Delivery systems for penetration enhancement of peptide and protein drugs: design considerations

Abstract

This paper discusses the challenges to be met in designing delivery systems that maximize the absorption of peptide and protein drugs from the gastrointestinal and respiratory tracts. The ideal delivery system for either route of administration is one that will release its contents only at a favorable region of absorption, where the delivery system attaches by virtue of specific interaction with surface determinants unique to that region and where the delivery system travels at a rate independent of the transitory constraints inherent of the route of administration. Such a delivery system, which is as yet unavailable, will benefit not only peptide and protein drugs, but other poorly absorbed drugs.

Introduction

Optimizing the delivery system to maximize the extent of peptide and protein absorption has not received as much attention as the use of penetration enhancers. The majority of delivery systems that have been investigated to date are those designed to protect peptide and protein drugs from the luminal proteases in the gastrointestinal tract [1], [2]. More recent applications of delivery systems seem to focus on their capability of being retained at the site of absorption or being actively phagocytosed by the epithelial cells. Of these systems, nano- and microparticles as well as nano- and microcapsules have received the most attention to date.

An impressive insulin bioavailability of 30% has been achieved following the administration of human monocomponent insulin in 45 μm lyophilized dextran starch microspheres to the nasal cavity of anesthetized rats [3]. Peak serum insulin concentration was achieved within 7–10 min, leading to maximal decrease in blood glucose level 20–30 min thereafter. This early peak time suggests that prolonged retention of the microspheres in the nasal cavity reported by Illum et al. [4] is not an important contributing factor in the enhancement of nasal insulin absorption. A more plausible mechanism is dehydration of the underlying mucus to which the microspheres adhered, followed by rapid release of insulin from the microspheres simultaneously with shrinkage of the underlying epithelial cells and opening of the tight junctions. At an insulin dose of 1 U kg⁻¹, an optimal microsphere load was attained at 3–7 mg kg⁻¹ over the range of 2.35–10 mg kg⁻¹ [5], suggesting that a proper balance between the degree of dehydration and solubility of insulin in the residual moisture in the mucus was essential to the extent of absorption enhancement attained. Differences in insulin release rate, as indicated by the time required to deplete 90% of the insulin load (t₉₀), did not affect blood glucose kinetics in rats following microsphere administration. The time course of hypoglycemia was the same whether the t₉₀ was 10 or 60 min.

A lower bioavailability, 2.7%, was obtained for human growth hormone (hGH) when administered to the nasal cavity of the sheep in the same dextran starch microsphere system (0.9 IU kg⁻¹) [6]. Poor absorption was attributed to the low solubility of human growth hormone (18 mg ml⁻¹). Its bioavailability was, however, increased to 14.4% upon incorporation of l-α-lysophosphatidylcholine at a dose of 0.2 mg kg⁻¹ into the delivery system. This was accompanied by a shortening of the peak time in plasma from 120 to 60 min. The long-term safety of microsphere administration to the nasal cavity of the rat has been discussed in the review by Edman and Björk in this theme volume (see first article in the subsequent issue).

Orally administered insulin has been reported to be absorbed from polyisobutylcyanoacrylate nanocapsules, mean size 220 nm, following intragastric administration in diabetic rats [7]. Fasted glucose levels were lowered to 50–60% by day 2 and were maintained for 6–20 days in a dose-dependent manner over the range of 12.5 and 50 U kg⁻¹ (Fig. 1). The investigators proposed that the nanocapsules passed through the epithelial mucosa via the paracellular pathway and were then transported to the liver, where insulin acts, even though the size of the nanocapsules exceeds the pore size of the portal capillaries (60–100 nm) [8]. While this study has raised the interesting possibility of using nanocapsules to circumvent the proteolytic and penetration barriers to orally administered peptide and protein drugs, it has not yet addressed the crucial issues of absorption enhancement mechanisms, biological fate of the nanocapsules and long-term safety.

Another application of orally administered microparticulate system is in targeting to the gut-associated lymphoid tissues represented by the Peyer’s patches [9]. This possibility was suggested in a study by Eldridge et al. [9], who administered 20 mg of microspheres containing the fluorescent dye, Coumarin-6, to non-anesthetized mice. At 48 h, the mice were killed and three representative Peyer’s patches, together with the first mesenteric lymph node proximal to the appendix and the spleen, were excised for microscopic observation. The number of absorbed microspheres was counted in frozen sections using a fluorescence microscope. The percentage of ingested dose was not determined, however.

Of the microspheres investigated, only those composed of polystyrene, poly(methylmethacrylate), poly(hydroxybutyrate), poly(d,l-lactide), poly(l-lactide) and poly(d,l-lactide-coglycolide) were absorbed into the Peyer’s patches of the small intestine, while those composed of ethyl cellulose, cellulose acetate hydrogen phthalate and cellulose triacetate were not. Microsphere uptake occurred only in the Peyer’s patches and was restricted to those microspheres up to 10 μm in diameter. Fig. 2 shows the tissue distribution of 50:50 poly(d,l-lactide-coglycolide) microspheres as a function of time. As can be seen, the total number of microspheres within the three Peyer’s patches reached a maximum on day 4 and then decreased as the particles smaller than 5 μm were removed from the tissue by the efferent lymphatics. By contrast, microspheres of diameter greater than 5 μm were retained in the Peyer’s patches even on day 35. In the mesenteric lymph nodes and the spleen, the number of microspheres reached a peak on day 7 for those smaller than 5 μm and on day 14 for those greater than 5 μm in diameter.

Following primary, secondary and tertiary oral immunization with 100 mg doses of microencapsulated or soluble staphylococcal enterotoxin B vaccine, plasma anti-toxin IgG antibody levels were shown to be several thousand times significantly higher with toxoid-containing poly(d,l-lactide-coglycolide) microspheres of a mean diameter of 4 μm [9]. A separate study by Ebel [10] using polystyrene latex particles, 2.65 μm in diameter, indicated that only 0.01% of doses in the 10⁶–10⁸ particle number range were absorbed in young adult female BALB/c mice given a single oral gavage. LeFevre et al. [11] found that the number of rhodamine B-labeled polystyrene latex particles, 1.8±0.13 μm, absorbed beyond the Peyer’s patches remained small even after gavage 5 days per week for 25 days. While such low absorption efficiency may be adequate for oral immunization, it would seem to preclude delivery of systemically active peptide and protein drugs.

The purpose of this paper is to discuss the challenges to be met in designing delivery systems for maximizing peptide and protein drug absorption from mucosal routes. The focus will be on the oral and pulmonary routes of administration.