diabetes Control and Complications Trial (DCCT) in-dicates the greatest pubic health benefit is seen in reducing A1C from 9% or higher to 7% or lower.

Their data showed that lowering A1C by 10% resulted in a 39% reduction in clinically significant retinopathy during the study. Because microvascular complications occurred at geometrically higher rates as the A1C increased, a decrease from 11% to 9.9% reduced retinopathy by 6.6 cases per 100 patient-years. However, the same 10% change from 8 to 7.2, reduced retinopathy by only .95 cases per 100 patient-years of treatment (Diabetes Control & Complications Research Group. Diabetes. 1996;45: 1289).

“From a clinical standpoint, getting hemoglobin A1Cs of 8 and 9 down into the range of 7 has a much bigger public health value than getting people from 7 to 6.5,” said Dr. Lebovitz. “The question we need to answer clinically is: What are practical goals that reduce complications?”

The Steno 2 study looked at outcomes associated with the aggressive management of all metabolic abnormalities of type 2 diabetes. They followed 160 type 2 diabetics with microalbuminuria for a mean of 7.8 years. Participants were randomized to either conventional or intensive metabolic treatment (Gaede P, et al. NEJM. 2003;348: 383).

One of the noticeable results was that almost no one was on dietary therapy alone at study’s end. In addition, insulin use was increased over the course of treatment in both groups.

The intensively treated group had more patients on a combination of oral agents and insulin compared to the conventionally treated group. However, even the intensively treated group had only 15% of their patients achieve the glycemic target of A1C < 6.5%.

“The message I take away from the Steno 2 study is that controlling blood pressure, lipids, and the procoagulant profile are very important in reducing macrovascular complications in patients with moderately good glycemic control,” said Dr. Lebovitz.

The United Kingdom Prospective Diabetes Study (UKPDS) reveals that less than 50% of patients have A1C levels under 8% after ten years of treatment (UKPDS Group. Lancet. 1998;352:837). This shows, according to Dr. Lebovitz, that current practice is not good at controlling glucose. Follow-up results from the DCCT showed that patients with good control (7.1%) were back to 8.1% 6 years following their leaving the study and returning to their regular health care system (DCCT Research Group. NEJM.2000;342;381).

“These were people who knew how to control their blood glucose,” said Dr. Lebovitz. “They had 6.5 years of really intensive control, but needed the support of a research center to maintain that level.”

Results from the DCCT and the UKPDS shows that as A1C status improved, so did the amount of severe hypoglycemia. Results from both also show that insulin therapy is associated with weight gain (UKPDS Group. Lancet.1998;352:837; Purnell JQ, et al. JAMA. 1998;280:140).

Analysis of the lipid profile of patients in the DCCT on insulin who gained weight was very worrisome to Dr. Lebovitz. There was a progressive increase in triglycerides and LDL cholesterol coupled with a decrease in HDL cholesterol. While A1C levels decreased to 7.1, the accompanying weight gain and change in the lipids created new cardiac risk in these patients.

“If getting improved glycemic control comes at the expense of hypoglycemia and weight gain, is this really a benefit to our patients?” asked Dr. Lebovitz. “While we need 10 and 20-year studies to say for sure, the data is certainly disconcerting.”

Multi-Hormonal Regulation of Plasma Glucose: It’s not just insulin

According to Matthew C. Riddle, MD, Professor of Medicine, the Oregon Health & Science University in Portland, diabetes is not just a disorder of insulin. Rather, it is a much more complex disease and requires doctors treat it that way.

“Endocrinologists should expand our thinking from the conventional ‘giblets’, as Fuller Albright used to refer to the glands that are our livelihood,” he said. “We need to think about other tissues that produce and secrete regulatory products.”

Conventional treatments for diabetes often correct blood glucose levels in only one dimension, the fasting glucose.

“To go the next step, we’re going to have to blunt the postprandial increments,” he noted. “Doing so is going to allow us to reduce the fasting and basal levels further without risking hypoglycemia.”

Two hormones, insulin and glucagon, regulate endogenous glucose production. Varying the ratio between them changes the endogenous production of glucose in the liver.

In a normal person, as the fasting glucose (FG) rises, insulin production ramps up. A person with type 2 diabetes showed a smaller increase of insulin and those with impaired glucose tolerance (IGT) fell in between. To get enough insulin secretion to control hepatic glucose production, a fasting glucose of 90 to 100 mg/dl is normally needed. In the type 2 diabetic patient, it takes much higher levels of fasting glucose to get an equal amount of secretion. Treating the patient with sulfonylureas reduces the insulin deficit and reduces FG without changing postprandial hyperglycemia.

Also, the liver is less sensitive to fasting insulin levels in diabetes. At moderate concentrations of insulin in the portal vein, there is nearly complete inhibition of hepatic glucose under ordinary conditions. In the diabetic person, however, it requires 2 or 3 times more insulin.

This physiology forms the basis for the other widely used treatment, metformin. It is a liver-specific insulin sensitizer that lowers fasting glucose, but has modest impact on postprandial levels.

“Conventional treatments for diabetes focus on basal glucose control,” said Dr. Riddle. “However, the postprandial problems remain. We have begun treating elevated postprandial glucose levels with alpha-glucosidase inhibitors and rapid-acting insulin analogues. The results are not as good as
we would like.”

Normal regulation of glucose following meals is quite complex. There is improved clearance of glucose by the liver, suppression of hepatic glucose production and enhanced clearance of glucose from the peripheral circulation. Abnormalities of type 2 diabetes affect all three mechanisms.

Insulin secretion is delayed and prolonged and glucagon, a hormone that opposes insulin’s effects in the liver, increases rather than decline like it normally does. In studies, normal insulin levels were achieved infusing intravenous insulin. But this did not suppress glucagon and only partly blunted the postprandial increase in glucose (Unger RH, NEJM. 1971;285:285).

“We see this in many situations where the best effort has been made to control postprandial hyperglycemia,” said Dr. Riddle. “I think that perfect replacement of insulin is not going to do this, there are other factors at work.”

Among these are the gastrointestinal peptide hormones. Some are familiar such as glucagon and gastrin. New to the list are amylin, glucagon-like peptide-1 and peptide YY. These hormones regulate food intake by providing feedback on appetite and satiety as well as the rate of food delivery from the stomach.

There is currently an amylin analogue, pramlintide, nearing completion of the regulatory process. It is hoped that it will be released to clinical use soon.
After a meal, insulin secretion is increased with an effect on the liver and glucagon/insulin ratio. In addition, amylin is co-located and co-secreted with insulin by the pancreas, binds to receptors in the brain and activates pathways to change the action of the glucagon-secreting cells. Studies have shown that amylin slows gastric emptying, prevents meal-related increases of glucagon and limits food intake (Young AA. Curr Opin Endocrinol Diabetes. 1997;4:282)

In type 1 diabetes, there is no insulin or amylin. In type 2, there are normal basal levels of insulin and amylin with blunted and prolonged secretion of both. This is a deficiency syndrome that might be amenable to treatment (Edleman SV, et al. Diabetes Technol Ther.2002;4: 175).

GLP-1 is a gut peptide secreted by the large and small intestines that might prove useful as well. Although its main effects are being debated, insulin secretion and potentiation of the growth of beta cells are among them.

“There is clearly a coordinated system of hormones affecting glucose levels following a meal; it is not just insulin postprandially,” said Dr. Riddle. “Gastrointestinal peptides such as amylin contribute to postprandial glucose reduction. Soon, we are going to be able take advantage of that as we begin to see analogues of these peptide hormones currently under development.”

Clinical Experience: Amylin Hormone Replacement in Patients with Diabetes

Type1 diabetics who are insulin deficient and type 2 patients requiring insulin are the initial populations targeted for possible treatment by pramlintide.
“When pramlintide is added to insulin therapy in type 1 diabetics, there is a dramatic blunting of postprandial glucose rise,” said Stephen L. Aronoff, MD, Clinical Endocrinologist, Endocrinologists of Dallas, TX.

Clinical trials by Fineman and associates have demonstrated that both type 1 and type 2 diabetics have a dramatic suppression of postprandial glucagon rise when pramlintide is given along with a patient’s insulin prior to the meal (Fineman MS, et al. Metabolism. 2002;51:636; Fineman MS, et al. Diabetologia. 1998;41:167A).

A study by Kong and others have shown a dose-dependent delay in gastric emptying. Participants were type 1 patients who were given their insulin dose with a placebo or 30, 60 or 90 micrograms of pramlintide. They were given a breakfast of radiolabled pancakes and milkshakes. There was a dose-dependent delay in gastric emptying in those patients given medication (Kong MF, et al. Diabetologia.1998:41:577).

The patients were given a follow-up meal about four hours later but no additional pramlintide or placebo. There were no differences in gastric emptying indicating the effect is short-lived. Another paper by Thompson and associates has shown suppression of postprandial glucose with amylin replacement in type 2 diabetes when pramlintide is used with insulin (Thompson RG, et al. Diabet Med.1997;14:547).

“This was also seen in type 1 patients, although the pattern was different,” said Dr. Aronoff. “In those getting the insulin/pramlintide combination, there was a fall in glucose for 30 to 60 minutes. After that, the glucose rises, but not above the baseline level.”

The effects of amylin replacement in patients using insulin analogues such as lispro have also been studied in type 1 and type 2 patients. Lispro was given immediately prior to the meal and the amylin given either 15 minutes prior to the meal, at the time of the meal, 15 minutes after feeding or 30 minutes postprandial (Weyer C, et al. Diabetologia. 2002;45 (suppl 2):A240) (Figure 1).

When pramlintide was given with the meal, glucose fell and remained suppressed for at least three hours. If given 15 minutes prior to the meal, there was an initial blunting, but glucose levels rose at 120 minutes to elevated levels. When administered at either time after the meal, there was an initial rise with a subsequent blunting of the glucose response. The same response was seen in type 2 diabetes (Maggs DL, et al. Diabetologia. 2002; 45 (suppl 2):A 264).
“It appears that the ideal time to administer pramlintide is at the meal,” said Dr. Aronoff. “Pramlintide, when added to lispro, has as pronounced an effect as that seen in patients receiving regular insulin and amylin replacement.”

The effects of pramlintide are persistent when it is given with each meal. Levetan and colleagues looked at patients taking pramlintide and insulin for 28 days. During the test, the patients were given their usual insulin and either pramlintide or a placebo prior to breakfast. There was no rise in glucose following the meal in those given amylin replacement. The same results were seen with each meal (Levetan C, et al. Diabetes Care. 2003;26:1).

There have been long-term trials into the effect of pramlintide on overall glucose control. Whitehouse and colleagues placed patients on insulin plus placebo or 60 micrograms of pramlintide four times a day. Upon entry, the hemoglobin A1C was 9% and the average length of diagnosis was 19 years (Whitehouse F, et al. Diabetes Care.2002;25:724).

Within four weeks, there was a significant drop in A1C levels. The difference in A1C, when compared to those on insulin-only, was significantly different throughout the 52-week trial.

“There was also an open-label extension where the patients on placebo were openly randomized to insulin plus pramlintide,” said Dr. Aronoff. “Over a short period of time, the A1C dropped to levels seen in the pramlintide arm during the previous year. The medication seems to work long term in these
patients.”

Whitehouse’s group then looked at body weight. Those on insulin only had a significant weight gain during the yearlong study when compared to no weight gain in those taking the insulin/pramlintide combination. In a sub-analysis of those with BMIs < 27 at entry, there was greater weight gain in those getting only insulin and a significant weight loss in those receiving pramlintide with insulin.

“Those trying to control with insulin only had an increase in insulin requirements and didn’t see as great a drop in A1C s as those taking insulin and amylin replacement,” said Dr. Aronoff. “Those taking both also had a tendency to decrease or have no significant change in insulin dosing. Furthermore, those receiving pramlintide actually lost weight in many cases.”

Studies have been completed with type 2 diabetic patients. Again, those receiving pramlintide had a significant decline in A1C that was apparent at 4 weeks and persistent through the entire year. There was also less change in insulin dosage in those taking both medicines and the weight loss seen in type 1 was also seen in the type 2 diabetic patients. Pivotal studies by Amylin found similar outcomes. (Fineman MS, et al. Dia-betes.2000; 49 (Suppl 1):A106; Data on file, Anylin Pharmaceuticals.) (Figure 2).

Two adverse events that must be considered with pramlintide are nausea and hypoglycemia. Both seem to be transient. Recently completed studies have assessed the effects of titration on these concerns.

Much of the worry about hypoglycemia was based on an initial study showing a significant increase in severe hypoglycemia in the pramlintide treated patients. The study design may have been at fault by not allowing for changes in insulin dosage when pramlintide was added. Once it was realized that there was an improvement in insulin efficiency with the added medication and the insulin dose was adjusted down, the hypoglycemia rate in those entered subsequently was reduced (FDA Advisory Committee Presentation, July 2001).

“Pramlintide itself does not cause hypoglycemia,” said Dr. Aronoff. “There have been studies in normal volunteers given 10 mgs (100 times the therapeutic dose) with no hypoglycemia. The addition of pramlintide to insulin therapy may impact the event rate of hypoglycemia by increasing insulin efficiency which calls for a decrease in the insulin dosage.”

Nausea has also been reported as an adverse effect. Most has been reported to be in the mild to moderate range and appears to be dose related. Titrating the medication appears to help.

Dr. Aronoff noted that there have been more than 4500 patients treated with no indications of cardiac, hepatic or renal toxicity. There have also been no clinically significant changes in lipid profile, EKGs, vital signs, or laboratory parameters.

“The multi-hormonal regulation of glycemic balance is disturbed in patients with diabetes,” said Dr. Aronoff. “Treatment with exogenous insulin and the amylin analogue pramlintide show promise for restoration of this balance.”