Langmuir 1996, 12, 2153-2161
2153
Polylactide-Poly(ethylene glycol) Copolymers as Drug Delivery Systems. 1. Characterization of Water Dispersible Micelle-Forming Systems S. A. Hagan, A. G. A. Coombes, M. C. Garnett, S. E. Dunn, M. C. Davies, L. Illum, and S. S. Davis* The Department of Pharmaceutical Sciences, The University of Nottingham, Nottingham NG7 2RD, U.K.
S. E. Harding Physical Biochemistry Laboratory, The Department of Applied Biochemistry and Food Science, The University of Nottingham, Sutton Bonnington LE12 5RD, U.K.
S. Purkiss and P. R. Gellert Zeneca Pharmaceuticals, Alderley Park, Macclesfield, U.K. Received August 1, 1995. In Final Form: January 31, 1996X Copolymers of polylactide and poly(ethylene glycol) (PLA-PEG), which self-disperse in water to form spherical nonionic micelles, have been investigated as a novel biodegradable drug delivery system. These copolymers are defined by the molecular weight ratios of their polylactide to poly(ethylene glycol) components (1.5:2 PLA-PEG and 2:5 PLA-PEG) and gave two peaks when purified by gel permeation chromatography (GPC). The first peak consisted of spherical micelles with a diameter of 15.6 nm for 1.5:2 PLA-PEG, and 18.9 nm for 2:5 PLA-PEG micelles after analysis by dynamic light scattering (DLS) and by transmission electron microscopy (TEM). The second peak was a PLA-depleted species resulting from the synthesis and did not form micelles. Testosterone and sudan black B (SBB), which have different hydrophobicities, were used as “model drugs” to evaluate the drug loading ability of the micelles. Ultracentrifugation sedimentation velocity studies confirmed that solubilization of the model drugs had occurred by micellar incorporation. Higher drug loading was obtained for the 1.5:2 PLA-PEG micelles (63.9% (w/w) of SBB, 0.74% (w/w) of testosterone) than for the 2:5 PLA-PEG micelles (59.0% (w/w) of SBB, 0.34% (w/w) of testosterone). The amount of testosterone solubilized was therefore significantly lower than SBB for both copolymers. Stability testing in the presence of salt suggested that the micelles had sterically stabilized surfaces. In vivo studies in the rat, using a radioactive marker, showed that PLA-PEG micelles demonstrated extended circulation times in the blood during the period of study (3 h). The 1.5:2 PLA-PEG showed increased blood levels and lower uptake of the micelles by the liver compared to the 2:5 PLA-PEG micelles. This is thought to be due to differences in the packing density of the copolymer molecules on the micelle surface.
Introduction Colloidal carriers display good potential as drug delivery systems due to the ease of both preparation and incorporation of drug molecules as well as a potential for high drug loading and possibilities for sustained systemic release. However, the effectiveness with which colloidal carriers are captured by the mononuclear phagocytic system (MPS) presents a major obstacle to the use of such vehicles for site-specific drug delivery. Extensive investigations have shown that this barrier can be overcome and particles can be directed away from the liver to other sites if the carrier surface is modified by hydrophilic poly(oxyethylene) chains.1,2 There is a significant interest in micelles or “selfassembling, supramolecular complexes” as microcontainers for drug targeting. For nonionic micelles produced from poloxamers, which are based on blocks of hydrophilic poly(oxyethylene) (PEO) and hydrophobic poly(oxypropylene) (PPO), molecules of the drug can be solubilized in the inner hydrophobic PPO core, with the PEO blocks forming the outer hydrophilic shell. It has been reported, for example, that the neuroleptic action of haloperidol, * To whom correspondence should be addressed. X Abstract published in Advance ACS Abstracts, April 1, 1996. (1) Illum, L.; Davis, S. S. Life Sci. 1987, 40, 1553-1560. (2) Gref, R.; Minamitake, Y.; Peracchia, M. T.; Trubetskoy, V.; Torchilin, V.; Langer, R. Science 1994, 263, 1600-1603.
S0743-7463(95)00649-4 CCC: $12.00
injected into mice in highly concentrated aqueous micellar solutions of PEO-PPO-PEO block copolymer surfactant (Pluronic P-85), was increased relative to aqueous haloperidol solutions.3 Targeting of so-called “microcontainers” to specific cells has been attempted by Kabanov et al.4 who conjugated the poloxamer molecules with antibodies against a target-specific antigen or with protein ligands selectively interacting with target cell receptors. These same authors also reported the ability of a low molecular weight compound (ATP), solubilized in poloxamer micelles, to penetrate an intact cell in vitro. Micellar complexes are, however, in dynamic exchange with free copolymer molecules in solution, continuously breaking and re-forming.5 This property will be of particular consequence in vivo under dilution and will have an important influence on the drug carrying capacity of amphiphilic polymers which show this micelle-type association behavior. (3) Kabanov, A. V.; Chekhonin, V. P.; Alakhov, V. Y.; Batrakova, E. V.; Lebedev, A. S.; Melik-Nubarov, N. S.; Arzakov, S. A.; Levashov, A. V.; Morozov, G. V.; Severin, E. S.; Kabanov, V. A. FEBS Lett. 1989, 258, 343-345. (4) Kabanov, A. V.; Batrakova, E. V.; Melik-Nubarov, N. S.; Fedoseev, N. A.; Dorodnich, T. Y.; Alakhov, V. Y.; Chekhonin, V. P.; Nazarova, I. R.; Kabanov, V. A. J. Controlled Release 1992, 22, 141-158. (5) Hall, D. G.; Pethica, B. A. In Nonionic Surfactants; Schick, M. J., Ed.; Marcel Dekker Inc.: New York, 1967; pp 516-557.
© 1996 American Chemical Society
2154 Langmuir, Vol. 12, No. 9, 1996
Figure 1. Structure of PLA-PEG copolymer.
Biodegradable block copolymers produced from poly(lactic acid) (PLA) and poly(ethylene glycol) (PEG) exhibit good potential for formulating drug delivery systems. PLA-PEG copolymers with various PLA:PEG ratios, molecular weights, structure, and solubility can be synthesized and have attracted interest for a variety of biomedical applications. Zhu et al.6 prepared “supermicrocapsules” of 40-100 nm diameter for drug delivery, by phase separation of poly(ethylene oxide)-polylactide star-type copolymers. PLA-PEG block copolymers containing relatively low weight percentages of PEG (5%) were synthesized by Deng et al.7 for preparation of drug delivery systems. Variation of the amount of PEG in the copolymer enabled the biodegradation rate of the copolymer to be controlled. Cohn and Younnes8 synthesized a family of poly(lactic acid)/poly(ethylene oxide) block copolymers in an attempt to provide biodegradable elastomers for cardiovascular implants. These copolymers possessed a novel range of properties since the lactic acid sequences form degradable “hard” blocks of the copolymer, while flexibility and elastic recovery properties were associated with the PEO “soft” segments. Higher molecular weight copolymers have been synthesized by Gref et al. for drug targeting applications, using poly(lactideco-glycolic acid) (PLGA), attached to PEG.2 Amphipathic, linear AB block copolymers composed of hydrophobic polylactide (PLA) blocks and hydrophilic poly(ethylene glycol) (PEG) blocks, with PEG content in excess of 50%, are the subject of the investigations in this paper.9 The triblock copolymer Poloxamer 407 has been included in some of these studies for comparison with PLA-PEG dispersions. The PLA-PEG copolymers discussed here are directly dispersible in aqueous media and self-associate to form complexes less than 50 nm in diameter and therefore offer advantages over water-insoluble copolymers in the formulation of delivery systems for pharmaceuticals. Control over the chain length, ratio of the hydrophilic and hydrophobic components, and the copolymer structure (branched or comb-type) is possible and offers the potential for modifying the size and surface characteristics of the association complexes formed in water. The complexes can be considered to be of a micellar type and in this form may be utilized directly for drug delivery. The nature and stability of the association complexes formed by PLA-PEG copolymers in aqueous dispersions are of fundamental importance as regards their potential drug carrying capacity. Materials Two linear (AB) block copolymers derived from poly(DL-lactic acid) and methoxy poly(ethylene glycol) (PEG 2000 M and PEG 5000 M) were supplied by Zeneca Pharmaceuticals (Alderley Park, Macclesfield, U.K.). These polymers were produced by ring opening polymerization of lactide using a stannous octoate catalyst. The copolymers have been designated 1.5:2 PLA-PEG and 2:5 PLA-PEG, respectively, which signifies the ratio of the molecular weights of PLA block to the PEG block in the copolymer. The structure of the PLA-PEG copolymer is described in Figure 1. (6) Zhu, K. J.; Bihai, Song, Shilin, Yang. J. Polym. Sci., Part A, Polym. Chem. 1989, 27, 2151-2159. (7) Deng, X. L.; Xiong, C. D.; Cheng, L. M.; Xu, R. P. J. Polym. Sci., Part C, Polym. Lett. 1990, 28, 411-416. (8) Cohn, D.; Younes, H. Biomaterials 1988, 10, 466-474. (9) Churchill, J. R.; Hutchinson, F. G. European Patent Application, No. 85304489.9, 1986.
Hagan et al. Poloxamer 407, a PEO-PPO-PEO triblock copolymer (Mw ∼ 12 600, as estimated by SEC, and 70% PEO content), was used as a reference amphipathic surfactant. This copolymer was obtained from the BASF Corp., Parsippany, NJ. PEG 8000 was also used as a reference and was obtained from Sigma Chemical Company, St. Louis, MO. Sudan black B and testosterone, used for the solubilization studies, were also both obtained from Sigma. The partition coefficient values in the form of log P values were calculated using the CLogP program, version 3.54.
Methods Preparation of Aqueous Dispersions. Aqueous dispersions of the PLA-PEG copolymer were prepared by dissolving the copolymer in double distilled water to give final concentrations in the range 0.1-5.4% (w/v). The copolymer/water system was generally retained at room temperature for several hours with occasional shaking, until clear dispersions were obtained. PLA-PEG dispersions were also produced after fractionation by gel permeation chromatography (GPC) (Sepharose CL-4B, 2.6 × 30 cm). After separation of a 5% (w/v) dispersion by GPC (10 mL), samples were freezedried and the copolymers redispersed as above. Gel Permeation Chromatography (GPC). Molecular weights of the copolymers and their polydispersity were determined by size exclusion chromatography (SEC) on three PL Gel 10E3 Å columns (30 cm × 7.5 mm), using dimethylformamide as mobile phase at a flow rate of 1 mL/min, with the columns thermostated at 70 °C. The peaks were detected by refractive index, with the signal captured by HP Chem Station software and analyzed using Polymer Laboratories GPC/SEC software. Data for aqueous gel permeation chromatography (GPC) were derived from preparative runs on a column of sepharose CL-4B (2.6 × 30 cm). Samples (10 mL) were loaded at concentrations ranging from 1 to 5% (w/v) and eluted at 80 mL/h with water. Peaks were detected using a refractive index detector (Gilson model 131). The column was calibrated using lactic acid, poly(ethylene glycol) standards, and PLGA (polylactide-co-glycolic acid) particles. Transmission Electron Microscopy (TEM) Studies. Aqueous dispersions of the fractionated and unfractionated 1.5:2 PLA-PEG and 2:5 PLA-PEG copolymers and Poloxamer 407 were examined at a concentration of 1% (w/v), in the air-dried condition using transmission electron microscopy (TEM). Specimens were prepared by dropping the dispersion onto carbon-coated EM grids. The grid was held horizontally for 20 s to allow the molecular aggregates to settle and then at 45° to allow excess fluid to drain. The grid was returned to the horizontal position and a drop of phosphotungstic acid (adjusted to a pH of 4 using potassium hydroxide) was added to give a negative stain. The grid was then left to stand for 20 s before removing excess stain as above. Specimens were air-dried before examination using a JEOL 1200 EX12 transmission electron microscope. Estimation of Effective Hydrodynamic Diameter and Polydispersity of PLA-PEG Dispersions by Dynamic Light Scattering. Dispersions of the fractionated 1.5:2 PLA-PEG and 2:5 PLA-PEG copolymers of concentration 1% (w/v) were produced as described above. A comparison was also made with 1% (w/v) aqueous dispersions prepared from freeze-dried unfractionated material and with 1% (w/v) Poloxamer 407. Samples were analyzed by dynamic light scattering, using a Malvern System 4700 dynamic light scattering photometer, with a Siemens helium/neon laser light source operating at a wavelength of 632.18 nm at 40 mW, with an assumed refractive index ratio of 1.60 and viscosity of 0.89. The sample cell was cleaned before each measuring run by flushing for at least 5 min with double distilled water,
PLA-PEG as a Drug Delivery System
using a filling apparatus similar to that described by Sanders and Cannell,10 and dried with ultrafiltered air. Traces of distilled water were removed from the cell by vacuum prior to injection of samples. Polymer dispersions prepared as outlined above were injected into the sample cell from a 1-mL disposable syringe through a 0.2 µm Millipore filter (type HA) and hypodermic needle. Ten measuring runs were performed to provide data for the effective hydrodynamic radius and polydispersity of each sample. The sample time for PLA-PEG dispersions was 5 µs and the experimental duration 90 s. All measurements were performed at 25 °C at a measurement angle of 90°. The Stability of PLA-PEG Aqueous Dispersions to Added Salt. The stability of dispersions to electrolyte addition was measured by determining the temperature of turbidity increase (the cloud point temperature (Tc)), for 1% (w/v) aqueous solutions of poly(ethylene glycol) molecular weight 8000 (PEG 8000), Poloxamer 407, 1.5:2 PLA-PEG copolymer and 2:5 PLA-PEG copolymer, as a function of molar concentration of added Na2SO4, using a turbidimeter. Salt concentration was varied between 0 and 0.5 M. In addition, the unfractionated PLA-PEG was compared with the fractionated copolymer. Also, for Poloxamer 407, and 1.5:2 and 2:5 PLA-PEG (both fractionated and unfractionated), a solid phase separation temperature (Ts) was measured, which was taken as the point of decrease in turbidity after a plateau region was observed. The test method was similar to that described by Cornet and Ballegooijen for measurement of θ temperatures of polymer solutions,11 in that the sample was illuminated using a light source and changes of turbidity with temperature were recorded by a change in light transmission detected by a photocell. The glass cuvette containing the solution was inserted into a thermostatically controlled water jacket. The sample temperature was recorded using a temperature probe inserted into the sample cell and changes in turbidity with increased temperature (10 °C/min) were monitored up to an upper limit of 80 °C (the limit of the apparatus). Photocell and temperature probe outputs were fed to a computer for data storage, handling, and display. Determination of Ethylene Glycol/Lactic Acid Ratio by NMR. The PLA-PEG copolymers were examined by NMR in CDCl3 solution, using a Bruker AC250 250 MHz Fourier transform NMR spectrometer. The areas at the resonance peaks were integrated and the integrals from the multiplets at 1.55 ppm δ shift (PLA methyl group) and 3.65 ppm δ shift (PEG methylene groups) were converted to a weight ratio assuming the MeOPEGs have weight average molecular weights of 2000 and 5000. Determination of the Critical Micelle Concentration (cmc) of PLA-PEG Micelles. Surface tension measurements (γ) of dilute aqueous solutions of fractionated PLA-PEG copolymer were measured using the Wilhelmy plate method by a dynamic contact angle (DCA) analyzer (DCA-322, Cahn Instruments, CA). All glassware was cleaned using chromic acid and a glass cover slip employed as the glass plate. The plate was suspended from a microbalance (calibrated with standard weights) and the surface tension of each copolymer solution measured at 20 °C. The required concentrations were obtained by diluting a stock solution of copolymer with water and each concentration was performed in triplicate, with distilled water as a control. Measurement of the (10) Sanders, A. H.; Connell, D. S. In Light Scattering in Liquids and Macromolecular Solutions; Degiorgio, V., Corti, M., Giglio, M., Eds.; Plenum Press: New York, 1980; pp 173-182. (11) Cornet, C. F.; van Ballegooijen, H. Polymer 1966, 7, 293-301.
Langmuir, Vol. 12, No. 9, 1996 2155
receding force was taken after 300 s, and the surface tension measurements were plotted against the logarithm of copolymer concentration. Sedimentation Velocity. A Beckman Optima XL-A analytical ultracentrifuge equipped with scanning absorption optics and a monochromator was used in sedimentation velocity experiments, in order to assay the homogeneity and determine the sedimentation coefficients of the samples.12 Solutions were prepared by fractionated material from the first elution peak (250 µL, 0.2% (w/v)) and scanned against distilled water (250 µL) loaded into a 10 mm path length cell. Sedimentation runs were then performed at 40 000 rpm and 20 °C, with UV scanning every 10 min at 231 nm for 1.5:2 PLA-PEG and 225 nm for 2:5 PLAPEG. Absorption profiles were subsequently obtained and overlaid to determine an apparent sedimentation coefficient, s20.12 The same experiment was also performed with testosterone and sudan black B (SBB) incorporated into 0.2% (w/v) solutions of material from the first GPC peak. PLAPEG/testosterone-loaded dispersions were prepared by adding a solution of testosterone in acetone (1 mL, 0.5 mg/mL) dropwise to 5 mL of a 0.2% (w/v) aqueous solution of fractionated PLA-PEG, followed by stirring overnight, at room temperature, to remove the solvent by evaporation. Sudan black B (10 mg) was also incorporated by addition to 5 mL of a 0.2% (w/v) aqueous solution of fractionated PLA-PEG and sonicating for 10 min at room temperature. The resulting suspensions were then filtered through a 0.2 µm filter to remove excess testosterone/sudan black B. The runs were performed at 20 °C, with the PLAPEG/testosterone system run at 40 000 rpm and the PLAPEG/SBB system at 30 000 rpm. Both these dispersions were scanned at two wavelengths, the first to detect the micelles plus model drug, and the second, where the micelles are transparent, to evaluate whether sedimentation boundaries were still produced. For the testosterone systems, 1.5:2 PLA-PEG was scanned at 232 and 258 nm, and 2:5 PLA-PEG was scanned at 226 and 258 nm. For the SBB system, 1.5:2 PLA-PEG was scanned at 231 and 510 nm and 2:5 PLA-PEG at 225 and 510 nm. Sedimentation coefficients (s20) were then determined for all the samples.12 Sedimentation Equilibrium. A Beckman Optima XLA analytical ultracentrifuge was used to determine the weight average molecular weight of solutions of the fractionated 1.5:2 and 2:5 PLA-PEG copolymers and Poloxamer 407, respectively, as described by Morgan et al.,13 under micelle forming (aqueous) conditions. These data were subsequently used to provide an estimate of the association number of copolymer molecules per micelle. The ultracentrifuge employed Rayleigh interference optics, a 5 mW He-Ne laser light source (632.18 nm), and an RTIC temperature control system. A low speed was used (5000 rpm) to avoid meniscus depletion conditions. The meniscus concentration therefore remained measurable and was obtained by mathematical manipulation of the fringe data.14 The solution was loaded into a 30 mm path length double sector cell (200 µL, 0.2% (w/v)) and the temperature was maintained at 20 °C. Thermodynamic nonideality was assumed to be negligible at the copolymer concentration used. Whole cell apparent weight average molecular weights (Mw,app) were obtained from the limiting value of the M* function at the cell base.15 The association number of (12) Harding, S. E. Methods Mol. Biol. 1984, 22, 61-73. (13) Morgan, P. J.; Harding, S. E.; Petrak, K. Macromolecules 1990, 23, 4461-4464. (14) Creeth, J. M. and Harding, S. E. J. Biochem. Biophys. Methods 1982, 7, 25-34.
2156 Langmuir, Vol. 12, No. 9, 1996
copolymer molecules in a micelle was then calculated by dividing by the copolymer molecular weight. Micellar Incorporation of Model Drugs, Extinction Coefficient, and Solubility. Solutions of testosterone in acetone (1 mg/mL) and sudan black B (SBB) (10 and 20 mg/mL) were added to fractionated dispersions of the 1.5:2 and 2:5 PLA-PEG micelles (0.25% (w/v)). The copolymer/drug dispersions were then centrifuged using a Centaur 2 (MSE, Sussex, U.K.) centrifuge (840g, 10 min), followed by a Micro Centaur (MSE, Sussex, U.K.) centrifuge (11600g, 5 min) and the precipitate of drug, which had not been incorporated into the micelles, dissolved in ethanol (SBB) or distilled water (testosterone). The absorbance of each solution was measured by UV spectroscopy at 257 nm to determine testosterone concentrations and 587 nm to determine SBB concentrations. The amount of drug incorporated was then calculated by subtraction. Aqueous solubility was determined similarly by adding a solution of drug in acetone to water (1 mL) and then centrifuging off the excess. The absorbance of the supernatant was then measured, and the samples were used as a control for the drug loading studies. Extinction coefficients for testosterone and SBB were calculated from the absorbance of aqueous dilutions of ethanolic stock drug solutions. Biodistribution Studies. Biodistribution studies were performed in rats, with the fractionated 1.5:2 and 2:5 PLA-PEG copolymers injected at concentrations of 0.2 and 5.4% (w/v). For the lower PLA-PEG copolymer concentration, an aqueous dispersion (1% (w/v)) was prepared by addition of PLA-PEG copolymer (1.5:2 and 2:5) (100 mg) to water (10 mL) containing a radiolabeled marker, hydrophobic indium-111-oxine complex (Amersham, 0.2 mL, 2.5 MBq) and shaken until the polymer dissolved. For the high PLA-PEG concentrations, indium-111 oxine solution (0.005-0.021 mL) was added to 1% (w/v) dispersions of fractionated PLA-PEG and left for 30 min to equilibrate. For both concentrations, free indium was separated from incorporated indium using a sepharose CL-4B gel permeation column (10 cm) equilibrated with distilled water. At the lower concentration, the PLA-PEG dispersions were collected and diluted to the required concentration. At the higher concentration, the eluent was monitored, and the most active fractions were pooled for the in vivo experiment. Radiolabeled dispersions were then mixed with concentrated fractionated PLA-PEG polymers to give a dispersion of 5.4% (w/ v). Radiolabeled PLA-PEG dispersions (0.2% (w/v), 0.1 mL; 5.4% (w/v), 0.5 mL) were injected into groups of three female Wistar rats (150 ( 10 g) via the tail vein. Blood samples (10 µL) were taken from the tail at 0.25, 0.5, 1, 2, and 3 h postinjection. At 3 h postinjection the rats were sacrificed and the liver, spleen, lungs, kidneys, and one femur were removed. The remaining activity in the organs and blood was counted using a γ spectrometer (Compugamma, LKB, Wallac, Finland), and the carcass activity was determined, for the higher concentration study, using a well counter (scintillation detector model SD1, Oakfield Instruments Ltd., Eynsham, U.K.). Free indium-111 oxine was similarly studied as a control.
Hagan et al. Table 1. Characterization of PLA-PEG Copolymers by Size Exclusion Chromatography (SEC) sample 1.5:2 PLA-PEG 1.5:2 PLA-PEG 1.5:2 PLA-PEG 2:5 PLA-PEG 2:5 PLA-PEG 2:5 PLA-PEG
elution peak
PEG chain (Mw)
Mw
polydispersity
2000 2000 2000 5000 5000 5000
3518 3947 2519 6307 6987 5547
1.09 1.07 1.03 1.09 1.10 1.09
1 2 1 2
Table 2. Comparison of NMR Peak Integration for Fractionated PLA-PEG Copolymers copolymer
elution peak
wt % of polymer
ethylene glycol/ lactic acid ratio
1.5:2 PLA-PEG 1.5:2 PLA-PEG 2:5 PLA-PEG 2:5 PLA-PEG
1 2 1 2
54.1 41.8 53.8 41.1
2.2 8.3 4.4 16.0
Purification and Analysis of PLA-PEG Copolymers. The PLA-PEG copolymers produced for this work
consisted of a PEG block of defined molecular weight, linked to a PLA block, the size of which is determined by the input of PLA to the reaction mixture. Characterization of these copolymers by SEC in dimethylformamide gave polydispersity indices (Table 1) that suggest a narrow distribution of PLA block sizes. However, initial work with these block copolymers indicated that they were not homogeneous, and gel permeation chromatography of the 1.5:2 and 2:5 PLA-PEG copolymer dispersions on a sepharose CL-4B matrix, in aqueous media, resulted in a clear separation into two peaks. A similar case has been reported for the Poloxamer 407, which also gave a bimodal distribution on GPC separation.16,17 Table 1 shows the weight average molecular weight of the block copolymer and the PEG component, relative to polystyrene standards, estimated by SEC, before and after purification by aqueous GPC. The material in each peak was quantitated by pooling the peak fractions and freezedrying to determine the percentage of each peak by weight and the total recovery (Table 2). NMR analysis of the freeze-dried material obtained from the two peaks produced by GPC (Table 2) showed that the first peak had a similar ratio of lactic acid groups to ethylene oxide groups to that predicted from the relative proportion of starting materials for the synthesis, but the second peak appeared to be low in PLA content (“PLAdepleted”) (the PEG chain is completely defined by the starting materials). These two polymer fractions therefore differ in both chemical composition and molecular weight. Similarly, SEC data showed that the molecular weight of material from the first GPC peak was close to that expected from the synthesis, whereas the molecular weight of material from the second GPC peak was significantly lower. This again indicates that the second peak is low in PLA. It was considered possible that the second peak was a decomposition product of the copolymer. However, when material isolated from the first peak was incubated in water at 37 °C for seven days and refractionated, only a single peak resulted (data not shown) suggesting that the second peak shown in the initial GPC analysis was a byproduct of the synthesis. The NMR data on the second peak indicated that the PLA chain length was about 5 PLA residues. As a result of this finding, the remaining characterization was carried out on the first micelle-forming fraction of material purified by GPC, except where the role of the second fraction in micelle formation was being addressed.
(15) Harding, S. E.; Horton, J. C.; Morgan, P. J. In Analytical ultracentrifugation in Biochemistry and Polymer Science; Harding, S. E., Rowe, A. J., Horton, J. C., Eds.; Royal Society of Chemistry: Cambridge, U.K., 1992; pp 276-279.
(16) Yu, Ga-Er; Deng, Y.; Dalton, S.; Wang, Q. G.; Attwood, D.; Price, C.; Booth, C. J. Chem. Soc., Faraday Trans. 1992, 88, 2537-2544. (17) Nicholas, C. V.; Luo, Y.; Deng, N.; Attwood, D.; Collett, J. H.; Price, C.; Booth, C. Polymer 1993, 34, 138-144.
Results
PLA-PEG as a Drug Delivery System
Langmuir, Vol. 12, No. 9, 1996 2157
Figure 2. (a, left) Transmission electron micrograph of a 1% (w/v) 1.5:2 PLA-PEG dispersion of material from the first GPC peak (magnification 100k) (scale bar ) 50 nm). (b, right) Transmission electron micrograph of a 1% (w/v) 2:5 PLA-PEG dispersion of material from the first GPC peak (magnification 100k) (scale bar ) 50 nm).
TEM Examination. Parts a and b of Figure 2 show transmission electron micrographs of fractionated 1.5:2 and 2:5 PLA-PEG copolymers, respectively, with the scale bar corresponding to 50 nm. These clearly indicate the presence of micelles in the material from the first peak for both copolymers, as shown by light spherical entities surrounded by the darker staining. With the material from the second peak, only the background of the coated grid is seen, with no micelles present for either copolymer (data not shown). The size of micelles in the material from the first GPC peak were 13.2 nm for 1.5:2 and 17.0 nm for 2:5 PLA-PEG systems and were smaller than micelles seen with the unfractionated copolymer (17.7 and 24.0 nm, respectively). The electron micrograph obtained for a 0.5% (w/v) solution of Poloxamer 407 exhibited similar features to those found in PLA-PEG specimens (data not shown) and showing micelles which were approximately 17.9 nm in diameter. Dynamic Light Scattering (DLS) Measurements. The fractionated 1.5:2 PLA-PEG and 2:5 PLA-PEG material from the first GPC peak resulted in micelles which were again smaller than those formed from freezedried unfractionated material (Table 3). This could be due to the PLA-depleted material from the second GPC peak associating with the palisade layer of the micelles, resulting in a larger size for the unfractionated micelles. Micelles were not detected by PCS measurement in material eluted in peak two, in agreement with the TEM results. The polydispersity is an indication of the nar-
Table 3. Dynamic Light Scattering and Sedimentation Equilibrium Results for Dispersions of 1.5:2 PLA-PEG, 2:5 PLA-PEG, and Poloxamer 407 Copolymersa
copolymer 1.5:2 PLA-PEG unfractionated 1.5:2 PLA-PEG peak one 2:5 PLA-PEG unfractionated 2:5 PLA-PEG peak one Poloxamer 407 a
micellar size (nm)
polydispersity
mol wt of micelle, Mw,app
association no. of micelle
19.8 ( 0.6 0.19 ( 0.02 15.6 ( 0.9 0.22 ( 0.07 520000 ( 20000
132 ( 5
25.3 ( 0.6 0.68 ( 0.01 18.9 ( 1.6 0.28 ( 0.01 330000 ( 30000
47 ( 4
29.1 ( 0.5 0.12 ( 0.06 200000 ( 10000
16 ( 1
No micelles observed for peak two material.
rowness of the particle size distribution, with increased values showing a greater spread of particle sizes, and was calculated by the Malvern 4700 software. This measures the variance of the distribution in decay times, and hence the particle size, and gives a log normal intensity distribution for the measured data from which a polydispersity value is calculated. Stability of PLA-PEG Dispersions to Added Salt. Poloxamer 407 was chosen as a comparable species to the “PLA-rich” fraction (peak one). PEG 8000 is comparable to the total PEG content of Poloxamer 407 and was chosen so that the contribution of PEG to the micelle stability could be determined. As the molecular weight of the PEG does not greatly influence the cloud point, this should
2158 Langmuir, Vol. 12, No. 9, 1996
Hagan et al.
Table 4. Cloud Point Temperatures (Tc) for 1% (w/v) PLA-PEG Copolymers in the Presence of Na2SO4 Na2SO4 concn (M) 0.1 0.2 0.3 0.4 0.5
1.5:2 PLA-PEG unfractionated Tc (°C) Ts (°C) 77.6 68.2 58.7 42.7 24.0
75.6 69.3 65.7 52.8
1.5:2 PLA-PEG fractionated Tc (°C) Ts (°C) 76.7 62.7 55.4 43.4 26.4
73.7 67.2 65.6 59.4
serve as a comparator to the PEG portion of the PLAPEG copolymers (Mw 2000 and 5000), and to the “PLAdepleted” fraction of the PLA-PEG copolymers. For PEG 8000 solutions, a sharp increase in turbidity was seen, as the temperature increased, followed by a plateau region, extending to the end point of the experiment (80 °C). This turbidity was reversible, clearing rapidly on cooling of the solution. Reversibility is generally noted for this type of incipient phase separation in polymer solutions. No further phase separation was observed for PEG 8000 solutions during testing to 80 °C. Poloxamer 407 showed a different pattern of behavior compared with PEG 8000. A sharp increase in absorbance at the cloud point (Tc) was followed by a plateau region, as with PEG 8000. However, a marked fall in absorbance was then observed corresponding to separation of a solid phase. Examination of samples at 80 °C, after testing, revealed a collection of solids at the surface of clear liquid, with reversibility exhibited at all stages of turbidity. The point of decrease in turbidity after the plateau region allowed the assignment of a solid phase separation temperature (Ts) for the surfactant. The unfractionated PLA-PEG materials showed similar behavior to Poloxamer 407, albeit with lower associated cloud point temperatures (Tc). Retained background turbidity gave rise to a second stable plateau region, which is considered to be due to the PLA-depleted species within the material and therefore appeared to behave similarly to the PEG 8000 homopolymer. The cloud point curves for the fractionated PLA-PEG micelle forming materials were again similar but showed a larger drop in turbidity to the second plateau region relative to the unfractionated PLA-PEG copolymers. Reversibility was displayed for both fractionated and unfractionated PLA-PEG copolymers. Table 4 shows that the cloud point temperatures (Tc) and the solid phase temperatures (Ts) for the fractionated PLA-PEG copolymers, the unfractionated 1.5:2 PLAPEG copolymer, and Poloxamer 407 decreased with increasing salt concentration in line with the expected behavior of PEG-containing copolymer solutions.18 Cloud point temperatures are lower than those measured for PEG 8000 by approximately 30 °C. It is interesting to note the very much reduced rate of decline of Ts compared with Tc with salt addition for the block copolymer. This could indicate that the thermodynamic factors which drive the separation of solid phase have a weaker temperature dependence than the “salting out” process of the hydrophilic PEG component which leads to the cloud point. The lower cloud point temperatures (Tc) seen with the 1.5:2 and 2:5 PLA-PEG micelles, compared with Poloxamer 407, indicate reduced solvency of the stabilizing PEG component associated with PLA-PEG relative to that in Poloxamer 407 (Table 4). This implies that the Poloxamer 407 micelles have greater “steric” stability than the 1.5:2 PLA-PEG micelles. This may be expected due to the higher proportion of the stabilizing PEG moiety in the Poloxamer 407, which has two PEG chains of Mw 4000, (18) Van den Boomgaard, Th. PhD Thesis, Agricultural University of Wageningen, The Netherlands, 1985.
2:5 PLA-PEG fractionated Tc (°C) Ts (°C) 81.8 69.4 56.6 39.8 27.3
80.5 74.6 70.3 65.0
poloxamer 407 Tc (°C) Ts (°C) 74.4 61.5 48.8 31.7
73.4 65.3 59.5
PEG 8000 Tc (°C)
73.1 58.2
Table 5. Sedimentation Velocity Analysis of Fractionated PLA-PEG Copolymers copolymer sample
sedimentation coefficient (S), s20
1.5:2 PLA-PEG 2:5 PLA-PEG 1.5:2 PLA-PEG with testosterone 2:5 PLA-PEG with testosterone 1.5:2 PLA-PEG with sudan black B 2:5 PLA-PEG with sudan black B
8.51 ( 0.14 7.99 ( 0.38 9.57 ( 0.11 7.36 ( 0.10 9.76 ( 0.20 10.88 ( 0.18; 6.55 ( 0.14
as opposed to the one PEG chain (Mw 2000 or 5000) on the PLA-PEG copolymers. There was little difference between the Tc’s of unfractionated and fractionated micelle forming species. Comparison of the 1.5:2 and 2:5 PLAPEG fractionated micelle forming material revealed that the difference between these copolymers was not significant. Determination of cmc of PLA-PEG Systems. The plot of surface tension against log concentration (% w/v) for aqueous solutions of fractionated 1.5:2 PLA-PEG copolymer showed a distinct change in slope at 0.0035% (w/v), which was taken to signify the formation of micelles (data not shown). Results for the 2:5 PLA-PEG copolymer were identical, also giving a cmc of 0.0035% (w/v). Sedimentation Velocity Experiments. The sedimentation coefficients (s20) (in svedbergs (S), where 1 S ) 1 × 10-13 s) calculated for the fractionated 1.5:2 and 2:5 PLA-PEG copolymers alone and in the presence of two model drugs, testosterone and SBB are listed in Table 5. For the 2:5 PLA-PEG material alone, a single sedimenting boundary was observed (Figure 3a) indicating homogeneity. The 1.5:2 PLA-PEG material showed similar results (data not shown). Both PLA-PEG dispersions with testosterone also showed single boundaries at both wavelengths scanned suggesting that the testosterone was incorporated into the micellar core and was sedimenting with the micelles. With these systems, a significant amount of material was found to be absorbing but not sedimenting. This is most likely due to free testosterone in solution or testosterone bound to free nonassociated copolymer, which had too low a molecular weight to sediment under the experimental conditions. Similar behavior was observed with PLAPEG dispersions incorporating SBB, except that there appeared to be a second sedimenting species with the 2:5 PLA-PEG system (Figure 3b). This behavior pattern may indicate polydispersity of the micelles. The sedimentation coefficients derived for both sedimenting species are shown in Table 5. Assuming the conformations of the two PLAPEG/SBB and the PLA-PEG systems are the same (spherical micelles), it can be speculated that the differences in sedimentation coefficient reflect differences in molecular weight, M (for spherical particles s20 ∼ M2/3). Sedimentation Equilibrium Analysis. Apparent whole-cell weight-average molecular weights, calculated for the fractionated 1.5:2 and 2:5 PLA-PEG copolymers, and also for Poloxamer 407, are presented in Table 3. The association numbers of the PLA-PEG micelles were calculated using the molecular weight values estimated from size exclusion chromatography analysis given in the
PLA-PEG as a Drug Delivery System
Figure 3. (a) Sedimentation velocity profiles for fractionated 0.2% (w/v) 2:5 PLA-PEG micelles at an absorbance of 225 nm. The direction of sedimentation is from left to right. Rotor speed ) 40 000 rpm; scan interval ) 10 min. (b) Sedimentation velocity profiles for fractionated 0.2% (w/v) 2:5 PLA-PEG micelles with sudan black B incorporated at an absorbance of 510 nm. Rotor speed ) 30 000 scan interval ) 10 min. Two clear sedimenting boundaries are visible (fast moving “A” (≈10.88 S) and slower moving “B” (≈6.55 S).
materials section, namely, 3947 for 1.5:2 PLA-PEG and 6987 for 2:5 PLA-PEG. Micellar Solubilization of Model “Drug” Compounds. Two compounds differing in hydrophobicity, sudan black B and testosterone, have been studied as model drugs in PLA-PEG dispersions. Sudan black B has previously been used in ultracentrifugation studies on micellar systems due to its ease of incorporation and useful absorbance spectrum. Testosterone was chosen as a second drug to give a pharmaceutically relevant drug of different hydrophobicity, which has previously been incorporated into micelles and particles by other groups.19 The hydrophobicity of testosterone was assessed by calculation of its octanol/water partition coefficient (log P), determined using the CLogP computer program (version 3.54). The calculated value of 3.35 compared well with an experimentally determined value of 3.32.20 A log P value of 7.621 has been calculated for SBB and is similar to another calculated value of 7.2 found in the litera(19) Malcolmson, C.; Lawrence, M. J. J. Pharm. Pharmacol. 1993, 45, 141-143. (20) Craig, P. Drug Compendium. In Comprehensive Medicinal Chemistry: the Rational Design, Mechanistic Study and Therapeutic Application of Chemical Compounds; Hansch, C., Sammes, P. G., Taylor, J. B., Eds.; Pergamon Press: Oxford, 1990; Vol. 6, p 890. (21) Saunders, Martin, SmithKline Beecham, Herts, personal communication.
Langmuir, Vol. 12, No. 9, 1996 2159
ture.22 It is not feasible to determine values of log P above 7.0 experimentally.20 Aqueous solubilities for testosterone and sudan black B were found to be 23 and 1.2 mg/L, respectively. The results in Table 6 show that higher drug loading was observed for the 1.5:2 PLA-PEG compared with the 2:5 PLA-PEG and that the sudan black B (SBB) incorporated by the micelles was significantly higher than that of testosterone. Biodistribution and Interaction with the Mononuclear Phagocytic System. Biodistribution and blood elimination studies were carried out in rats. Micelles were labeled with radioactive indium-111 oxine, which acts as a hydrophobic drug model and can thus be readily incorporated into micellar forms of the copolymer dispersions. Two concentrations were studied. On dilution in the blood, the lower concentration system would be expected to revert to a nonassociated copolymer, while the higher concentration system would be expected to remain above the cmc for a significant period in vivo. (Assuming a blood volume of 7.5% of body weight and a 150 g rat, the lower concentration (0.2% (w/v), 0.1 mL) in blood would be at 0.0008% (w/v) and the higher concentration (5.4% (w/v), 0.5 mL) would be at 0.24% (w/v) in the blood initially (cmc ) 0.0035% (w/v)); Vd is not considered in this case since it is not known.) The biodistribution results are presented in Figure 4 and the blood clearance is given in Table 7. Control values for free indium oxine are also included. The free indium oxine was partly taken up by the liver, but also circulated in the blood at a significant level for the 3 h period of the study. At low concentrations, both of the PLA-PEG copolymers appeared to be directed specifically to the liver, while there is no increase in spleen uptake levels. Also, blood levels dropped quite rapidly. In contrast, at the higher concentrations of copolymer in the blood, there were marked differences between the two copolymers. The 2:5 PLA-PEG dispersion showed a similar biodistribution profile to that seen with the lower concentrations, but the 1.5:2 PLA-PEG exhibited an improved biodistribution, namely, a greatly reduced liver uptake in comparison to the lower concentration and a corresponding increase in the circulating blood levels. The 2:5 PLA-PEG copolymer at the higher concentration differed from the lower concentration only by a slightly higher kidney uptake and attainment of a stable blood activity at around 10% of injected dose. Discussion Novel water-dispersible PLA-PEG copolymers have been synthesized as potential self-associating, biodegradable micellar drug delivery systems. The copolymer preparations were readily separated into two fractions by aqueous gel permeation chromatography, despite their characterization as a single peak by gel permeation chromatography in organic media. This behavior may be explained by the composition of the PLA-PEG copolymers. Synthesis was achieved by polymerization of lactide onto poly(ethylene glycol) polymers of defined chain length, and the ratio of PLA to PEG in the copolymer was determined by the initial amount of lactide. Little unpolymerized lactide was present at the termination of the reaction, and characterization of the resulting polymer suggested satisfactory polydispersities of 1.09 for both PLA-PEG copolymers. It is considered that the technique of aqueous GPC results in the separation of a fraction of the PLA-PEG which has a PLA chain length too short to allow a stable incorporation into multimolecular micellar (22) Juarranz, A.; Horobin, R. W.; Proctor, G. B. Histochemistry 1986, 84, 426-431.
2160 Langmuir, Vol. 12, No. 9, 1996
Hagan et al.
Table 6. Drug Incorporation Studies of Testosterone and Sudan Black B into Fractionated 1.5:2 and 2:5 PLA-PEG Micelles testosterone
sudan black B
copolymer
% w/w drug in micelle
% incorporation efficiency
% w/w drug in micelle
% incorporation efficiency
1.5:2 PLA-PEG 2:5 PLA-PEG
0.74 ( 0.01 0.34 ( 0.03
73.9 ( 1.1 34.2 ( 3.3
63.9 ( 0.1 59.0 ( 1.2
79.8 ( 0.0 73.7 ( 1.5
Figure 4. Data from biodistribution studies of purified 1.5:2 and 2:5 PLA-PEG micelles.
complexes (a “PLA-depleted” fraction) and that in the micelle forming fraction, a distribution of copolymers with longer PLA chain lengths is present. From NMR and SEC data on fractions from the second peak, the critical PLA chain length for aggregation into a micelle appears to be between five and seven residues. The cmc of the PLA-PEG micelles was found to be 0.0035% (w/v). With Poloxamer 407, below a concentration of 0.1% (w/v), multimolecular aggregates are not believed to be formed, but instead a smaller monomolecular unit (unimer) exists where the molecule has undergone a conformational change, such that the hydrophobic chain segment is coiled in its interior, shielded by the PEO units.23-25 Below a concentration of 0.004% (w/v), these monomolecular units cease to be formed and Poloxamer 407 exists as extended copolymer molecules. This latter concentration is very similar to the cmc of the PLA-PEG micelles, but in the case of the PLA-PEG copolymers, multimolecular aggregates are believed to be formed down to a concentration of 0.0035% (w/v). PLA-PEG micelles have been detected by DLS at 0.1% (w/v), where multimolecular aggregates are thought to become unstable with Poloxamer 407. The PLA-PEG micelles therefore show enhanced stability on dilution, relative to Poloxamer 407 micelles. The size, in terms of equivalent hydrodynamic radius of the hydrated association complexes, measured by DLS in 1% (w/v) aqueous dispersions of Poloxamer 407 and 2:5 PLA-PEG, gave diameters of approximately 29 and 19 nm, respectively, compared with 16 nm for the more hydrophobic 1.5:2 PLA-PEG copolymer. TEM micrographs revealed similar size and shape features and analysis of the sedimentation equilibrium data in Table 3 supports these results. These data indicate that the copolymer molecules of Poloxamer 407 and 2:5 PLA-PEG formed less densely packed micelles relative to the smaller micelles observed with 1.5:2 PLA-PEG copolymer, as the association numbers of the micelles indicate less copolymer molecules present per micelle for the 2:5 PLA-PEG and Poloxamer 407 compared with the 1.5:2 PLA-PEG, yet (23) Wanka, G.; Hoffman, H.; Ulbricht, W. Colloid Polym. Sci. 1990, 268, 101-117. (24) Alexandridis, P.; Athanassiou, V.; Fukuda, S.; Hatton, T. A. Langmuir 1994, 10, 2604-2612. (25) Alexandridis, P.; Hatton, T. A. Colloids Surf., A: Physicochem. Eng. Aspects 1995, 96, 1-46.
with larger aggregate sizes being detected. The decrease in size with increased hydrophilic polymer content is in accordance with that expected.26 Micellization in PLA-PEG dispersions is signified by the sedimentation coefficient values presented in Table 5, with the suggestion of two micellar species in the sample of 2:5 PLA-PEG incorporating SBB. Although as noted above with the 2:5 PLA-PEG system, the difference between the sedimentation coefficients of the polymer alone and with drug incorporated probably represents differences in size, it is unlikely that the differences between these s20 values for the 1.5:2 PLA-PEG systems are significant, and they are probably due to thermodynamic nonideality (arising from coexclusion or unsuppressed polyelectrolyte phenomena). The values also verify the solubilization of the model drugs testosterone and SBB had occurred by incorporation into the micelles. Many lipophilic drugs, including steroids, barbiturates, and water-insoluble vitamins, have been formulated using solubilization, whereby the substances are brought into solution by incorporation into the hydrophobic core of micelles.27-29 Solubilization in micelles is generally considered to be a poor delivery system for parenteral administration, since micelles are unstable upon dilution, resulting in disassociation of drug and carrier, and in low circulation times. Covalent bonding of an antitumor drug (adriamycin) to micelle forming surfactant molecules has, however, been applied to increase in vitro and in vivo stability and to diminish drug toxicity.30 Drug loading in micelles is usually low, and when Barry and El Eini28 investigated the solubility of several steroids, including testosterone, in long chain poly(oxyethylene) surfactants (polyoxyethylated cetyl alcohols), they calculated that only two to nine molecules were associated with each micelle at 25 °C, representing a maximum of 3% of the micellar weight. This is similar in magnitude to the results observed for the incorporation of testosterone into the PLA-PEG micelles, but the values for SBB are considerably greater. Improved incorporation was seen with the 1.5:2 PLAPEG micelles compared with the 2:5 PLA-PEG micelles, despite their smaller size. This is most likely due to the increased hydrophobic content of the 1.5:2 PLA-PEG, which is a factor known to improve solubilization.31-33 It should be remembered that the number of drug molecules per micelle is an average value for the micellar solution. The micelle acts as a dynamic entity, with a fluctuating size and drug molecules can therefore pass back and forth between the micelle and the aqueous solution.33 Noticeably more SBB is solubilized than testosterone in PLAPEG dispersions. This can partly be accounted for by the greater hydrophobicity of SBB (log P ) 7.6) when compared (26) Elworthy, P. H.; Macfarlane, C. B. J. Chem. Soc. 1963, 907, 537. (27) Ismail, A. A.; Gouda, M. W.; Geneidi, A. S. Pharm. Ind. 1974, 36, 108. (28) Barry, B. W.; El Eini, D. I. D. J. Pharm. Pharmac. 1976, 28, 210-218. (29) Gstirner and Tata. Mitt. dt. pharm. Ges. 1958, 28, 191. (30) Yokoyama, M.; Miyauchi, M.; Yamada, N.; Okano, T.; Sakoura, Y.; Kataoka, K.; Inoue, S. J. Controlled Release 1990, 11, 269-278. (31) Al-Saden, A. A.; Whately, T. L.; Florence, A. T. J. Colloid Interface Sci. 1982, 90, 303-309. (32) Anton, P.; Ko¨berle, P.; Laschewsky, A. Makromol. Chem. 1993, 194, 1-27. (33) Jacobs, P. T.; Geer, R. D. and Anacker, E. W. J. Colloid Interface Sci. 1972, 39, 611-620.
PLA-PEG as a Drug Delivery System
Langmuir, Vol. 12, No. 9, 1996 2161
Table 7. Blood Clearance Values for High and Low Concentration PLA-PEG in Vivo Studies percentage of total dose injected (%) blood activity
indium-111 oxine
1.5:2 PLA-PEG 5.4% (w/v)
2:5 PLA-PEG 5.4% (w/v)
1.5:2 PLA-PEG 0.2% (w/v)
2:5 PLA-PEG 0.2% (w/v)
after 1 h after 2 h after 3 h
23.7 ( 14.9 11.3 ( 6.1 12.1 ( 9.7
22.7 ( 8.4 19.1 ( 5.3 19.3 ( 7.6
14.1 ( 4.9 10.9 ( 4.2 10.5 ( 1.8
0.0 ( 3.0 0.0 ( 0.0 2.1 ( 0.0
14.2 ( 9.3 0.0 ( 0.0 0.0 ( 0.0
with testosterone (log P ) 3.35) and by the more aromatic nature of SBB. It has been shown that aromatic compounds are solubilized in greater quantities34 and selectively over more aliphatic compounds,34,35 and whereas SBB contains five aromatic rings, testosterone only contains unsaturated bonds. However, this is unlikely to be the full explanation and further investigations are necessary. From Table 4, the higher Tc values seen for the 2:5 PLA-PEG copolymer became less pronounced with increasing Na2SO4 concentration and closer to the Tc values for the 1.5:2 PLA-PEG copolymer. This suggests that the increased PEG molecular weight with the 2:5 PLAPEG does not greatly enhance the steric stability of the micelles, and therefore the biodistribution profile would not be expected to differ much from the 1.5:2 PLA-PEG.36 This might be expected since Napper has shown that the Tc is insensitive to changing the molecular weight of the stabilizing moieties (Mw > 1000) and changes in the particle size.37 Tc is also unaffected by surface coverage of PEO chains above approximately two-thirds of the particle surface.38 The reduced stability of the PLA-PEG copolymers to added electrolyte, relative to the Poloxamer 407, reflects differences in molecular structure, chain conformation, and association. Two different concentrations of PLA-PEG copolymer dispersions were investigated as delivery systems in vivo in the rat model. Micelles were expected to be present in the blood at the higher concentration. The biodistribution results showed improved results for the 1.5:2 PLA-PEG compared with the 2:5 PLA-PEG at the higher concentration. ANOVA statistical analysis was performed on the results and showed that the liver uptake of the higher concentration 1.5:2 PLA-PEG copolymer was significantly improved over both 2:5 PLA-PEG systems and the lower concentration 1.5:2 PLA-PEG. This difference between the two copolymers probably reflects differences in the stability and packing of the copolymer molecules within the micelle structure. The closer packing of the 1.5:2 PLA-PEG may lead to an increased surface density of the stabilizing PEG at the surface of the micelle. Dunn et al.39 have shown recently a direct relationship between surface density of PEG and the uptake of a colloidal delivery system by the MPS, with increased surface density leading to a reduction in uptake. The similarity in Tc values between the two PLA-PEG copolymers at higher Na2SO4 concentrations also supports this. Although the 2:5 PLA-PEG has a longer PEG chain length, which has been shown to reduce uptake by the MPS,36 the increased surface density conferred on the 1.5:2 PLAPEG micelles due to the higher aggregation number and closer packing allows comparable steric stabilization. Free indium-111 oxine is partly eliminated from the circulation by the kidneys and partly complexed to plasma proteins such as transferrin. This accounts for the different (34) Nagarajan, R.; Barry, M.; Ruckenstein, E. Langmuir 1986, 2, 210-215. (35) Kumar, S. and Singh, H. N. Colloids Surf. 1992, 69, 1-4. (36) Illum, L.; Jacobsen, L. O.; Mu¨ller, R. H.; Mak, E.; Davis, S. S. Biomaterials 1987, 8, 113-117. (37) Napper, D. H. J. Colloid Interface Sci. 1970, 32, 106-114. (38) Napper, D. H. Trans. Faraday Soc. 1968, 64, 1701-1711. (39) Dunn, S. E.; Brindley, A.; Davis, S. S.; Davies, M. C.; Illum, L. Pharm. Res. 1994, 11, 1016-1022.
biodistribution profile to that seen with the PLA-PEG copolymers, even though the blood levels at the 3 h time point are not significantly different. At the lower concentrations, removal of the delivery system from the circulation to the liver is more rapid than that for free oxine suggesting that the oxine remains complexed, but in a nonmicellar form. This is possibly a unimer form where the steric shielding would be less effective. ANOVA showed that the 1.5:2 PLA-PEG blood circulation result was significantly better at 3 h with the higher concentration (5.4% (w/v) injected) but that there was no significant difference between the two concentrations of 2:5 PLA-PEG. No significant differences were found with ANOVA between any of the spleen values. Although the results for the 1.5:2 PLA-PEG copolymer are encouraging, longer circulation times have been obtained with micelles elsewhere. Stable micelles of less than 100 nm in size were prepared from an ABA block copolymer, poly(oxyethylene-b-isoprene-oxyethylene) (POE-PI-POE), by cross linking the chains in the hydrophobic (polyisoprene) core. These nonbiodegradable micelles were shown to remain in circulation in mice with a half-life in excess of 50 h.40 The above results indicated that the association complexes formed in aqueous dispersions of the water-soluble PLA-PEG copolymers were similar to those formed by the Poloxamer 407 PEO-PPO-PEO block copolymer; i.e., micellar structures were present. The PLA-PEG copolymers therefore present opportunities for replacing PEO-PPO-PEO copolymers in drug delivery applications. In this respect, the biodegradability and biocompatibility of the hydrophobic PLA component present distinct advantages over PEO-PPO-PEO copolymers. The two copolymers discussed in this paper are part of a range of PLA-PEG copolymers which are currently being evaluated for use as drug delivery systems. The studies presented above outline work performed on micellar PLA-PEG systems, which self-disperse in water both before and after freeze-drying, and form a promising basis for future research on novel PLA-PEG drug delivery systems, due to their extended circulation times and potential high drug loading, and we shall shortly describe the use of further block copolymers from this series which form solid colloidal particulate systems. The PLA-PEG micellar solutions are also being successfully deployed as a surfactant coating system for PLGA particles.41 Acknowledgment. Susan A. Hagan is grateful to SERC for the financial assistance provided. Thanks are due to Trevor Gray from the Department of Pathology and Wu Lin of the Department of Pharmaceutical Sciences for their help with the electron micrographs in Figure 2. Also thanks to Immo Fiebrig and Peter Morgan for their help with the ultracentrifuge measurements summarized in Tables 5 and 6. LA950649V (40) Rolland, A.; O’Mullane, J.; Goddard, P.; Brookman, L.; Petrak, K. J. Appl. Polym. Sci. 1992, 44, 1195-1203. (41) Stolnick, S.; Dunn, S. E.; Garnett, M. C.; Davies, M. C.; Coombes, A. G. A.; Taylor, D. C.; Irving, M.; Purkis, S. C.; Tadros, Th.; Illum, L.; Davis, S. S. Pharm. Res. 1994, 11, 1800-1808.
