New Human Insulin Analogs and Intensive Insulin Therapy: Mimicking the Body’s Response

At a symposium held in conjunction with the American Association of Clinical Endocrinologists annual meeting in Chicago, Illinois, three leaders in diabetes presented the latest research on intensive insulin therapy (IIT), which has been proven to reduce the incidence of long-term complications such as retinopathy, neuropathy, and nephropathy in diabetic patients. Topics included how insulin analogs are being developed to mimic the body’s response during normal and stressful conditions, as well as recent developments in infusion and monitoring apparatuses that are becoming available.

This program was supported by an unrestricted educational grant from Novo Nordisk Pharmaceuticals.

Insulin Analogs as a Focus

The primary goal of intensive insulin therapy (IIT) is to provide diabetic patients with near-normal glycemia. The goals of IIT should also include avoiding short-term crises (i.e., hypoglycemia, hyperglycemia), minimizing long-term complications, and improving quality of life.

“The most powerful agent we have to control glucose is insulin, and with the recent recombinant DNA technology the development of insulin analogs has greatly improved the likelihood that patients will live longer and have a greater quality of life,” stated Bruce W. Bode, MD, medical director, Diabetes Resource Center, Piedmont Hospital, Atlanta.

Insulin Analogs
Regular insulin is a highly effective treatment available in several different formulations. First among these were modified insulins such as PZI and NPH, followed by lente and ultralente. “These were not really basal insulins, they were peaking insulins,” said Dr. Bode, adding that “they had problems with marked variability in insulin absorption and not lasting 24 hours.” These problematic time limitations lead to the development of several insulin analogs.

Most of the concern with mealtime insulins centers on the need to inject 30 minutes prior to meals. As a result, rapid-acting insulin analogs (lispro and aspart) were developed that could be taken with a meal instead of before a meal. The rationale for using these rapid-acting insulins is fourfold: they can be administered at mealtime, they mimic the physiologic insulin profile, they improve postprandial glycemic control, and they lower the risk of late hypoglycemia. In addition to these rapid-acting insulins, there is a long-acting insulin analog (glargine) that provides peakless basal insulin coverage lasting up to 24 hours (Table 1).

Rapid-Acting Insulin Analogs
Rapid-acting insulin analogs are valuable because they best mimic normal physiologic response to a meal. The first analog developed was insulin lispro which is similar to normal insulin except for the switching of amino acids at positions B28 and B29 (lysine and proline). Insulin aspart was the second analog developed and had a single amino acid substitution at position B28, proline changed to aspartic acid. Both analogs have their maximum effect within 30 to 60 minutes, and in several trials they have been shown to control postprandial glucose better than regular human insulin (Figure 1).

According to Dr Bode, rapid-acting insulins are more predictable, and they reduce the number of hypoglycemia events. Although quite similar, the two rapid-acting insulins do have a few differences. In a study by Hedman and colleagues, insulin aspart was found to reach a maximum serum insulin concentration about 5 minutes slower than insulin lispro. However, the decrease in serum insulin from maximum to 50% of maximum was significantly longer (~40 min) for insulin aspart which may lead to a longer duration of action compared to insulin lispro.

A comparison of the two rapid-acting insulins was made in a multicenter study involving 146 subjects with type 1 diabetes. Dr. Bode and colleagues found both rapid-acting insulins to control postprandial glycemia better than regular insulin. In regard to reported episodes of hypoglycemic symptoms, insulin aspart had a 30% relative reduction compared to regular insulin or insulin lispro. Finally, an examination of pump compatibility found insulin aspart to be as effective as buffered regular insulin or insulin lispro.

Dr. Bode summarized the data by stating “insulin aspart has faster onset and shorter duration of action (than regular insulin). This has resulted in convenience and flexibility with mealtime dosing. It has also resulted in improved postprandial glucose control without increased hypoglycemia. It’s approved for insulin pump therapy and has benefits over the other current insulins on the market for use in insulin pump therapy.”

Long-Acting Insulin Analogs
The medical rationale for using a long-acting insulin analog is to better mimic the basal physiological insulin profile of normal individuals. “The limitations of NPH, lente, and ultralente are that they do not mimic basal insulin levels,” said Dr. Bode. They are all peaking insulins and have problems with variation in absorption rates and in sustained insulin concentrations. NPH and glargine vary about 25% and ultralente 52%. These modified insulins also have unpredictable hypoglycemia.

Overall, glargine is the preferred long-acting analog with once-daily dosing providing a smooth, peakless action profile, which lowers risk of nocturnal hypoglycemic events, has better glycemic control, and has a safety profile similar to regular insulin.

Currently in development is insulin detemir which binds to serum albumin to prolong its circulation in the body and extend its duration of action. While still in phase III development, insulin detemir appears to be very effective for basal insulin control.

Conclusion
The use of a rapid-acting insulin analog that optimizes postprandial glycemic control, in combination with a long-acting insulin analog that provides consistent basal insulin needs indicates that “recombinant DNA technology has fulfilled our promise of giving us better insulins,” concluded Dr. Bode.

Intensive Insulin Therapy

The logo for the American Diabetes Association is a triangle that symbolizes the need to balance three factors: diet, exercise, and insulin. According to Lois Jovanovic, MD, Director and Chief Scientific Officer of the Sansum Medical Research Institute, Santa Barbara, California and Clinical Professor of Medicine, University of Southern California, this triangle should be replaced by a square to acknowledge stress as a factor. “Stress comes in three forms: psychological, physical, and hormonal,” said Dr. Jovanovic. “The body responds to all three stresses the exact same way, by outpouring anti-insulin hormones. So as stress swells up, our blood sugars rise.”

Psychological stress can have several levels and is generally difficult to study. Physical stress such as trauma, inflammation, or infection can greatly affect glucose levels. Hormonal stress has a great impact on daily glycemic control. Hormonal changes occur throughout a patient’s lifespan, ranging from early growth and development to puberty, menstrual cycles, and pregnancy, and it can also be affected by exogenous hormone use (i.e., HRT, corticosteroids).

Calculating Insulin Requirements
To illustrate how stress can affect a patient’s 24-hour insulin requirement, Dr. Jovanovic used the formula I = ‘wgt in kilograms’ x ‘constant’, where the constant was an indication of stress. For mild stress, the constant is 0.7–0.8, for moderate stress it is 0.8–1.0, and for severe stress it is 1.0–2.0. Dr. Jovanovic provided the audience with a list of various patient descriptions with different constant values. For example, a trained athlete has a constant of 0.5, whereas a female during the luteal phase of menstruation would have a constant of 0.7. A constant of 1.0 would represent a person with a bacterial infection or during term pregnancy. A constant of 1.5–2.0 would represent a severely stressed person or an adolescent in the peak of pubescence. To illustrate, Dr. Jovanovic used the example of puberty showing insulin requirements to almost quadruple in the early teens from a value of 0.5 to 2.0 units per kilogram for 24 hours.

The ‘I’ in the equation must accommodate both basal (B) and meal-related (M) needs. Generally, both B and M are equal and each represents 1¼2 I. In a healthy adult, daily basal insulin needs (B) vary during the 24-hour cycle. In most adults, B is fairly stable from 10 a.m. to midnight. From 12 a.m. to 4 a.m., insulin requirements can drop to half; from 4 a.m. to 10 a.m., they can double.

To illustrate how to calculate insulin requirements, a sedentary (constant = 0.6) 80 kilogram person would need 48 units (0.6 x 80). “Those 48 units then need to be divided such that they provide the basal need and the meal-related need,” said Dr. Jovanovic. As a result, the basal need would be 24 units over 24 hours or 1 unit/hour. When basal insulin needs are stable (10 a.m. to midnight) 1 unit/hr is required. When hormones drop after midnight, the insulin need would also drop to about 0.5 units/hour. Finally, when hormones increase between 4 a.m. and 10 a.m., the insulin need would go up to 1.5 units/hour in this
example.
Dr. Jovanovic also pointed out that basal insulin needs are complicated by behavioral factors such as large bedtime snacks or irregular sleeping habits that can greatly alter the basal insulin requirements. Clinicians must always remember to treat each patient individually and data should be collected to better understand the patient’s basal glucose patterns.

Meal-related Insulin Requirements
To understand the carbohydrate content of a meal plan and cover it with a perfect dose of insulin requires almost a graduate degree in nutritional science, joked Dr. Jovanovic. Fortunately, with the introduction of rapid-acting insulin analogs, the need to “feed the insulin” with snacks is no longer necessary and that these insulins do not remain long enough to create hypoglycemic problems in the middle of the night. Instead, with the new rapid-acting insulin analogs, we can now “perfectly match the insulin to be about one unit of insulin per ten grams of carbohydrate in the meal plan,” she said.

Insulin During Pregnancy
Dr. Jovanovic ended her presentation with a brief summary of insulin use during pregnancy. Of great concern to all pregnant women with diabetes is the risk of macrosomia, or ‘big bad baby syndrome,’ that increases dramatically with elevated postprandial blood sugar. Fortunately, Dr. Jovanovic said that the introduction of insulin analogs has greatly improved postprandial glucose control in pregnant women with diabetes.

Dr. Jovanovic said to date, “there is no data on the use of long-acting insulin analogs during pregnancy and, although they are likely safe and effective, until clinical trial data becomes available they should be used cautiously.”

Devices: Monitoring and Treatment

In the 1920’s, animal-derived insulin was taken 6 to 8 times a day. These products were highly contaminated solutions that had to be administered with large syringes and led to severe local reactions. “You didn’t hear a mother and you didn’t hear a child complain because it kept them alive. That’s not the case in our clinics today. Today, people complain,” said Robert E. Ratner, MD, of Medstar Research Institute and George Washington University Medical School in Washington, DC. To illustrate, Dr. Ratner showed the audience a cartoon by one of his patients showing her interpretation of insulin injections. The cartoon depicted a syringe that was bigger than the frightened patient.

Insulin Pens and Pumps
There are currently a large variety of devices becoming available. “…whether we’re looking at reusable pens or disposable pens, we’re now down to pen juniors in half-unit increments for kids,” said Dr. Ratner, adding that “the ability to carry your insulin around with you in a pen device is the greatest motivator to moving people onto basal/bolus therapy. If you have to carry around a vial and a syringe, you’re not going to do it.”

In addition to concerns about compliance, trials have shown that pumps do improve glucose control compared to multiple injections. Dr. Ratner mentioned that pumps continue to get smaller and there is even a disposable insulin infusion pump undergoing FDA approval that can be placed on the patient’s abdomen to infuse insulin for up to 3 days after which it can be thrown away.

Ultimately, the best way to control insulin is to ‘close the loop,’ and implantable pumps are currently being developed to accomplish this. “This pump is implanted underneath the skin, typically in the abdominal wall, with a catheter going into the peritoneum and free-floating within the peritoneum,” said Dr. Ratner. “You then load the pump with U-400 insulin transcutaneously.” Compared to peripheral pumps and multiple injections, they are equally effective in controlling hemoglobin A1C and they appear to lower the risk of hypoglycemia.

In order to close the loop completely, a glucose sensor is also required. Originally, the Bio-Stater was available but it was costly, heavy, and required three intravenous lines. To date, Dr. Ratner said, better sensors are still not available.

Oral and Pulmonary Insulin
Dr. Ratner said that there are efforts underway to develop oral insulin but this may not be available in the near future. Oral insulin is problematic because gastric juices and the pancreatic enzymes break down the insulin polypeptide.

Another route of administration currently in development involves pulmonary insulin. A pulmonary delivery system is noninvasive and insulin delivered via the lungs is rapidly absorbed into the circulation. A pulmonary system called Exubera is about the size of a soda can and involves inhaling the dry powdered insulin. Because the insulin is a dry powder, it is not measured in units but by milligram, so Dr. Ratner cautioned that “we’re going to have to change our paradigm when we begin to look at this.” One disadvantage to this system is the dry powder is administered in increments of 3 mg, 6 mg, or 9 mg, limiting its dosing flexibility.

Another device in development is the AERx system, a device that uses a small blister pack that holds ten units of insulin. The patient can set the amount of insulin to be delivered within one unit increments. Pharmacokinetic studies have shown that it is absorbed much faster than a subcutaneous dose, making it ideal for bolus insulin requirements. A major concern with pulmonary insulin devices is that it is difficult to control the patient’s breathing. The AERx system will only deliver insulin if proper breathing techniques are used, thus ensuring consistent dosing of insulin to the deep lung (Figure 2).

Glucose Sensors
Glucose sensors that are noninvasive or implantable are being developed, but there are several problems with each. Noninvasive sensors, according to Dr. Ratner, have been “right around the corner for the last 15 years,” but are unlikely to be available soon. Implantable sensors are also being tested. Unfortunately, they are reliable for only short periods of time. The body tends to develop scar tissue around the sensor making it difficult to accurately measure glucose in the long term.

Most likely, the best glucose sensors will be ‘minimally invasive.’ Minimally invasive sensors measure interstitial fluid glucose not plasma glucose, so Dr. Ratner cautioned that this can be an inaccurate assessment because glucose levels in the plasma will change more quickly than in the interstitium when a person exercises or eats.

Conclusions
“As our understanding of diabetes continues to grow, so does our understanding of the need to improve the devices that control insulin and measure glucose,” said Dr. Ratner. He is optimistic that in the near future, “we’re going to have a whole lot more exciting devices for the treatment of our patients with diabetes.”