OBJECTIVE—To comprehensively review the progress to date on the development of alternative routes for insulin delivery.

RESEARCH DESIGN AND METHODS—Study data were collected through a Medline review.

RESULTS—Proof of principle has been established for many routes of administration including dermal, nasal, oral, buccal, and pulmonary insulin delivery.

CONCLUSIONS—Of all the approaches to date, pulmonary delivery appears to be most feasible. Ongoing phase III studies will ultimately determine safety, tolerability, and efficacy before approval for clinical use.

The therapeutic insulin era began on 11 January 1922 with the first clinical use of insulin by Banting and Best. In <2 years, insulin moved “from the realm of hypothesis to the reality of treatment” (1). In the ensuing 80 years, scientists uncovered the basic pathophysiology of diabetes, gradually elucidated insulin’s structure, and focused their attention on developing better insulin formulations (e.g., NPH, lente). In this regard, the development and availability of rapid-acting (e.g., lispro, aspart) and basal insulins (e.g., glargine) have allowed for the routine use of very physiological insulin regimens.

However, despite significant research, the first effective noninvasive delivery systems for insulin are only now in development, marking a new milestone in effective management of diabetes. There is a clinical need for such because insulin therapy is central to the treatment of people with type 1 diabetes. Intensive insulin therapy, in particular, is associated with better long-term clinical outcomes. The progressive decrease in β-cell function evident in the pathogenesis of type 2 diabetes means that the majority of patients will eventually fail on oral therapy and will require insulin therapy at some point. Furthermore, oral agent failure is just one of a number of situations where insulin therapy might be indicated in patients with type 2 diabetes (2).

Despite the benefits of tight glycemic control, which ultimately can only be achieved with insulin in type 2 diabetic patients, there is reluctance on the part of physicians and patients to initiate insulin therapy (2). Factors such as weight gain, social stigma, lifestyle restriction, injection anxiety, sense of guilt or failure, and perception of worsening disease may affect patients’ attitudes. Although hypoglycemia and weight gain are also perceived as barriers, these factors should not discourage the start of insulin. This reluctance contributes to poor glycemic control, and any advances that can help to minimize such concerns clearly have a place in diabetes management.

Originally, insulin was administered intramuscularly. It soon became clear that subcutaneous injections were just as effective but considerably less traumatic (3). Over the years, researchers have suggested transdermal, ocular, oral, buccal, nasal, rectal, vaginal, and uterine delivery systems (3,4). However, subcutaneous administration has remained the route of choice, although the original researchers recognized the limitations. Best wrote the following in 1974: “We hope for and expect more physiological methods of giving insulin” (1).

Until recently, many believed the low bioavailability of insulin delivered noninvasively meant such methods would not be clinically realistic. This view is no longer sustainable in light of recent history, and during the next few years, several novel delivery devices could be launched that may result in improved glycemic control for many patients.

The following exemplify the different approaches tried during the history of insulin delivery and the current status of development (Table 1).

Jet injectors

Jet injectors are devices that administer insulin without needles by delivering a high-pressure stream of insulin into subcutaneous tissue and were first proposed over 40 years ago (3,5). It was suggested that insulin doses delivered may be more precise and more quickly absorbed than subcutaneous needle delivery; but in reality, the time-action profile of intermediate-acting insulin may be altered with these devices (5). Further, discomfort associated with this technique may be comparable to, if not greater than, that associated with injections. This method may benefit selected cases, such as those with severe insulin-induced lipoatrophy (3). However, it has been suggested that jet injectors should be advocated only for people who develop problems with other injection routes (5).

Transdermal

Human skin is an extremely effective barrier in protecting against the entry of foreign proteins. Attempts to overcome this skin barrier to allow insulin transfer have included techniques to weaken the barrier with chemical substances, ultrasound techniques, or electric current (6).

Iontophoresis.

Iontophoresis refers to the concept of achieving transdermal delivery of a protein (e.g., insulin) by direct electric current. An analogous concept may be the passive transdermal medication patches currently used to deliver nicotine for smoking cessation programs or hormone therapy for postmenopausal women. However, iontopheresis differs by using a low-level electrical current in the process, enhancing the delivery of drug ions into the skin and surrounding tissues. Depending on the net charge of the insulin molecule, the applied electrical potential has been shown to increase the rate of insulin transfer across skin (6,7). Proof of concept has been shown in studies demonstrating that iontophoretic delivery of bovine insulin in a study of depilated diabetic rats produced a concentration-dependent reduction in plasma glucose levels (8). Interestingly, efficacy did not appear as high in nondepilated rats. This result suggests that the creams used acted as a penetration enhancer only in animals treated in advance with a depilatory cream, rather than having their hair removed with scissors, or that the depilation per se was effective in reducing the skin’s barrier function. However, iontophoretic delivery of a monomeric human insulin analog produced a significant fall in plasma glucose in the rats, suggesting that the type of insulin may be a factor (8). With this technique, the rate of insulin transfer may be appropriate for the coverage of basal insulin requirements (6).

Low-frequency ultrasound.

Low-frequency ultrasound has been demonstrated to increase, by severalfold, the permeability of human skin to macromolecules. Researchers estimate that permeability achieved by 1 h of sonophoresis performed three times daily would allow a typical daily dose of insulin (∼36 units) to be delivered via a transdermal patch (9). Although this approach is potentially feasible, the insulin delivery rate may not provide for physiological replacement.

Transfersomes.

Transfersomes are lipid vesicles made of soybean phosphatidylcholine that are flexible enough to pass through pores much smaller than themselves with a similar permeability to water, despite being much larger (10). Animal and human data demonstrate that when the vesicles are loaded with insulin and applied to intact skin in a sufficient quantity, glucose levels are significantly reduced because it is observed that transfersomes transport the insulin with at least 50% of the bioefficiency of a subcutaneous injection (10). The application of insulin-laden transfersomes over a skin area ≤40 cm2 would provide the daily basal insulin needs of a typical patient with type 1 diabetes according to research estimates (10).

Intranasal

Nasal administration of certain proteins—such as oxytocin, desmopressin, and calcitonin—is well established. These are highly potent molecules that are smaller than insulin (4). Insulin delivered using nasal administration reaches the systemic circulation, although various factors—including dose, timing, and frequency of administrations—influence its bioavailability. For this reason, most nasal formulations incorporate permeability enhancers to augment the low bioavailability to levels typically between 8 and 15% (11,12). However, these can cause nasal irritation—a formulation based on lecithin was associated with nasal irritation, and up to 100% of patients report irritation with formulations containing bile salts, whereas 25–50% experienced irritation with laureth-9 (1214).

A study in 17 subjects with type 2 diabetes concluded that intranasal insulin could be effective in reducing postprandial glycemia by 50–70% (15). However, a study involving 31 subjects with type 1 diabetes found that the intranasal insulin dose needed to reach given markers of glycemic control was some 20 times higher than that of subcutaneous administration. Moreover, serum insulin concentrations increased more rapidly and declined more quickly during nasal administration than during subcutaneous administration (16). The authors concluded that intranasal insulin treatment was not a realistic alternative to subcutaneous insulin because of the low bioavailability and high rate of treatment failure.

Two recent studies in small numbers of people with diabetes report more encouraging results. One of ten severely hyperglycemic subjects with type 2 diabetes, who had failed on oral agents, reported that three preprandial doses of intranasal insulin (with or without one bedtime injection of NPH) gave comparable glycemic control to NPH b.i.d., although mean HbA1c levels were high (9.4 ± 0.5 vs. 8.8 ± 0.2%). Nasal insulin alone achieved adequate glycemic control in 3 of the 10 subjects (7.6 ± 0.3%) (17). A 6-month study of a gelified sprayed nasal insulin containing absorption promoters in 16 subjects with type 1 diabetes showed that twice-daily NPH plus three preprandial nasal insulin doses (plus nasal supplementation in case of unexpected hyperglycemia) was as efficient as NPH b.i.d. plus three preprandial regular insulin injections. Although nasal tolerance appeared better than that with the previously reported lyophilized form, 4 of the 16 subjects discontinued because of treatment-related side effects. It was suggested that the absorption promoters played an additional role in intolerance (18). Nasal administration appears promising, but intolerance and high rates of treatment failure may prove difficult to overcome.

Oral enteric

Insulin’s oral bioavailability is limited because 1) insulin is too large and hydrophilic to readily cross the intestinal mucosa and 2) polypeptides undergo extensive enzymatic and chemical degradation, in particular by α-chymotrypsin (19,20). In the case of insulin, this enzymatic barrier is more important than that posed by the mucosa (19). Another major barrier for oral insulin administration, besides gastrointestinal proteolysis, is that no selective transport mechanism exists (6). The epithelial cells of the intestine do not normally transport macromolecules such as insulin and therefore may require extremely high doses to achieve some measurable insulin absorption (6,21). Other barriers that exist include the unpredictable transit time and the delayed absorption of encapsulated insulin (6). These factors may explain why ∼0.5% of an oral insulin dose may reach the systemic circulation (22). Because of these obstacles, it would be extremely difficult to consider oral insulin therapy as a physiological option for premeal dosing (6). However, researchers have tried several steps to promote the bioavailability of oral insulin, including attaching caproic acid molecules and coating with chitosan, which stabilize degradation and improve permeability; facilitating absorption (e.g., with salicylates); concurrent administration with protease inhibitors; and entrapping insulin within microparticles (2326). Other approaches have demonstrated that the chemical modification of insulin with fatty acids could improve insulin absorption from the intestine (27). In addition, engineered polymer microspheres were demonstrated to increase gastrointestinal absorption of insulin (28).

The most promising oral insulin to date is hexyl-insulin-monoconjugate-2 (HIM2), a native recombinant insulin with a small polyethylene glycol 7-hexyl group attached to the position B29 amino acid lysine, currently in development (Nobex Corporation, Research Triangle Park, NC). Preclinical pharmacokinetic and safety data for HIM2 are beginning to emerge. Ongoing phase I and II clinical trials suggest that oral HIM2 has a bioavailability of ∼5% and may result in an acceptable glucose-lowering effect (29).

Oral buccal

A buccal system delivering a liquid aerosol formulation of insulin via a metered dose inhaler (Oralin) has been developed by Generex Biotechnology (Toronto, Canada). The buccal insulin preparation is a human recombinant insulin with added enhancers, stabilizers, and a non-chlorofluorocarbon propellant. To date, however, efficacy studies have only been presented as abstracts, safety reports of buccal insulin are scarce, and the majority of the previously mentioned abstracts did not assess side effects. Further review must await the results of the peer-reviewed reports.

Pulmonary delivery

Pulmonary delivery of insulin appears to be the first reported alternative to injections. Specifically, 3 years after the initial clinical use of insulin, it was reported that insulin can be administered as an aerosol and could produce a decrease in blood glucose (30). However, it was 46 years later that the pivotal study of inhaled insulin offered proof of principle (31). Specifically, Wigley et al. (31) showed that delivering pork-beef insulin using a nebulizer produced a prompt increase in plasma immunoreactive insulin in three subjects without diabetes and four subjects with diabetes and that hypoglycemia showed a temporal relationship with the increase in plasma immunoreactive insulin. During the next 22 years, several studies confirmed that pulmonary delivery was possible (32,33). Bioavailability was low—between 20 and 25% of that associated with subcutaneous insulin—but the predictability of the blood glucose response was at least as good. It was suggested that technological modifications, such as optimizing particle size and regulating the effect of breathing, might enhance absorption and improve biological response. This proved to be an accurate prediction (33).

The reality of pulmonary insulin becoming a viable alternative to injections is in large measure due to the inherent anatomic advantages that make it an ideal route for the administration of insulin. Specifically, the lung provides a vast (50–140 m2, ∼500 million alveoli) and well-perfused absorptive surface (34). The lung lacks certain peptidases that are present in the gastrointestinal tract, and “first pass metabolism” (i.e., immediate hepatic degradation by the absorbed insulin) is not a concern (34). In addition, the presence of a very thin alveolar-capillary barrier allows for a rapid uptake of peptides in the bloodstream and a rapid onset of action after inhalation (34). The exact mechanism of insulin absorption across the pulmonary epithelium remains unclear, although it is likely to involve transcytotic and paracellular mechanisms (35). Nevertheless, it was evident from the early studies that after inhalation into the lungs, insulin is rapidly absorbed with peak plasma concentrations being reached after 15–40 min (31,33). The rapid absorption of inhaled insulin has been consistently demonstrated with recent devices and makes inhaled insulin appropriate for premeal use.

Experience gained with asthma treatment has facilitated the development of pulmonary delivery systems for insulin. Early asthma drug delivery devices for pulmonary absorption included nebulizers, metered dose inhalers, and dry-powder inhalers (36). These devices proved useful in the management of local respiratory diseases such as asthma but are not designed to deliver drugs deep into the alveoli for systemic drug delivery. Conventional asthma inhalers have poor lung delivery efficiencies, and even in well-controlled human studies, there is considerable variation. Such systems are not clinically effective for pulmonary insulin delivery. The early challenge therefore was to develop an inhaler that delivers insulin to the “deep lung.”

Despite anatomic advantages offered by the pulmonary bed to insulin uptake, potential limitations must be overcome for effective drug deposition. These factors include type of propellants used, air flow speed, losses within the device and the environment, particle size, drug clearance, drug deposition into the throat and bronchial tubes, drug absorption, patient compliance, and the potential impact of concomitant diseases, all of which can influence insulin delivery (3739). Size and particle velocity are recognized as two major physical factors affecting optimal deep-lung deposition of an inhaled drug (35). For optimal deposition efficacy, particles should have low velocity and be between 1 and 3 μm in diameter. After the aerosolized drug reaches the deep lung, it must be absorbed with high enough bioavailability to make the delivery system feasible. Most of the devices currently in development will use a liquid or dry-powder insulin formulation for the generation of the insulin aerosol (Fig. 1). In addition, the development of inhaled insulin with these formulations varies from subcutaneous delivery by the fact that both components (the insulin formulation and the device) need to be studied as a unit (6). In this regard, the specific device has a major impact on the metabolic effect of the applied insulin. As stated by Heinemann, “Even the best insulin formulation becomes ineffective if the inhaler does not generate an aerosol with appropriate features” (6).

Exubera

The Exubera system is one of the more extensively investigated in terms of clinical development. Developed by Nektar Therapeutics (San Carlos, CA), it is currently being clinically assessed by Pfizer (Groton, CT) and Aventis Pharma (Frankfurt, Germany). This system delivers a fine dry-powder formulation (<5 μm in diameter) of regular short-acting human insulin to the deep lung in a reproducible and efficient manner. The number of individual blister packs inhaled controls the dose. The insulin dry powder is packaged into a single-dose blister containing 1 or 3 mg, with the 1-mg blister corresponding to ∼3 units of insulin. The pack is inserted into the inhalation device (similar to a nebulizer) where a pneumatic mechanism punctures it and disperses the powder in a discrete cloud into the air chamber, which ensures consistent delivery (40). The insulin cloud is inhaled slowly at the beginning of a deep breath; the volume of the holding chamber is <20% of a deep breath. Early studies demonstrated reproducible pharmacokinetics and postprandial control with this device comparable to subcutaneous regular insulin (41).

Inhaled insulin therapy in insulin-treated patients with type 2 diabetes has been assessed (42). Patients received one or two inhalations of insulin before each meal plus a bedtime Ultralente injection (Humulin U; Eli Lilly, Indianapolis, IN). Glycemic control, as assessed by HbA1c levels, was significantly improved compared with baseline. When compared with patients who were randomly assigned to receive continued subcutaneous insulin, HbA1c levels were comparable between groups. A subsequent 24-week phase III study involving 298 subjects also showed comparable HbA1c control in patients with type 2 diabetes treated with an inhaled insulin regimen versus a subcutaneous insulin regimen (41).

The study by Skyler et al. (43) enrolled 73 patients with type 1 diabetes. Patients received their usual daily insulin regimen of two to three injections or premeal inhaled insulin plus a bedtime Ultralente injection for 12 weeks. Inhaled insulin provided comparable HbA1c, fasting, and postprandial glucose levels compared with subcutaneous injection. A 24-week phase III study, involving 334 subjects with type 1 diabetes, compared treatment with premeal inhaled insulin plus a single bedtime injection of Ultralente to treatment with a conventional subcutaneous insulin regimen consisting of two to three injections of regular/NPH insulin per day (41). Subjects in the inhaled insulin group achieved comparable changes in HbA1c to those in the subcutaneous group. Preliminary reports also show that this system may be effective in patients with type 2 diabetes who were inadequately controlled on oral hypoglycemic agents. Those receiving inhaled insulin monotherapy or inhaled insulin combined with an oral agent achieved greater improvements in HbA1c than those on oral agent therapy alone (41).

AERx

This insulin diabetes management system (iDMS), developed by Aradigm and now in collaboration with Novo Nordisk, creates aerosols (1- to 3-μm diameter particles) from liquid insulin. It uses microprocessors to guide the correct rate and depth of breathing, triggering insulin at the optimum moment in the inspiration-expiration cycle. This aims at consistent delivery of insulin, regardless of a patient’s breathing ability (44). In a study of 18 subjects with type 1 diabetes, insulin administered by the AERx iDMS shows faster absorption and onset of the metabolic effect than subcutaneous insulin (45). A study in healthy volunteers suggested that changes in the inhaled volume and depth of inspiration influenced pharmacokinetics of insulin delivered using this system (44). Upper respiratory tract infections do not seem to induce any clinically relevant changes in pharmacokinetics or pharmacodynamics (although tmax is significantly reduced) (46), whereas the dose may need to be adjusted upward in asthmatics (47).

Preliminary clinical efficacy and safety data are now emerging for this system. A 12-week study in relatively poorly controlled insulin-treated subjects with type 2 diabetes suggests that pulmonary insulin administered via the AERx iDMS immediately before meals (plus bedtime NPH) is as efficacious in terms of HbA1c reduction as premeal subcutaneous insulin (also with bedtime NPH) (48).

AIR (Advanced Inhalation Research)

The AIR system uses particles formulated with a low-density porous structure of between 5 and 30 μm in diameter. The device aerosolizes the particles from blister packs and is breath-activated. The system is being developed by Alkermes and Eli Lilly in two forms: a sustained-release AIR insulin with a pharmacokinetic profile similar to Humulin L in rats and a fast-release formulation that shows pharmacokinetics similar to Humulin R (49).

Dura’s Spiros

This pulmonary delivery system (currently on hold) delivers a dry-powder formulation through a handheld battery-powered multidose system. The device delivers consistent doses to the lung across a wide range of inhalation flow rates. The company’s latest development is a new powder dispersion system (Spiros-S2) that should achieve efficient delivery at low inspiratory flow rates (15–30 l/min) and without Spiros’ electromechanical components.

Aerogen’s Aerodose

Aerogen’s Aerodose is a breath-activated multidose inhaler that delivers liquid insulin to the systemic circulation. A study in 13 healthy volunteers suggested that aerosolization time, but not particle size, had a significant impact on the metabolic effect of insulin delivered using the Aerodose device (50). In a study of 15 subjects with type 2 diabetes, insulin delivered using the Aerodose device had a shorter time to peak action than subcutaneous insulin and gave comparable reproducibility of dose (51). A further study in 24 insulin-naive subjects with type 2 diabetes revealed a linear dose-response relationship for pharmacodynamic parameters, suggesting pharmacological predictability with this system (52). It is currently on hold after initial phase II clinical trials.

Pharmaceutical Discovery Corporation’s Technosphere/Insulin

The ordered lattice array of Technosphere dry-powder microparticles and recombinant human insulin provides yet another inhaled insulin system demonstrating “proof of concept.” Several studies using the euglycemic clamp have been performed in healthy volunteers and patients with type 2 diabetes. The data demonstrate that this system will provide for a rapid systemic insulin uptake (tmax ∼12–14 min), a fast onset of action (∼20–30 min), a short duration of action (∼2–3 h), and a low within-subject variability (53,54). The approach appears appropriate for prandial insulin supplementation.

Adverse events and benefits

The frequency and nature of adverse events reported with inhaled insulins appear, in general, to be comparable with subcutaneous insulin, with the exception of cough (although this decreases in incidence and prevalence with continued use) (43). Subjects treated with inhaled insulin may develop an increase in serum insulin antibody levels, but these levels thus far have not been related to any significant clinical change (41,55). Smoking, however, appears to greatly enhance insulin absorption (56). Pulmonary function tests, including forced expiratory volume in 1 s, forced vital capacity, total lung capacity, and carbon monoxide diffusing capacity (DLCO) have been conducted for all inhaled insulin studies. Some studies reported a statistically significant decrease in the more variable DLCO relative to subcutaneous insulin (41). Further studies are ongoing to characterize the insulin antibody and DLCO changes, including whether a difference of this magnitude has any clinical significance, as well as any mechanistic or methodologic basis. One of the most important and beneficial findings to date, however, is that patient satisfaction is enhanced with inhaled insulin treatment compared with subcutaneous insulin injections (57,58).

There is a long history of attempts to develop novel routes of insulin delivery that are both clinically effective and tolerable. The various approaches that have been studied to date have involved strategies that are designed to overcome the inherent barriers that exist for protein uptake across the skin, gastrointestinal tract, and nasal mucosa. Nevertheless, the “proof of concept” for many of these approaches appears to have been established. However, it does appear that the most clinically viable system to date may be pulmonary delivery. Two of the devices for pulmonary delivery are in phase III testing, and the results thus far demonstrate comparable efficacy to that of subcutaneous insulin. The pulmonary safety and tolerability data will need to be established before these devices are clinically available. Thus, 25 years after Best hoped for and expected “more physiological methods for giving insulin,” systems are in place to achieve his hopes and expectations.

Figure 1—

Pulmonary insulin delivery systems in development. *Dura Pharmaceuticals is now part of Elan Pharmaceuticals. Development is on hold. sc, subcutaneous.

Figure 1—

Pulmonary insulin delivery systems in development. *Dura Pharmaceuticals is now part of Elan Pharmaceuticals. Development is on hold. sc, subcutaneous.

Close modal
Table 1—

Potential noninvasive insulin delivery options

• Jet injectors 
 Deliver high pressure stream of insulin 
 May benefit selected cases (e.g., severe lipoatrophy); suggested that use be advocated only for those who develop problems with other injection routes 
• Transdermal 
 Iontophoresis 
  Electrical current used to enhance transdermal insulin delivery; proof-of-principle from animal studies; human studies needed 
 Low-frequency ultrasound 
  Use of low-frequency sound wave to augment delivery of insulin and other macromolecules across human skin 
 Transfersomes 
  Composite phosphatidylcholine-based vesicles with similar permeability to water, although 1,000 times larger 
  Transdermal delivery of insulin with bioefficiency ≥50% of the subcutaneous dose 
  Administration over a 40-cm2 skin area may supply sufficient basal insulin to a typical type 1 diabetic patient 
• Intranasal 
 Nasal administration of certain proteins (e.g., oxytocin, desmopressin, and calcitonin) is now well established 
 Permeability enhancers are generally required to augment insulin bioavailability; insulin bioavailability is typically in the range of 8–15% with enhancers 
 Nasal irritation is common (e.g., with lecithin, bile salts, or laureth-9 as enhancers). 
 Nasal tolerance and high rates of treatment failure are major limitations 
 Recent clinical studies have shown more promising results (e.g., with gelified nasal insulin) 
• Oral 
 Enteric 
  Oral enteric insulin delivery has limited bioavailability 
  Insulin is too large and hydrophillic to readily cross the intestinal mucosa 
  Polypeptides undergo extensive enzymatic and chemical degardation 
  Only around 0.5% of a dose of oral insulin reaches the systemic circulation 
  Several methods used to promote bioavailability 
  Ongoing phase I and II clinical trials with new formulation suggest a bioavailability of ∼5%, which may result in an acceptable  glucose-lowering effect 
 Buccal 
  Liquid aerosol insulin is sprayed into the buccal cavity without entering the airways 
  A liquid formulation of human recombinant insulin with added enhancers, stabilizers, and a non-chlorofluorocarbon propellant delivered via a metered dose inhaler in clinical trials 
  Efficacy studies are preliminary and safety reports are scarce 
• Pulmonary 
 High permeability and large surface area provide a favorable anatomy for protein/drug uptake 
 Very rapid absorption of insulin after inhalation mimics time-activity profile of fast-acting insulin; appropriate for premeal delivery 
 Appears comparable to subcutaneous insulin on glycemic parameters for both type 1 and type 2 diabetic patients 
 Several pulmonary insulin delivery systems are in development and in phase III testing (see Fig. 1
• Jet injectors 
 Deliver high pressure stream of insulin 
 May benefit selected cases (e.g., severe lipoatrophy); suggested that use be advocated only for those who develop problems with other injection routes 
• Transdermal 
 Iontophoresis 
  Electrical current used to enhance transdermal insulin delivery; proof-of-principle from animal studies; human studies needed 
 Low-frequency ultrasound 
  Use of low-frequency sound wave to augment delivery of insulin and other macromolecules across human skin 
 Transfersomes 
  Composite phosphatidylcholine-based vesicles with similar permeability to water, although 1,000 times larger 
  Transdermal delivery of insulin with bioefficiency ≥50% of the subcutaneous dose 
  Administration over a 40-cm2 skin area may supply sufficient basal insulin to a typical type 1 diabetic patient 
• Intranasal 
 Nasal administration of certain proteins (e.g., oxytocin, desmopressin, and calcitonin) is now well established 
 Permeability enhancers are generally required to augment insulin bioavailability; insulin bioavailability is typically in the range of 8–15% with enhancers 
 Nasal irritation is common (e.g., with lecithin, bile salts, or laureth-9 as enhancers). 
 Nasal tolerance and high rates of treatment failure are major limitations 
 Recent clinical studies have shown more promising results (e.g., with gelified nasal insulin) 
• Oral 
 Enteric 
  Oral enteric insulin delivery has limited bioavailability 
  Insulin is too large and hydrophillic to readily cross the intestinal mucosa 
  Polypeptides undergo extensive enzymatic and chemical degardation 
  Only around 0.5% of a dose of oral insulin reaches the systemic circulation 
  Several methods used to promote bioavailability 
  Ongoing phase I and II clinical trials with new formulation suggest a bioavailability of ∼5%, which may result in an acceptable  glucose-lowering effect 
 Buccal 
  Liquid aerosol insulin is sprayed into the buccal cavity without entering the airways 
  A liquid formulation of human recombinant insulin with added enhancers, stabilizers, and a non-chlorofluorocarbon propellant delivered via a metered dose inhaler in clinical trials 
  Efficacy studies are preliminary and safety reports are scarce 
• Pulmonary 
 High permeability and large surface area provide a favorable anatomy for protein/drug uptake 
 Very rapid absorption of insulin after inhalation mimics time-activity profile of fast-acting insulin; appropriate for premeal delivery 
 Appears comparable to subcutaneous insulin on glycemic parameters for both type 1 and type 2 diabetic patients 
 Several pulmonary insulin delivery systems are in development and in phase III testing (see Fig. 1
1.
Peterson CM: Symposium on optimal insulin delivery. Introduction: history and goals of insulin treatment.
Diabetes Care
5
:
1
–5,
1982
2.
Korytkowski M: When oral agents fail; practical barriers to starting insulin.
Int J Obes
26 (Suppl. 3)
:
S18
–S24,
2002
3.
Saudek CD: Novel forms of insulin delivery.
Endocrinol Metab Clin North Am
26
:
599
–610,
1997
4.
Chetty DJ, Chien YW: Novel methods of insulin delivery: an update.
Crit Rev Ther Drug Carrier Syst
15
:
629
–670,
1998
5.
American Diabetes Association. Insulin administration (Position Statement).
Diabetes Care
25 (Suppl. 1)
:
S112
–S115,
2002
6.
Heinemann L, Pfutzner A, Heise T: Alternative routes of administration as an approach to improve insulin therapy: update on dermal, oral, nasal and pulmonary insulin delivery.
Curr Pharm Des
7
:
1327
–1351,
2001
7.
Langkjaer L, Brange J, Grodsky GM, Guy RH: Iontophoresis of monomeric insulin analogues in vitro: effects of insulin charge and skin pretreatment.
J Control Release
51
:
47
–56,
1998
8.
Kanikkannan N, Singh J, Ramarao P: Transdermal iontophoretic delivery of bovine insulin and monomeric human insulin analogue.
J Control Release
59
:
99
–105,
1999
9.
Mitragotri S, Blankschtein D, Langer R: Ultrasound-mediated transdermal protein delivery.
Science
269
:
850
–853,
1995
10.
Cevc G, Gebauer D, Stieber J, Schatzlein A, Blume G: Ultraflexible vesicles, Transfersomes, have an extremely low pore penetration resistance and transport therapeutic amounts of insulin across the intact mammalian skin.
Biochim Biophys Acta
1368
:
201
–215,
1998
11.
Jacobs MA, Schreuder RH, Jap-A-Joe K, Nauta JJ, Andersen PM, Heine RJ: The pharmacodynamics and activity of intranasal administered insulin in healthy male volunteers.
Diabetes
42
:
1649
–1655,
1993
12.
Frauman AG, Jerums G, Louis WJ: Effects of intranasal insulin in nonobese type II diabetics.
Diabetes Res Clin Pract
3
:
197
–202,
1987
13.
Gordon GS, Moses AC, Silver RD, Flier JS, Carey MC: Nasal absorption of insulin: enhancement by hydrophobic bile salts.
Proc Natl Acad Sci U S A
82
:
7419
–7423,
1985
14.
Salzman R, Manson JE, Griffing GT, Kimmerle R, Ruderman N, McCall A, Stoltz EI, Mullin C, Small D, Armstrong J, et al.: Intranasal aerosolized insulin: mixed meal studies and long-term use in type 1 diabetes.
N Engl J Med
312
:
1078
–1084,
1985
15.
Coates PA, Ismail IS, Luzio SD, Griffiths I, Ollerton RL, Volund A, Owens DR: Intranasal insulin: the effects of three dose regimens on postprandial glycaemic profiles in type II diabetic subjects.
Diabet Med
12
:
235
–239,
1995
16.
Hilsted J, Madsbad S, Hvidberg A, Rasmussen MH, Krarup T, Ipsen H, Hansen B, Pedersen M, Djurup R, Oxenboll B: Intranasal insulin therapy: the clinical realities.
Diabetologia
38
:
680
–684,
1995
17.
Lalej-Bennis D, Boillot J, Bardin C, Zirinis P, Coste A, Escudier E, Chast F, Peynegre R, Selam JL, Slama G: Efficacy and tolerance of intranasal insulin administered during 4 months in severely hyperglycaemic type 2 diabetic patients with oral drug failure: a cross-over study.
Diabet Med
18
:
614
–618,
2001
18.
Lalej-Bennis D, Boillot J, Bardin C, Zirinis P, Coste A, Escudier E, Chast F, Peynegre R, Slama G, Selam JL: Six month administration of gelified intranasal insulin in 16 type 1 diabetic patients under multiple injections: efficacy vs subcutaneous injections and local tolerance.
Diabetes Metab
27
:
372
–377,
2001
19.
Davis SS: Overcoming barriers to the oral administration of peptide drugs.
Trends Pharmacological Sci
11
:
353
–355,
1990
20.
Schilling RJ, Mitra AK: Degradation of insulin by trypsin and alpha-chymotrypsin.
Pharm Res
8
:
721
–727,
1991
21.
Olsen CL, Liu G, Charles A: Novel routes of insulin delivery. In
The Diabetes Annual
. Marshall SM, Home PD, Eds. Amsterdam, Elsevier,
1994
, p.
243
–276
22.
Crane CW, Path MC, Luntz GR: Absorption of insulin from the human small intestine.
Diabetes
17
:
625
–627,
1968
23.
Ramadas M, Paul W, Dileep KJ, Anitha Y, Sharma CP: Lipoinsulin encapsulated alginate-chitosan capsules: intestinal delivery in diabetic rats.
J Microencapsul
17
:
405
–411,
2000
24.
Nishihata T, Rytting JH, Kamada A, Higuchi T: Enhanced intestinal absorption of insulin in rats in the presence of sodium 5-methoxysalicylate.
Diabetes
30
:
1065
–1067,
1981
25.
Ziv E, Kidron M, Raz I, Krausz M, Blatt Y, Rotman A, Bar-On H: Oral administration of insulin in solid form to nondiabetic and diabetic dogs.
J Pharm Sci
83
:
792
–794,
1994
26.
Kimura T, Sato K, Sugimoto K, Tao R, Murakami T, Kurosaki Y, Nakayama T: Oral administration of insulin as poly(vinyl alcohol)-gel spheres in diabetic rats.
Biol Pharm Bull
19
:
897
–900,
1996
27.
Asada H, Douen T, Waki M, Adachi S, Fujita T, Yamamoto A, Muranishi S: Absorption characteristics of chemically modified-insulin derivatives with various fatty acids in the small and large intestine.
J Pharm Sci
84
:
682
–687,
1995
28.
Mathiowitz E, Jacob JS, Jong YS, Carino GP, Chickering DE, Chaturvedi P, Santos CA, Vijayaraghavan K, Montgomery S, Bassett M, Morrell C: Biologically erodable microspheres as potential oral drug delivery systems.
Nature
386
:
410
–414,
1997
29.
Still JG: Development of oral insulin: progress and current status.
Diabetes Metab Res Rev
18 (Suppl. 1)
:
S29
–S37,
2002
30.
Gänsslen M: Über inhalation von insulin.
Klin Wochensubcutaneoushr
4
:
71
,
1925
31.
Wigley FW, Londono JH, Wood SH, Shipp JC, Waldman RH: Insulin across respiratory mucosa by aerosol delivery.
Diabetes
20
:
552
–556,
1971
32.
Elliott RB, Edgar BW, Pilcher CC, Quested C, McMaster J: Parenteral absorption of insulin from the lung in diabetic children.
Aust Paediatr J
23
:
293
–297,
1987
33.
Laube BL, Georgopoulus A, Adams GK III: Preliminary study of the efficacy of insulin aerosol delivered by oral inhalation in diabetic subjects.
JAMA
269
:
2106
–2109,
1993
34.
Heinemann L: Alternative delivery routes: inhaled insulin.
Diabetes Nutr Metab
15
:
417
–222,
2002
35.
Patton JS: Mechanisms of macromolecule absorption by the lungs.
Adv Drug Deliv Rev
19
:
3
–36,
1996
36.
Gonda I: The asubcutaneousent of pulmonary drug delivery.
J Pharm Sci
89
:
940
–945,
2000
37.
Klonoff DC: Inhaled insulin.
Diabetes Technol Ther
1
:
307
–313,
1999
38.
Patton JS: Deep-lung delivery of therapeutic proteins.
Chemtech
27
:
34
–38,
1997
39.
Laube BL, Benedict GW, Dobs AS: Time to peak insulin level, relative bioavailability, and effect of site of deposition of nebulized insulin in patients with noninsulin-dependent diabetes mellitus.
J Aerosol Med
11
:
153
–173,
1998
40.
Patton JS, Bukar J, Nagarajan S: Inhaled insulin.
Adv Drug Deliv Rev
35
:
235
–247,
1999
41.
Bindra S, Rosenstock J, Cefalu WT: Inhaled insulin: a novel route for insulin delivery.
Expert Opin Investig Drugs
11
:
687
–691,
2002
42.
Cefalu WT, Skyler JS, Kourides IA, Landschulz WH, Balagtas CC, Cheng, Gelfand RA, for the Inhaled Insulin Study Group: Inhaled human insulin treatment in patients with type 2 diabetes mellitus.
Ann Intern Med
134
:
203
–207,
2001
43.
Skyler JS, Cefalu WT, Kourides IA, Landschulz WH, Balagtas CC, Cheng SL, Gelfand RA: Efficacy of inhaled human insulin in type 1 diabetes mellitus: a randomised proof-of-concept study.
Lancet
357
:
331
–335,
2001
44.
Farr SJ, McElduff A, Mather LE, Okikawa J, Ward ME, Gonda I, Licko V, Rubsamen RM: Pulmonary insulin administration using the AERx system: physiological and physiochemical factors influencing insulin effectiveness in healthy fasting subjects.
Diabetes Tech Ther
2
:
185
–197,
2000
45.
Brunner GA, Balent B, Ellmerer M, Schaupp L, Siebenhofer A, Jendle JH, Okikawa J, Pieber TR: Dose-response relation of liquid aerosol inhaled insulin in type 1 diabetic patients.
Diabetologia
44
:
305
–308,
2001
46.
McElduff A, Clauson P, Uy C, Kam P, Mather LE: Pulmonary absorption profiles of insulin during and after an upper respiratory tract infection in healthy volunteers using the AERx insulin diabetes management system: an open labelled cross-over study in healthy volunteers (Abstract).
Diabetes
51 (Suppl. 2)
:
A107
,
2002
47.
Henry RR, Mudaliar SR, Howland III WC, Chu N, Kim D, An B, Reinhardt RR: Inhaled insulin using the AERx insulin diabetes management system in healthy and asthmatic subjects.
Diabetes Care
26
:
764
–769,
2003
48.
Adamson U, Rönnemaa T, Petersen AH, Hermansen K: Inhaled human insulin via the AERx iDMS insulin diabetes management system in combination with NPH insulin offers the same metabolic control as intensive subcutaneous therapy: a proof of concept trial in type 2 diabetic patients.
Diabetologia
45 (Suppl. 2)
:
A255
–A256,
2002
49.
Hrkach J, Batycky R, Chen D, Deaver D, Elbert K, Johnston L, Kovalesky MA, Nice J, Schmitke J, Stapleton K, Edwards D: AIR insulin: complete diabetes therapy via inhalation of fast-acting and slow-acting dry powder aerosols (Abstract).
Diabetes
49 (Suppl. 1)
:
A9
,
2000
50.
Heinemann L, Kapitza C, Heise T, Shapiro DA, Gopalakrishnan V, Fishman RS: Impact of particle size and aerosolisation time on the metabolic effect of an inhaled insulin aerosol.
Diabetologia
44 (Suppl. 1)
:
A5
,
2001
51.
Perera AD, Kapitza C, Nosek L, Fishman RS, Shapiro DA, Heise T, Heinemann L: Absorption and metabolic effect of inhaled insulin: intrapatient variability after inhalation via the Aerodose insulin inhaler in patients with type 2 diabetes.
Diabetes Care
25
:
2276
–2281,
2002
52.
Kim D, Mudaliar S, Plodkowski R, Perera A, Fishman R, Shapiro D, Henry R: Dose-response relationships of inhaled and subcutaneous insulin in type 2 diabetic patients (Abstract).
Diabetes
51 (Suppl. 2)
:
A47
,
2002
53.
Steiner S, Pfutzner A, Wilson BR, Harzer O, Heinemann L, Rave K: Technosphere/Insulin: proof of concept study with a new insulin formulation for pulmonary delivery.
Exp Clin Endocrinol Diabetes
110
:
17
–21,
2002
54.
Pfutzner A, Mann SE, Steiner SS: Technosphere/Insulin: a new approach for effective delivery of human insulin via the pulmonary route.
Diabetes Technol Ther
4
:
589
–594,
2002
55.
Stoever JA, Palmer JP: Inhaled insulin and insulin antibodies: a new twist to an old debate.
Diabetes Technol Ther
4
:
157
–161,
2002
56.
Himmelman A, Jendle J, Mellen A, Petersen AH, Dahl UL, Wollmer P: The impact of smoking on inhaled insulin.
Diabetes Care
26
:
266
–282,
2003
57.
Gerber RA, Cappelleri JC, Kourides IA, Gelfand RA: Treatment satisfaction with inhaled insulin in patients with type 1 diabetes mellitus: a randomized controlled trial.
Diabetes Care
24
:
1556
–1559,
2001
58.
Cappelleri JC, Cefalu WT, Rosenstock J, Kourides IA, Gerber RA: Treatment satisfaction in type 2 diabetes: a comparison between an inhaled insulin regimen and a subcutaneous insulin regimen.
Clin Ther
24
:
552
–564,
2002

W.T.C. has received honoraria for speaking engagements from Pfizer, Merck, Lilly, Takeda, Aventis, and Novartis and has received grant support from Pfizer, Merck, Novartis, RW Johnson, Dexcom, and Bristol-Myers Squibb.

W.T.C. is currently affiliated with the Pennington Biomedical Research Center, Baton Rouge, Louisiana.

A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.