Development of a Nanoparticle Controlled-Release Formulation for Human Use

205 of Controlled Release, 3 (1986) 206-210 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands Journal DEVELOPMENT OF A NANOPARTICLE FOR HUMAN USE CONTROLLED-RELEASE FORMULATION C. Verdun”, P. Couvreur+, H. Vranckx”, V. Lenaerts+ and M. Roland+ *Soper S.A. Company, ‘Laboratoire (Received 6328 de Pharmacie Sart-Dames-Avelines GalBnique, January 28, 1985; accepted University (Belgium) Catholique de Louvain, in revised form October 1200 Bruxelles (Belgium) 3,1985) Prior to the first clinical trials of doxorubicin-loaded nanoparticles, it was necessary to prepare this formulation in such a way as to meet the requirementsgenerally associated with parenteral administration. This paper describes the conditions under which nanoparticles should be prepared and lyophilized in order to be sterile and free of bacterial endotoxins. These nanoparticles were also subjected to a resuspension test and their size and drug adsorption capacity were found practically unchanged. A gel permeation chromatographic method allowed both an estimation of the molecular weight distribution of the cyanoacrylic polymer and the detection of the possible presence of monomeric residues. Additional data have shown that the doxorubicin adsorbed on nanoparticles was released well after intravenous injection in mice. Blood clearance of the drug was observed to be disminished when it was linked to nanoparticles, whereas its cardiac concentration was considerably reduced. Finally, preliminary stability assays, carried out after 6 and 12 months of storage, showed no modification in nanoparticle size, drug con tent and relative moisture con tent. INTRODUCTION In previous papers, we have described the preparation of submicroscopic particles by polymerization of alkylcyanoacrylates [ 1, 21. These polymers were chosen because of their biodegradability [3, 41. Their frequent use in surgery constitutes a favourable toxicological a priori [ 51. Furthermore, the anionic polymerization of cyanoacrylates in aqueous media requires no energy source for its initiation, thus avoiding the alteration of drug molecules that can occur when such means as gamma or UV radiation or elevated temperatures are needed. These particles of about 200 nm diameter are able to adsorb a large variety of drugs with high efficiency and to modify the tissue distribution of these drugs after intravenous ad- ministration [6] . The alteration of the distribution profile of drugs by linkage to nanoparticles can, in the case of some anticancer drugs, considerably reduce the toxicity of the drug [7]. This phenomenon can probably be explained by the observed reduction in the accumulation of the drug in organs where the most acute toxic effects are exerted [8]. In other in vivo experiments, an increased efficacy was observed for anticancer drugs on different cancer models when these drugs were adsorbed on nanoparticles [9] . Finally, preliminary data were collected concerning the acute and sub-acute toxicity of nanoparticles, revealing a remarkable safety in their use. Owing to these considerations, and in order to make clinical assays possible, it was necessary to develop, at a semi-industrial level, a 206 solid nanoparticles formulation consistent with the requirements of an intravenous administration. The aim of this paper is to describe the experimental conditions required to prepare doxorubucin or dactinomycinloaded nanoparticles. Information is also given concerning drug release in viva. MATERIALS AND METHODS Preparation and freeze-dn/ing of nanoparticles Isobutyl cyanoacrylate monomer (20 ml) was added under mechanical stirring to 2 1 of an aqueous polymerization medium (glucose, 5%; dextran 70, 1%; citric acid, 0.5%) containing doxorubicin (750 pg/ml) or dactinomycin (50 pg/ml). After polymerization of the monomer during 4 h, the two liters of nanoparticle suspension were divided into 148 vials of 13.3 ml each. Freeze-drying took place in a Lyovac GT2 freeze-drier (Leybold Heraeus) during 90 h under vacuum (6 X lo-’ mbar). Resuspension of solid nanoparticle formulations was carried out by simple addition of distilled water (5 ml) to the vial. Doxorubicin and dactinomycin nanopartitle preparation as well as freeze-drying were done in an aseptic room under laminar flow. Materials were sterilized by dry heat before use. A sterile filtration of the polymerization medium was also carried out using a 0.22 +rn Millipore filter. All chemicals were pyrogen free. Determinations of size and molecular weight of nanoparticles Five nanoparticles samples were taken and each lyophilisate was resuspended in 5 ml distilled water. In each suspension, the size of the nanoparticles was determined using a laser light scattering method (Nanosizer-Coulter). Molecular weights and residual monomer contents [12] were evaluated by gel permea- tion liquid chromatography (Waters Ass., M-45 solvent delivery system, U6-K universal LC injector), using a refractive index detector (Waters Ass., R-401 differential refractometer). Columns of p Styragel of 100, 500 and 1000 A were used simultaneously (Waters Ass.). Tetrahydrofuran, with a solvent flow of 2 ml/min, was used as eluant. After centrifugation at 20,000 rpm, nanoparticles were dissolved in tetrahydrofuran at a concentration of 5 mg/ml. Aliquots of 150 ~1 of tetrahydrofuran solution were filtered through a 0.45 pm filter and injected into the chromatograph. The chromatograms were registered and the peak surfaces integrated on a printer fitted with GPC calculation capacity (Waters Ass., M 730 Data Module). Polyethyleneglycol standards were used for column calibration. Determinations of drug content and drug release Measurement of doxorubicin- and dactinomycin- loaded nanoparticles Five samples of doxorubicin-loaded nanoparticles were taken out and resuspended in 5 ml distilled water. Nanoparticle suspensions were centrifuged at 20,000 rpm for 1 h. Sediments were then separated and dissolved in 5 ml of dioxan-water (4:l). Doxorubicin contents were determined in both supernatant and sediment by fluorimetric measurements according to a previously published method [71. Five samples of dactinomycin-loaded nanoparticles were taken out and resuspended in 5 ml distilled water. After centrifugation (20,000 rpm, 1 h), dactinomycin contents were measured in both sediment (bound dactinomycin) and supernatant (free dactinomycin) by scintillation counting, as previously described [13]. The level of drug binding is expressed as the percentage of drug associated with the carrier in comparison with the initial amount of drug which was previously dissolved in the polymerization medium. 207 Drug release of doxorubicin from nanoparticles after intravenous administration to mice Amounts of 7 mg/kg of free and nanoparticle-bound doxorubicin were administered intravenously to NMRI mice. Animals were killed at various time intervals and blood samples were collected in duplicate. At the same time, samples of heart tissue were taken out, in view of the well-known cardiac toxicity of the drug. Both plasmatic and cardiac doxorubicin concentrations were determined by a modified HPLC method [14] . Errors due to blood contamination of the cardiac tissue were avoided by measuring both cardiac and plasma proteins in the cardiac tissue homogenate. Corrections were made following the method described by Mancini [15]. To 500 ~1 of the blood sample, 15 ~1 of an aqueous daunorubicin solution (2 pg/ml) was added as internal standard. After the addition of borate buffer pH 9.2 (100 pl), the drugs were extracted by 1 ml of chloroformmethanol (6:l); 50 ~1 of the organic layer were then injected into the chromatograph (Waters Ass.) fitted with a luminescence spectrometer (Perkin-Elmer LS-5) and connected to a Lichrosorb (Si-60) 5 ym column. Elution was performed at a flow rate of 0.9 ml/min using a mixture of chloroformmethanol- anhydrous acetic acid-MgClz water solution (72:21:4:3). Measurements were made fluorimetrically (480-560 nm). Peak areas were computed using an integrator (Waters Ass., M 730 Data Module). Evaluation of the sterility and the pyrogenicity of the nanoparticle preparations Sterility assays were realised according to the European Pharmacopoeia by a membrane filtration technique. Ten samples were tested and filtered through a membrane with a nominal porosity of 0.45 pm. Membranes were transferred to culture media (thioglycolate and caseinsoya). Incubation took place for one week at 37°C for thioglycolate medium and at room temperature for caseinsoya medium. According to the European Pharmacopoeia, bacterial endotoxin tests were carried out by measuring the rise in temperature of 3 healthy adult rabbits following intravenous administration of the nanoparticle samples. Stabiiity of nanoparticle preparations Nanoparticle samples were stored at -30°C in a deep-freezer for periods of 6 and 12 months. Stability was controlled both for nanoparticle size and for drug adsorption rate. Furthermore, moisture content of the samples was estimated by the Karl Fischer technique before and after storage. RESULTS AND DISCUSSION After the freeze-drying process, both doxorubicin- and dactinomycin-loaded nanopartitles were easily resuspended in water. comparative particle size Furthermore, measurements showed no significant modification of the carrier dimensions (Table 1). Likewise, the level of drug binding to nanoparticles was not altered by the freeze-drying process for both doxorubicin and dactinomytin (Table 1). It was shown that gel permeation chromatography is a satisfactory method to determine the molecular weight of the polymer and to examine the presence of possible monomeric residues. Compared with unloaded nanoparticles [ 121, the molecular weight of the polymer was dramatically increased by the addition of doxorubicin to the polymerization medium. Indeed, Fig. 1 shows a bimodal distribution with a peak corresponding to a high molecular weight of about 45,000. Furthermore, for the sample presenta peak corresponding to monomeric ed, residues was seen after a retention time of 13.3 min. Quantitative evaluation of monomeric residues was, however, hazardous because of partial polymerization of the monomer in the column. For the samples tested, reproducibility of molecular weight values was satisfactory. 208 polymer 50 760 a / -_ monomeric polymer r&.idue 1743 b:T 0.5 RETENTION TIME Fig. 1. Gel permeation chromatographic profiles of (a) unloaded polyisobutylcyanoacrylate nanoparticles, and (b) nanoparticles loaded with doxorubicin. After intravenous administration to mice, plasma levels of doxorubicin were higher when the drug was adsorbed on the nanoparticulate carrier (Fig. 2). At the same tune, the cardiac concentration of the drug was dramatically reduced (Fig. 3). Concerning the sterility assays, none of the tested nanoparticles samples showed any germ growth, neither in the thioglycolate nor in the caseinsoya medium. Likewise, no bacterial TABLE c 0 ‘*_ -_*_____---______f 1 30 1 60 1 90 TIME 1 120 (min) Fig. 2. Plasma concentrations of free doxorubicin (A) and nanoparticle-bound doxorubicin (0) after intravenous administration to mice (dose of doxorubicin administred: 7 mg/kg). endotoxin was found in the analysed nanoparticle lyophilisates. Furthermore, after storage for 6 and 12 months in a freezer, nanoparticles remained unchanged with respect to size and level of drug binding (Table 2). Finally, the moisture content of the samples was generally between 3% and 4%. 1 Particle size and level of drug binding before and after freeze-drying of doxorubicin-loaded nomycin-loaded (DACT-NP) polyisobutylcyanoacrylate nanoparticles DOX-NP (DOX-NP) and dacti- DACT-NP Size (nm) Drug binding level (%) Size (nm) Drug binding level (%) Before freeze-drying 180 + 10 89 i 5 215 + 10 95 + 3 After freeze-drying 190 fr 10 88 f 5 242 i 10 95 * 3 205 ‘? I L o.*c P- \ ?\ \ 2 0.1 - \ \ \ 5f ti ‘\ -& B 2 5 4 0 F --__ I I 4 6 --__ --t_ I 2 0 -.__ 12 18 24 (h) Fig. 3. Cardiac concentration of free doxorubicin (A) and nanoparticle-bound doxorubicin (0) after intravenous administration to mice (dose of doxorubicin administred: 7 mg/kg). TIME CONCLUSION Before starting clinical assays with anticancer drug-loaded nanoparticles it was imperative to develop a pharmaceutical formulation suitable for intravenous administration to humans under safe conditions, Such a formulation would ideally be presented in a solid form, for improved stability and ease of manipulation. In the present study we determined experimental conditions that are necessary to lyophilize nanoparticle formulations that can TABLE easily be resuspended in water and are able to adsorb doxorubicin and dactinomycin efficiently. When prepared under sterile conditions, this formulation meets the usual requirements needed for intravenous administration, like sterility and lack of bacterial endotoxins. Furthermore, it has been shown that such a formulation was very reproducible regarding size and drug adsorption rate, even after prepration at a semi-industrial level. Nanoparticles were developed mainly with the intention of using them as drug carriers. It is therefore important to take into account their molecular weight and their drugadsorption capacity. Indeed, substantial modification in the molecular weight could probably induce modifications in the tissue distribution and the elimination rate of the nanoparticles. The described gel permeation chromatographic method has proved efficient in the estimation of the molecular weight of the nanoparticles. Another important feature of this method is the possibility to detect even small quantities of unreacted monomer. Using an autofluorographic histological technique, we have previously observed a lack of doxorubicin-related fluorescence in the cardiac muscle of mice after administration of various doxorubicin-loaded nanoparticle formulations [7] . The data presented in this paper quantitatively confirm the observations and illustrate the possibility of avoiding a high cardiac level of doxorubicin by linkage 2 Particle size and level of drug binding before tin-loaded (DOX-NP) and dactinomycin-loaded and after storage at -30°C (6 months (DACT-NP) polyisobutylcyanoacrylate DOX-NP and 12 months) nanoparticles of doxorubi- DACT-NP Size (nm) 185 * 9 Before storage Drug binding @) level 91 f 5 211+ 10 95* 3 3 Size (nm) Drug binding (%) After storage (6 months) 180 * 8 89 + 5 205 + 10 95t After storage (12 months) 190 * 9 90 f 5 210 * 10 95 * 3 level 210 to nanoparticles. Such a modified drug distribution is therefore likely to reduce the cardiotoxicity of doxorubicin. Finally, the blood clearance of the drug has been observed to be reduced when bound to nanoparticles. This study, involved with the development of a nanoparticle formulation, was suggested by the lack of published data concerning drug-carrier formulations acceptable for human use. 6 7 8 ACKNOWLEDGEMENTS 9 The authors wish to thank Dr. Scouvart (Laboratoires Simon S.A.) and Dr. Tricot (Sopar S.A. Company) for helpful discussions and valuable technical assistance, The excellent technical aid of Mr. Bulckens was greatly appreciated. The authors are also thankful to Mr. Vandiest for the drawings and to Mrs. D’Heur for the secretarial work. 10 11 REFERENCES P. Couvreur, B. Kante, M. Roland, P. Guiot, P. Baudhuin and P. Speiser, Polycyanoacrylate nanocapsules as potential lysosomotropic carriers: Preparation, morphological and sorptive properties, J. Pharm. Pharmacol., 31 (19’79) 331-332. P. Couvreur, M. Roland and P. Speiser, Biodegradable submicroscopic particles containing a biologically active substance and compositions containing them, U.S. Patent 4,329,332, May 11, 1982. F. Leonard, R. Kulkarni, G. Brandes, J. Nelson and J. Cameron, Synthesis and degradation of poly(alkyl-~-cyanoa~rylates), J. Appl. Polym. Sci., 10 (1966) 259-272. W.R. Vezin and A.T. Florence, In vitro heterogenous degradation of poly(n-alkylcyanoacrylates), J Biomed. Mater. Res., 14 (1980) 93106. F. Leonard, R. Kulkarni, J. Nelson and G. Bran- 12 13 14 15 des, Tissue-adhesives and hemostasis-inducing compounds: The alkylcyanoacrylates, J. Biomed. Mater. Res., l(lQ67) 3-Q. B. Kante, P. Couvreur, V. Lenaerts, P. Guiot, M. Roland, P. Baudhuin and P. Speiser, Tissue distribution of [‘Hlactinomycin D adsorbed on polybutylcyanoacrylate nanoparticles, Int. J. Pharm., 7 (1980) 45-53. P. Couweur, B. Kante, L. Grislain, M, Roland and P. Speiser, Toxicity of poIyalkylcyanoa~rylate nanoparticles. II. Doxorubicin loaded nanoparticles, J. Pharm. Sci., 71 (1982) 790-792. P. Couvreur, Mise au point d’un nouveau vecteur de medicament, These d’Agregation de l’Enseignement Superieur, Universite Catholique de Louvain, 1983. F. Brasseur, P. Couvreur, B. Kante, L. DeckersPassau, M. Roland, C. Deckers and P. Speiser, Actinomycin D adsorbed on polymethylcyanoacrylate nanoparticles: Increased efficiency against an experimental tumor, Eur. J. Cancer, 16 (1980) 1441-1445. B. Kante, P. Couvreur, G. Dubois-Krack, C. De Meester, P. Guiot, M. Roland, M. Mercier and P. Speiser, Toxicity of polyalkylcyanoacrylate nanoparticles. I. Free nanoparticles, J. Pharm. Sic, 71(1982) 786-790. F. Brasseur, A. Biernacki, V. Lenaerts, L. Galanti, P. Couvreur, C. Deckers and M. Roland, Etude de la toxic&e des nanoparticules de polycyanoacrylate d’alkyle, in: Proceedings of the 3rd International Conference on Pharmaceutical Technology, APGI, Paris, II, 1983, pp. 194202. L. Vansnick, P. and M. Roland, Couvreur, Molecular D. Christiaens-Leyh weights of free and drug-loaded nanoparticles, Pharm. Res., 1 (1985) 36- 41. P. Couvreur, B. Kante, M. Roland and P. Speiser, Adsorption of antineoplastic drugs to polyalkylcyanoacrylate nanoparticles and their release in calf serum, J. Pharm. Sci., 68 (1979) 1521.-1524. R. Baurain, D. Deprez-De Campeneere and A. Trouet, Rapid determination of doxorubicin and its fluorescent metabolites by high pressure liquid chromatography, Anal. Biochem., 94 (1979) 112-116. G. Mancini, A.O. Carbonara and J.F. Heremans, Immunochemical quantitation of antigens by single radial immunodiffusion., Immunochemistry, 2 (1965) 235-254.