WO1998022089A1

WO1998022089A1 - Pharmaceutical composition based on a lipid matrix system

Info

Publication number: WO1998022089A1
Application number: PCT/SE1997/001956
Authority: WO
Inventors: Brita Sjöström; Birgitta Olsson
Original assignee: Pharmacia & Upjohn Ab
Priority date: 1996-11-22
Filing date: 1997-11-21
Publication date: 1998-05-28
Also published as: AU5142398A

Abstract

The present invention relates to a pharmaceutical composition based on a lipid matrix system having at least two lipid components and at least one bioactive compound. The two lipid components are chosen from classes of different polarity in which at least one of the lipid components is bilayer forming. The bioactive compound is a low-molecular weight peptide which preferably stimulates the release of growth hormone and comprises 2-alkyl-D tryptophan and most preferably the peptide is hexarelin. The composition can include polar solvents and preferably a water-containing solvent.

Description

PHARMACEUTICAL COMPOSITION BASED ON A LIPID MATRIX SYSTEM

Field of invention The present invention relates to a pharmaceutical composition based on a lipid matrix system having at least two lipid components and at least one bioactive compound. The two lipid components are chosen from classes of different polarity in which at least one of the lipid components is bilayer forming. The bioactive compound is a low molecular weight peptide which preferably stimulates the release of growth hormone and comprises 2-alkyl-D tryptophan and most preferably the peptide is hexarelin.

The composition can include polar solvents and preferably a water containing solvent.

Introduction

Peptides are usually given parenterally but since parenteral administration often needs to be carried out by physicians or nurses and the fact that many patients find such therapy uncomfortable, a lot of effort is made in developing drug delivery forms applicable for other routes of administration.

However, the absorption of bioactive drug through biological membranes is a very complex process due to the nature and complex structure of the different membranes to be passed. Many orally or enterally administered drugs show a restricted absorption due to their physicochemical properties, molecular size, degradation in the gut lumen or due to specific absorption mechanism in the gastrointestinal tract. Also, many bioactive materials administered nasally or topically show erratic and variable absorption due to limited membrane permeability. Thus oral, rectal , nasal, pulmonal, buccal, sublingual and topical administration of drugs often need the addition of absorption enhancers which in many cases have been shown to be detrimental to the biological membranes. There is a medical need for delivery system which enhances the absorption and also the stability of the drug.

Many peptides and proteins are very potent drugs, but when administered orally, the bioavailability and hence the pharmacological activity obtained is often very low (Lee and Yamamoto, 1990; Drewe et al, 1993; Taki et al, 1995). This is because of the metabolic and/ or permeability barriers to intestinal absorption, and/ or extensive liver extraction and/ or complex formation with food components (Bai et al, 1995; Taki et al, 1995; Lee, 1995). Various approaches have been investigated to overcome these barriers, such as synthesis of more stable and lipophilic analogues, the use of absorption enhancers and peptidase/ protease inhibitors, and delivery to a site in the gastrointestinal tract where the enzyme activity is minimal (Tomita et al, 1988; Aungst et al, 1991; Drewe et al, 1993; Yodoya et al, 1994; Muranishi and Yamamoto, 1994; Lee, 1995; Taki et al, 1995; Krishnamoorthy and Mitra, 1995).

Chemical modification of peptides and proteins is a potential approach because this method can change physico-chemical properties of these compounds, such as a reduction of the molecular size and an increased lipophilicity (Yodoya et al., 1994; Stewart and Taylor, 1995). Furthermore, it has been demonstrated that this also might reduce the enzymatic degradation (Yodoya et al, 1994). These modification strategies include identification of a minimal fragment with maintained pharmacological activity, substitution of L-amino acids to unnatural (e.g. D-amino acids), substitution of hydrogen bonding groups, cyclonization, and targeting for carrier-mediated transport across the intestinal mucosa (Stewart and Taylor, 1995). Several papers have been published demonstrating the influence of lipids on drug absorption. Results have been obtained showing an enhanced oral absorption either in man or animals, for example:

- griseofulvin in an oil-in-water emulsion (Bates and Sequeria, J. Pharm. Sci., 1975, 64, 793), cefoxitin in an oil-in-water emulsion (Palin et al, Int. J. Pharm., 1986, 33, 99), insulin in liposomes of phosphatidyl choline/ cholesterol, as well as in water-in- oil microemulsion (Patel and Ryman, FEBS Letters, 1976, 62, 60; Cho and Flynn, Lancet, 1989, Dec. 23/30), cyclosporine in microemulsion (Tarr and Yalkowsky, Pharm. Res. 1989, 6, 40),

Enhanced nasal absorption of insulin is shown in rats in solution with lyso phosphatidyl choline (Ilium et al., Int. J. Pharm., 1989, 57, 49).

In the present invention it is suggested that combinations of lipids can be used as

(R) absorption enhancers for low molecular weight peptides. Biosomes is an example of such an enhancer, here reported to enhance the intestinal permeability and oral bioavailability of both hydrophilic and lipid soluble compounds. This drug delivery systems consists of a lipid matrix, which in excess of water spontaneously form lipid particles (Bruce, 1994; Bohlinder et al, 1994; Betageri et al, 1993). The system and its principles are disclosed in WO 92/05771. The mechanism(s) by which Biosomes^® increase the intestinal uptake is not fully understood. The toxicity appears to be minimal (Betageri et al, 1993). The low risk associated with administration of the lipid matrix based system, such as Biosomes^®, is a great advantage, as local irritation and damage to the intestinal mucosa might limit the clinical use of additives used for absorption enhancement (van Hoogdalem et al, 1990; Swenson et al, 1994; Yodama et al, 1994). This aspect must also be considered when clinical use of protease inhibitors is considered. Pefabloc® SC (Pentapharm Ltd.) may be one product of choice. The protease profile of the gastrointestinal tract would suggest that colon and rectum would be attractive for delivery of peptides and proteins that are labile towards enzymatic degradation. (Bai et al, 1995; Rubinstein, 1995). These regions have also been shown to be most sensitive to absorption enhancers (Muranishi, 1990). However, it must be taken into account that the colonic microflora, which has a broad spectrum of metabolic reactions, might extensively degrade peptides and proteins, and moreover, the colonic and rectal mucosa has a permeability for hydrophilic compounds that is > 2 times lower than in the small intestine (Chadwick et al, 1977; Schultz and Winne, 1987; Bai et al, 1995; Lennernas et al, 1995).

The invention

The present invention relates to composition and processes as defined in the present claims.

The examples illustrate the invention by the use of hexarelin as a low molecular weight peptide and a lipid matrix comprising soybean phosphatidyl choline and medium-chain monoacylglycerol (30:70 w:w), which is suitable for effective enhancement of intestinal absorption of incompletely absorbed drugs.

By medium chain is here meant a chain containing 8 to 12, preferably 8-10 carbon atoms.

The growth hormone release of hexarelin after oral administration is only about 0.3+0.1% as compared to after i.v. administration. (Ghigo et al., 1994).

One aim of the present study was thus to investigate if the presence of the lipid matrix, with or without Pefabloc^® SC, is an effective approach to enhance the effective permeability (Peff) of hexarelin in various regions of rat intestine in situ. The local effects on the intestinal mucosa of these additives was assessed by monitoring the uptake of a non-absorbable marker, polyethylene glycol, PEG 4000, and the release of lactate dehydrogenase (LDH) into the lumen. In order to better understand the mechanism(s) of action of the lipid matrix we also investigated how this delivery system influences the uptake of a small and hydrophilic compound, atenolol. Brief description of the drawing.

Figure 1. Mean ± SD ( standard deviation) effective permeability coefficients (Peff ) of hexarelin in the absence (grey square) and presence (black square) of soybean phosphatidyl choline and medium chain monoacylglycerol matrix in the single-pass in situ perfused rat jejunum, ileum and colon. Estimates were obtained in the presence of the protease inhibitor Pefabloc^® SC.

EXAMPLES Materials and Methods

Hexarelin is a low molecular weight peptide with six amino acids:

His - Trp - Ala - Trp - Phe - Lys in which Trp at position 2 is D-2 methyl-Trp, Phe is D-Phe and Lys is Lys-NH₂. Hexarelin is a synthetic growth hormone-releasing peptide, shown to produce a substantial increase of growth hormone plasma levels in humans (Imbimbo et al, 1994; Ghigo et al., 1994). The compound is disclosed in the patent application WO

91/18061. log KD is 2.2, log D is -2.3 at pH 6.5 and MW is 887.

Lipid matrix A lipid based drug delivery system composed of medium chain monoacylglycerol and soy bean phosphatidyl cholin in a ratio of 70:30 w:w.

(R) Pefabloc SC is a serine protease inhibitor.

Atenolol is used as a reference substance with the permeability value and partition coefficient given in the literature . log D is -1.8 at pH 7.4 and MW is 266. log KD and log D are partition coefficients in octanol and a buffer, known in the art. The perfusion solution (pH 6.5, 290 mOsm/L) contained 48 mM NaCl, 5.4 mM KCl, 28 mM Na2HPθ4, 43 mM aH2Pθ4, 35 mM mannitol, 1 g/L polyethylene glycol

4000 (PEG 4000), 2.5 μCi/L [¹ C]PEG 4000 (non-absorbable fluid flux and functional viability marker) (Amersham Labs., Buckinghamshire, England) and 10 mM D- glucose. Functional viability tests. The local toxicity of the lipid matrix and Pefabloc^® SC was assessed by employing functional viability tests throughout each experiment. This was done by monitoring the recovery of a non-absorbable marker molecule, PEG 4000, and the release of lactate dehydrogenase (LDH) into the intestinal lumen (Fagerholm et al, 1996; Swenson et al, 1994). A considerable intestinal uptake of PEG 4000 and increased release of LDH is assumed to be associated with intestinal damage and lost intestinal barrier function (viability) (Fagerholm et al., 1996; Swenson et al, 1994). LDH measurements were performed in all experiments, except for the initial jejunal perfusions in parts I and II.

Stability. The stability of hexarelin was assessed in perfusate at 37°C for 180 min, and by incubation in rat GI fluids at 37°C for 180 min. The stability of atenolol has previously been investigated, and there was no sign of degradation of this compound after 180 min at 37°C.

Calculations and predictions. The recovery of [¹⁴C]PEG 4000 (PEG) _rec was estimated from equation 1:

PEG rec = ∑ PEG _out/ Σ PEG _m (equation 1)

where Σ PEG i_n and Σ PEG out are the accumulated amounts of [¹⁴C]PEG 4000 entering and leaving the intestinal segment during equilibrium, respectively. The parallel-tube model was used to estimate the effective intestinal permeability coefficients (P_eff cm/s) for hexarelin and atenolol (Amidon et al., 1980; Komiya et al, 1980; Fagerholm et al, 1996):

Peff = (equation 2) A where C_zn and C_oui are the inlet and fluid- transport corrected outlet solute concentrations, respectively. [PEG]_zn and [PEG]_ouf. are the inlet and outlet

concentrations of the water flux marker [ C]-PEG 4000, respectively. Q_zn is the perfusion flow rate, and A is the mass transfer surface area within the intestinal segment assumed to be a cylinder area with the length (L) of 7-12 cm in the jejunum and ileum, and 1.5-6 cm in the colon, and radius (r) 0.18, 0.18 and 0.25 cm in the jejunum, ileum and colon, respectively (Komiya et al 1980; Kararli 1995). Rat jejunal Peff -estimates for highly soluble, stable, passively absorbed compounds (Peff,rat) can be used to predict the corresponding human coefficient (Peff,man) arid the extent of intestinal absorption in vivo in man (fa) (Fagerholm et al, 1996):

Peff,man= 3.6 • Pef f, rat + 0.03 • 10"⁴ (equation 3)

fa = 1 . e"(^{2 *} Peff,man ^* *res / r ^• 2.8) _{(equation )}

where tres and r are the average small intestinal transit time and radius in humans, assumed to be 3 hrs and 1.75 cm, respectively (Fagerholm et al, 1996). 2.8 is a correction factor (determined by non-linear regression) which compensates for the difference between P_eff and intestinal absorption after oral ingestion of the drug (Fagerholm et al, 1996). However, the estimated P_eff and predicted fa of metabolically instable substances, such as hexarelin, must be corrected for degradation, and furthermore, as the case in the present study, a period of enhanced absorption must be taken into account when predicting the fa. To solve this problem the simulation application Stella^® was used. It was assumed that the degradation rate (Kdegr) ^or hexarelin in humans in vivo equals that observed in intestinal fluids of rats, and that it is a first order process. Based upon jejunal P_eff -values of hexarelin in the absence and presence of the lipid matrix and Pefabloc^® SC, the fa of this drug during normal conditions, and conditions with an enhanced P_eff over a period of 5, 10 and 30 min, with and without total inhibition of enzymatic degradation was predicted. The Student's paired --test was used to determine differences in P_eff. Peff coefficients are presented as mean ± standard deviation (SD).

Formulations containing hexarelin were prepared as follows:

A. Formulation with hexarelin, atenolol and lipids

Hexarelin acetate (Bachem Feinchemikalien AG, Bubendorf, Switzerland) was dissolved in sterile water at a concentration of 26% (w/w), i.e. pure peptide 22.1

%(w/w). The hexarelin acetate solution was subsequently blended with a mixture of medium chain monoglyceride (MG), (Scotia LipidTeknik AB, Stockholm, Sweden) and soybean phosphatidyl choline (PC), (Scotia LipidTeknik AB, Stockholm, Sweden), (70% MG/30% PC, w/w). The amount of hexarelin solution in the mixture was 11,76 %(w/w). The tube containing the mixture was filled with nitrogen whereafter it was sealed. Thereafter the tube was shaken moderately until the mixture appeared to be homogenous and clear ( about 1.5 h at 37 °C). The hexarelin/ lipid mixture was than stored at -20°C until the perfusion study was performed. Upon starting the perfusion study an isotonic NaCl-solution (0.9 % w/w) was added to the thawed hexarelin/ lipid mixture in order to swell the lipids (for 10 minutes at 20 °C). The amount of the NaCl-solution in the mixture was 82 %(w/ w). Then perfusion buffer containing atenolol was added to the hexarelin/ lipid mixture. The final dispersion contained hexarelin at a concentration of 0.15 mM, atenolol at a concentration of 0.83 mM and lipids at a concentration of 15 mM.

B. Solution containing hexarelin and atenolol

Hexarelin (0.15 mM) and atenolol (0.83 mM) were dissolved in perfusion buffer. C. Formulation with hexarelin, atenolol, Pef abloc® SC and lipids

Hexarelin acetate was dissolved in sterile water to give 26% (w/w) (i.e. 22.1 % (w/w) pure peptide). The hexarelin acetate solution was thereafter mixed with a mixture of MG and PC (70/30 w/w). The amount of hexarelin solution in the mixture was 11.76 %(w/w). The tube containing the mixture was filled with nitrogen whereafter it was sealed. The tube was shaken moderately until the mixture appeared to be homogenous and clear (1,5 h at 37 °C). The hexarelin/ lipid mixture was then stored at -20°C until the perfusion study was performed. When starting the perfusion study, an isotonic NaCl-solution (0.9 % w/w) was added to the thawed hexarelin/ lipid mixture to allow for swelling of the lipids for 10 minutes at 20 °C. The amount NaCl-solution in the mixture was 82% (w/w). Then perfusion buffer containing atenolol and Pefabloc® SC (Pentapharm LTD, Basel, Switzerland) was added to the hexarelin/ lipid mixture giving the following final concentrations: 0.15 mM hexarelin, 0.83 mM atenolol, 1.25 mM Pefabloc® SC and 15 mM lipids.

D. Solution containing hexarelin, atenolol and Pefabloc® SC

Hexarelin (0.15 mM), atenolol (0.83 mM) and Pefabloc® SC (1.25 mM) were dissolved in perfusion buffer.

E. Formulation with atenolol, hexarelin and lipids

Atenolol was dissolved in sterile water to give 1.1% (w/w). The atenolol solution was blended with a mixture of MG and PC (70/30 w/w). The amount of atenolol solution in the mixture was 11.9 %(w/w). The tube containing the mixture was filled with nitrogen whereafter it was sealed. The mixture was then shaken moderately to give a clear dispersion (i.e. about 1.5 h at 37 °C). The atenolol/ lipid mixture was then stored at -20°C until the perfusion study was performed. When starting the perfusion study, an isotonic NaCl-solution (0.9 % w/w) was added to the thawed atenolol/ lipid mixture to allow for swelling of the lipids for 10 minutes at 20 °C. The amount of NaCl-solution in the mixture was 82 % (w/w). Thereafter, perfusion buffer containing hexarelin was added to the atenolol/ lipid mixture giving a final atenolol concentration of 0.028 mM, hexarelin concentration of 0.15 mM and lipid concentration of 15 mM. Compositions A, C and E are compositions according to the invention.

The study was divided into 3 parts (I, II and III).

In part I, the P_eff of hexarelin and atenolol was estimated in the absence and presence of 15 mM of the lipid matrix (A and B) in the rat jejunum (n=6), ileum (n=5) and colon (n=5), respectively.

Study part II was identical to part I, with the exception that the serine protease inhibitor Pefabloc ^® SC (0.3 g/L) was added to the perfusion solution, and that jejunal experiments were performed in 5 animals ( C and D).

In part III, atenolol but not hexarelin, was mixed with the lipid matrix prior to addition of perfusion buffer. Pefabloc ^®SC was not included in these experiments (E).

The atenolol concentration was lower than in study parts I and II. These experiments were performed in the jejunum of 5 rats. Basal P_eff-estimates of atenolol have previously been determined in the rat (Fagerholm et al., 1996).

Intestinal single-pass in situ perfusion experiments. Fasted male Sprague- Dawley rats weighing 200-280 g were anesthetizised with an i.p. injection of Inactin^®-Byk (thiobutabarbital sodium: 120 mg/kg), an anesthetic shown to have little or no influence on the P_eff (Yuasa et al, 1993; Fagerholm et al., 1996). After the surgical procedure (a more detailed description is presented in Fagerholm et al., 1996), and administration of a perfusion solution bolus dose of 4 ml via the inlet plastic tube, the isolated jejunal, ileal and colonic segments were perfused at 0.2 ml/ min (Harvard Apparatus 22) during 180 (parts I and II) or 90 min (part III), respectively. Perfusions in parts I and II were divided into two separate experimental periods of 90 min (first 90 min without and thereafter 90 min with lipid matrix), both of which were initiated with a bolus of 4 ml. Perfusate was quantitatively collected in test tubes kept on ice in the following intervals during each period; 0-45, 45-60, 60-75 and 75-90 min. At the end of the perfusion the intestinal segments were rinsed for 2-5 minutes with approximately 15 ml saline. This was undertaken in order to collect the remaining amounts of substances from the perfusion system. All perfusion syringes and perfusate samples were weighed, and the samples were frozen immediately and stored at -70°C until analysis.

In calculations in this report it was assumed that enzymatic degradation of hexarelin is a first-order process and that the metabolic half-life in the human intestine is approximately 3 hours. This corresponds to a metabolic rate constant (k_met) of 0.23

hr . There were no significant hexarelin P_eff differences with and without Pefabloc SC and between intestinal segments. This enabled pooling of data obtained in the presence and absence of Pefabloc SC.

Results and Discussion

The PEG recovery (PEGrec) ^was complete in all perfusion experiments, and the release of LDH into the intestinal lumen was generally not increased during exposure of the lipid matrix or Pefabloc^® SC (Table 1). This indicates a maintained intestinal viability, and absence of a pronounced mucosal damaging effect of these additives. The LDH activity declined along the intestine, which could be explained by a decreased anatomical intestinal surface area, and a shorter residence time in the colon, where the perfused segments were shorter. It might also be explained by a lower secretion rate of LDH from intestinal cells, since the fluidity of the apical cell membranes decreases from the proximal to distal intestine (Kararli, 1995). It is, however, difficult to interpret the implication of the LDH levels obtained in perfusate in this study, since no reference values of LDH release during in vivo conditions are available. The stability tests demonstrated that hexarelin is stable in perfusate during 180 min in body tempered fluids, which validates the estimated Peff- coefficients of hexarelin. In rat small intestinal fluids, however, the degradation of hexarelin was relatively pronounced. After 180 min about 30% of intact hexarelin was found in ileal fluids.

Peff -coefficients of hexarelin and atenolol, obtained in the absence and presence of the lipid matrix and/ or Pefabloc^® SC, are presented in Table 2. The estimated Peff -values of hexarelin demonstrated that this drug has a low intestinal permeability, which was neither different between intestinal regions nor influenced by Pefabloc^® SC (Table 2). The low P_eff agrees with its hydrophilic character at physiological pH (log D -2.3 at pH 6.5), and high molecular weight (887 g/mole). Otherwise, hexarelin possesses highly lipophilic properties in its unionized form (log

KD 2.2). In the presence of the lipid matrix, independently of the way of mixing hexarelin with the lipid matrix and whether or not Pefabloc^® SC was present, its jejunal Peff increased several-fold, on average 20 times (See Table 2 and Figure 1). In contrast, the uptake of atenolol was unaffected by the presence of the lipid matrix (See Table 2). One plausible mechanism of action of the lipid matrix might be an intermembrane transfer mainly of monoacylglycerol to the intestinal cells (Betageri, 1993). Fatty acids have previously been shown to cause a disordering in the interior of cell membranes, and thereby an increased membrane fluidity and permeability (Muranishi and Yamamoto, 1994). Using this drug delivery system, an enhanced membrane permeability seems therefore most likely to be observed for substances with high partitioning into apical cell membranes, such as lipophilic solutes and compounds with lipid soluble regions. The increases in the ileal and colonic Peff of hexarelin in the presence of the lipid matrix were insignificant (See Table 2 and Figure 1). Another mechanism that might be involved in the intestinal absorption of hexarelin, and which also might have been inactivated during the altered membrane characteristics, is the P-glycoprotein efflux pump (Karlsson et al, 1993; Bai et al, 1995; Saitoh and Aungst, 1995; Fricker et al, 1996). This mechanism is demonstrated to be responsible for secretion of drugs and peptides back into the intestinal lumen, and hexarelin has the properties (amphiphilic, partitioning coefficient > 2, protonated at physiological pH) to be a substrate for these transport systems (Karlsson et al, 1993; Saitoh and Aungst, 1995; Fricker et al, 1996). It is however difficult to tell whether this mechanism is of importance for hexarelin or not, because the specificity, capacity, and the distribution of P-glycoproteins in the animal and human intestine are not fully clarified.

Based upon an average P_eff of hexarelin in the rat jejunum of 0.14 • 10"⁴ cm/s (pooled jejunal P_eff data), and a degradation half -life of 3 hrs (kdegr ⁼ 0.23 hr¹) in the small intestine in vivo, a fa-value of 67% was predicted. For a metabolically stable compound with equal jejunal P_eff the estimated extent of absorption would have been 84%, which indicates that for hexarelin, and during these assumptions, the absorption rate exceeds the degradation rate. According to the predictions, at the times 5, 10 and 30 min absorption enhancement (jejunal P_eff = 2.7 • 10"⁴ cm/s) and total metabolic inhibition, would increase the extent of absorption of hexarelin to 88, 95 and 100%, respectively. The corresponding estimates assuming no inhibition of intestinal proteases were 86, 94 and 98%, respectively.

In summary, these results and predictions demonstrate that the lipid matrix effectively enhances the extent of absorption and hence oral bioavailability of amphiphilic peptide analogues, such as hexarelin.

Conclusion

We have clearly demonstrated that a drug delivery system of lipids can increase the effective permeability (Peff) of a hexapeptide, hexarelin, approximately 20 times using the in situ perfusion model in the rat jejunum.

Hexarelin might thus be suitable for oral delivery. It is predicted that hexarelin has an intermediate fractional absorption in humans after oral administration and that this is significantly enhanced in the presence of the lipid matrix. These additives seemed not to have a damaging effect on the intestinal mucosa.

Claims

1. A pharmaceutical composition based on a defined lipid matrix system having at least two lipid components and at least one bioactive compound characterized in that at least two of the lipid components are chosen from classes with different polarity in which at least one is bilayer forming and the bioactive compound is a low molecular weight peptide.

2. A composition according to claim 1 characterized in that the composition includes polar solvents, preferably a water-containing solvent.

3. A composition according to claim 1 or 2 characterized in that the polar solvent, preferably water-containing, is present in an amount such that individual and discrete lipid particles are formed.

4. A composition according to any of claims 1-3 characterized in that it is adapted for oral, rectal, buccal, sublingual, nasal, pulmonary, subcutaneous or transdermal use and therewith contains suitable excipients for each appropriate administration route.

5. A composition according to any of claims 1-4 characterized in that the bilayer forming lipid component is chosen from phosphatidyl choline, lyso phosphatidyl choline, phosphatidyl glycerol, phosphatidyl inositol, phosphatidyl serine or mixtures thereof or other phospholipids and/ or glycolipids and is present in an amount of 1 to 90% of the lipid system and preferably in the range of 1 to 50 % and more preferably 5-50 % (w/w).

6. A composition according to any of claims 1-5 characterized in that one lipid component is chosen from the classes of mono-, di- or triglycerides, protonated fatty acids or mixtures thereof, preferably of medium chain, and more preferably with 8-12 carbons .

7. A composition according to claim 5 and 6 characterized in that the bilayer forming lipid component is a phosphatidyl choline and the other lipid component is a monoglyceride.

8. A composition according to any of claims 1-7 characterized in that the low molecular weight peptide is present in an amount of 1-90% with respect to the lipid system, preferably in an amount of 1-70 % and more preferably in an amount of 2-50 % (w/w).

9. A composition according to any of claims 1-8 characterized in that the low molecular weight peptide has a molecular mass of less than 5000 Da.

10. A composition according to any of claims 1-9 characterized in that the low molecular weight peptide is stimulating the release of growth hormone.

11. A composition according to claim 10 characterized in that the low molecular weight peptide is stimulating the release of growth hormone and comprises 2- alkyl-D tryptophan.

12. A composition according to claim 11 characterized in that the low molecular weight peptide is hexarelin.

13. A composition according to any of claims 1-12 characterized in that it further comprises a peptidase and/ or protease inhibiting excipient or excipient which decreases the activity of the peptidase and/ or protease.

14. An oral preparation containing the composition according to any of claims 1- 13 and suitable excipients.

15. A method for preparing the compositions according to any of claims 1- 14 characterized by adding the low molecular weight peptide or a solution thereof to a stirred lipid mixture comprising at least two lipid components of which at least two of the lipid components are chosen from classes with different polarity in which at least one is bilayer forming.

16. A method for preparing the compositions according to any of claims 1- 14 characterized by adding a solution of low molecular weight peptide to a stirred mixture of a lipid mixture comprising at least two lipid components of which at least two of the lipid components are chosen from classes with different polarity in which at least one is bilayer forming, thereafter adding polar solvent and treating the resultant mixture mechanically or alternatively mixing the peptide with a polar solvent before adding the mixture to the stirred lipid mixture.

17. A method for stimulation of the release of growth hormone characterized by administering a therapeutically effective amount of the composition according to any of claims 1-14 and suitable excipients to a patient.

18. Method for increasing effective permeability of a low molecular peptide in jejunum by incorporation of the peptide in a lipid system having at least two lipid components of which at least one is bilayer forming.

19. Use of a defined lipid system having at least two lipid components of which at least one is bilayer forming for increasing the effective permeability of a low molecular peptide in jejunum. 1/1

Fig.l

B A B A B A