Journal of Controlled Release, 2
`0 1992
`
`72
`
`Elsevier Science Publishers B.V. All rights reserved 01%3659/92/$05.00 165 COREL 00753 Microspheres as a nasal delivery system for peptide drugs P. EdmanaTb, E. Bjiirka>b and L. RydCnb
`
`“Kabi Pharmacia Therapeutics, Uppsala, Sweden
`bCrppsala University, Department of Pharma~~t~cs, BMC, Uppsa~a, Sweden
`
`(Received 7 July 199 1; accepted in revised form 9 March 1992 ) Microspheres of starch and dextran, cross-linked with epichlorohydrine, function as an enhancer system for the absorption of insulin in rats. The effect on the glucose level is rapid and maximal reduc- tion of plasma glucose is seen within 30-40 min. Starch microspheres are more effective than dextran spheres in inducing a decrease in blood sugar. The starch microspheres have been evaluated from a toxicological point of view in rabbits. The spheres were administered 2 times per day for 8 weeks and in two dosages, 10 and 20 mg. Scanning electron microscopy of the nasal mucosa showed no altera- tions. The only finding observed in light microscopy was a small hyperplasia in the septum wall. A preliminary test on healthy volunteers with starch microspheres given nasally for 1 week shows good acceptability. A temporary widening of the tight junctions in a monolayer of human epithelial (Caco- 2) cells was seen in the presence of dry starch microspheres. The widening of the tight junctions coin- cided with the increased absorption rate of insulin. A conceivable hypothesis with regard to the mech- anism of action of DSM can be that the epithelial mucosa is dehydrated, with a reversible “shrinkage” of the cells, thus giving a physical separation of the intercellular junctions. Key words: Insulin absorption; Microspheres Introduction The idea of giving peptides and proteins na- sally is not new. Immediately after the discovery of insulin, 70 years ago, attempts were made to give insulin by the nasal route [ 11. Since then, several peptides have been administered intra- nasally. In the treatment of diabetes insipidus, the nasal administration of vasopressin or its an- alogues is well known [ 2,3 1, Some of the reasons why nasal administration is attractive are (a) rapid absorption and quick pharmacodynamic effect of the drug; (b) avoidance of first pass me-
`
`P. Edman, Uppsala University, Dept. of Pharmaceutics, BMC, Box 580, S-75 1 23 Uppsala, Sweden. tabolism; (c) an easy administration route, par- ticularly suitable for selfmedication. However, even though the advantages of in- tranasal administration are clear, there are dis- advantages, such as different drug distribution in the nose, giving variable absorption, low perme- ability of the peptide through the mucosa [4] and degradation of the peptide drug by proteolytic enzymes in the nose [ 51. This often results in a low bioavailability of nasally given peptides. Consequently, a great deal of effort has been put into facilitating peptide absorption. Variations of pH [ 61, ionic strength f 71, and inhibition of proteolytic enzymes in the nose [ 8 ] are some ex- amples of parameters which can be varied in or- der to promote the nasal absorption of peptides.
`
`Correspondence to:
`
`AQUESTIVE EXHIBIT 1032 page 0001
`
`1 ( 1992) 165-l
`

`

`166 Often these steps are insufficient to achieve an acceptable bioavailability of the peptide. An en- hancer is therefore usually needed to reach an ac- ceptable absorption of larger peptides. Several enhancers have been tested, such as surfactants [ 91, bile salts [ 10 1, gels [ 111 and fu- sidic acid derivatives [ 121. When working with enhancers giving increased nasal absorption of peptides, one must carefully address the ques- tion of nasal irritation and cell damage. Re- cently, microspheres given as a nasal powder have been shown to promote the absorption of peptides/proteins [ 13 1. The enhancement of drug absorption is obtained if the spheres are water-insoluble and able to take up water and swell [ 141. This paper will give a survey of the use of dif- ferent microspheres as a nasal absorption pro- moting system. Not only the technical aspects but also biological/clinical and toxicological issues regarding microspheres will be considered. In vitro Characteristics The particle systems evaluated in this paper are based on starch and dextran cross-linked with epichlorohydrine. The degradable starch micro- spheres (DSM) are degradable with cr-amylase whereas Sephadex and DEAE-Sephadex are non degradable in biological compartments such as the nasal cavity. DEAE-Sephadex microspheres are substituted with 2-ethylaminoethyl groups giving anion-exchange properties. All three par- ticle systems fulfil the prerequisite for absorp- tion enhancing properties, i.e. they are water-ab- sorptive and water-insoluble. According to the manufacturer, the dextran spheres will exclude all molecules with a molec- ular weight higher than 5,000. Consequently, in- sulin will not be incorporated in the spheres, but will instead be situated on the surface. The starch microspheres (DSM) have a cut-off at approxi- mately 30,000-50,000, so insulin can in this case be situated both on the surface and inside the spheres. The two systems are of different particle size. The starch spheres have a mean particle size of 45 pm whereas the dextran spheres have a par- ticle size of between 50 and 180 pm. Preparation of spheres The dextran spheres and the degradable starch microspheres were mixed with human mono- component insulin 100 III/ml in a ratio of 100 mg spheres per ml insulin solution and 100 mg spheres per 800 ~1 insulin solution respectively. The obtained gels were freeze dried. The dry dextran spheres were passed through a 180 pm sieve and the starch powder through a 63 pm sieve. DEAE-sephadex was suspended in a buffer so- lution to equilibrate the ion-exchange groups. After sedimentation the gel that had been formed was mixed with insulin and buffer. The suspen- sion was again allowed to sediment and the gel formed was freeze dried. The dried spheres were passed through a 180 pm sieve. Insulin release The release kinetics were measured by placing 10 mg of the different sphere preparations in a diffusion chamber. The receiving compartment contained 0.15 M or 0.86 M sodium chloride so- lution. The experiments were performed as re- ported earlier [ 141. The protein content was de- termined with the Folin-Lowry assay [ 15 1. Starch microspheres with different swelling factors, i.e. different degrees of cross-linking, re- lease insulin at different rates. The swelling fac- tor, defined as the bead volume in cm3 obtained when 1 g DSM is allowed to swell in a buffer, is low when the degree of cross-linking is high. The swelling of the spheres and the release of insulin from the spheres with the low swelling factor of 5 is rapid: 90% of the incorporated amount is re- leased within 10 min. The rate of insulin release from loosely cross-linked starch with a swelling factor of 17 is slower: 60% of the incorporated amount is released within 10 min. The high swelling factor indicates that more liquid is needed for the spheres to be completely swollen. It will take a longer time for the incorporated in-
`
`AQUESTIVE EXHIBIT 1032 page 0002
`
`

`

`167
`
`C
`._
`
`( -s - P PI 3 -G B 3 z
`& 3 c .z s lO( L I 0 30 I 60 I 120 Time (min) I 4 180 240 Fig. 1. Change in blood glucose in rats after intranasal administration of insulin. Empty DSM 0.5 mg/kg (
`(n=6-8). (From BjGrk and Edman [ 131, reproduced by permission of Elsevier. ) sulin to be dissolved and to diffuse out of the ma- trix [ 141. The swelling factor of the dextran particles (Sephadex) is 4-6, i.e. more or less the same as the starch microspheres with a high degree of cross-linking. The release rate of insulin is also the same: 90% of the incorporated amount is re- leased within 10 min. DEAE-Sephadex requires a high saline concentration in the receiving com- partment before the insulin is released. This sug- gests that the insulin is bound to the DEAE- groups and can only be displaced by high ionic strength [ 161. In vivo Biological effect Rats have been widely used for intranasal drug
`
`);
`x
`(A);
`soluble insulin 2.0 W/kg
`SD
`
`AQUESTIVE EXHIBIT 1032 page 0003
`
`DSM+insulin 1.7 IU/kg (0). Thedataareexpressed asmeanf
`

`

`168 TABLE 1 Total decrease (P?) and maximal decrease of plasma glu- cose level after intranasal administration of insulin in three different sphere systems with the same swelling factor Dose D Maximal Number (IU/kg) (O/o) decrease of animals (%I DSM 1 16.4 26 6 Sephadex 1 7.5 24 3 DEAE-Sephadex 1 3.1 8 3 delivery studies. Three main models, with minor modifications, have been described in the litera- ture [ 17 1. The model used in these experiments is a modification of the “in vivo surgical model”. Instead of preventing drainage of the drug from the nasal cavity by sealing the nasopalatine tract leading from the nasal cavity to the oral cavity with an adhesive glue [ 18 1, the normal function was maintained to the fullest possible extent. The rats were anesthetized and operated on as earlier reported, [ 13 1. The insulin preparations were given through the nostril with a polyethyl- ene tube 30 min post-operation. The spheres were weighed in the tube and administered by blow- ing air from a syringe through the tube. Insulin was administered in DSM (swelling factor 8-10) with an amount of 3.5 mg DSM/ kg, This resulted in a rapid decrease in blood glu- cose (Fig. I). The maximal reduction of the blood glucose level after a dose of 1.7 IU/kg body weight was 64%. The maximal decrease was reached approximately 40 min after administra- tion and the glucose level was normalized after 4 h. The decrease in blood glucose was found to be dose-dependent. However, the time to maximal decrease and the time for the glucose level to normalize were independent of the dose [ 131. A later study by Bjiirk and Edman [ 141 showed that starch microspheres with different swelling factors are biologically equivalent in reducing the blood glucose level, in spite of the differences in in vitro release rate. Insulin 1 IV/kg administered together with Sephadex particles, 5 mg/kg, also induced a rapid decrease in blood glucose but not to the same ex- tent as DSM. The maximal decrease in plasma glucose, 25%, was seen 40-60 min after admin- istration and the level was normalized within 3 h. DEAE-Sephadex, on the other hand had no ef- fect on the glucose level. The total decrease in plasma glucose level (0%) was calculated ac- cording to Hirai: Dy _AUG-AUG x 1oo 0- AUC, AUC, and AUCj = area under plasma glucose vs time curve from 0 to 4 h after administration of control and insulin preparation. The total de- crease values (0%) and the maximal decrease are summarized in Table 1. A plausible explanation of the lack of effect with DEAE-Sephadex can be that the available ionic strength in the nasal cavity is too low to displace the bound insulin. Therefore saline of high ionic strength was administered nasally 5 min after instillation of the powder. However, no effect on the plasma glucose level was seen [ 16 1, Insulin must probably be available for absorp- tion during the first minutes after administra- tion of the microspheres to have any effect on the blood glucose, i.e. during the time the spheres will absorb water and swell to a gel. This result supports the hypothesis launched by Edman and BjSrk [ 141 that during the time they swell the spheres will draw water from mucus and the un- derlying epithelial barrier and thereby cause widening of the tight junctions, resulting in en- hanced transport of hydrophilic molecules. Toxicological evaluation The effect of DSM on the nasal mucosa after repeated administration was studied in rabbits [ 19 1. Degradable starch microspheres (DSM) were administered twice daily for two, four or eight weeks. Two different doses were used, 10 and 20 mg. The spheres were administered with an insufflator which gave a good spreading of the powder in the nasal cavity of the rabbit. Only one
`
`AQUESTIVE EXHIBIT 1032 page 0004
`
`

`

`I69
`
`Fig. 2. Cross-section of the septum of one rabbit administered 20 mg DSM for 8 weeks. Mild hyperplasia of the columnar epithelium with an increased number of goblet cells on the administration side LM x 200 (From Bj6rk et al. [ 191, reproduced by permission of Elsevier). nostril was used for dosing, the other served as a control. After the end of the administration pe- riod the noses were prepared for either scanning electron microscopy (SEM ) or light microscopy (LM) study. The cilia were intact in all groups and no changes in the amount and location of the ciliated cells were seen with SEM. Only in some animals given DSM for 8 weeks did the na- sal septum show mild, focal hyperplasia of the epithelium. It was characterized by a mild in- crease in the number of goblet cells and mild hy- perplasia of the columnar epithelium (Fig. 2 ) . The effect was more obvious with the higher dose. This is probably due to the water holding capac- ity of the microspheres giving a desiccated mucosa. Interaction with cells The cell model (Caco-2 ), originating from a human colorectal carcinoma, was used for the permeability studies [20]. The cell model is a human intestinal epithelial cell line that differ- entiates to enterocyte-like cells under cell culture conditions. The cells are grown on a special filter in a chamber allowing measurement of the transport of different marker molecules across the cell layer. In this model we have applied DSM on the top of the cells for a certain time and after that time the test substance 3H-mannitol (MW 180) was added.
`
`AQUESTIVE EXHIBIT 1032 page 0005
`
`

`

`Fig. 3. ‘EM picture of open tight junction between two cells. The cells were fixed 15 min after the application of mar&tot. Degradable starch microspheres increased the transport ofmannitol across the cell layer during a certain period of time. The permeability coef- ficient returned to normal at the end of the ex- periment, showing a reversible effect of DSM on the permeation of mannitol. The same was done with insulin (MW 6000) and an increased trans- port with the same kinetic profile was seen. Transmission electron microscope (TEM ) pictures were taken of Caco-2 cells exposed to DSM. TEM revealed that the cells and cilia were not damaged and were intact. The only change that could be seen was an opening of the tight junctions to a higher degree in cells exposed to DSM (Fig. 3 ) and that this effect was reversibie. They were closed again after 180 min [ 2 11. This finding indicates that the mode of action of DSM on the epitheiial cell barrier, giving in- creased permeability of hydrophiIic compounds, is a temporary widening of the tight junctions,
`
`AQUESTIVE EXHIBIT 1032 page 0006
`
`

`

`probably induced by the water holding of the spheres. Acceptability in humans The effect of degradable starch microspheres (DSM) on human nasal tissue was evaluated in a pilot clinical study. Fifteen volunteers were given 10 mg DSM in each nostril once daily for seven days. Their physiological status was checked before, during and after administration of the spheres. The spheres were well accepted by the test persons. The mucociliary clearance time did not seem to be affected by the spheres, as studied with the saccharin test developed by An- dersen et al. [ 22 1. This suggests that even if the DSM adhere to the mucosa for a longer time than a solution or particle they do not inhibit the mu- cosal ciliary function during this test period [ 23 1. Conclusions The main advantage of microspheres given nasally is their rapid promotion of insulin ab- sorption, almost simulating intravenous injec- tion. This situation mimics the physiological be- haviour of insulin when the blood sugar is raised during a meal. From the present paper it is clear that the microspheres must be insoluble in water but should absorb water in order to promote the ab- sorption of insulin given together with the spheres. Furthermore, a difference in efficacy exists be- tween Sephadex@’ and DSM, probably depen- dent on size and porosity (cut-off limit). The DSM are biologically acceptable in healthy vol- unteers, giving no effect on the mucociliary transport function and no inflammation at least not in the time frame studied. A temporary wid- ening of the tight junctions between the epithe- lial cells induced by DSM probably contributes to the effect seen. Even though a lot of informa- tion has been gathered about DSM, there are several questions which have to be answered be- fore the microspheres can be considered as a re- 171 alistic alternative for the nasal formulation of peptide drugs. An extended toxicological evaluation of microspheres given nasally has to be performed before the spheres can be classified as a tolerant dosage form. A prolonged desiccation of cells is not harmless. Immunological aspects must also be addressed. Classical parameters in pharma- ceutics such as dose variation, stability, compat- ibility, up-scaling, bacterial contamination etc must be investigated. Even if there are many questions unanswered with regard to micro- spheres they are still an interesting concept for the enhancement of drug absorption through na- sal mucosa. 1 2 3 4 5 6 7 8 9 10 References R.T. Woodyatt, The clinical use of insulin, J. Metabol. Res. 2 (1922) 793-801. AS. Harris, I.M. Nilsson, Z.G. Wagner and U. Alkner, Intranasal administration of peptides: nasal deposition biological response, and absorption of desmopressin, J. Pharm. Sci. 75 (1986) 1085-1088. K. Morimoto, H. Yamaguchi, Y. Iwakura, K. Morisaka, Y. Ohashi and Y. Nakai, Effects of viscous hyaluronate- sodium solution on the nasal absorption of vasopressin and an analogue, Pharm. Res. 8 ( 199 1) 47 l-474. A.N. Fischer, K. Brown, S.S. Davis, G.D. Parr and D.A. Smith, The effect of molecular size on the nasal absorp- tion of water-soluble compounds in the albino rat, J. Pharm. Pharrnacol. 39 (1987) 357-362. R.E. Stratford, Jr. and V.H.L. Lee, Aminopeptidase ac- tivity in homogenates of various absorptive mucosae in the albino rabbit: implications in peptide delivery, Int. J. Pharm. 30 (1986) 73-82. T. Ohwaki, H. Ando, S. Watanabe and Y. Miyake, Ef- fects of dose, pH and osmolarity on nasal absorption of secretin in rats, J. Pharm. Sci. 74 (1985) 550-552. P. Tengamnuay and A.K. Mitra, Bile salt-fatty acid mixed micelles as nasal absorption promotors of pep- tides. I. Effects of ionic strength, adjuvant composition, and lipid structure on the nasal absorption of [ D-Arg’] kyotorphin, Pharm. Res. 7 (1990) 127-133. M.A. Hussain, C.A. Koval, A.B. Shenvi and B.J. Aungst, Recovery of rat nasal mucosa from the effects of ami- nopeptidase inhibitors, J. Pharm. Sci. 79 ( 1990) 398- 400. S. Hirai, T. Yashiki and H. Mima, Effect of surfactants on the nasal absorption of insulin in rats, Int. J. Pharm. 9(1981)165-172. G.S.M.J.E.Duchateau,J.ZuidemaandF.W.H.M.Mer- kus, Bile salts and intranasal drug absorption, Int. J. Pharm. 31 (1986) 193-199.
`
`AQUESTIVE EXHIBIT 1032 page 0007
`
`

`

`172 11 12 13 14 15 16 K. Morimoto, K. Morisaka and A. Kamada, Enhance- ment of nasal absorption of insulin and calcitonin using polyacrylic acid gel, J. Pharm. Pharmacol. 37 ( 1985 ) 134-136. J.P. Longenecker, A.C. Moses, J.S. Flier, R.D. Silver, M.C. Carey and E.J. Dubovi, Effects of sodium tauro- dihydrofusidate on nasal absorption of insulin in sheep, J. Pharm. Sci. 76 (1987) 351-355. E. Bjork and P. Edman, Degradable starch micro- spheres as a nasal delivery system for insulin, Int. J. Pharm. 47 (1988) 233-238. E. Bjork and P. Edman, Characterization of degradable starch microspheres as a nasal delivery system for drugs, Int. J. Pharm. 62 (1990) 187-192. O.H. Lowry, N.J. Rosebrough, A.L. Farr and R.J. Ran- dall, Protein measurement with the Folin phenol re- agent, J. Biol. Chem. 193 (1951) 265-275. L. RydCn and P. Edman, Effect of polymers and micro- spheres on the nasal absorption of insulin in rats. Int. J. Pharm., in press. 17 18 19 20 21 22 23 S. Gizurarson, Animal models for intranasal drug deliv- ery studies. A review article, Acta Pharm. Nord. 2 (1990) 105-122. S. Hirai, T. Yashiki, T. Matsuzawa and H. Mima, Ab- sorption of drugs from the nasal mucosa of rat, Int. J. Pharm. 7 (1981) 317-325. E. Bjork, S. Bjurstrijm and P. Edman, Morphologic ex- amination of rabbit nasal mucosa after nasal adminis- tration of degradable starch microspheres, Int. J. Pharm., in press. P. Artursson, Epithelial transport of drugs in cell cul- ture. I. A model for studying the passive diffusion of drugs over intestinal absorptive (CaCo-2) cells, J. Pharm. Sci. 79 (1990) 476-482. E. Bjork, U. Isaksson, P. Edman and P. Artursson, unpublished. I. Andersen, P. Camner, P.L. Jensen, K. Philipson and D.F. Proctor, Nasal clearance in monozygotic twins, Am. Rev. Respir. Dis. 110 (1974) 301-305. E. Bjork, K. Holmberg and P. Edman, unpublished.
`
`AQUESTIVE EXHIBIT 1032 page 0008
`
`