- Email: [email protected]
Intratracheal Delivery of Peptide and Protein Agents: Absorption from Solution and Dry Powder by Rat Lung †
Intratracheal Delivery of Peptide and Protein Agents: Absorption from Solution and Dry Powder by Rat Lungt FUSAOKOMADA', SEIGO IWAKAWA'~, NAKAYUKIYAMAMOTO~, HIDEOSAKAKIBARA~, AND KATSUHIKOOKUMURA'~ Received July 26, 1993, from the 'Department of Hospital Pharmacy, School of Medicine, Kobe University, Chuo-ku, Kobe 650, Japan, and #Institute for Life Science Research, Asahi Chemical Industry Co., Ltd., 632- 1, Mifuku, Ohito-cho, Tagata-gun, Accepted for publication February 2, 1994@. §Present address: Kobe Pharmaceutical University, Shizuoka 4 70-23, Japan. Motoyamakita, Higashinada-ku, Kobe 658, Japan. Abstract Proteins of high molecular weight and low lipophilicity must be administered parenterally to achieve the desired therapeutic blood levels. We investigated the absorption of peptide and protein agents by rat lung following their intratracheal administration, expressing it as percent bioavailability. An aqueous solution and/or a dry powder of calcitonin. insulin, thyrotropin stimulating hormone (TSH), follicle stimulating hormone (FSH), and human chorionic gonadotropin (HCG) was delivered into the exposed trachea of anesthetized rats, and blood was sampled from the jugular vein at specified intervals. The bioavailabilities of TSH, FSH, and HCG delivered in a solution of neutral pH were 2.5,2.3, and 0.2 % ,respectively. Transpulmonary absorption of a solution of these agents, administered with a surfactant or under acidic conditions, was 2-30 times greater than the values obtained in controls. The bioavailabilities of calcitonin, insulin, TSH, FSH, and HCG, given intratracheally as a dry powder, were 11.5, 6.5, 1.6, 0.6, and 0.1 YO,respectively. Following intratracheal administration, we noted a negative association between molecular weight and bioavailabili. The intratracheal route may thus be useful for delivering peptide and protein agents.
Introduction Peptide and protein therapeutic agents have been prepared for clinical use by new biotechnology techniques. Their high molecular weight and low lipophilicity require that they m u s t be administered by injection. We believe the lung may be a good site for transferring high molecular weight compounds such as peptides and proteins into systemic circulation. The thickness of the air-blood pathway measures less than 0.5 pml in parts of the alveolar region. The entire cardiacoutput perfuses the lungs via a network of fine capillaries.' The surface area available for absorption slightly exceeds that of t h e small intestine. We previously described the distribution and metabolism of several agents administered intratracheally to the lungs of rats and rabbits.2-6 The bioavailability of insulin after aerosol inhalation in rabbit was approximately 40% ? T h e bioavailability of insulin following an intratracheal aerosol in rats resembled that after subcutaneous administration.8 W e conducted this experimental study to determine t h e effect of surfactants, pH of t h e solutions, and d r y powder on t h e absorption of various peptides and proteins from the rat lung.
Experimental Section Materials-We used eel [Asu1J]calcitonin (4000 IU/mg, Asahi ChemicalIndustry Co., Ltd., Tokyo, Japan), recombinant human insulin (26.3 IU/mg, Eli Lilly, Indianapolis, IN), human thyrotropin stimulating hormone (TSH, 6 IU/mg, Bioscan Continental Inc., QuBbec, Canada), human follicle stimulating hormone (FSH, 174 IU/mg, Mochida +This paper was presented in part at Advances in Delivery of Therapeutic and Diagnostic Agents, '92, on 9-11 December, 1992, in Sydney, Australia. Abstract published in Advance ACS Abstracfs. March 15, 1994. @
c 1994, American Chemical Society and American Pharmaceutical Associa tion
Pharmaceutical Co. Ltd., Tokyo, Japan), and human chorionic gonadotropin (HCG, 4500 IU/mg, Mochida Pharmaceutical Co. Ltd., Tokyo, Japan). Sodium glycocholateand sodium citrate were supplied by Wako Pure Chemical (Osaka, Japan). All other chemicals were of analyticalreagent grade. Animal Experiments-Male Wistar rats weighing 250-300 g were used. They were anesthetized with pentobarbital (40 mg/kg, ip) during the experiments. After the trachea was exposed, a microsyringe with silicon tubing (0.9 mm 0.d.) was inserted through an incision made between the fifth and sixth tracheal rings caudal to the thyroid cartilage to a depth of 12-15 mm according to the method of Enna and S ~ h a n k e r . ~ TSH (40 mIU/kg), FSH (75 IU/kg), and HCG (2500 IU/kg) in 10 pL of isotonic buffer of either pH 7.0 (phosphate buffer), with or without 50 mM glycocholate, or pH 3.0 (isotonic citrate buffer) was injected into the trachea via the silicon tubing. Rats were maintained in the head-up position at an angle of 90' to the horizontal plane for 30 s after the administration of peptide and protein drugs in solution, and then at 15' during the experiment.8 Calcitonin (1.0 IU/kg), TSH (40 mIU/kg), FSH (7.5 IUlkg), and HCG (1800 IUlkg) in saline solution were injected intravenously using a microsyringe. Blood samples (100-150 pL) were collected from the jugular vein at 0,5,15,30,60,90,120,180,240,300, and 360 min after the intravenous and intratracheal administration of FSH (assaylimit,0.2mIU/mL) and HCG (assaylimit, 2 mIU/mL). Blood was collected from the jugular vein at 0,5,15,30,60,90,120,180,240, and 300 min after the intravenous and intratracheal administration of TSH. After 300 min the TSH concentration in plasma was below the limit of detection (0.03 pIUlmL). Blood samples were collected from the jugular vein at 0, 30, 60, and 120 rnin after the intratracheal administration of calcitonin. After 120 min the plasma calcitonin concentration was less than the limit of the radioimmunoassay for calcitonin (35 pg/mL). Plasma levels of immunoreactive TSH, FSH, and HCG were estimated using the IMX system for enzymeimmunoassay (Abbott, North Chicago, IL). Plasma calcitonin levels were measured by radioimmunoassay.1° Dataon plasma insulin levels after intratracheal administration were obtained from our previous study.8 We used a powder inhaler to administer the peptide or protein as a dry powder (Figure 1). To saturate the adhesion of dry powder to apparatus, the dry powder was passed through the inhaler three times before the experiment. The amount of dry powder remaining in the capsule after inhalation was less than 10%. Insulin, TSH, FSH, and HCG were dissolved in distilled water with lactose and lyophilized.Each dry powder was then grained using an agate mill. The biological activity of a dry powder with lactose for TSH, FSH, and HCG was 10 mIU/mg, 30 IU/mg, and 600 IU/mg, respectively. A dry powder of calcitonin with mannitol(4.5 IU/mg) was prepared with a jet mill. The diameter of the dry powder was estimated using a centrifugal particle size analyzer, (SA-CP2, Shimadzu, Kyoto, Japan). The weight percentage of diatribution values of calcitonin dry powder particles measuring (1 and 1-2 pm, estimated using a centrifugal particle size analyzer, were 11.4 and 22.1 % , respectively. The trachea was exposed and the inhaler (1.5 mm 0.d.) was inserted via a tracheal incision to a depth of 10 mm. A dry powder of 20 IU/kg calcitonin, 3.0 IU/kg insulin, 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG was each administered to rats using the dry powder inhaler (Figure 1). The weight of dry powder doses of approximately 1 mg/rat was inhalated. Citric acid, 0.5 mg/dose, was added to the dry powder to provide an acidic environment. After the inhalation of each agent, the rats were maintained in the head-up position at an angle of 15' to the horizontal plane for the experiment. Blood samples (10G150 rL) were collected as described before. Bioavailability-For each agent we calculated the area under the concentration-versus-time curve (AUC) from 0 to the final sampling
0022-3549/94/ 1200-863$04.50/0
Journal of Pharmaceutical Sciences / 863 Vol. 83, No. 6, June 1994
-
A)
point using the trapezoidal method. The bioavailability of TSH, FSH, and HCG was calculated by comparing the AUC followingintratracheal administration with that following intravenous administration. Bioavailabilitywas expressedas a percentage of intratracheal to intravenous administration. Data are reportedas means f SE. The AUC for insulin was obtained from our previous study.8 Statistical analysis was performed using ANOVA. A value of p < 0.05 was considered to be statistically significant.
Results
.
Transpulmonary Absorption from Solution-Drug concentrations were measured after the intratracheal administration of TSH, FSH, and HCG in 10 pL of isotonic phosphate buffer at pH 7.0 with or without 50 mM glycocholate, or in isotonic citrate buffer at pH 3.0. Plasma protein concentration vs time profiles are shown in Figure 2. A lag in absorption was observed followingintratracheal administration of a high molecular weight protein such as FSH and HCG from the pH 7.0 solution. However, no lag time in absorption was observed after administering TSH in the pH 7.0 solution. Each agent was absorbed by the lung as shown by the AUC values given in Table 1. The bioavailability following intratracheal administration in a pH 7.0 solution was generally higher for TSH and FSH than for HCG, as shown in Table 1. The plasma concentration of each agent after intratracheal administration of a pH 7.0 solution with 50 mM glycocholate was 2-3 times higher than that for the pH 7.0 solution without glycocholate (Figure 2). In addition, the plasma concentration of each agent after the intratracheal administration of a pH 3.0 solution exceeded that obtained in the pH 7.0 solution (Figure 2). The lag in the absorption of FSH and HCG decreased for the pH 3.0 and 7.0 solutions with 50 mM
Figure 1-Apparatus used to administer inhalation of dry powder to rat. A hard gelatin capsule contained a single dose of dry powder of peptide and protein agents. Following insertion of the capsule into the inhaler, it was opened, and the contents were dispensed into the airstream. The airstream was made using a 2.0-mL injector. The trachea was exposed and the inhaler (outer diameter 1.5 mm) inserted via the tracheal incision to a depth of 10 mm. Key: (A) apparatus for dry powder inhalation; (B) hard gelatin capsule containing peptiie and protein agents, (C) needle to open the two pores, (D)2.0-mL injector.
:I :" 0
2
0
4
2
'"1
4
6
0
2
140 120 160
TSH(1V)
100
4
6
hr
hr
hr
"1
HCGUV)
-
80-
6040-
20 0 0
2
4
hr
6
1
r
.
1
.
i
0
2
4
6
hr
Figure 2-Effects of pH and glycocholate on plasma drug levels after intratrachealadministration of 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG in 10 pL of an aqueous solution and intravenous administration of 40 mIU/kg, TSH, 7.5 IUlkg FSH. and 1800 IUlkg HCG: vertical bars indicate SE (n = 3-5); (0)intratracheal administration in isotonic phosphate buffer pH 7.0, (0)intratracheal administration in isotonic citrate buffer pH 3.0, (A)intratrachealadministration in isotonic pH 7.0 phosphate buffer with 50 mM sodium glycocholate, (0)intravenous administration.
864 /Journal of Pharmaceutical Sciences Vol. 83, No. 6, June 1994
Table 1-Area under Concentration vs Time Curve and Bioavaliablllties ( % ) following Intracheal Admlnlstratlon of Peptlde and Protein Agents to Rats* Dry Powder with Citrate
pH 7.0 with
Drug
pH 7.0 Solution
Dosee
Glycocholate
pH 3.0 Solution
Dry Powder
Calcitonin (h pg/mL) 20 IU/kg Insulin (h pIU/mL) (bioavailability)b TSH (h pIU/mL) (bioavailability) FSH (h mIU/mL) (bioavailability) HCG (h IU/mL) (bioavailability)
3.0 IU/kg
253.4 f 45.6 (13.1 f 2.4) 0.98 f 0.12 (2.5 f 0.3) 90.0 f 73.9 (2.3 f 1.0) 0.62 f 3.24 (0.2 f 0.1)
40 mIU/kg 75 IU/kg 2500 IU/kg
1303.8 f 186.8'" (67.2 f 9.6) 6.18 f 0.42" (16.2 f 1.1) 528.4 f 73.9" (13.6 f 1.9) 12.09 f 3.24' (4.8 f 1.3)
194.6 f 25.6 (11.5 f 2.6) 125.2 f 12.4 339.0 f 12.4 (6.5 f 0.6) (17.5 f 0.7) 0.57 f 0.08 (1.6 f 0.2) 24.8 f 4.6 (0.6 f 0.1) 0.17 f 0.04 (0.1 f 0.0)
807.7 f 156.0" (41.6 f 8.0) 3.10 f 1.41' (8.0 f 1.9) 351.1 f 112.6" (4.3 f 3.0) 15.87 f 1.90" (6.3 f 0.8)
Intravenous Injection [Dose] 90.8 f 9.2 [ 1.O IU/kg] 194.1 f 17.4d [0.3 IU/kglc 38.6 f 4.5 [40 mIU/kg] 389.4 f 38.7 [7.5 IU/kg] 181.57 f 7.08 [ 1800 IU/kg)1
Intratracheal administration. Bioavailabili in percent. Previously published data.8 Subcutaneous administration. Values are means < 0.05 and * * p < 0.01 compared with pH 7.0 solution.
f SE ( n = 3-8). Statistical significance: ' p
TSH
ij 0.2 6
0.0
0
2
1
0
hr 50-7
40
hr 0.2
FSH
1
2
4
hr
HCG
-
0
2
4
6
0.0
hr
hr
Figure 3-Change in plasma concentration after the intratracheal administration of 20 IU/kg calcitonin, 3.0 IU/kg insulin, 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG given in dry powder and in aqueous solution, and intravenous administration of 1.0 IU/kg calcitonin: vertical bars indicate SE (n = 3-5); (0) isotonic pH 7.0 phosphate buffer, (M) control dry powder, (A)dry powder with citrate, (0)intravenous administration. Data for the
insulin solution was obtained from our previous report.8
glycocholate compared with the pH 7.0 solution. Either the addition of glycocholate to the solution or an acidic solution increased the transpulmonary absorption of these agents by 2-30 times as compared with a solution of neutral pH (Table 1). Transpulmonary Absorption from Dry Powder-We observed lower rates of drug absorption when a dry powder containing calcitonin, insulin, TSH, FSH,and HCG with or without citric acid was delivered into the trachea compared with their administration in an aqueous solution (Table 1and Figure 3). The bioavailability of the dry powder was about one-half
that obtained with a solution of neutral pH. The bioavailability of insulin followingintratracheal administration as a dry powder containing citric acid was more than twice that seen with the control dry powder.
Discussion A previous study showed the bioavailability of insulin given as an inhaled aerosol to rabbits to be about 40% of that obtained Journal of Pharmaceutical Sciences / 885 Vol. 83, No. 6, June 1994
after an intravenous injection? The bioavailability of insulin in rats after administration in 10 p L of pH 7.0 isotonic phosphate buffer and pH 3.0 isotonic citrate buffer was 13% and 42%, respectively. The absorption of insulin in the presence of a surfactant such as glycocholate, surfactin, and Span 85 was 3-4 times greater than without a surfactant. The bioavailability of insulin given as an intratracheal aerosol resembled that following a subcutaneous iqjection.8 In this study, we found that a surfactant such as glycocholate enhanced the absorption by the lung of peptide and protein agents in solution (Figure 2). Previously, we found that the surfactant Span 85 enhanced pulmonary absorption of insulin from an aqueous solution.8 Niven and Byron" reported that Span 85 also enhanced the absorption of fluorescein from the airways. In short, Span 85 not only stabilizes insulin in aerosol form12 but also effectively enhances insulin absorption after intratracheal administration of the aerosol. Other surfactants also increase the absorption of insulin administered via the lung, nasally, buccally, or rectally.8J3-18 Hirai and coauthors14reported that glycocholate suppresses the degradation of insulin after nasal administration by inhibiting of aminopeptidase activity. In our previous work,8weshowed that nafamostat and bacitracin, both peptidase inhibitors, did not affect the bioavailability of insulin following intratracheal administration. Those agents inhibited the degradation of insulin given by subcutaneous inje~tion.lS*~* These findings demonstrated that the ability of proteolytic enzymes in the lung to degrade insulin is weaker than that in recta121and subcutaneous t i s s u e ~ . ~ Although ~,~3 no macroscopic injury has been found, this finding suggests that somehow the lung tissue barrier and/or the endogenous pulmonary surfactant is damaged by exogenous surfactant. We determined that the concentration-time profiles of drugs following their intratracheal administration in a solution of pH 3.0 rapidly increased with a decrease in lag time. The bioavailability was 2-30 times greater than that with a solution of pH 7.0 (Figure 2). Citric acid also improved the transpulmonary absorption of insulin given as an inhaled dry powder (Figure 3). EDTA and salicylate each increase paracellular transport by affecting the permeability of tight junctions as a consequence of calcium removal.24 However, salicylate and EDTA, which enhance the rectal absorption of insulin, did not increase its transpulmonary absorption.8 Aungst and Rogers18showed that EDTA and salicylate also had no effect on the buccal or nasal absorption of insulin. Therefore, the chelating effect of citric acid to calcium may not contribute to the enhanced mechanism of absorption by lung. These results indicate that the lung tissue barrier and/or endogenous surfactant on alveoli may be damaged by acidic conditions. The rate of transpulmonary absorption was slower following the intratracheal administration of these drugs given as a dry powder than as an aqueous solution. For the transpulmonary absorption of a drug to occur from a dry powder, it must first dissolve at the site before the drug is absorbed. The bioavailabilities of these agents were lower after their intratracheal administration as a dry powder than as an aqueous solution (Table 1). The size of the particles contained in the aerosol influences their distribution of the lung; particles of about 1 pm in diameter constitute the greatest sediment in the alveoli.25 Adjei and Garren26showed the importance of particle size in the pulmonary delivery of leuprolide to humans. In this present study, the mean size of the calcitonin dry powder was 4 pm; about 11.4%of the particles measured less than 1pm in diameter. The bioavailability of calcitonin following its inhalation as a dry powder directly into the rat tracheal incision was 11.5%. Therefore, a dry powder that contains particles smaller than 1 pm in diameter may reach the alveoli and be absorbed. The relationship between molecular weight and bioavailability followingthe intratracheal administration of peptide and protein agents under various conditions appears in Figure 4. The data 888 /
Journal of Pharmaceutical Sciences Vol. 83, No. 6, June 1994
Insulin 8o
1
T
R Calcitonin
I--...
HCG
Ib
4
1; 5
Molecular weight
Figure 4-Relationship
between bioavailability (%) and molecular weight following the intratracheal administration of peptide and protein agents under various conditions: vertical bars indicate SE (n = 3-7);(0)isotonic pH 7.0 phosphate buffer, (0)isotonic citrate buffer pH 3.0,(A)isotonic pH 7.0 phosphate buffer with 50 mM sodium glycocholate. (I control ) dry powder, (A)dry powder with citrate. The area under concentration vs time curves,AUCs, of insulin solutionunder various conditionswere obtained from our previous report.* on bioavailability following the intratracheal administration of peptide and protein agents of molecular weights ranging from about 3600 to 40 000 showed an inverse relationship to molecular weight. A hydrophilic drug is usually absorbed via the aqueous pores. Schanker and Hemberger27suggest that at least three different populations of pore size are required to describe the permeability of the rat lung to a range of hydrophilic compounds. The absorption of hydrophilic compounds of molecular weights ranging from 60 to 75 000 is inverselyrelated to molecular weight. Numerous small pores are permeable to compounds of molecular weights > 122 while their pores exclude compounds with molecular weights of 5250 and higher, while very few large pores are permeable to compounds with molecular weights exceeding 5250.28 High molecular weight and hydrophilic peptides and proteins are thought to be absorbed via the junction between the cells, and/or via pinocytosis. Rojanasakul et al.29 demonstrated that various epithelia possess a high selectivity for the absorption of positively charged solutes. In present study, the relationship between molecular weight and bioavailability following the transpulmonary absorption of these agents (isoelectric points are ca. 5.5-8) under both neutral and acidic pH was observed. Therefore, for the absorption via alveolar epithelia, the molecular weight of these agents may be more influential than the charge of the drug. In conclusion, our observations indicate that the intratracheal route may be useful in delivering peptide and protein agents.
References and Notes 1. Forrest, J. B. In Aerosols in Medicine; Morh, F.; Newhouse, M. T.; Dolovich, B. M., Eds.; Elsevier: Amsterdam, 1985; pp 21-52. 2. Okumura, K.; Yoshida, H.; Hori, R. J . Pharmacobio-Dyn. 1978, I , 230-237. 3. Hori, R.; Okumura, K.; Yoshida, H. Pharm. Res. 1987,4,142-146. 4. Okumura, K.; Yoshida, H.; Kamiya, A.; Hori, R. Chem. Pharm. Bull. 1989.37. 1109-1111. 5. Yoshida, H.; Okumura, K.; Kamiya, A.; Hori, R. Chem. Pharm. Bull. 1989, 37, 450-453. 6. Yoshida, H.; Okumura, K.; Hori, R. Pharm. Res. 1990,7,398-401. 7. Yoshida, H.;Okumura, K.; Hori, R.; Anmo, T.; Yamaguchi,H. J. Pharm. Sci. 1979.68. 670-671. 8. Okumura, K.; Iwakawa, S.; Yoshida, T.;Seki, T.;Komada, F. Int. J. Pharm. 1992,88,63-73. 9. Enna, S . J.; Schanker, L. S. J. Physiol. 1972, 222, 409-414. 10. Yamauchi, H.; Shiraki, M.; Orimo, H.; Tsukamoto, T.; Sakurada, T.; Watabe, S. Kotsutaisha 1979,12, 378-384.
11. Niven, R. W.; Byron, P. R. Pharrn. Res. 1990, 7,8-13. 12. MorBn, F. In Aerosols in Medicine; Morkn, F., Newhouse, M. T.; Dnlnvich. M. Amsterdam. 1985 DD 261-287. -~B.. Eds.:, Elsevier: - 13. Hirai, S.f IkenagGT.; Matsuzawa, T. Diabetes 1978;27,296-299. 14. Hirai, S.; Yashiki, T.; Mima, H. Int. J. Pharrn. 1981,9, 165-172. 15. Gorden, G. S.;Moses, A. C.; Filier, P. D.; Carey, M. C. Proc. Natl. Acad. Sci. U.S.A. 1985,82, 7419-7423. 16. Aungst, B. J.; Rogers, N. J. Pharrn. Res. 1988,5, 305-308. 17. Ichikawa, K.; Ohata, I.; Mitomi, M.; Kawamura, S.; Maeno, H.; Kawata, H. J. Pharrn. Pharmacol. 1980,32,314-318. 18. Moses,A. C.; Gordon, G. S.; Carey, M. C.; Flier, J. S. Diabetes 1983, 32, 1040-1047. 19. Komada. F.: Okumura, K.; Hori, R. J. Pharmacobio-Dyn. 1985,8, 33-40. 20. Takeyama, M.; Ishida, T.; Kokubu, N.; Komada, F.; Iwakawa, S.; Okumura, K.; Hori, R. Pharrn. Res. 1991,8, 60-64. 21. Hayakawa, E.; Chien, D.-S.; Lee, V. H. L. Pharm. Res. 1987,4,s-39. 22. Hori, R.; Komada, F.; Okumura, K. J. Pharrn. Sci. 1983,72,435439. 23. Okumura, K.; Komada, F.; Hori, R. J. Pharmacobio-Dyn. 1985,8, 25-32.
24. Cassidy, M. M.; Tidball, C. S. J. Cell Biol. 1967, 32,685498. 25. Brain, J. D.; Valberg, P. A.; Sneddon, S. In Aerosols in Medicine; Morkn, F.; Newhouse, M. T.; Dolovich, M. B., Eds.; Elsevier: Amsterdam, 1985; pp 261-287. 26. Adjei, A.; Garren, J. Pharrn. Res. 1990, 7 , 565-569. 27. Schanker, L. S.; Hemberger, J. A. Biochern. Pharrnacol. 1983,32, 2599-2601. 28. Taylor, G. Adv. Drug Delivery Rev. 1990,5, 37-61. 29. Rojanasakul, Y.; Wang, L.-Y.; Bhat, M.; Glover, D. D.; Malanga, C. J.; Ma., J. K. H. Pharm. Res. 1992,9, 1029-1034.
Acknowledgments This work was supported in part by a Grant-in Aid for Scientific Research (B)(No. 02921039) from the Ministry of Education, Science and Culture, Japan, and by the Japan Research Foundation for Clinical Pharmacology (No. 16).
Journal of Pharmaceutical Sciences / 867 Vol. 83. No. 6, June 1994
Introduction Peptide and protein therapeutic agents have been prepared for clinical use by new biotechnology techniques. Their high molecular weight and low lipophilicity require that they m u s t be administered by injection. We believe the lung may be a good site for transferring high molecular weight compounds such as peptides and proteins into systemic circulation. The thickness of the air-blood pathway measures less than 0.5 pml in parts of the alveolar region. The entire cardiacoutput perfuses the lungs via a network of fine capillaries.' The surface area available for absorption slightly exceeds that of t h e small intestine. We previously described the distribution and metabolism of several agents administered intratracheally to the lungs of rats and rabbits.2-6 The bioavailability of insulin after aerosol inhalation in rabbit was approximately 40% ? T h e bioavailability of insulin following an intratracheal aerosol in rats resembled that after subcutaneous administration.8 W e conducted this experimental study to determine t h e effect of surfactants, pH of t h e solutions, and d r y powder on t h e absorption of various peptides and proteins from the rat lung.
Experimental Section Materials-We used eel [Asu1J]calcitonin (4000 IU/mg, Asahi ChemicalIndustry Co., Ltd., Tokyo, Japan), recombinant human insulin (26.3 IU/mg, Eli Lilly, Indianapolis, IN), human thyrotropin stimulating hormone (TSH, 6 IU/mg, Bioscan Continental Inc., QuBbec, Canada), human follicle stimulating hormone (FSH, 174 IU/mg, Mochida +This paper was presented in part at Advances in Delivery of Therapeutic and Diagnostic Agents, '92, on 9-11 December, 1992, in Sydney, Australia. Abstract published in Advance ACS Abstracfs. March 15, 1994. @
c 1994, American Chemical Society and American Pharmaceutical Associa tion
Pharmaceutical Co. Ltd., Tokyo, Japan), and human chorionic gonadotropin (HCG, 4500 IU/mg, Mochida Pharmaceutical Co. Ltd., Tokyo, Japan). Sodium glycocholateand sodium citrate were supplied by Wako Pure Chemical (Osaka, Japan). All other chemicals were of analyticalreagent grade. Animal Experiments-Male Wistar rats weighing 250-300 g were used. They were anesthetized with pentobarbital (40 mg/kg, ip) during the experiments. After the trachea was exposed, a microsyringe with silicon tubing (0.9 mm 0.d.) was inserted through an incision made between the fifth and sixth tracheal rings caudal to the thyroid cartilage to a depth of 12-15 mm according to the method of Enna and S ~ h a n k e r . ~ TSH (40 mIU/kg), FSH (75 IU/kg), and HCG (2500 IU/kg) in 10 pL of isotonic buffer of either pH 7.0 (phosphate buffer), with or without 50 mM glycocholate, or pH 3.0 (isotonic citrate buffer) was injected into the trachea via the silicon tubing. Rats were maintained in the head-up position at an angle of 90' to the horizontal plane for 30 s after the administration of peptide and protein drugs in solution, and then at 15' during the experiment.8 Calcitonin (1.0 IU/kg), TSH (40 mIU/kg), FSH (7.5 IUlkg), and HCG (1800 IUlkg) in saline solution were injected intravenously using a microsyringe. Blood samples (100-150 pL) were collected from the jugular vein at 0,5,15,30,60,90,120,180,240,300, and 360 min after the intravenous and intratracheal administration of FSH (assaylimit,0.2mIU/mL) and HCG (assaylimit, 2 mIU/mL). Blood was collected from the jugular vein at 0,5,15,30,60,90,120,180,240, and 300 min after the intravenous and intratracheal administration of TSH. After 300 min the TSH concentration in plasma was below the limit of detection (0.03 pIUlmL). Blood samples were collected from the jugular vein at 0, 30, 60, and 120 rnin after the intratracheal administration of calcitonin. After 120 min the plasma calcitonin concentration was less than the limit of the radioimmunoassay for calcitonin (35 pg/mL). Plasma levels of immunoreactive TSH, FSH, and HCG were estimated using the IMX system for enzymeimmunoassay (Abbott, North Chicago, IL). Plasma calcitonin levels were measured by radioimmunoassay.1° Dataon plasma insulin levels after intratracheal administration were obtained from our previous study.8 We used a powder inhaler to administer the peptide or protein as a dry powder (Figure 1). To saturate the adhesion of dry powder to apparatus, the dry powder was passed through the inhaler three times before the experiment. The amount of dry powder remaining in the capsule after inhalation was less than 10%. Insulin, TSH, FSH, and HCG were dissolved in distilled water with lactose and lyophilized.Each dry powder was then grained using an agate mill. The biological activity of a dry powder with lactose for TSH, FSH, and HCG was 10 mIU/mg, 30 IU/mg, and 600 IU/mg, respectively. A dry powder of calcitonin with mannitol(4.5 IU/mg) was prepared with a jet mill. The diameter of the dry powder was estimated using a centrifugal particle size analyzer, (SA-CP2, Shimadzu, Kyoto, Japan). The weight percentage of diatribution values of calcitonin dry powder particles measuring (1 and 1-2 pm, estimated using a centrifugal particle size analyzer, were 11.4 and 22.1 % , respectively. The trachea was exposed and the inhaler (1.5 mm 0.d.) was inserted via a tracheal incision to a depth of 10 mm. A dry powder of 20 IU/kg calcitonin, 3.0 IU/kg insulin, 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG was each administered to rats using the dry powder inhaler (Figure 1). The weight of dry powder doses of approximately 1 mg/rat was inhalated. Citric acid, 0.5 mg/dose, was added to the dry powder to provide an acidic environment. After the inhalation of each agent, the rats were maintained in the head-up position at an angle of 15' to the horizontal plane for the experiment. Blood samples (10G150 rL) were collected as described before. Bioavailability-For each agent we calculated the area under the concentration-versus-time curve (AUC) from 0 to the final sampling
0022-3549/94/ 1200-863$04.50/0
Journal of Pharmaceutical Sciences / 863 Vol. 83, No. 6, June 1994
-
A)
point using the trapezoidal method. The bioavailability of TSH, FSH, and HCG was calculated by comparing the AUC followingintratracheal administration with that following intravenous administration. Bioavailabilitywas expressedas a percentage of intratracheal to intravenous administration. Data are reportedas means f SE. The AUC for insulin was obtained from our previous study.8 Statistical analysis was performed using ANOVA. A value of p < 0.05 was considered to be statistically significant.
Results
.
Transpulmonary Absorption from Solution-Drug concentrations were measured after the intratracheal administration of TSH, FSH, and HCG in 10 pL of isotonic phosphate buffer at pH 7.0 with or without 50 mM glycocholate, or in isotonic citrate buffer at pH 3.0. Plasma protein concentration vs time profiles are shown in Figure 2. A lag in absorption was observed followingintratracheal administration of a high molecular weight protein such as FSH and HCG from the pH 7.0 solution. However, no lag time in absorption was observed after administering TSH in the pH 7.0 solution. Each agent was absorbed by the lung as shown by the AUC values given in Table 1. The bioavailability following intratracheal administration in a pH 7.0 solution was generally higher for TSH and FSH than for HCG, as shown in Table 1. The plasma concentration of each agent after intratracheal administration of a pH 7.0 solution with 50 mM glycocholate was 2-3 times higher than that for the pH 7.0 solution without glycocholate (Figure 2). In addition, the plasma concentration of each agent after the intratracheal administration of a pH 3.0 solution exceeded that obtained in the pH 7.0 solution (Figure 2). The lag in the absorption of FSH and HCG decreased for the pH 3.0 and 7.0 solutions with 50 mM
Figure 1-Apparatus used to administer inhalation of dry powder to rat. A hard gelatin capsule contained a single dose of dry powder of peptide and protein agents. Following insertion of the capsule into the inhaler, it was opened, and the contents were dispensed into the airstream. The airstream was made using a 2.0-mL injector. The trachea was exposed and the inhaler (outer diameter 1.5 mm) inserted via the tracheal incision to a depth of 10 mm. Key: (A) apparatus for dry powder inhalation; (B) hard gelatin capsule containing peptiie and protein agents, (C) needle to open the two pores, (D)2.0-mL injector.
:I :" 0
2
0
4
2
'"1
4
6
0
2
140 120 160
TSH(1V)
100
4
6
hr
hr
hr
"1
HCGUV)
-
80-
6040-
20 0 0
2
4
hr
6
1
r
.
1
.
i
0
2
4
6
hr
Figure 2-Effects of pH and glycocholate on plasma drug levels after intratrachealadministration of 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG in 10 pL of an aqueous solution and intravenous administration of 40 mIU/kg, TSH, 7.5 IUlkg FSH. and 1800 IUlkg HCG: vertical bars indicate SE (n = 3-5); (0)intratracheal administration in isotonic phosphate buffer pH 7.0, (0)intratracheal administration in isotonic citrate buffer pH 3.0, (A)intratrachealadministration in isotonic pH 7.0 phosphate buffer with 50 mM sodium glycocholate, (0)intravenous administration.
864 /Journal of Pharmaceutical Sciences Vol. 83, No. 6, June 1994
Table 1-Area under Concentration vs Time Curve and Bioavaliablllties ( % ) following Intracheal Admlnlstratlon of Peptlde and Protein Agents to Rats* Dry Powder with Citrate
pH 7.0 with
Drug
pH 7.0 Solution
Dosee
Glycocholate
pH 3.0 Solution
Dry Powder
Calcitonin (h pg/mL) 20 IU/kg Insulin (h pIU/mL) (bioavailability)b TSH (h pIU/mL) (bioavailability) FSH (h mIU/mL) (bioavailability) HCG (h IU/mL) (bioavailability)
3.0 IU/kg
253.4 f 45.6 (13.1 f 2.4) 0.98 f 0.12 (2.5 f 0.3) 90.0 f 73.9 (2.3 f 1.0) 0.62 f 3.24 (0.2 f 0.1)
40 mIU/kg 75 IU/kg 2500 IU/kg
1303.8 f 186.8'" (67.2 f 9.6) 6.18 f 0.42" (16.2 f 1.1) 528.4 f 73.9" (13.6 f 1.9) 12.09 f 3.24' (4.8 f 1.3)
194.6 f 25.6 (11.5 f 2.6) 125.2 f 12.4 339.0 f 12.4 (6.5 f 0.6) (17.5 f 0.7) 0.57 f 0.08 (1.6 f 0.2) 24.8 f 4.6 (0.6 f 0.1) 0.17 f 0.04 (0.1 f 0.0)
807.7 f 156.0" (41.6 f 8.0) 3.10 f 1.41' (8.0 f 1.9) 351.1 f 112.6" (4.3 f 3.0) 15.87 f 1.90" (6.3 f 0.8)
Intravenous Injection [Dose] 90.8 f 9.2 [ 1.O IU/kg] 194.1 f 17.4d [0.3 IU/kglc 38.6 f 4.5 [40 mIU/kg] 389.4 f 38.7 [7.5 IU/kg] 181.57 f 7.08 [ 1800 IU/kg)1
Intratracheal administration. Bioavailabili in percent. Previously published data.8 Subcutaneous administration. Values are means < 0.05 and * * p < 0.01 compared with pH 7.0 solution.
f SE ( n = 3-8). Statistical significance: ' p
TSH
ij 0.2 6
0.0
0
2
1
0
hr 50-7
40
hr 0.2
FSH
1
2
4
hr
HCG
-
0
2
4
6
0.0
hr
hr
Figure 3-Change in plasma concentration after the intratracheal administration of 20 IU/kg calcitonin, 3.0 IU/kg insulin, 40 mIU/kg TSH, 75 IU/kg FSH, and 2500 IU/kg HCG given in dry powder and in aqueous solution, and intravenous administration of 1.0 IU/kg calcitonin: vertical bars indicate SE (n = 3-5); (0) isotonic pH 7.0 phosphate buffer, (M) control dry powder, (A)dry powder with citrate, (0)intravenous administration. Data for the
insulin solution was obtained from our previous report.8
glycocholate compared with the pH 7.0 solution. Either the addition of glycocholate to the solution or an acidic solution increased the transpulmonary absorption of these agents by 2-30 times as compared with a solution of neutral pH (Table 1). Transpulmonary Absorption from Dry Powder-We observed lower rates of drug absorption when a dry powder containing calcitonin, insulin, TSH, FSH,and HCG with or without citric acid was delivered into the trachea compared with their administration in an aqueous solution (Table 1and Figure 3). The bioavailability of the dry powder was about one-half
that obtained with a solution of neutral pH. The bioavailability of insulin followingintratracheal administration as a dry powder containing citric acid was more than twice that seen with the control dry powder.
Discussion A previous study showed the bioavailability of insulin given as an inhaled aerosol to rabbits to be about 40% of that obtained Journal of Pharmaceutical Sciences / 885 Vol. 83, No. 6, June 1994
after an intravenous injection? The bioavailability of insulin in rats after administration in 10 p L of pH 7.0 isotonic phosphate buffer and pH 3.0 isotonic citrate buffer was 13% and 42%, respectively. The absorption of insulin in the presence of a surfactant such as glycocholate, surfactin, and Span 85 was 3-4 times greater than without a surfactant. The bioavailability of insulin given as an intratracheal aerosol resembled that following a subcutaneous iqjection.8 In this study, we found that a surfactant such as glycocholate enhanced the absorption by the lung of peptide and protein agents in solution (Figure 2). Previously, we found that the surfactant Span 85 enhanced pulmonary absorption of insulin from an aqueous solution.8 Niven and Byron" reported that Span 85 also enhanced the absorption of fluorescein from the airways. In short, Span 85 not only stabilizes insulin in aerosol form12 but also effectively enhances insulin absorption after intratracheal administration of the aerosol. Other surfactants also increase the absorption of insulin administered via the lung, nasally, buccally, or rectally.8J3-18 Hirai and coauthors14reported that glycocholate suppresses the degradation of insulin after nasal administration by inhibiting of aminopeptidase activity. In our previous work,8weshowed that nafamostat and bacitracin, both peptidase inhibitors, did not affect the bioavailability of insulin following intratracheal administration. Those agents inhibited the degradation of insulin given by subcutaneous inje~tion.lS*~* These findings demonstrated that the ability of proteolytic enzymes in the lung to degrade insulin is weaker than that in recta121and subcutaneous t i s s u e ~ . ~ Although ~,~3 no macroscopic injury has been found, this finding suggests that somehow the lung tissue barrier and/or the endogenous pulmonary surfactant is damaged by exogenous surfactant. We determined that the concentration-time profiles of drugs following their intratracheal administration in a solution of pH 3.0 rapidly increased with a decrease in lag time. The bioavailability was 2-30 times greater than that with a solution of pH 7.0 (Figure 2). Citric acid also improved the transpulmonary absorption of insulin given as an inhaled dry powder (Figure 3). EDTA and salicylate each increase paracellular transport by affecting the permeability of tight junctions as a consequence of calcium removal.24 However, salicylate and EDTA, which enhance the rectal absorption of insulin, did not increase its transpulmonary absorption.8 Aungst and Rogers18showed that EDTA and salicylate also had no effect on the buccal or nasal absorption of insulin. Therefore, the chelating effect of citric acid to calcium may not contribute to the enhanced mechanism of absorption by lung. These results indicate that the lung tissue barrier and/or endogenous surfactant on alveoli may be damaged by acidic conditions. The rate of transpulmonary absorption was slower following the intratracheal administration of these drugs given as a dry powder than as an aqueous solution. For the transpulmonary absorption of a drug to occur from a dry powder, it must first dissolve at the site before the drug is absorbed. The bioavailabilities of these agents were lower after their intratracheal administration as a dry powder than as an aqueous solution (Table 1). The size of the particles contained in the aerosol influences their distribution of the lung; particles of about 1 pm in diameter constitute the greatest sediment in the alveoli.25 Adjei and Garren26showed the importance of particle size in the pulmonary delivery of leuprolide to humans. In this present study, the mean size of the calcitonin dry powder was 4 pm; about 11.4%of the particles measured less than 1pm in diameter. The bioavailability of calcitonin following its inhalation as a dry powder directly into the rat tracheal incision was 11.5%. Therefore, a dry powder that contains particles smaller than 1 pm in diameter may reach the alveoli and be absorbed. The relationship between molecular weight and bioavailability followingthe intratracheal administration of peptide and protein agents under various conditions appears in Figure 4. The data 888 /
Journal of Pharmaceutical Sciences Vol. 83, No. 6, June 1994
Insulin 8o
1
T
R Calcitonin
I--...
HCG
Ib
4
1; 5
Molecular weight
Figure 4-Relationship
between bioavailability (%) and molecular weight following the intratracheal administration of peptide and protein agents under various conditions: vertical bars indicate SE (n = 3-7);(0)isotonic pH 7.0 phosphate buffer, (0)isotonic citrate buffer pH 3.0,(A)isotonic pH 7.0 phosphate buffer with 50 mM sodium glycocholate. (I control ) dry powder, (A)dry powder with citrate. The area under concentration vs time curves,AUCs, of insulin solutionunder various conditionswere obtained from our previous report.* on bioavailability following the intratracheal administration of peptide and protein agents of molecular weights ranging from about 3600 to 40 000 showed an inverse relationship to molecular weight. A hydrophilic drug is usually absorbed via the aqueous pores. Schanker and Hemberger27suggest that at least three different populations of pore size are required to describe the permeability of the rat lung to a range of hydrophilic compounds. The absorption of hydrophilic compounds of molecular weights ranging from 60 to 75 000 is inverselyrelated to molecular weight. Numerous small pores are permeable to compounds of molecular weights > 122 while their pores exclude compounds with molecular weights of 5250 and higher, while very few large pores are permeable to compounds with molecular weights exceeding 5250.28 High molecular weight and hydrophilic peptides and proteins are thought to be absorbed via the junction between the cells, and/or via pinocytosis. Rojanasakul et al.29 demonstrated that various epithelia possess a high selectivity for the absorption of positively charged solutes. In present study, the relationship between molecular weight and bioavailability following the transpulmonary absorption of these agents (isoelectric points are ca. 5.5-8) under both neutral and acidic pH was observed. Therefore, for the absorption via alveolar epithelia, the molecular weight of these agents may be more influential than the charge of the drug. In conclusion, our observations indicate that the intratracheal route may be useful in delivering peptide and protein agents.
References and Notes 1. Forrest, J. B. In Aerosols in Medicine; Morh, F.; Newhouse, M. T.; Dolovich, B. M., Eds.; Elsevier: Amsterdam, 1985; pp 21-52. 2. Okumura, K.; Yoshida, H.; Hori, R. J . Pharmacobio-Dyn. 1978, I , 230-237. 3. Hori, R.; Okumura, K.; Yoshida, H. Pharm. Res. 1987,4,142-146. 4. Okumura, K.; Yoshida, H.; Kamiya, A.; Hori, R. Chem. Pharm. Bull. 1989.37. 1109-1111. 5. Yoshida, H.; Okumura, K.; Kamiya, A.; Hori, R. Chem. Pharm. Bull. 1989, 37, 450-453. 6. Yoshida, H.; Okumura, K.; Hori, R. Pharm. Res. 1990,7,398-401. 7. Yoshida, H.;Okumura, K.; Hori, R.; Anmo, T.; Yamaguchi,H. J. Pharm. Sci. 1979.68. 670-671. 8. Okumura, K.; Iwakawa, S.; Yoshida, T.;Seki, T.;Komada, F. Int. J. Pharm. 1992,88,63-73. 9. Enna, S . J.; Schanker, L. S. J. Physiol. 1972, 222, 409-414. 10. Yamauchi, H.; Shiraki, M.; Orimo, H.; Tsukamoto, T.; Sakurada, T.; Watabe, S. Kotsutaisha 1979,12, 378-384.
11. Niven, R. W.; Byron, P. R. Pharrn. Res. 1990, 7,8-13. 12. MorBn, F. In Aerosols in Medicine; Morkn, F., Newhouse, M. T.; Dnlnvich. M. Amsterdam. 1985 DD 261-287. -~B.. Eds.:, Elsevier: - 13. Hirai, S.f IkenagGT.; Matsuzawa, T. Diabetes 1978;27,296-299. 14. Hirai, S.; Yashiki, T.; Mima, H. Int. J. Pharrn. 1981,9, 165-172. 15. Gorden, G. S.;Moses, A. C.; Filier, P. D.; Carey, M. C. Proc. Natl. Acad. Sci. U.S.A. 1985,82, 7419-7423. 16. Aungst, B. J.; Rogers, N. J. Pharrn. Res. 1988,5, 305-308. 17. Ichikawa, K.; Ohata, I.; Mitomi, M.; Kawamura, S.; Maeno, H.; Kawata, H. J. Pharrn. Pharmacol. 1980,32,314-318. 18. Moses,A. C.; Gordon, G. S.; Carey, M. C.; Flier, J. S. Diabetes 1983, 32, 1040-1047. 19. Komada. F.: Okumura, K.; Hori, R. J. Pharmacobio-Dyn. 1985,8, 33-40. 20. Takeyama, M.; Ishida, T.; Kokubu, N.; Komada, F.; Iwakawa, S.; Okumura, K.; Hori, R. Pharrn. Res. 1991,8, 60-64. 21. Hayakawa, E.; Chien, D.-S.; Lee, V. H. L. Pharm. Res. 1987,4,s-39. 22. Hori, R.; Komada, F.; Okumura, K. J. Pharrn. Sci. 1983,72,435439. 23. Okumura, K.; Komada, F.; Hori, R. J. Pharmacobio-Dyn. 1985,8, 25-32.
24. Cassidy, M. M.; Tidball, C. S. J. Cell Biol. 1967, 32,685498. 25. Brain, J. D.; Valberg, P. A.; Sneddon, S. In Aerosols in Medicine; Morkn, F.; Newhouse, M. T.; Dolovich, M. B., Eds.; Elsevier: Amsterdam, 1985; pp 261-287. 26. Adjei, A.; Garren, J. Pharrn. Res. 1990, 7 , 565-569. 27. Schanker, L. S.; Hemberger, J. A. Biochern. Pharrnacol. 1983,32, 2599-2601. 28. Taylor, G. Adv. Drug Delivery Rev. 1990,5, 37-61. 29. Rojanasakul, Y.; Wang, L.-Y.; Bhat, M.; Glover, D. D.; Malanga, C. J.; Ma., J. K. H. Pharm. Res. 1992,9, 1029-1034.
Acknowledgments This work was supported in part by a Grant-in Aid for Scientific Research (B)(No. 02921039) from the Ministry of Education, Science and Culture, Japan, and by the Japan Research Foundation for Clinical Pharmacology (No. 16).
Journal of Pharmaceutical Sciences / 867 Vol. 83. No. 6, June 1994