In Vitro Model for the degradation of Alkylcyanoacrylate Nanoparticle

In vitro modelfor the degradation nanoparticles of aJ&ylcyanoacrylate Rainer H. ~~er, ~a~er~e Lherm,Jens He&ox% Patrick Co~e~ and UnticeMt6 de Paris-&d, Lab&at&? de Pharmacie GalMque France (Received 2 7 December et Biopharmacie. 5 rua J. 8. Cl&ment. F-92290 Chatenay-Malabry. 1989: eccepted 3 May 1990) A photometric assay was developed to study the surface erosion of polymeric nanoparticles. The hydrolytic deg~dation of polyalkylcyanoac~late particles was studied in different environments (NaOH, buffer, cell culture medium and serum). The influence of particle modification on the degradation rate was assessed. Particularly, the effect of polymer coating for particle targeting and fluorescence labelling was investigated. From the absorption data, a t50# and tlOO% can be calculated for fast degrading particles and obtained by an extrapolation in case of a slow degradation process. The degradation rate was found to decrease with increasing alkyl chain length from methyl-, ethyl-, isobutyl- to isohexylcyanoacrylate particles. Polymer coating and fluorescent labelling had little effect on the rate of degradation. Kevords: 8i~egr~atjon, drug deikery, ~anupa~icles Polymeric devices, such as implants” ‘, microparticles and nanoparticles4 are employed for controlled release and sitespecific drug delivery. Relatively long degradation times have been described for implants made from polyesters, e.g. 1 yr for poly(lactic acid) (PLA) and 3-6 wk for its copolymer with glycolic acid (PLA/GA)‘. The degradation of larger surgical PLA/GA implants has been reported to take more than 6 month”. Particles from these PLA/GA copolymers were, however, found to degrade almost completely within 4 wk’. Possible reasons are the absence of a lag time before weight loss occurs’, the larger surface area adsorbing hydrolysis accelerating proteins’ and a more porous particle structure’. Incorporation of additives accelerated drug release but had little effect on the degradation rate”.Fast-degrading alkylcyanoacrylate polymers should be more suitable for accumulation in the body for the production of particulate drug carriers’ ‘* ‘*. However, a too rapid degradation can lead to a burst release of degradation products, possibly causing cytotoxic effects l3 . Therefore, particle degradation rate is determined to provide basic information for the comprehension of possible cytotoxic effects of such products. Indeed, cytotoxicity was observed even with degradation products which are non-toxic metabolic compounds (e.g. lactic acid) but which appear in a very high local concentration’4. In addition, the degradation kinetics could change after modification of the particles. Coating nanoparticles for site-specific drug delivery (drug targeting) changed the Correspondence to Or P. Couvreur. surface properties such as hydrophobicity’5 which, in turn, may be supposed to modify the degradation profile of the polymer. To assess these effects, this paper proposes an in vitro model based on the measurement of the degradation of the nanoparticles themselves rather than, as generally proposed, on assaying the concentration of the final degradation products after total chemical decomposition of the polymer’*-‘*. MATERIALS Biomaterials 1990, Vol 11 October METHODS Materials Methylcyanoacrylate and ethylcyanoacrylate monomers (MCA and ECA) were provided by Loctite Ltd (Ireland), isobutylcyanoacrylate (IBCA) was purchased from Sigma (USA), isohexylcyan~c~late (IHCA) was a gift from Sopar (Belgium). The other compounds used for particle preparation were reagent or pharmacopeia grade and used without further purification. Propidium iodide, the fluorescent marker for the nanoparticles, was purchased from Sigma (USA). The polymers for the coating of nanoparticles Poloxamer 407 and Poloxamine 908 were obtained from BASF-Wyandotte (USA). Cell culture medium and related chemicals were obtained from Flobio (Paris, France), fetal calf serum from IBF (Paris, France). 0 590 AND 1990 Butterworth-Heinemann Ltd. 0142-96 1 Z/90/080590-06 Oegrauon of aiky’cyenoacrykte METHODS Nanoparticle preparation. Nanoparticles were prepared by polymerization of the alkylcyanoacrylate monomers in an aqueous solution as described previously”. The polymerization medium contained HCI (0.01 N), Tween 20 (0.36%) for the production of polymeth~~yanoac~late particles (PMCA) or dextran 70 (1%) and glucose (5%) for the polymerization of polyethyf- (PECA), polyisobutyl- (PISCA) and polyisohexylcyanoacrylate particles (PIHCA). The particle radii were 102 nm (PMCA), 155 nm (PECA), 165 (PIBCA) and 68 nm (PIHCA). The nanoparticles were used in the experiments dispersed in their polymerization medium (nanopa~icle suspension f % w/w). Nanoparticle modification. PIHCA nanoparticfe suspension (1% w/w) was mixed with an equal volume of a Poloxamer 407 or Poloxamine 908 solution (1% w/w polymer in distilled water) and incubated overnight. The polymers adsorbed on the particles, forming a hydrophilic coat. These coating layers were measured by photon correlation spectroscopy (PCS) and found to be 6.4 nm (+ 0.3 nm) for Poloxamer 407 and 7.5 nm (F 0.4 nm) for Poloxamine 908. The coating layers increased the PCS panicle radius by 6.47.5 nm, which is equivalent to about 5%, which can be a significant increase, detected by PCS. Aggregation of the particles was monitored using the PCS polydispersity index. Swelling effects were excluded by using uncoated particles as a control. No swelling period was observed with uncoated particles during the incu~tion, because equilibrated aqueous particle suspensions were used. This avoided swelling effects occurring after redispersion of freeze-dried particles. Nanoparticles were ffuorescently labelled by addition of propidium iodide during the polymerization process of the cyanoacrylate monomers*‘, leading to a propidium iodide content of 0.3% (w/w). The content was determined by measuring the remaining free propidium iodide in the dispersion medium by spe~trophotomet~ (absorption at 285 nm). Degradation experiments. Degradation of the nanoparticles was followed by spectrophotometric measurements using a Uvikon 8 10 (Kontron Instruments, FRG). A multiple cuvette carriage allowed the performance of five degradation studies simultaneously. The reduction in light transmission caused by the nanoparticles was determined at 400 nm, a wavelength with a linear relationship between r~uction in light transmission and nanoparticle concentrated (0.2-0.8 mg/ ml). Although the reduction in light transmission is caused by light scattering and not by absorption, the term absorption is used in this paper. The standard deviation of the absorption data was generally found to be between 5 and 10%. Degradation was studied during incubation of nanoparticles in aquaous solutions with a pH adjusted by NaOH (pH 10 and 12) and in isotonic phosphate buffer at pH 7.4. Experiments were also performed using MEM cell culture medium and fetal calf serum as the incubation medium. THEORETICAL Spectrophotometric assay Absorption measurements with a spectrometer can be employad when the degradation takes place by surface erosion rather than by bulk hydrolysis. Polyalkylcyano- nanoparticles: R.H. MiNer er al. acrylate nanoparticles degrade mainly by surface erosion, as indicated by the size reduction during the degradation process2’. An immediate, continuous decrease in particle diameter was shown by size measurements using PCS”‘. PCS can be used to follow particle degradation as a function of size, but only for the initial phase of the degradation process. After the initial drop in particle size, a plateau and a size increase were observed”. This effect occurred when most small particles had disappeared. Larger particles and aggregates remained in the sample. Owing to the measuring principle of PCS, which is based on the scattered light intensity, the larger particles contribute more to the calculated mean PCS diameter. This led to the size increase found by PCS. Although the disappearance of particles by degradation can be monitored using the PCS count rate, this is a relatively expensive apparatus for particle concentration measurement and limited in its availability. It was therefore replaced by a spectrophotometric assay. The measured absorption depended on particle size but also on particle concentration. At the beginning and at the intermediate phase of the degradation, the reduction in the measured absorption was only due to the size decrease. Towards the end, the concentration was reduced, due to the dissolution of particles contributing to the observed decay in absorption. This decreased particle concentration is detected by an absorption measurement, but leads to an apparent particle size increase in PCS measurements2’. For an ideal monodisperse particle population, the reduction in absorption will be determined by the size decrease until final dissolution of the degraded particles. Data analysis Absorption measurements provided information about the time required for 100% degradation of the particles. For fast degrading nanoparticles (e.g. PMCA and PECA), a decay to zero absorption can be measured within minutes to a few hours, representing the time required for 100% degradation t(deg) 100%. Degradation media containing physiol~ical salt concentrations (e.g. 0.14 M NaCl) caused some particle aggregation after a few hours. For slowly degrading nanoparticles (e.g. PIHCA) a decrease in absorption occurred over larger time periods (e.g. 24 h) and therefore did not permit us to determine directly t (deg) 100%. To characterize slow degrading particles in physiological salt concentrations, a t 50% of the absorption can be used instead (r(abs)50%). The semilogarithmi~ plot of the absorption data yields a straight line, indicating pseudo-first order kinetics for the hydrolytic degradation of the polymer. From this plot, a t50% value of absorption (t(abs)50%) based on measured data can be obtained tocharacterize the degradation velocity. If a 50% degradation does not occur during the observation period or if a particle aggregation takes place, a t(abs)50% can be extrapolated. It should be noted that the absorption decays, not linearly, but exponentially with decreasing particle size. For example, the r(abs)50% of a 200 nm particles corresponds to a size decrease of < 50% when considering the relation between scattering intensity and particle size for particles up to 200 nm. Calculating the reduction in the volume of spherical particles when reducing the size from 200 to 130 nm, this is equivalent to a decrease in polymer mass of 60%. For all particles, a theoretical t(abs)O% which is equivalent to a r(deg)lOO% can be extrapolated from the semil~ar~thmic plot. From the slope of the semilogarithmic plot, a degradation rate constant Biomaterials 1990. Vol 7 I October 591 Begration of al~l~affo~~late nanopartictes: Rff. artier et al. 0 0,O 40 20 0 20 60 TIME (min) I I 0,2 0,4 1 I 0,s 0,8 I,0 I TiME(hours) Figure ;I Degradation of PACA nanoparticles in NaOH, the fast degrading PMCA cm) and PECA (0) at pH 10, the slower degrading P&Y (+) and PitfCA (@I_) pfi 12. at kfabs) can be obtained or alternatively f (abs)50% (K = inZ/f (abs)50%). calculated using Figure 2 Degrsdatidn of PMCA fmj and PECA {O) isotonic, physiotogica~ phosphate buffer (pH 7.4). nanoparticles in This yielded 20 and 75 min for PMCA and PECA, respectively. These values were close to the measured zero absorption values in Figure 2. The degradation rate constants k(abs) calculated from the slope were -0.071 and -0.223 min-’ for PMCA and PECA, respectively. The semilogarithmic plot of the absorption values obtained with PIBCA and PIHCA nanoparticles in NaOH 2,o RESULTS AND DISCUSSION Influence of alkyl~hain on particle degradation Polyalkylcyanoacrylates can be degraded by hydrolysis in sodium hydroxide whereby the rate of degradation decreases with increasing length 0; the alkyl chain”. PMCA and PECA nanoparticles degrade very quickly in NaOH at pH 10 (Figure I) with corresponding t(deg)lOO% values of C. 3 min. The degradation velocity of PIBCA and PIHCA is too slow at pH 10 to measure a distinct decrease in absorption over a period of 1 h. At pH 12, a fast initial decay was observed. The decrease in absorption decelerated after some time, due to particle aggregation caused by the ionic strength of the dispersion medium. In such cases, a direct d~ermination of t(deQ) 100% was not possible and could only be extrapolated. The differences in the hydrolytic degradation of the nanoparticles corresponded to the relative decomposition rate of the polymers’ ‘. There was a large difference between the fast-degrading PMCA and PECA and the slow-degrading PIBCA and PIHCA, as indicated by the one hundred fold higher hydroxyl ion concentration required. This should result in different kinetics of drug liberation and release of possibly cytotoxic degradation products. The faster release of degradation products affects the cytotoxicity of the nanoparticles’3. The hydroi~ic degradation of PMCA and PECA in NaOH at pH 10 was too fast to differentiate between the two. Studies in physiological phosphate buffer at pH 7.4 showed a faster degradation for PMCA and yielded information about the hydrolytic contribution to the degradation under physiological conditions (figure 2). The absorption decaystozero(f(abs)O%)in20 minforPMCAandin60 min for PECA. From these experiments, nanoparticles can be placed in order of decreasing hydrolytic degradation velocity: PMCA > PECA > PIBCA > PIHCA. By plotting the absorption data in a semilogarithmic manner and a straightline fit, the theoretical r(d~)~OO% was obtained by the interception with thex-axis (Figure 3s). 592 Homaterials t990, Vol 11 October 0 20 80 40 80 TIME (min) Figure3e Semilogarithmic plot of absorption dare for PMCA (ml and PECA (0) nanopariicles in PBS at pH 7.4. The interception of tl;e fitted straight Iine with the x-axis yields the theoretical t (degl 1DO%. 2,o 196 03 090 0,5 1,o 1,s 2,o 2,s TIME (hours) Figure 3b Semilogarithmic plot of absorption data for PfECA {8I and PIHCA fl/ nanoparticles in NaOfi a? pH 12. The extrapolation of the fitted straight line u&g the data~jnts of the initial degradation phase yields the theoreticai tfdegj t 00%. &g&ion in different media The influence of the medium on the degradation of PECA nanoparticles at pH 7.4 is shown in Figure 4. Compared to the hydrolytic degradation in phosphate buffer, acceleration was observed in the cell culture medium. This acceleration was attributed to the presence of esterases because of the 5% FCS in the medium. Esterases are still active in serum, e.g. the activity of choline esterase in FCS was found to be 770 munit/ml. However, inactivation of the serum reduced the esterase activity. Heating to 40°C for 30 min reduced the activity moderately to 650 munit/ml, whereas inactivation at 80°C (2 min) removed the activity (18 mu&/ml). Therefore active FCS was used in the studies. The contribution of esterases was described for the in vivu degradation of alkyl~yanoac~iates’6. The further accelerated particle degradation observed in FCS supported this assumption (Figure nanopart~cies: R.H. Miiller et al. loo yielded only a straight line for the initial phase of the degradation. interference of the decreasing absorption caused by particle degradation and increasing absorption due to particle aggregation led to the positive deviation (Figure 3b). The theoretical t(deg)lOO% in NaOH (pH 12) obtained by extrapolation were 18 min (PIBCA) and 127 min (PIHCA), the velocity constants -15.14 and -2.1 7 h-‘, respectively. Degradation of ai~l~anoac~tare s 20 1 8 6 12 I I 18 24 TIME (hours) Figure 5 Degradation of Plt?CA in cell culture medium fetal calf serum) (-G-j and in fetal calf serum l-t). (containing 5% 4). The enzymatic degradation played a more important role for the slowly degrading PIBCA and PIHCA. In buffer at pH 7.4, no reduction in absorption could be observed for PIBCA over a period of 10 h, but a decrease was found in cell culture medium and, still more marked, in serum (Figure 5). This was in agreement with in viva degradation studies describing the excretion of 80% of radioactive PIBCA polymer within 3 d’*. For PIBCA, the semilogarithmic plot of the absorption data in cell culture medium showed a positive deviation from the straight line (Figure 6). The straight line in Figure 6 was fitted using only data points obtained during the first 2 h, when distortion by aggregation was minor. PIBCA data points measured between 2 and 18 h are above the fitted line (positive deviation) due to aggregation and must not be included in the fit. This is an identical effect to that discussed above (Figure 36). In serum, the particles were sterically 8 18 12 TIME 24 (hours) Figure 6 Semiiogarifhmicplot of absorption data from PlsCA {Figure 51 in ceil culture medium (0) and in fetal calf serum (mJ. Only data points obtained during the first 2 h of degradation were used for the fit (not all data points used for the calculation are plotted in the figure): for explanation compare text. stabilized against aggregation by an adsorbed serum albumin layer. This excluded a positive deviation (Figure 6). On the contrary, a tendency towards a negative deviation was observed when applying the same fit procedure as for the culture medium (fit through data points obtained within the first 2 h). This might be possibly due to some superposition of the pseudo-first order degradation kinetic in aqueous media with the zero order kinetics resulting from enzymatic degradation. This effect was only observed with slowly degrading nanoparticles where the enzymatic degradation played a more important role. Degradation TIME (min) Figure (-II-). 4 Degradation of PECA nanopanicles atpH 7.4inphosphate cell culture medium (+) and fetal calf serum (-+-I. buffer of surface-modified nanoparticles To determine whether polymer coating layers on nanoparticles could influence the degradation velocity, Poloxamer 407 and Poloxamine 908-coated PIHCA nanoparticles were degraded in NaDH at pH 12 (Figure 7). No distinct differences were found over the initial degradation period of 1 h. During this time, particle aggregation interferred little with the assay of the uncoated particles (deceleration of decrease in absorption); aggregation cannot take place for Biomaterials 1990, Voi 11 October 593 Degration of af~~~anoa~~iate nanoparticles: ff. H. ~~tlef et a/. 40 20 0,o 0;2 0,s 0,8 0.8. 1 0 2 3 TIME TIME (hours) Figure 7 Degradation of uncoated PM&A fi3) particles and particles coated with PIHCA 407 (0) and PIHCA 908 /8) in NaOH (pH 121 over a period of 1 h. 1 *m 1,O (hours) 4 5 Figure8 Degradation of uncoated PIHCA (m) particles and particles coated with PIHCA 407 (0) and PIHCA 908 (+) in setum over a period ofdh. the sterically stabilized Poloxamer- and Poloxamine-coated particles l5 . The similar decay in absorption for coated and uncoated nanoparticles indicated no effect of the coating layers on degradation velocity. After 60 min increased aggregation was observed for the uncoated particles as shown in Figure 36. However, after 145 min the coated particles were fully degrade which is close to the theoretical value of 127 min calculated from the fit in Figure 36. ’ An identical decay for uncoated PIHCA and coated PIHCA particles was also observed in serum (Figure 8). In serum, the uncoated particles are sterically stabilized by adsorbed serum proteins, minimizing interference byaggreg&ion. In serum the Poloxamer and Poloxamine coating also had no effect on the enzymatic d~radation process. This allows the use of Poloxamer 407 and Poloxamine 908 as aggregation-avoiding agents, e.g. in long-term degradation studies where particle aggregation can cause distortions. Degradation Of flUOr88C8ntly labelled nanOpartiCleS The incor~ration of fluorescent markers can change particle properties such as surface hydrophobicityor particle charge15. However, no effect was observed on the degradation when incubating PIHCA nanoparticles labelled with propidium iodide in NaOH and in cell culture medium (Figures 9 and 70). This result permits the use of fluorescently labelled cyanoacrylate nanoparticles for easy detection in cell culture studies. The degradation rate, an important parameter determining the cytotoxicity of the nanoparticles’, was therefore not affected by the labelling. This spectrophotometric method could also be employed to measure the effects of drug incorporation on to nanoparticle degradation. The dependence of the release rate of incorporated compounds on the degradation can be assessed by a simultaneous assay of particle degradation and drug release rate. The photometric assay was found to be simple, fast and easy for the determination of nanoparticle degradation. It required no sophisticated equipment as for PCS. Basic information can be obtained about the relative Biomaterials ) OS0 r 6 I I 0,6 0.4 0.2 0.8 I 180 TIME (hours) Figure 9 Degradation of PlHCA (OJ nanoparticles and particles fiuorescently labelled with PIHCAIP (+) in NaOH (pH 12) within 1 h. 60 1 60 I 0 I I 6 12 I 18 I 24 TIME (hours) Figure 10 Degradation of PIHCA (0) nanoparticles and particles fluoresce&y labekl with PlHCAiP/eJ in cell wfture medium within apet&i of 24 h. CONCLUSIONS 594 0 1990, Vol I 1 October hydrolytic degradation of different polymeric particles in aqueous media. Furthermore, hydrolytic degradation of slowly degrading particles can be measured in accelerated assays, e.g. increasing hydroxyl ion concentration. Degrarion of alkylcyanoacrylate In contrast to the chemical methods which measure sometimes complex degradation products, the drug releasedetermining process of particle erosion is directly monitored in our method. Finally, the degradation can be studied in different media to simulate degradation under in vitro or in viva conditions (cell cultures, incubation with blood). Assessment of the effect of nanoparticle modification (surface coating, labelling, drug incorporation) is also possible. 8 9 10 11 ACKNOWLEDGEMENT 12 The research was supported within the framework of the Biotechnology programme by the Commission of the European Community, to whom we would like to express our sincere thanks. 13 14 REFERENCES Holland, S.J.. Tighe. 8.J. and Gould, P.L.. Polymers for biodegradable medical devices. 1. The potential of polyesters as controlled macromolecular release systems, J. Confr. Rel. 1986, 4, 155-l 80 Wise, D.L., Rosenkrantz, H., Gregory, J.8. and Esber, H.J., Longterm controlled delivery of levonorgestrel in rats by means of small biodegradable cylinders, .I. Pharm. Pharmacol. 1980, 32, 399403 Davis. S.S., Illum, L.. McVie. J.G. andTomlinson. E. (Eds), Microspheres and drug therapy, in Pharmaceutical, Immunological and Medical Aspects, Elsevier Science Publishers, Amsterdam, 1984 Verdun. C.. Couvreur. P.. Vranckx, H., Lenaerts. V. and Roland, M.. Development of a nanoparticle controlled release formulation for human use, J. Confr. &I. 1986, 3. 205-2 10 Tice. T.R. and Cowsar. DR.. Biodegradable controlled-release parenteral systems, Pharm. Tech. 1984, 11, 26-36 Vert. M.. Structure et comportement des polymeres. Exemples des polymeres bioresorbables. S.TIP. Pharma 1987. 3. 216-222 Ogawa. Y., Okada. H., Yamamoto, M. and Shimamoto, T., In viva 15 16 17 18 19 20 21 nanopanicles: R. H. Miller et al. release profiles of leuprolide acetate from microcapsules with polylactic acids or copoly(lactic/glycolic) acids and in viva degradation of these polymers. Chem. Pharm. Bull. 1988, 36, 2576-258 1 Makino, K.. Ohshima. H. and Kondo, T.. Effects of plasma proteins on degradation properties of poly(L-lactide) micro-capsules, Pharm. Res. 1987,4, 62-65 Koosha, F., Preparation and characterisation of biodegradable polymeric drug carriers. PhD Thesis, Nottingham University. 1989 Kubota, M.. Nakano, M. and Juni. K.. Mechanism of enhancement of the release rate of aclarubicin from poly-p-hydroxybutyric acid microspheres by fatty acid esters, Chem. Pharm. Bull. 1988, 36, 333-337 Leonard, F.. Kulkarni, R.K.. Brandes. G., Nelson, J. and Cameron, J.J.. Synthesis and degradation of poly(alkyl-acyano acrylates). J. Appl. Polym. Sci. 1966, 10, 259-272 Grislain, L.. Couvreur, P., Lenaerts.V., Roland, M.. Deprez- Decampeneere, D. and Speiser, P., F’harmacokinetics and distribution of a biodegradable drug-carrier, Inr. J. Pharm. 1983, 16, 335-345 Lherm, C.. Miiller, R.H. and Couvreur, P., Alkylcyanoacrylate drug carriers II: Cytotoxicity of cyanoacrylate nanoparticles with different alkyl chain length, J. Pharm. Sci. (Submitted) Smith, A. and Hunneyball, I.M., Evaluation of poly(lactic acid) as a biodegradable drug delivery system for parenteral administration. lnt J. Pharm. 1986, 30. 215-220 Muller, R.H., Colloidal Carriers for controlled drug delivery: modification. characterization and in viva distribution, Habilitationsschrift (German DS). University of Kiel (FRG), 1989 Lenaerts, V., Couvreur, P., Christiaens-Leyh, D., Joiris, E., Roland, M., Rollman. 8. and Spaiser. P.. Degradation of poly(isobutylcyanoacrylate) nanoparticles, Biomaterials 1984, 5, 65-70 Vezin. W.R. and Florence, A.T.. In vitro heterogenous degradation of poly(n-alkyl a-cyanoactylates). J. Biomea. Mafer. Res. 1960. 14, 93 Leyh, D., Couvreur. P., Lenaerts, V., Roland, M. and Speiser, P., Etude du mChanisme de degradation des nanopanicules de polycyanoacrylate d’alkyle, Labo-Pharma - Probl. Tech. 1984, 32, 1OO- 104 Kubiak. C., Manil, L.andCouvreur. P.,Sorptive propertiesofantibodies onto cyanoacrylic nanopanicles. lnt J. Pharm. 1988, 41, 18 1- 187 Milller, R.H., Lherm, C., Herbort. J. and Couvreur, P., Propidium iodide loaded polyalkylcyanoacrylate particles: labelling conditions and loading capacity, Colloid B Polymer Sci. (Submitted) Mtiller, R.H., Lherm, C.. Herbert, J. and Couvreur, P.. Alkylcyanoacrylate drug carriers I: Physicochemical characterization of nanoparticles with different alkyl chain 1ength.J. Pharm. Sci. (Submitted) Biomaterials 1990, Vol 7 1 October 595