ered essentially unbeatable, was one of the major medical breakthroughs of the 20th century. As a result, the Nobel Prize in 1923 went to Banting and department head John Macleod. Although Best's and Col lip's roles were unrecognized by the Nobel committee, the four researchers shared the award.
Shorliy after Banting and Best's breakthrough, insulin became widely available. Since that initial treatment over 80 years ago, there have been incredible advances in the development of formulations with different onset of action, duration, and roles in therapy. Advances have also been made in our understanding of the pathophysiology of both type 1 and type 2 diabetes, as well as of the molecular Structure of insulin. Subsequently, techniques of recombinant DNA technology have been developed and synthetic human insulin has become a reality, Advances made by manipulating insulin's structure have led to a number of short- and long-acting analogs that provide tremendous options for patients. In particular, the short-acting analogs represent an important therapeutic advance. These analogs, intended for use just prior to ingesting a meal, more closely mimic the pharmacokinetic and pharmacodynamic properties of endogenous insulin and overcome the limitations of injected Regular insulin.
However, despite the remarkable advances made in understanding insulin's structure, activity, and physiological role, a major limitation that exists even today is that the only available route of insulin administration is still by injection. Attempts have been made to develop alternative means of delivery, and significant technological innovations have madeself-injection easier. For example, the insulin pen devices use smaller-gauge needles, are convenient, offer more accurate delivery, are less painful to use than conventional needles, and may enhance compliance. Unfortunately, injection with a pen is still an injection, and self-injection is required several times a day to optimize glucose control. In addition to the goal of noninvasive insulin delivery to eliminate the injection, broader and more important aims of alternative insulin delivery should include mimicking the normal physiological insulin time-concentration profile as closely as possible. As discussed in Chapter 1, a 24-hour normal endogenous profile can be divided into basal levels of insulin in postabsorptive states and the postprandial rise after meal ingestion. The physiological insulin profile peaks within 30-60 minutes of eating and returns to baseline within 2-4 hours. Unfortunately, following a subcutaneous injection, Regular human insulin has a slow absorption, meaning that patients may have to inject 30-60 minutes prior to meals and insulin may persist in the circulation for 4—6 hours. Additional concerns for the intermediate- and long-acting human insulin preparations, e.g., NPH and Ultralente, is that they may have an undesirable peak profile after subcutaneous injection while still lacking sufficient duration in action. It is for this reason that the short- and long-acting analogs of human insulin were developed that offer a more physiological insulin pharmacokinetic profile than that achieved with previous formulations. Unfortunately, despite the ability of the new formulations to closely mimic a normal i
physiological profile, all these preparations still have the drawback that they re-, quire an injection.
THE SEARCH FOR A PRACTICAL NONINVASIVE INSULIN-DELIVERY SYSTEM
The major goals when considering a practical noninvasive insulin-delivery system are 10: 1) overcome the major limitation associated with conventional insulin injections and 2) preserve a more physiological insulin profile. Achieving these goals will allow for more intensive insulin delivery, a regimen clinically proven to significantly improve glycemia and reduce complications, while enhancing patient compliance. Such goals are lofty, but they are receiving increased research attention. Current areas of ongoing research are outlined in Table I.
Jet injectors are devices (hat administer insulin without needles by delivering a high-pressure stream of insulin into subcutaneous tissue. Even though needles are not used, (he discomfort associated with jet injectors is reported to be comparable to thai with injections, fn subjects with great anxiety about needles, however, these devices can obviously have a benefit. Unfortunately, these devices may negatively impact the pharmacokinetics of long-acting insulin, forc-
Table 1 Current Approaches to Penetrating the Skin Barrier
Jet injectors: Devices th8t administer insulin without needles by delivering a high-pressure stream of insulirt into subcutaneous tissue lonotophoresis: Process using low-level electrical current to speed the delivery of drug ions into the skin and surrounding tissues Low-frequency ultrasound: Process by which ultrasound increases, by several-fold, the permeability of human skin to macromolecules such as insulin
Transfersomes: Composites of pharmaceutical^ accepted ingredients designed to deliver drugs across the skin barrier by opening temporary channels between adjoining cells -g
Intranasal administration: Delivery of insulin to nasal mucosa for absorp- |
tion; bioavailability is low Oral insulin: Via the buccal mucosa in the mouth or mucosa in the gastro- 'Si intestinal tract ^
Pulmonary delivery: Takes advantage of favorable pulmonary anatomy j
(e.g., permeability, increased surface area) for absorption of insulin for rapid systemic effect 3
@ •S, ing more frequent administration . Although they are not widely used, they have been recommended only for people intolerant of conventional needle injection systems.
Iontophoresis is a process very similar to that of the passive transdermal medication patches that deliver nicotine for smoking cessation or hormone therapy for postmenopausal women, iontophoresis uses low-level electrical current to speed the delivery of drug ions into the skin and surrounding tissues. It appears to he an effective and rapid method of delivering medication into the skin. Such a process has been evaluated for insulin in diabetic rats that had been depilated; transdermal delivery of bovine insulin using iontophoresis produced aconcentra-tion-dependent reduction in plasma glucose levels. However, the method did not appear to be as effective in rats that had not been depilated, which suggests that either the depilation was effective in reducing the skin's barrier function or the creams used with the iontophoretic device acted as a penetration enhancer. Although this is an area of great interest, the factors critical to transdermal absorption and delivery of insulin will need to be defined in further studies.
The use of low-frequency ultrasound has also been evaluated as a means of noninvasive! y administering insulin. It has been demonstrated to increase, by several-fold, the permeability of human skin to macro molecules. This technique may, augment delivery of drugs such as interferon-gamma and erythropoietin. However, given that a goal of noninvasive delivery is to mimic physiological insulinl patterns, this approach provides a rate of delivery of insulin that is too slow to! be a viable clinical option.
being some 1000 times larger. With particular regard to insulin, transfersomes coupled with insulin allow insulin to cross intact skin cells with a bioeffi-ciency of approximately 50% of the subcutaneous dose. It has been estimated that, with present technology, a skin surface area of approximately 40 cm2 is sufficient to cover the basal daily insulin requirements of most patients with type 1 diabetes. However, limitations regarding delay in onset, making prandial use problematic, are of concern. Additional studies will be needed to determine the clinical utility of this approach.
The use of intranasal insulin as a viable clinical option has received considerable attention. The major problem associated with nasal administration of insulin is poor bioavailability, which is typically observed to be between 8 and 15%. Factors affecting bioavailability include timing of dose and frequency of administration. Nasal formulations have used permeability enhancers to augment bioavailability, but these result in nasal irritation in many patients. Nasal insulin, however, has been shown to have clinical effect. The effectiveness of intranasal insulin in 31 patients was evaluated hy Hilsted and colleagues in 1995. They reported that the insulin dose needed to reach given markers of glycemic control by intranasal administration was 20 times higher than that given by subcutaneous injection. Although no difference in the number of hypoglycemic episodes between subcutaneous and nasal delivery was observed, markers of metabolic control worsened slightly but significantly during nasal insulin treatment. Nevertheless, nasal insulin has been shown to reduce postprandial glycemia. The reduction is especially marked following repeated dosing. Therefore, intranasal administration of insulin is promising, and has shown to be effective in lowering glucose. However, further studies to establish long-term safety, patient acceptance, and effectiveness are needed.
of caproic acid molecules and coating with chitosan stabilizes degradation and improves permeability, 2) concurrent administration with protease inhibitors; experiments in dogs suggest that protease inhibitors may augment the oral bioavailability of insulin, but this has not been tested in humans and the long-term side effects are unknown, 3) enclosing insulin within microspheres, protecting against hydrolization or enzymatic degradation by attaching insulin to carrier molecules, and 4) use of absorption enhancers, an approach similar to nasal formulation to enhance buccal delivery and that avoids reliance on lower gastrointestinal absorption.
Of all the suggested approaches, it appears that only buccal delivery has recently advanced to more widespread testing in patients with diabetes, and reports suggest clinical benefits.
Many of the limitations outlined above that exist for other routes of noninvasive insulin delivery appear to be nicely addressed because of Lhe favorable anatomy of the lung. First, the lung's large surface area is highly permeable, and it is suggested that peptides such as insulin will cross the alveolar cells in a process called transcytosis, although this has not been conclusively proven. The process is believed to occur as follows: 1) the inhaled insulin molecules, once deposited in the alveoli, are believed to be taken up into vesicles; 2) the insulin particles are then transported across the epithelial cells and released into the interstitial fluid between the epithelium and the alveolar capillary endothelium; and 3) the insulin is then taken up into vesicles, transported across the capillary endothelium, and released into the bloodstream (Figure 1). This process results in a rapid systemic effect as pulmonary inhalation of insulin results in peak levels after 1520 minutes. A second major advantage for pulmonary delivery is the large surface area of the lung—the normal human lung may have a surface area of 100 m2. This large surface area, combined with the favorable permeability properties, makes the lung a very attractive route for the administration of insulin.
The feasibility of inhaling insulin is not a new idea; it was first reported three years after Banting and Best's extraction of insulin. In 1925, Gansslen observed a lowering of blood glucose approximately 2 hours after administering pulmonary insulin in five diabetic patients. Despite Gansslen's finding, this noninvasive approach was not pursued further until Wigley and colleagues revisited pulmonary insulin delivery over 45 years later. They reported increased levels of plasma immunoreactive insulin in both normal volunteers and patients with diabetes after delivering pork-beef insulin using a nebubzer. The observation that hypoglycemia closely followed the rise of plasma insulin confirmed proof of concept for the approach.
Once the feasibility of pulmonary delivery had been established, the next step was to design an effective pulmonary delivery system appropriate for patient use. In this regard, the vast experience gained in the management of asthma has greatly aided the development of pulmonary delivery systems for insulin. The asthma drug-delivery devices—e.g., nebulizers, metered-dose inhalers, and dry-powder inhalers—although very useful in the management of local respiratory diseases such as asthma, are not appropriately designed to deliver drugs deep into the alveoli for systemic drug delivery. For inhaled insulin to work effectively, a device would need to allow insulin to reach the alveloli, and to do so in a very reproducible manner. The pulmonary delivery systems currently in clinical development have been designed to deliver both dry-powdered and soluble insulin to the alveoli (for reviews, see Klonoff, 1999, and Greener, 2000).
In addition to the requirement that the device be able to administer insulin to the deep reaches of the lung, a number of other potential limitations must be overcome for effective pulmonary drug delivery (Table 2). One factor is particle size—it has been suggested that the lungs very effectively filter particles with -g die exception of very small (mean aerodynamic diameter <5 |im) particles. These |
observations established that particle size is a main determinant of efficient pul- |
monary insulin delivery. Studies have established that particles should have low velocity and an aerodynamic diameter of 1-3 |im. Larger particles are observed ?
to deposit in the bronchial tubes and smaller panicles are exhaled. Thus, the early a attempts to develop pulmonary delivery systems, by not sufficiently controlling js particle size, failed to reliably and sufficiently deliver drug. Only in the recent q
Table 2 Factors Affecting Pulmonary Drug Delivery
Type of propellants utilized Airflow speed
Losses within the device and the environment
Particle size and velocity
Drug clearance and absorption
Drug deposition into the throat and bronchial tubes
Potential impsct of concomitant diseases past have technological advances in manufacturing, by now having the ability to control particle size, enabled the development of a range of inhaled formulations and delivery systems including the dry-powder insulin inhalation systems and aqueous insulin aerosol. These systems are undergoing extensive clinical evaluations, but, as described, the dry-powder formulations may have several advantages over conventional liquid formulations, including product and formulation Stability, and high-drug-volume delivery.
Once the aerosolized drug reaches the alveloi, its bioavailability must be high enough to make the delivery system feasible. This was the apparent problem with the earlier studies. Despite ihe successful demonstration of the principle of insulin inhalation for the treatment of diabetes in 1925, the bioavailability was low (<3%). However, more recent inhalation studies that have compared insulin administration by inhalation devices with subcutaneous injection for reproducibility of dosing have shown that the variability in glucose response to the two methods was equivalent. Recent observations suggest that the bioavailability with aerosol insulin is approximately 20%, supporting the use of pulmonary delivery as a method of insulin administration.
Although studies have not evaluated pulmonary delivery in pulmonary diseases such as emphysema or COPD, it has been suggested that acdve smoking greatly influences the pharmacokinetic profile. It appears that smoking, while not altering the time of peak effect (approximately 15 minutes), can greatly elevate the concentrations obtained.
One of the major advantages of inhaled insulin is that its pharmacokinetics appear to be similar to those of the new insulin analogs and may be closer to a physiological profile. In support of such observations is the evidence suggesting that the time—concentration profile of inhaled human insulin in blood is similar to that of physiological insulin secretion in healthy volunteers. Several studies have eval-
Was this article helpful?